hexsha
stringlengths 40
40
| size
int64 6
14.9M
| ext
stringclasses 1
value | lang
stringclasses 1
value | max_stars_repo_path
stringlengths 6
260
| max_stars_repo_name
stringlengths 6
119
| max_stars_repo_head_hexsha
stringlengths 40
41
| max_stars_repo_licenses
list | max_stars_count
int64 1
191k
⌀ | max_stars_repo_stars_event_min_datetime
stringlengths 24
24
⌀ | max_stars_repo_stars_event_max_datetime
stringlengths 24
24
⌀ | max_issues_repo_path
stringlengths 6
260
| max_issues_repo_name
stringlengths 6
119
| max_issues_repo_head_hexsha
stringlengths 40
41
| max_issues_repo_licenses
list | max_issues_count
int64 1
67k
⌀ | max_issues_repo_issues_event_min_datetime
stringlengths 24
24
⌀ | max_issues_repo_issues_event_max_datetime
stringlengths 24
24
⌀ | max_forks_repo_path
stringlengths 6
260
| max_forks_repo_name
stringlengths 6
119
| max_forks_repo_head_hexsha
stringlengths 40
41
| max_forks_repo_licenses
list | max_forks_count
int64 1
105k
⌀ | max_forks_repo_forks_event_min_datetime
stringlengths 24
24
⌀ | max_forks_repo_forks_event_max_datetime
stringlengths 24
24
⌀ | avg_line_length
float64 2
1.04M
| max_line_length
int64 2
11.2M
| alphanum_fraction
float64 0
1
| cells
list | cell_types
list | cell_type_groups
list |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
cb905ed8631284e37fd8cda6a9207a0dddc8a603
| 2,083 |
ipynb
|
Jupyter Notebook
|
docs/src/notebooks/06_Strings.ipynb
|
diegozea/JuliaForBioinformatics
|
259236533283c4e582c1d01c32141a9f38930abd
|
[
"MIT"
] | 1 |
2020-12-28T01:10:51.000Z
|
2020-12-28T01:10:51.000Z
|
docs/src/notebooks/06_Strings.ipynb
|
diegozea/JuliaForBioinformatics
|
259236533283c4e582c1d01c32141a9f38930abd
|
[
"MIT"
] | null | null | null |
docs/src/notebooks/06_Strings.ipynb
|
diegozea/JuliaForBioinformatics
|
259236533283c4e582c1d01c32141a9f38930abd
|
[
"MIT"
] | null | null | null | 20.223301 | 120 | 0.513202 |
[
[
[
"empty"
]
]
] |
[
"empty"
] |
[
[
"empty"
]
] |
cb905f8d78655707e84899a1ab0d62036ccf6e59
| 161,385 |
ipynb
|
Jupyter Notebook
|
notebook/Data_read.ipynb
|
maxi1571/USGS-LIDAR-on-AgriTech
|
c3064449822d68abdd2c933dfa4c02a142122f2a
|
[
"MIT"
] | null | null | null |
notebook/Data_read.ipynb
|
maxi1571/USGS-LIDAR-on-AgriTech
|
c3064449822d68abdd2c933dfa4c02a142122f2a
|
[
"MIT"
] | 7 |
2021-08-24T13:34:27.000Z
|
2021-08-24T13:40:46.000Z
|
notebook/Data_read.ipynb
|
maxi1571/USGS-LIDAR-on-AgriTech
|
c3064449822d68abdd2c933dfa4c02a142122f2a
|
[
"MIT"
] | null | null | null | 349.318182 | 147,292 | 0.931245 |
[
[
[
"import pdal",
"_____no_output_____"
],
[
"import warnings\nwarnings.filterwarnings('ignore')\n# import geoplot as gplt\n# import geoplot.crs as gcrs\nimport geopandas as gpd\nimport imageio\nimport pathlib\nimport mapclassify as mc\nimport numpy as np\nimport laspy\nimport rasterio\nfrom rasterio import mask\nimport folium\nimport matplotlib.pyplot as plt\nimport logging\nimport sys\nfrom shapely.geometry import Polygon\nsys.path.append(\"/home/michael/USGS-LIDAR-on-AgriTech\")\nfrom fetch_data import FetchData\nfrom script.visualize import Visualize",
"_____no_output_____"
],
[
"MINX, MINY, MAXX, MAXY = [-93.756155, 41.918015, -93.756055, 42.918115]\npolygon = Polygon(((MINX, MINY), (MINX, MAXY), (MAXX, MAXY), (MAXX, MINY), (MINX, MINY)))\nregion = \"IA_FullState\"",
"_____no_output_____"
],
[
"data_fetcher = FetchData(polygon, region)",
"_____no_output_____"
],
[
"data=data_fetcher.run_pipeline()\nprint(type(data))\ndf = data_fetcher.get_elevation(data)\nprint(df.info())\nprint(df)",
"_____no_output_____"
],
[
"## Plot raster/tif image\n# --------------------\ndef plot_raster(rast_data, title=''):\n \n \"\"\"\n Plots raster tif image both in log scale(+1) and original verion\n \"\"\"\n fig, (axlog, axorg) = plt.subplots(1, 2, figsize=(14,7))\n im1 = axlog.imshow(np.log1p(rast_data)) # vmin=0, vmax=2.1)\n# im2 = axorg.imshow(rast_data)\n\n plt.title(\"{}\".format(title), fontdict = {'fontsize': 15}) \n plt.axis('off')\n plt.colorbar(im1, fraction=0.03)",
"_____no_output_____"
],
[
"# Read raster/tif file\n# --------------------\niowa_tif = '../data/tif/iowa.tif'\nraster_iowa = rasterio.open(iowa_tif)\niowa_data = raster_iowa.read(1)",
"_____no_output_____"
],
[
"type(iowa_data)",
"_____no_output_____"
],
[
"count = iowa_data[iowa_data > 0].sum()\ncount",
"_____no_output_____"
],
[
"title = 'Log scaled (+1) and No Scale Raster plots'.format(count)\nplot_raster(iowa_data, title)",
"_____no_output_____"
],
[
"ep.hist(array, title=[\"Band 1\", \"Band 2\", \"Band 3\", 'Band 4'])\nplt.show()",
"_____no_output_____"
],
[
"# get shp from tif\nfrom glob import glob\ndef get_shp_from_tif(tif_path:str, shp_file_path:str) -> None:\n raster = rasterio.open(tif_path)\n bounds = raster.bounds\n\n df = gpd.GeoDataFrame({\"id\":1,\"geometry\":[box(*bounds)]})\n \n # save to file\n df.to_file(shp_file_path)\n return df\n print('Saved..')",
"_____no_output_____"
],
[
"shp_df = get_shp_from_tif(\"../data/tif/iowa.tif\", \"../data/shp\")",
"_____no_output_____"
],
[
"shp_df",
"_____no_output_____"
],
[
"def select_name(name:str):\n name_ls = []\n names_list = io.open('../data/filename.txt', encoding='UTF-8').read().strip().split('\\n')\n if name in names_list:\n return name\n if name == 'all':\n return names_list\n else:\n for words in names_list:\n words_ls = words.split('_')\n if name in words_ls:\n name_ls.append(words)\n else: continue\n if name_ls == []:\n print(f\"Name - ({name}) not found, input a valid name\")\n return None\n else: return name_ls",
"_____no_output_____"
],
[
"y = select_name('all')",
"_____no_output_____"
],
[
"import json",
"_____no_output_____"
],
[
"with open('../pipeline.json', 'r')as json_file:\n dict_ob = json.load(json_file)",
"_____no_output_____"
],
[
"dict_ob",
"_____no_output_____"
],
[
"piplines = pdal.Pipeline(json.dumps(dict_ob))",
"_____no_output_____"
],
[
"piplines",
"_____no_output_____"
],
[
"piplines.execute()\nmetadata = piplines.metadata\nlog = piplines.log",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb906dbe6b699aa1cf1415141623eab636d7013d
| 14,581 |
ipynb
|
Jupyter Notebook
|
03-Methods and Functions/03-Function Practice Exercises.ipynb
|
vjsniper/Python
|
5331fb5c4d922d3eca884ac4704dbf8efa3e25d5
|
[
"Apache-2.0"
] | 4 |
2019-11-09T13:46:44.000Z
|
2021-09-23T17:58:02.000Z
|
03-Methods and Functions/03-Function Practice Exercises.ipynb
|
vjsniper/Python
|
5331fb5c4d922d3eca884ac4704dbf8efa3e25d5
|
[
"Apache-2.0"
] | 1 |
2021-05-10T10:49:44.000Z
|
2021-05-10T10:49:44.000Z
|
03-Methods and Functions/03-Function Practice Exercises.ipynb
|
vjsniper/Python
|
5331fb5c4d922d3eca884ac4704dbf8efa3e25d5
|
[
"Apache-2.0"
] | 6 |
2019-11-13T13:33:30.000Z
|
2021-10-06T09:56:43.000Z
| 20.364525 | 270 | 0.50312 |
[
[
[
"# Function Practice Exercises\n\nProblems are arranged in increasing difficulty:\n* Warmup - these can be solved using basic comparisons and methods\n* Level 1 - these may involve if/then conditional statements and simple methods\n* Level 2 - these may require iterating over sequences, usually with some kind of loop\n* Challenging - these will take some creativity to solve",
"_____no_output_____"
],
[
"## WARMUP SECTION:",
"_____no_output_____"
],
[
"#### LESSER OF TWO EVENS: Write a function that returns the lesser of two given numbers *if* both numbers are even, but returns the greater if one or both numbers are odd\n lesser_of_two_evens(2,4) --> 2\n lesser_of_two_evens(2,5) --> 5",
"_____no_output_____"
]
],
[
[
"def lesser_of_two_evens(a,b):\n pass",
"_____no_output_____"
],
[
"# Check\nlesser_of_two_evens(2,4)",
"_____no_output_____"
],
[
"# Check\nlesser_of_two_evens(2,5)",
"_____no_output_____"
]
],
[
[
"#### ANIMAL CRACKERS: Write a function takes a two-word string and returns True if both words begin with same letter\n animal_crackers('Levelheaded Llama') --> True\n animal_crackers('Crazy Kangaroo') --> False",
"_____no_output_____"
]
],
[
[
"def animal_crackers(text):\n pass",
"_____no_output_____"
],
[
"# Check\nanimal_crackers('Levelheaded Llama')",
"_____no_output_____"
],
[
"# Check\nanimal_crackers('Crazy Kangaroo')",
"_____no_output_____"
]
],
[
[
"#### MAKES TWENTY: Given two integers, return True if the sum of the integers is 20 *or* if one of the integers is 20. If not, return False\n\n makes_twenty(20,10) --> True\n makes_twenty(12,8) --> True\n makes_twenty(2,3) --> False",
"_____no_output_____"
]
],
[
[
"def makes_twenty(n1,n2):\n pass",
"_____no_output_____"
],
[
"# Check\nmakes_twenty(20,10)",
"_____no_output_____"
],
[
"# Check\nmakes_twenty(2,3)",
"_____no_output_____"
]
],
[
[
"# LEVEL 1 PROBLEMS",
"_____no_output_____"
],
[
"#### OLD MACDONALD: Write a function that capitalizes the first and fourth letters of a name\n \n old_macdonald('macdonald') --> MacDonald\n \nNote: `'macdonald'.capitalize()` returns `'Macdonald'`",
"_____no_output_____"
]
],
[
[
"def old_macdonald(name):\n pass",
"_____no_output_____"
],
[
"# Check\nold_macdonald('macdonald')",
"_____no_output_____"
]
],
[
[
"#### MASTER YODA: Given a sentence, return a sentence with the words reversed\n\n master_yoda('I am home') --> 'home am I'\n master_yoda('We are ready') --> 'ready are We'\n \nNote: The .join() method may be useful here. The .join() method allows you to join together strings in a list with some connector string. For example, some uses of the .join() method:\n\n >>> \"--\".join(['a','b','c'])\n >>> 'a--b--c'\n\nThis means if you had a list of words you wanted to turn back into a sentence, you could just join them with a single space string:\n\n >>> \" \".join(['Hello','world'])\n >>> \"Hello world\"",
"_____no_output_____"
]
],
[
[
"def master_yoda(text):\n pass",
"_____no_output_____"
],
[
"# Check\nmaster_yoda('I am home')",
"_____no_output_____"
],
[
"# Check\nmaster_yoda('We are ready')",
"_____no_output_____"
]
],
[
[
"#### ALMOST THERE: Given an integer n, return True if n is within 10 of either 100 or 200\n\n almost_there(90) --> True\n almost_there(104) --> True\n almost_there(150) --> False\n almost_there(209) --> True\n \nNOTE: `abs(num)` returns the absolute value of a number",
"_____no_output_____"
]
],
[
[
"def almost_there(n):\n pass",
"_____no_output_____"
],
[
"# Check\nalmost_there(104)",
"_____no_output_____"
],
[
"# Check\nalmost_there(150)",
"_____no_output_____"
],
[
"# Check\nalmost_there(209)",
"_____no_output_____"
]
],
[
[
"# LEVEL 2 PROBLEMS",
"_____no_output_____"
],
[
"#### FIND 33: \n\nGiven a list of ints, return True if the array contains a 3 next to a 3 somewhere.\n\n has_33([1, 3, 3]) → True\n has_33([1, 3, 1, 3]) → False\n has_33([3, 1, 3]) → False",
"_____no_output_____"
]
],
[
[
"def has_33(nums):\n pass",
"_____no_output_____"
],
[
"# Check\nhas_33([1, 3, 3])",
"_____no_output_____"
],
[
"# Check\nhas_33([1, 3, 1, 3])",
"_____no_output_____"
],
[
"# Check\nhas_33([3, 1, 3])",
"_____no_output_____"
]
],
[
[
"#### PAPER DOLL: Given a string, return a string where for every character in the original there are three characters\n paper_doll('Hello') --> 'HHHeeellllllooo'\n paper_doll('Mississippi') --> 'MMMiiissssssiiippppppiii'",
"_____no_output_____"
]
],
[
[
"def paper_doll(text):\n pass",
"_____no_output_____"
],
[
"# Check\npaper_doll('Hello')",
"_____no_output_____"
],
[
"# Check\npaper_doll('Mississippi')",
"_____no_output_____"
]
],
[
[
"#### BLACKJACK: Given three integers between 1 and 11, if their sum is less than or equal to 21, return their sum. If their sum exceeds 21 *and* there's an eleven, reduce the total sum by 10. Finally, if the sum (even after adjustment) exceeds 21, return 'BUST'\n blackjack(5,6,7) --> 18\n blackjack(9,9,9) --> 'BUST'\n blackjack(9,9,11) --> 19",
"_____no_output_____"
]
],
[
[
"def blackjack(a,b,c):\n pass",
"_____no_output_____"
],
[
"# Check\nblackjack(5,6,7)",
"_____no_output_____"
],
[
"# Check\nblackjack(9,9,9)",
"_____no_output_____"
],
[
"# Check\nblackjack(9,9,11)",
"_____no_output_____"
]
],
[
[
"#### SUMMER OF '69: Return the sum of the numbers in the array, except ignore sections of numbers starting with a 6 and extending to the next 9 (every 6 will be followed by at least one 9). Return 0 for no numbers.\n \n summer_69([1, 3, 5]) --> 9\n summer_69([4, 5, 6, 7, 8, 9]) --> 9\n summer_69([2, 1, 6, 9, 11]) --> 14",
"_____no_output_____"
]
],
[
[
"def summer_69(arr):\n pass",
"_____no_output_____"
],
[
"# Check\nsummer_69([1, 3, 5])",
"_____no_output_____"
],
[
"# Check\nsummer_69([4, 5, 6, 7, 8, 9])",
"_____no_output_____"
],
[
"# Check\nsummer_69([2, 1, 6, 9, 11])",
"_____no_output_____"
]
],
[
[
"# CHALLENGING PROBLEMS",
"_____no_output_____"
],
[
"#### SPY GAME: Write a function that takes in a list of integers and returns True if it contains 007 in order\n\n spy_game([1,2,4,0,0,7,5]) --> True\n spy_game([1,0,2,4,0,5,7]) --> True\n spy_game([1,7,2,0,4,5,0]) --> False\n",
"_____no_output_____"
]
],
[
[
"def spy_game(nums):\n pass",
"_____no_output_____"
],
[
"# Check\nspy_game([1,2,4,0,0,7,5])",
"_____no_output_____"
],
[
"# Check\nspy_game([1,0,2,4,0,5,7])",
"_____no_output_____"
],
[
"# Check\nspy_game([1,7,2,0,4,5,0])",
"_____no_output_____"
]
],
[
[
"#### COUNT PRIMES: Write a function that returns the *number* of prime numbers that exist up to and including a given number\n count_primes(100) --> 25\n\nBy convention, 0 and 1 are not prime.",
"_____no_output_____"
]
],
[
[
"def count_primes(num):\n pass\n ",
"_____no_output_____"
],
[
"# Check\ncount_primes(100)",
"_____no_output_____"
]
],
[
[
"### Just for fun:\n#### PRINT BIG: Write a function that takes in a single letter, and returns a 5x5 representation of that letter\n print_big('a')\n \n out: * \n * *\n *****\n * *\n * *\nHINT: Consider making a dictionary of possible patterns, and mapping the alphabet to specific 5-line combinations of patterns. <br>For purposes of this exercise, it's ok if your dictionary stops at \"E\".",
"_____no_output_____"
]
],
[
[
"def print_big(letter):\n pass",
"_____no_output_____"
],
[
"print_big('a')",
"_____no_output_____"
]
],
[
[
"## Great Job!",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
]
] |
cb908431b720d535ec1e5288351587433265d223
| 15,968 |
ipynb
|
Jupyter Notebook
|
03-introduction_variables.ipynb
|
hanisaf/advanced-data-management-and-analytics
|
e7bffda5cad91374a14df1a65f95e6a25f72cc41
|
[
"MIT"
] | 6 |
2020-04-13T19:22:18.000Z
|
2021-04-20T18:20:13.000Z
|
03-introduction_variables.ipynb
|
hanisaf/advanced-data-management-and-analytics
|
e7bffda5cad91374a14df1a65f95e6a25f72cc41
|
[
"MIT"
] | null | null | null |
03-introduction_variables.ipynb
|
hanisaf/advanced-data-management-and-analytics
|
e7bffda5cad91374a14df1a65f95e6a25f72cc41
|
[
"MIT"
] | 10 |
2020-05-12T01:02:32.000Z
|
2022-02-28T17:04:37.000Z
| 21.72517 | 303 | 0.51622 |
[
[
[
"*This notebook contains an excerpt from the [Whirlwind Tour of Python](http://www.oreilly.com/programming/free/a-whirlwind-tour-of-python.csp) by Jake VanderPlas; the content is available [on GitHub](https://github.com/jakevdp/WhirlwindTourOfPython).*\n\n*The text and code are released under the [CC0](https://github.com/jakevdp/WhirlwindTourOfPython/blob/master/LICENSE) license; see also the companion project, the [Python Data Science Handbook](https://github.com/jakevdp/PythonDataScienceHandbook).*\n",
"_____no_output_____"
],
[
"# A Quick Tour of Python Language Syntax",
"_____no_output_____"
]
],
[
[
"x = 1\ny = 4\nz = x + y\nz",
"_____no_output_____"
],
[
"# set the midpoint\nmidpoint = 5\n\n# make two empty lists\nlower = []; upper = []\n\n# split the numbers into lower and upper\nfor i in range(10):\n if (i < midpoint):\n lower.append(i)\n else:\n upper.append(i)\n \nprint(\"lower:\", lower)\nprint(\"upper:\", upper)",
"lower: [0, 1, 2, 3, 4]\nupper: [5, 6, 7, 8, 9]\n"
]
],
[
[
"## Comments Are Marked by ``#``\nThe script starts with a comment:\n``` python\n# set the midpoint\n```\nComments in Python are indicated by a pound sign (``#``), and anything on the line following the pound sign is ignored by the interpreter.\nThis means, for example, that you can have stand-alone comments like the one just shown, as well as inline comments that follow a statement. For example:\n``` python\nx += 2 # shorthand for x = x + 2\n```\nPython does not have any syntax for multi-line comments, such as the ``/* ... */`` syntax used in C and C++, though multi-line strings are often used as a replacement for multi-line comments (more on this in [String Manipulation and Regular Expressions](14-Strings-and-Regular-Expressions.ipynb)).",
"_____no_output_____"
],
[
"## End-of-Line Terminates a Statement\nThe next line in the script is\n``` python\nmidpoint = 5\n```\nThis is an assignment operation, where we've created a variable named ``midpoint`` and assigned it the value ``5``.\nNotice that the end of this statement is simply marked by the end of the line.\nThis is in contrast to languages like C and C++, where every statement must end with a semicolon (``;``).\n\nIn Python, if you'd like a statement to continue to the next line, it is possible to use the \"``\\``\" marker to indicate this:",
"_____no_output_____"
],
[
"## Semicolon Can Optionally Terminate a Statement\nSometimes it can be useful to put multiple statements on a single line.\nThe next portion of the script is\n``` python\nlower = []; upper = []\n```\nThis shows the example of how the semicolon (``;``) familiar in C can be used optionally in Python to put two statements on a single line.\nFunctionally, this is entirely equivalent to writing\n``` python\nlower = []\nupper = []\n```\nUsing a semicolon to put multiple statements on a single line is generally discouraged by most Python style guides, though occasionally it proves convenient.",
"_____no_output_____"
],
[
"## Indentation: Whitespace Matters!\nNext, we get to the main block of code:\n``` Python\nfor i in range(10):\n if i < midpoint:\n lower.append(i)\n else:\n upper.append(i)\n```\nThis is a compound control-flow statement including a loop and a conditional – we'll look at these types of statements in a moment.\nFor now, consider that this demonstrates what is perhaps the most controversial feature of Python's syntax: whitespace is meaningful!",
"_____no_output_____"
],
[
"## Parentheses Are for Grouping or Calling\n\nIn the previous code snippet, we see two uses of parentheses.\nFirst, they can be used in the typical way to group statements or mathematical operations:",
"_____no_output_____"
]
],
[
[
"2 * (3 + 4)",
"_____no_output_____"
]
],
[
[
"They can also be used to indicate that a *function* is being called.\nIn the next snippet, the ``print()`` function is used to display the contents of a variable (see the sidebar).\nThe function call is indicated by a pair of opening and closing parentheses, with the *arguments* to the function contained within:",
"_____no_output_____"
]
],
[
[
"print('first value:', 1)",
"first value: 1\n"
]
],
[
[
"# Basic Python Semantics: Variables and Objects",
"_____no_output_____"
],
[
"## Python Variables Are Pointers\n\nAssigning variables in Python is as easy as putting a variable name to the left of the equals (``=``) sign:\n\n```python\n# assign 4 to the variable x\nx = 4\n```",
"_____no_output_____"
]
],
[
[
"x = 1 # x is an integer\nx = 'hello' # now x is a string\nx = [1, 2, 3] # now x is a list",
"_____no_output_____"
],
[
"x = [1, 2, 3]\ny = x",
"_____no_output_____"
]
],
[
[
"We've created two variables ``x`` and ``y`` which both point to the same object.\nBecause of this, if we modify the list via one of its names, we'll see that the \"other\" list will be modified as well:",
"_____no_output_____"
]
],
[
[
"print(y)",
"[1, 2, 3]\n"
],
[
"x.append(4) # append 4 to the list pointed to by x\nprint(y) # y's list is modified as well!",
"[1, 2, 3, 4]\n"
],
[
"print(y)",
"[1, 2, 3]\n"
],
[
"x = 10\ny = x\nx += 5 # add 5 to x's value, and assign it to x\nprint(\"x =\", x)\nprint(\"y =\", y)",
"x = 15\ny = 10\n"
]
],
[
[
"When we call ``x += 5``, we are not modifying the value of the ``10`` object pointed to by ``x``; we are rather changing the variable ``x`` so that it points to a new integer object with value ``15``.\nFor this reason, the value of ``y`` is not affected by the operation.",
"_____no_output_____"
],
[
"## Everything Is an Object\n\nPython is an object-oriented programming language, and in Python everything is an object.\n\nLet's flesh-out what this means. Earlier we saw that variables are simply pointers, and the variable names themselves have no attached type information.\nThis leads some to claim erroneously that Python is a type-free language. But this is not the case!\nConsider the following:",
"_____no_output_____"
]
],
[
[
"x = 4\ntype(x)",
"_____no_output_____"
],
[
"x.bit_length()",
"_____no_output_____"
],
[
"x.real",
"_____no_output_____"
],
[
"x = 'hello'\ntype(x)",
"_____no_output_____"
],
[
"x.upper()",
"_____no_output_____"
],
[
"x = 3.14159\ntype(x)",
"_____no_output_____"
],
[
"x.as_integer_ratio()",
"_____no_output_____"
]
],
[
[
"Python has types; however, the types are linked not to the variable names but *to the objects themselves*.\n\nIn object-oriented programming languages like Python, an *object* is an entity that contains data along with associated metadata and/or functionality.",
"_____no_output_____"
]
],
[
[
"L = [1, 2, 3]\nL.append(100)\nprint(L)",
"[1, 2, 3, 100]\n"
]
],
[
[
"While it might be expected for compound objects like lists to have attributes and methods, what is sometimes unexpected is that in Python even simple types have attached attributes and methods.\nFor example, numerical types have a ``real`` and ``imag`` attribute that returns the real and imaginary part of the value, if viewed as a complex number:",
"_____no_output_____"
]
],
[
[
"x = 4.5\nprint(x.real, \"+\", x.imag, 'i')",
"4.5 + 0.0 i\n"
]
],
[
[
"Methods are like attributes, except they are functions that you can call using opening and closing parentheses.\nFor example, floating point numbers have a method called ``is_integer`` that checks whether the value is an integer:",
"_____no_output_____"
]
],
[
[
"x = 4.5\nx.is_integer()",
"_____no_output_____"
],
[
"x = 4.0\nx.is_integer()",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
]
] |
cb90aa712f7ff42de4e9564b59bd9100285f56eb
| 19,301 |
ipynb
|
Jupyter Notebook
|
de_sim/examples/jupyter_examples/3. Ordering simultaneous events in DE-Sim.ipynb
|
KarrLab/desim
|
6f189d8c8e850e092d816f6be3d6f87b4f983ac2
|
[
"MIT"
] | 16 |
2019-12-12T15:49:17.000Z
|
2022-03-31T20:34:36.000Z
|
de_sim/examples/jupyter_examples/3. Ordering simultaneous events in DE-Sim.ipynb
|
KarrLab/desim
|
6f189d8c8e850e092d816f6be3d6f87b4f983ac2
|
[
"MIT"
] | 65 |
2019-08-15T14:50:38.000Z
|
2020-12-17T14:36:04.000Z
|
de_sim/examples/jupyter_examples/3. Ordering simultaneous events in DE-Sim.ipynb
|
KarrLab/desim
|
6f189d8c8e850e092d816f6be3d6f87b4f983ac2
|
[
"MIT"
] | 5 |
2020-07-16T22:15:47.000Z
|
2021-08-16T02:16:17.000Z
| 43.470721 | 349 | 0.640122 |
[
[
[
"<!-- :Author: Arthur Goldberg <[email protected]> -->\n<!-- :Date: 2020-08-02 -->\n<!-- :Copyright: 2020, Karr Lab -->\n<!-- :License: MIT -->\n# DE-Sim: Ordering simultaneous events\n\nDE-Sim makes it easy to build and simulate discrete-event models.\nThis notebook discusses DE-Sim's methods for controlling the execution order of simultaneous messages.\n\n## Installation\nUse `pip` to install `de_sim`.",
"_____no_output_____"
],
[
"## Scheduling events with equal simulation times\n\nA discrete-event simulation may execute multiple events simultaneously, that is, at a particular simulation time.\nTo ensure that simulation runs are reproducible and deterministic, a simulator must provide mechanisms that deterministically control the execution order of simultaneous events.\n\nTwo types of situations arise, *local* and *global*.\nA local situation arises when a simulation object receives multiple event messages simultaneously, while a global situation arises when multiple simulation objects execute events simultaneously.\n\nSeparate *local* and *global* mechanisms ensure that these situations are simulated deterministically.\nThe local mechanism ensures that simultaneous events are handled deterministically at a single simulation object, while the global mechanism ensures that simultaneous events are handled deterministically across all objects in a simulation.\n\n### Local mechanism: simultaneous event messages at a simulation object\nThe local mechanism, called *event superposition* after the [physics concept of superposition](https://en.wikipedia.org/wiki/Superposition_principle), involves two components:\n\n1. When a simulation object receives multiple event messages at the same time, the simulator passes all of the event messages to the object's event handler in a list.\n(However, if simultaneous event messages have different handlers then the simulator raises a `SimulatorError` exception.)\n2. The simulator sorts the events in the list so that any given list of events will always be arranged in the same order.\n\nEvent messages are sorted by the pair (event message priority, event message content).\nSorting costs O(n log n), but since simultaneous events are usually rare, sorting event lists is unlikely to slow down simulations.",
"_____no_output_____"
]
],
[
[
"\"\"\" This example illustrates the local mechanism that handles simultaneous\n event messages received by a simulation object\n\"\"\"\nimport random\n\nimport de_sim\nfrom de_sim.event import Event\n\nclass Double(de_sim.EventMessage):\n 'Double value'\n\nclass Increment(de_sim.EventMessage):\n 'Increment value'\n\nclass IncrementThenDoubleSimObject(de_sim.SimulationObject):\n \"\"\" Execute Increment before Double, demonstrating superposition \"\"\"\n\n def __init__(self, name):\n super().__init__(name)\n self.value = 0\n\n def init_before_run(self):\n self.send_events()\n\n def handle_superposed_events(self, event_list):\n \"\"\" Process superposed events in an event list\n\n Each Increment message increments value, and each Double message doubles value.\n Assumes that `event_list` contains an Increment event followed by a Double event.\n\n Args:\n event_list (:obj:`event_list` of :obj:`de_sim.Event`): list of events\n \"\"\"\n for event in event_list:\n if isinstance(event.message, Increment):\n self.value += 1\n elif isinstance(event.message, Double):\n self.value *= 2\n self.send_events()\n\n # The order of the message types in event_handlers, (Increment, Double), determines\n # the sort order of messages in `event_list` received by `handle_superposed_events`\n event_handlers = [(Increment, 'handle_superposed_events'),\n (Double, 'handle_superposed_events')]\n\n def send_events(self):\n # To show that the simulator delivers event messages to `handle_superposed_events`\n # sorted into the order (Increment, Double), send them in a random order.\n if random.randrange(2):\n self.send_event(1, self, Double())\n self.send_event(1, self, Increment())\n else:\n self.send_event(1, self, Increment())\n self.send_event(1, self, Double())\n\n # Register the message types sent\n messages_sent = (Increment, Double)\n\n\nclass TestSuperposition(object):\n\n def increment_then_double_from_0(self, iterations):\n v = 0\n for _ in range(iterations):\n v += 1\n v *= 2\n return v\n\n def test_superposition(self, max_time):\n simulator = de_sim.Simulator()\n simulator.add_object(IncrementThenDoubleSimObject('name'))\n simulator.initialize()\n simulator.simulate(max_time)\n for sim_obj in simulator.get_objects():\n assert sim_obj.value, self.increment_then_double_from_0(max_time)\n print(f'Simulation to {max_time} executed all messages in the order (Increment, Double).')\n\nTestSuperposition().test_superposition(20)",
"Simulation to 20 executed all messages in the order (Increment, Double).\n"
]
],
[
[
"This example shows how event superposition handles simultaneous events.\nAn `IncrementThenDoubleSimObject` simulation object stores an integer value.\nIt receives two events every time unit, one carrying an `Increment` message and another containing a `Double` message.\nExecuting an `Increment` event increments the value, while executing a `Double` message event doubles the value.\nThe design for `IncrementThenDoubleSimObject` requires that it increments before doubling.\n\nSeveral features of DE-Sim and `IncrementThenDoubleSimObject` ensure this behavior:\n\n1. The mapping between event message types and event handlers, stored in the list `event_handlers`, contains `Increment` before `Double`. This gives events containing an `Increment` message a higher priority than events containing `Double`.\n2. Under the covers, when DE-Sim passes superposed events to a subclass of [`SimulationObject`](https://docs.karrlab.org/de_sim/master/source/de_sim.html#de_sim.simulation_object.SimulationObject), it sorts the messages by their (event message priority, event message content), which sorts events with higher priority message types earlier.\n3. The message handler `handle_superposed_events` receives a list of events and executes them in order.\n\nTo challenge and test this superposition mechanism, the `send_events()` method in `IncrementThenDoubleSimObject` randomizes the order in which it sends `Increment` and `Double` events.\nFinally, `TestSuperposition().test_superposition()` runs a simulation of `IncrementThenDoubleSimObject` and asserts that the value it computes equals the correct value for a sequence of increment and double operations.",
"_____no_output_____"
],
[
"### Global mechanism: simultaneous event messages at multiple simulation objects\nA *global* mechanism is needed to ensure that simultaneous events which occur at distinct objects in a simulation are executed in a deterministic order.\nOtherwise, the discrete-event simulator might execute simultaneous events at distinct simulation objects in a different order in different simulation runs that use the same input.\nWhen using a simulator that allows 0-delay event messages or global state shared between simulation objects -- both of which DE-Sim supports -- this can alter the simulation's predictions and thereby imperil debugging efforts, statistical analyses of predictions and other essential uses of simulation results.\n\nThe global mechanism employed by DE-Sim conceives of the simulation time as a pair -- the event time, and a *sub-time* which breaks event time ties.\nSub-time values within a particular simulation time must be distinct.\nGiven that constraint, many approaches for selecting the sub-time would achieve the objective.\nDE-Sim creates a distinct sub-time from the state of the simulation object receiving an event.\n\nThe sub-time is a pair composed of a priority assigned to the simulation class and a unique identifier for each class instance.\nEach simulation class defines a `class_priority` attribute that determines the relative execution order of simultaneous events by different simulation classes.\nAmong multiple instances of a simulation class, the attribute `event_time_tiebreaker`, which defaults to a simulation instance's unique name, breaks ties.\nAll classes have the same default priority of `LOW`. If class priorities are not set and `event_time_tiebreaker`s are not set for individual simulation objects, then an object's global priority is given by its name.",
"_____no_output_____"
]
],
[
[
"from de_sim.simulation_object import SimObjClassPriority\n\n\nclass ExampleMsg(de_sim.EventMessage):\n 'Example message'\n\n\nclass NoPrioritySimObj(de_sim.SimulationObject):\n\n def init_before_run(self):\n self.send_event(0., self, ExampleMsg())\n\n # register the message types sent\n messages_sent = (ExampleMsg, )\n\n\nclass LowPrioritySimObj(NoPrioritySimObj):\n\n def handler(self, event):\n print(f\"{self.time}: LowPrioritySimObj {self.name} running\")\n self.send_event(1., self, ExampleMsg())\n\n event_handlers = [(ExampleMsg, 'handler')]\n\n # have `LowPrioritySimObj`s execute at low priority\n class_priority = SimObjClassPriority.LOW\n\n\nclass MediumPrioritySimObj(NoPrioritySimObj):\n\n def handler(self, event):\n print(f\"{self.time}: MediumPrioritySimObj {self.name} running\")\n self.send_event(1., self, ExampleMsg())\n\n event_handlers = [(ExampleMsg, 'handler')]\n\n # have `MediumPrioritySimObj`s execute at medium priority\n class_priority = SimObjClassPriority.MEDIUM\n\nsimulator = de_sim.Simulator()\nsimulator.add_object(LowPrioritySimObj('A'))\nsimulator.add_object(MediumPrioritySimObj('B'))\nsimulator.initialize()\nprint(simulator.simulate(2).num_events, 'events executed')",
"0.0: MediumPrioritySimObj B running\n0.0: LowPrioritySimObj A running\n1.0: MediumPrioritySimObj B running\n1.0: LowPrioritySimObj A running\n2.0: MediumPrioritySimObj B running\n2.0: LowPrioritySimObj A running\n6 events executed\n"
]
],
[
[
"This example illustrates the scheduling of simultaneous event messages.\n`SimObjClassPriority` is an `IntEnum` that provides simulation object class priorities, including `LOW`, `MEDIUM`, and `HIGH`.\nWe create two classes, `LowPrioritySimObj` and `MediumPrioritySimObj`, with `LOW` and `MEDIUM` priorities, respectively, and execute them simultaneously at simulation times 0, 1, 2, ...\nAt each time, the `MediumPrioritySimObj` object runs before the `LowPrioritySimObj` one.\n\n#### Execution order of objects without an assigned `class_priority`\nThe next example shows the ordering of simultaneous events executed by objects that don't have assigned priorities.",
"_____no_output_____"
]
],
[
[
"class DefaultPrioritySimObj(NoPrioritySimObj):\n\n def handler(self, event):\n print(f\"{self.time}: DefaultPrioritySimObj {self.name} running\")\n self.send_event(1., self, ExampleMsg())\n\n event_handlers = [(ExampleMsg, 'handler')]\n\n\nsimulator = de_sim.Simulator()\nfor name in random.sample(range(10), k=3):\n sim_obj = DefaultPrioritySimObj(str(name))\n print(f\"{sim_obj.name} priority: {sim_obj.class_event_priority.name}\")\n simulator.add_object(sim_obj)\nsimulator.initialize()\nprint(simulator.simulate(2).num_events, 'events executed')",
"9 priority: LOW\n6 priority: LOW\n2 priority: LOW\n0.0: DefaultPrioritySimObj 2 running\n0.0: DefaultPrioritySimObj 6 running\n0.0: DefaultPrioritySimObj 9 running\n1.0: DefaultPrioritySimObj 2 running\n1.0: DefaultPrioritySimObj 6 running\n1.0: DefaultPrioritySimObj 9 running\n2.0: DefaultPrioritySimObj 2 running\n2.0: DefaultPrioritySimObj 6 running\n2.0: DefaultPrioritySimObj 9 running\n9 events executed\n"
]
],
[
[
"In this example, the [`SimulationObject`s](https://docs.karrlab.org/de_sim/master/source/de_sim.html#de_sim.simulation_object.SimulationObject) have no priorities assigned, so their default priorities are `LOW`. (The `class_event_priority` attribute of a simulation object is a `SimObjClassPriority`)\n\nThree objects with names randomly selected from '0', '1', ..., '9', are created.\nWhen they execute simultaneously, events are ordered by the sort order of the objects' names.\n\n#### Execution order of instances of simulation object classes with relative priorities\nOften, a modeler wants to control the *relative* simultaneous priorities of simulation objects, but does not care about their absolute priorities.\nThe next example shows how to specify relative priorities.",
"_____no_output_____"
]
],
[
[
"class FirstNoPrioritySimObj(NoPrioritySimObj):\n\n def handler(self, event):\n print(f\"{self.time}: FirstNoPrioritySimObj {self.name} running\")\n self.send_event(1., self, ExampleMsg())\n\n event_handlers = [(ExampleMsg, 'handler')]\n\n\nclass SecondNoPrioritySimObj(NoPrioritySimObj):\n\n def handler(self, event):\n print(f\"{self.time}: SecondNoPrioritySimObj {self.name} running\")\n self.send_event(1., self, ExampleMsg())\n\n event_handlers = [(ExampleMsg, 'handler')]\n\n\n# Assign decreasing priorities to classes in [FirstNoPrioritySimObj, SecondNoPrioritySimObj]\nSimObjClassPriority.assign_decreasing_priority([FirstNoPrioritySimObj,\n SecondNoPrioritySimObj])\n\nsimulator = de_sim.Simulator()\nsimulator.add_object(SecondNoPrioritySimObj('A'))\nsimulator.add_object(FirstNoPrioritySimObj('B'))\nfor sim_obj in simulator.simulation_objects.values():\n print(f\"{type(sim_obj).__name__}: {sim_obj.name}; \"\n f\"priority: {sim_obj.class_event_priority.name}\")\nsimulator.initialize()\nprint(simulator.simulate(2).num_events, 'events executed')",
"SecondNoPrioritySimObj: A; priority: SECOND\nFirstNoPrioritySimObj: B; priority: HIGH\n0.0: FirstNoPrioritySimObj B running\n0.0: SecondNoPrioritySimObj A running\n1.0: FirstNoPrioritySimObj B running\n1.0: SecondNoPrioritySimObj A running\n2.0: FirstNoPrioritySimObj B running\n2.0: SecondNoPrioritySimObj A running\n6 events executed\n"
]
],
[
[
"The `assign_decreasing_priority` method of `SimObjClassPriority` takes an iterator over `SimulationObject` subclasses, and assigns them decreasing simultaneous event priorities.\nThe `FirstNoPrioritySimObj` instance therefore executes before the `SecondNoPrioritySimObj` instance at each discrete simulation time.",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
]
] |
cb90b3b0aeec4f5a76cd9eafb2f24d89779ee3c5
| 428,318 |
ipynb
|
Jupyter Notebook
|
[MLDA@EEE] Linear and Logistic Regression/Linear_&_Logistic_Regression_Notebook_v2.ipynb
|
leephilipx/workshops
|
6294ad6d54069e095e2e0373d811c5f7df7fdfaa
|
[
"MIT"
] | null | null | null |
[MLDA@EEE] Linear and Logistic Regression/Linear_&_Logistic_Regression_Notebook_v2.ipynb
|
leephilipx/workshops
|
6294ad6d54069e095e2e0373d811c5f7df7fdfaa
|
[
"MIT"
] | null | null | null |
[MLDA@EEE] Linear and Logistic Regression/Linear_&_Logistic_Regression_Notebook_v2.ipynb
|
leephilipx/workshops
|
6294ad6d54069e095e2e0373d811c5f7df7fdfaa
|
[
"MIT"
] | null | null | null | 184.064461 | 147,683 | 0.903894 |
[
[
[
"# Essential Libraries\n\n- ```Pandas```: Library for data acquisition and preparation\n- ```NumPy```: Library for numeric computations in python\n- ```Matplotlib```: Low-level library for data visualisation\n- ```Seaborn```: Higher-level library for data visualisation\n- ```scikit-learn```: Basic library for machine learning\n- ```mpl_toolkits.mplot3d```: Library to handle 3D data visualisation",
"_____no_output_____"
]
],
[
[
"import pandas as pd\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport seaborn as sns\n%matplotlib inline\nsns.set()",
"_____no_output_____"
]
],
[
[
"# Simple Linear Regression\n\nNBA Players Dataset from: https://www.kaggle.com/justinas/nba-height-and-weight-analysis/?select=all_seasons.csv\n\nQuick download link: https://drive.google.com/file/d/19u4DF8b5YT4ODbo1SCdmpsosRtBgzhVo/view?usp=sharing\n\nColumns include:\n* player_name\n* team_abbreviation\n* age\n* player_height - in kg\n* player_weight - in kg\n* college - Name of the college the player attended\n* country - Name of the country the player was born in\n* draft year - The year the player was drafted\n* draft round - The draft round the player was picked\n* draft number - The number at which the player was picked in his draft round\n* gp - Games played throughout the season\n* pts - Average number of points scored\n* reb - Average number of rebounds grabbed\n* ast - Average number of assists distributed\n* net_rating - Team's point differential per 100 possessions while the player is on the court\n* oreb_pct - Percentage of available offensive rebounds the player grabbed while he was on the floor\n* dreb_pct - Percentage of available defensive rebounds the player grabbed while he was on the floor\n* usg_pct - Percentage of team plays used by the player while he was on the floor (FGA + Possession Ending FTA + TO) / POSS)\n* ts_pct - Measure of the player's shooting efficiency that takes into account free throws, 2 and 3 point shots (PTS / (2*(FGA + 0.44 * FTA)))\n* ast_pct - Percentage of teammate field goals the player assisted while he was on the floor\n* season - NBA season\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n",
"_____no_output_____"
]
],
[
[
"df = pd.read_csv('all_seasons.csv')\ndf.head()",
"_____no_output_____"
]
],
[
[
"We will start by analysing whether the weight of a NBA player is a good predicting factor for the height of that particular player.\n",
"_____no_output_____"
]
],
[
[
"simple_df = df[['player_weight', 'player_height']]\nsimple_df",
"_____no_output_____"
]
],
[
[
"Plotting and Analysing Correlation",
"_____no_output_____"
]
],
[
[
"sns.scatterplot(data=simple_df, x='player_weight', y='player_height');",
"_____no_output_____"
],
[
"simple_df.player_weight.corr(simple_df.player_height, method='pearson')",
"_____no_output_____"
]
],
[
[
"Using the formula:\n\n\\\\[b_0=\\bar{y}-b_1\\bar{x}; ~~~ b_1=\\frac{\\sum_{i=1}^n(x_i-\\bar{x})(y_i-\\bar{y})}{\\sum_{i=1}^n(x_i-\\bar{x})^2}\\\\]",
"_____no_output_____"
]
],
[
[
"X = simple_df['player_weight'].values\nY = simple_df['player_height'].values\nmean_x = np.mean(X)\nmean_y = np.mean(Y)\n\nn = len(X)\n\nnumer = 0\ndenom = 0\nfor i in range(n):\n numer += (X[i] - mean_x) * (Y[i] - mean_y)\n denom += (X[i] - mean_x) ** 2\n \nb1 = numer/denom\nb0 = mean_y - b1 * mean_x\n\nprint(\"coefficient:\", b1, \"\\nintercept:\", b0)",
"_____no_output_____"
]
],
[
[
"Documentation for sklearn's linear regression model (normal equation): \n\nhttps://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LinearRegression.html\n\n---\n\n",
"_____no_output_____"
]
],
[
[
"X = X.reshape((len(X), 1)) # reshaping the array to fit as input for sklearn framework \nfrom sklearn.linear_model import LinearRegression\n\n# instantiate linear regression object\nreg = LinearRegression()\n# train the model\nreg.fit(X, Y)",
"_____no_output_____"
],
[
"print(\"coefficient: \", reg.coef_[0])\nprint(\"intercept: \", reg.intercept_)",
"_____no_output_____"
],
[
"# evaluate with built-in r2 score function\nreg.score(X, Y)",
"_____no_output_____"
],
[
"sns.set(rc={'figure.figsize':(4,3)})\nsns.heatmap(data=simple_df.corr(), annot=True);",
"_____no_output_____"
]
],
[
[
"## Lasso Regression\n\n[Documentation](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.Lasso.html) for sklearn's lasso regression model:",
"_____no_output_____"
]
],
[
[
"from sklearn.linear_model import Lasso\nlasso_reg = Lasso(alpha=2)\nlasso_reg.fit(X, Y)\nlasso_reg.score(X, Y)",
"_____no_output_____"
]
],
[
[
"## Ridge Regression\n\n[Documentation](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.Ridge.html) for sklearn's ridge regression model:",
"_____no_output_____"
]
],
[
[
"from sklearn.linear_model import Ridge\nridge_reg = Ridge(alpha=2)\nridge_reg.fit(X, Y)\nridge_reg.score(X, Y)",
"_____no_output_____"
]
],
[
[
"## Combined Plots",
"_____no_output_____"
]
],
[
[
"plt.figure(figsize=(10, 10))\n\nmax_x = np.max(X)\nmin_x = np.min(X)\n\nx = np.linspace(min_x, max_x, 1000)\ny_lin = reg.coef_[0]*x + reg.intercept_\ny_las = lasso_reg.coef_[0]*x + lasso_reg.intercept_\ny_rid = ridge_reg.coef_[0]*x + ridge_reg.intercept_\n\n# plotting the regression line\nplt.plot(x, y_lin, 'r-', label='linear regression')\nplt.plot(x, y_las, 'g-', label='lasso regression')\nplt.plot(x, y_rid, 'b-', label='ridge regression')\n\n# plotting the training samples\nplt.scatter(X, Y, label='scatter plot')\n\nplt.xlabel('player_weight')\nplt.ylabel('player_height')\nplt.legend()\nplt.show()",
"_____no_output_____"
]
],
[
[
"## SGDRegressor\n\nFor more insights on Stochastic Gradient Descent: \n\nhttps://rumankhan1.medium.com/understanding-optimization-in-ml-with-gradient-descent-implement-sgd-regressor-from-scratch-4e11dac74c9\n\nDocumentation for sklearn's SGDRegressor(gradient descent):\n\nhttps://scikit-learn.org/stable/modules/generated/sklearn.linear_model.SGDRegressor.html",
"_____no_output_____"
]
],
[
[
"from sklearn.linear_model import SGDRegressor\n\n# Try changing verbose to 1 instead to see the loss for each epoch\nSGDReg = SGDRegressor(eta0 = 0.000001, verbose=0, max_iter=1000, learning_rate='adaptive')",
"_____no_output_____"
]
],
[
[
"Documentation for sklearn's standard scaler:\n\nhttps://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html",
"_____no_output_____"
]
],
[
[
"# scaling the data according to standard normal distribution ==> z = (x - u) / s\nfrom sklearn.preprocessing import StandardScaler\nscaler = StandardScaler().fit(X)\nX_scaled = scaler.transform(X)",
"_____no_output_____"
],
[
"SGDReg.fit(X_scaled, Y)",
"_____no_output_____"
],
[
"# note that coefficient and intercept of regressor cannot be compared to sklearn linear regression model \n# as the training samples in regressor has been scaled\n\nprint(\"coefficient: \", SGDReg.coef_)\nprint(\"intercept: \", SGDReg.intercept_)",
"_____no_output_____"
],
[
"# evaluate with built-in r2 score function\nSGDReg.score(X_scaled, Y)",
"_____no_output_____"
]
],
[
[
"## Mutiple Variable Linear Regression",
"_____no_output_____"
],
[
"Maybe weight itself is not a good enough predictor for a player's height. For this part, we shall add in another predictor variable called 'reb', which is the average number of rebounds grabbed per game by a player. Rebound means ball or shot that bounces back after striking a the frame or the rim.",
"_____no_output_____"
]
],
[
[
"df_multi = df[['player_weight', 'reb', 'player_height']]\ndf_multi",
"_____no_output_____"
],
[
"df_multi.describe()",
"_____no_output_____"
],
[
"X = df_multi[['player_weight', 'reb']].values.reshape((-1, 2))\nY = df_multi['player_height']",
"_____no_output_____"
],
[
"from sklearn.linear_model import LinearRegression\nreg = LinearRegression()\nreg.fit(X, Y)",
"_____no_output_____"
],
[
"print(\"coefficients: \", reg.coef_)\nprint(\"intercept: \", reg.intercept_)",
"_____no_output_____"
],
[
"# abit of improvement from the previous model with only 1 predictor variable\nreg.score(X, Y)",
"_____no_output_____"
],
[
"sns.set(rc={'figure.figsize':(4,3)})\nsns.heatmap(data=df_multi.corr(), annot=True);",
"_____no_output_____"
]
],
[
[
"3D Visualisation of the Multivariate Linear Regression Model\n",
"_____no_output_____"
]
],
[
[
"from mpl_toolkits.mplot3d import Axes3D",
"_____no_output_____"
],
[
"# setup\nx = X[:, 0]\ny = X[:, 1]\nz = Y\n\nx_pred = np.linspace(60, 180, 100) # range of porosity values\ny_pred = np.linspace(0, 20, 100) # range of brittleness values\nxx_pred, yy_pred = np.meshgrid(x_pred, y_pred)\nmodel_viz = np.array([xx_pred.flatten(), yy_pred.flatten()]).T\n\n# train model\nmulti_reg = LinearRegression()\nmodel = multi_reg.fit(X, Y)\npredicted = model.predict(model_viz)",
"_____no_output_____"
],
[
"#plotting 3D models\nplt.style.use('default')\n\nfig = plt.figure(figsize=(12, 5.5))\n\nax1 = fig.add_subplot(131, projection='3d')\nax2 = fig.add_subplot(132, projection='3d')\nax3 = fig.add_subplot(133, projection='3d')\n\naxes = [ax1, ax2, ax3]\n\nfor ax in axes:\n ax.plot(x, y, z, color='k', zorder=15, linestyle='none', marker='o', alpha=0.5)\n ax.scatter(xx_pred.flatten(), yy_pred.flatten(), predicted, facecolor=(0,0,0,0), s=20, edgecolor='#70b3f0')\n ax.set_xlabel('player_weight', fontsize=12)\n ax.set_ylabel('reb', fontsize=12)\n ax.set_zlabel('player_height', fontsize=12)\n ax.locator_params(nbins=4, axis='x')\n ax.locator_params(nbins=5, axis='x')\n\n# credits\nax1.text2D(0.2, 0.32, 'aegis4048.github.io', fontsize=13, ha='center', va='center',\n transform=ax1.transAxes, color='grey', alpha=0.5)\nax2.text2D(0.3, 0.42, 'aegis4048.github.io', fontsize=13, ha='center', va='center',\n transform=ax2.transAxes, color='grey', alpha=0.5)\nax3.text2D(0.85, 0.85, 'aegis4048.github.io', fontsize=13, ha='center', va='center',\n transform=ax3.transAxes, color='grey', alpha=0.5)\n\n# ‘elev’ stores the elevation angle in the z plane. ‘azim’ stores the azimuth angle in the x,y plane.\nax1.view_init(elev=20, azim=30)\nax2.view_init(elev=4, azim=100)\nax3.view_init(elev=60, azim=170)\n\nfig.tight_layout()",
"_____no_output_____"
]
],
[
[
"# Full Pipeline Walkthrough (Linear Regression)",
"_____no_output_____"
],
[
"### Import Libraries",
"_____no_output_____"
]
],
[
[
"import pandas as pd\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport seaborn as sns\n%matplotlib inline",
"_____no_output_____"
]
],
[
[
"### Exploratory Data Analysis",
"_____no_output_____"
],
[
"Dataset of House Sales in King County, USA: https://www.kaggle.com/harlfoxem/housesalesprediction\n\nQuick Download Link: https://drive.google.com/file/d/1HREApvec84pUxXuGSONaRUgDVDxWjoNq/view?usp=sharing\n",
"_____no_output_____"
],
[
"* id - Unique ID for each home sold\n* date - Date of the home sale\n* price - Price of each home sold\n* bedrooms - Number of bedrooms\n* bathrooms - Number of bathrooms, where .5 accounts for a room with a toilet but no shower\n* sqft_living - Square footage of the apartments interior living space\n* sqft_lot - Square footage of the land space\n* floors - Number of floors\n* waterfront - A dummy variable for whether the apartment was overlooking the waterfront or not\n* view - An index from 0 to 4 of how good the view of the property was\n* condition - An index from 1 to 5 on the condition of the apartment,\n* grade - An index from 1 to 13, where 1-3 falls short of building construction and design, 7 has an average level of construction and design, and 11-13 have a high quality level of construction and design.\n* sqft_above - The square footage of the interior housing space that is above ground level\n* sqft_basement - The square footage of the interior housing space that is below ground level\n* yr_built - The year the house was initially built\n* yr_renovated - The year of the house’s last renovation\n* zipcode - What zipcode area the house is in\n* lat - Lattitude\n* long - Longitude\n* sqft_living15 - The square footage of interior housing living space for the nearest 15 neighbors\n* sqft_lot15 - The square footage of the land lots of the nearest 15 neighbors",
"_____no_output_____"
],
[
"#### Extracting Info",
"_____no_output_____"
]
],
[
[
"df = pd.read_csv('kc_house_data.csv')\ndf.head()",
"_____no_output_____"
],
[
"df.describe()",
"_____no_output_____"
],
[
"df.info()",
"_____no_output_____"
],
[
"# Checking for null values\ndf.isnull().sum()",
"_____no_output_____"
]
],
[
[
"Checking on discrete varaiables",
"_____no_output_____"
]
],
[
[
"df['waterfront'].value_counts()",
"_____no_output_____"
],
[
"df['view'].value_counts()",
"_____no_output_____"
]
],
[
[
"#### Visualisation and Cleaning\n",
"_____no_output_____"
]
],
[
[
"# Plot heatmap\nsns.set(rc={'figure.figsize':(16,10)}) #read at startup to configure\ncorr_matrix = df.corr().round(2)\n##Your Code Here##",
"_____no_output_____"
],
[
"# Plotting out the relationship between the 2 discrete variables and the price\nfig, ax= plt.subplots(ncols=2, nrows=1)\nsns.scatterplot(x=\"view\", y=\"price\",data=df, ax=ax[0])\nsns.scatterplot(x=\"waterfront\", y=\"price\",data=df, ax=ax[1])\nplt.tight_layout()",
"_____no_output_____"
],
[
"# Drop all that has magnitude of correlation coefficients less than 0.1\ndf.drop(['id','date','zipcode','condition','long','sqft_lot15','yr_built','sqft_lot','view','waterfront','yr_renovated']\n ,axis=1,inplace=True)",
"_____no_output_____"
]
],
[
[
"<img width=\"600px\" src=\"data:image/png;base64,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\"/>",
"_____no_output_____"
]
],
[
[
"## Adjustments\nf,ax = plt.subplots(figsize=(15, 15), nrows=5, ncols=2)\nplt.subplots_adjust(hspace=0.8)\n\n## Features\nsns.boxplot(x='price',data=df,ax=ax[0][0])\nsns.boxplot(x='bedrooms',data=df,ax=ax[0][1])\nsns.boxplot(x='bathrooms',data=df,ax=ax[1][0])\nsns.boxplot(x='sqft_living',data=df,ax=ax[1][1])\nsns.boxplot(x='floors',data=df,ax=ax[2][0])\nsns.boxplot(x='grade',data=df,ax=ax[2][1])\nsns.boxplot(x='sqft_above',data=df,ax=ax[3][0])\nsns.boxplot(x='sqft_basement',data=df,ax=ax[3][1])\nsns.boxplot(x='lat',data=df,ax=ax[4][0])\nsns.boxplot(x='sqft_living15',data=df,ax=ax[4][1]);",
"_____no_output_____"
],
[
"columnsList = df.columns\ncolumnsList",
"_____no_output_____"
],
[
"# Cleaning Up\nfor column in columnsList:\n\n # Q1 & Q2 definition\n Q1 = df[column].quantile(0.25)\n Q3 = df[column].quantile(0.75)\n\n IQR = Q3-Q1\n\n lower_limit = Q1 - 1.5*IQR\n upper_limit = Q3 + 1.5*IQR\n\n df[column] = np.where(df[column]>upper_limit,upper_limit,df[column])\n df[column] = np.where(df[column]<lower_limit,lower_limit,df[column])",
"_____no_output_____"
],
[
"## Adjustments\nf,ax = plt.subplots(figsize=(15, 15), nrows=5, ncols=2)\nplt.subplots_adjust(hspace=0.8)\n\n## Features\nsns.boxplot(x='price',data=df,ax=ax[0][0])\nsns.boxplot(x='bedrooms',data=df,ax=ax[0][1])\nsns.boxplot(x='bathrooms',data=df,ax=ax[1][0])\nsns.boxplot(x='sqft_living',data=df,ax=ax[1][1])\nsns.boxplot(x='floors',data=df,ax=ax[2][0])\nsns.boxplot(x='grade',data=df,ax=ax[2][1])\nsns.boxplot(x='sqft_above',data=df,ax=ax[3][0])\nsns.boxplot(x='sqft_basement',data=df,ax=ax[3][1])\nsns.boxplot(x='lat',data=df,ax=ax[4][0])\nsns.boxplot(x='sqft_living15',data=df,ax=ax[4][1]);",
"_____no_output_____"
]
],
[
[
"### Training A Model",
"_____no_output_____"
]
],
[
[
"from sklearn.linear_model import LinearRegression",
"_____no_output_____"
],
[
"##Set your predictor variables to be all columns other than 'Price'##\nX = ##Your Code Here##\n\n##Set your response variable to be 'Price'##\ny = ##Your Code Here##",
"_____no_output_____"
]
],
[
[
"Documentation for sklearn's train_test_split function: \n\nhttps://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html",
"_____no_output_____"
]
],
[
[
"# Perform train test split\nfrom sklearn.model_selection import train_test_split\n##Your Code Here## ",
"_____no_output_____"
],
[
"# Scale data according to standard normal distribution\nfrom sklearn.preprocessing import StandardScaler\n##Your Code Here##",
"_____no_output_____"
],
[
"# Instantiate and train linear regression model\n##Your Code Here##",
"_____no_output_____"
]
],
[
[
"### Model Evaluation",
"_____no_output_____"
]
],
[
[
"# Make predictions with new model and assign them to a variable named 'predictions'\n##Your Code Here##",
"_____no_output_____"
],
[
"from sklearn.metrics import r2_score, mean_squared_error, mean_absolute_error",
"_____no_output_____"
],
[
"print('MAE:', mean_absolute_error(y_test, predictions))\nprint('MSE:', mean_squared_error(y_test, predictions))\nprint('RMSE:', ##Your Code Here##)\nprint('R2_score: ', r2_score(y_test, predictions))",
"_____no_output_____"
]
],
[
[
"===== End of Linear Regression Part =====\n\n<br><hr>",
"_____no_output_____"
],
[
"# Feature Engineering\n\nWe will add some potentially useful features to the previous dataset. \n\n\n*Log Transform*\n- Helps to handle skewed data and after transformation, the distribution becomes more approximate to normal\n- Decreases the effect of the outliers due to the normalization of magnitude differences and the model become more robust",
"_____no_output_____"
]
],
[
[
"## Re-run this code-block if there is any error below\n\ndf = pd.read_csv('kc_house_data.csv')\ndf.drop(['id','date','zipcode','condition','long','sqft_lot15','yr_built','sqft_lot','view','waterfront','yr_renovated'], axis=1, inplace=True)\n\nfor column in columnsList:\n Q1 = df[column].quantile(0.25)\n Q3 = df[column].quantile(0.75)\n IQR = Q3-Q1\n lower_limit = Q1 - 1.5*IQR\n upper_limit = Q3 + 1.5*IQR\n df[column] = np.where(df[column]>upper_limit,upper_limit,df[column])\n df[column] = np.where(df[column]<lower_limit,lower_limit,df[column])",
"_____no_output_____"
],
[
"plt.figure(figsize=(10, 8))\nsns.scatterplot(data=df, x='sqft_basement', y='price');",
"_____no_output_____"
],
[
"df['log_sqft'] = np.log(df['sqft_basement']+1)\nplt.figure(figsize=(10, 8))\nsns.scatterplot(data=df, x='log_sqft', y='price');",
"_____no_output_____"
]
],
[
[
"Observe that the dataframe has a new column ```log_sqft```.",
"_____no_output_____"
]
],
[
[
"df.head()",
"_____no_output_____"
]
],
[
[
"*Dummy Variable Encoding*\n- Convert categorical features into multiple binary features\n- Observe that new columns are added \n\n<img width=\"600px\" src=\"data:image/png;base64,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\"/>",
"_____no_output_____"
]
],
[
[
"encoded_columns = pd.get_dummies(df['floors'], prefix='floors')\ndf = df.join(encoded_columns).drop('floors', axis=1)\ndf.head()",
"_____no_output_____"
]
],
[
[
"### Train Another Model and Evaluate It",
"_____no_output_____"
]
],
[
[
"from sklearn.model_selection import train_test_split\nfrom sklearn.preprocessing import StandardScaler\nfrom sklearn.metrics import r2_score, mean_squared_error, mean_absolute_error\n\nprint('Columns:', ', '.join(col for col in df.columns), '\\n')\n\n# Set your predictor variables to be all columns other than 'Price'\nX = df.drop(['price'], axis=1)\n\n# Set your response variable to be 'Price'\ny = df['price']\n\n# Perform train test split\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=21) \n\n# Scale data according to standard normal distribution\nscaler = StandardScaler()\nX_train = scaler.fit_transform(X_train, y_train)\nX_test = scaler.transform(X_test)\n\n# instantiate and train linear regression model\nLinReg = LinearRegression()\nLinReg.fit(X_train, y_train)\n\n# Make predictions with new model and assign them to a variable named 'predictions'\npredictions = LinReg.predict(X_test)\n\nprint('MAE:', mean_absolute_error(y_test, predictions))\nprint('MSE:', mean_squared_error(y_test, predictions))\nprint('RMSE:', np.sqrt(mean_squared_error(y_test, predictions)))\nprint('R2_score: ', r2_score(y_test, predictions))",
"_____no_output_____"
]
],
[
[
"### Important Predictors\n\nThe code block below will display which features are important and to 'extract' as features.\n\n> Replace all instance of ```None``` below with your own code.",
"_____no_output_____"
]
],
[
[
"from sklearn.linear_model import Lasso, Ridge\n\nmodel = None (alpha=1e-2, max_iter=1000) ## Insert code here\nmodel.fit(X_train, y_train)\n\nprint('\\nFeatures with non-zero weight (sorted by absolute magnitude):')\nfor e in sorted (list(zip(list(X.columns), model.coef_)), key = lambda e: -abs(e[1])):\n if e[1] != 0: print('• {}, {:.3f}'.format(e[0], e[1]))\nprint()",
"_____no_output_____"
]
],
[
[
"# Polynomial Regression\n\nDefine a custom function for easy data plotting",
"_____no_output_____"
]
],
[
[
"def plot_data(X, y, title='', X_line=[], y_line=[]):\n plt.figure(figsize=(8, 5))\n plt.scatter(X, y, marker='o', s=50, alpha=0.8)\n if len(y_line) > 0: # Only plot line when given\n plt.plot(X_line, y_line, 'r-')\n plt.title(title)\n plt.xlabel('Feature value (x)')\n plt.ylabel('Target value (y)')\n plt.show()",
"_____no_output_____"
]
],
[
[
"### 1-variable Polynomial Regression but \\\\(\\text{deg}(p(x))\\geq1\\\\)\n\nWe first generate the whole set of features using ```PolynomialFeatures```. \nThen, we apply the same multi-variable linear regression on these features.\n\n> Replace all instance of ```None``` below with your own code. \n> *Try changing the polynomial degree number and re-run the code, observe the graph plotted!*",
"_____no_output_____"
]
],
[
[
"import numpy as np\nfrom sklearn.preprocessing import PolynomialFeatures\n\nnp.random.seed(0)\nn = 15\nX_R2 = np.linspace(0, 10, n) + np.random.randn(n)/5\ny_R2 = np.sin(X_R2) + X_R2/6 + np.random.randn(n)/10\n\ntrain_scores = []\ntest_scores = []\nend_deg = 11\n\nfor deg in range(1, end_deg):\n \n poly = None (degree=deg) ## Insert code here\n X_poly = poly.fit_transform(X_R2.reshape(-1, 1))\n\n X_train, X_test, y_train, y_test = None (X_poly, y_R2, random_state=0) ## Insert code here\n\n linreg = LinearRegression().fit(None, None) ## Insert code here\n train_scores.append(linreg.score(X_train, y_train))\n test_scores.append(linreg.score(X_test, y_test))\n\n ## Try changing this to see different curves of best fit ##\n if deg == 7:\n x_fit = np.linspace(0, 10, 1000)\n y_fit = None.predict(None.fit_transform(x_fit.reshape(-1, 1))) ## Insert code here\n plot_data(X_R2, y_R2, 'Least squares polynomial regression of best deg=7', X_line=x_fit, y_line=y_fit)\n print(f'Features: {poly.get_feature_names()}')\n\n # Prematurely break for-loop when test R2 is negative\n if (test_scores[-1] < 0 and deg > 2) or (deg == end_deg): \n end_deg = deg\n break",
"_____no_output_____"
]
],
[
[
"The best fit curve is chosen by observing that the highest R2-score occurs at a polynomial degree of 7, as seen below. However, do be careful of overfitting!",
"_____no_output_____"
]
],
[
[
"plt.figure(figsize=(8, 5))\nplt.plot(np.arange(1, end_deg+1), train_scores, 'b-')\nplt.plot(np.arange(1, end_deg+1), test_scores, 'r--')\nplt.title('Train & test R2 scores')\nplt.xlabel('Polynomial degree')\nplt.ylabel('R-squared scores')\nplt.legend(['Train', 'Test'])\nplt.show()",
"_____no_output_____"
]
],
[
[
"### 2-variable Polynomial Regression\n\nModel features are generated from the original variables. \nEg. The multiplicative combination of variables $x_0$ and $x_1$ will form the set $\\{1, x_0, x_1, x_0^2, x_0x_1,x_1^2,...\\}$.",
"_____no_output_____"
]
],
[
[
"np.random.seed(0)\nn = 15\nX1_R3 = np.linspace(0, 10, n) + np.random.randn(n)/5\nX2_R3 = np.linspace(10, 0, n) + np.random.randn(n)/5\nX_R3 = np.array([X1_R3, X2_R3])\ny_R3 = np.sin(X1_R3) + np.sin(X2_R3) + X1_R3/6 + np.random.randn(n)/10\n\nfor deg in range(1, 5):\n poly = PolynomialFeatures(degree=deg)\n X_poly = poly.fit_transform(X_R3.reshape(-1, 2))\n linreg = LinearRegression().fit(X_poly, y_R3)\n print(f'Degree {deg} features: {poly.get_feature_names()}')",
"_____no_output_____"
]
],
[
[
"===== End of Polynomial Regression Part =====\n\n<br><hr>",
"_____no_output_____"
],
[
"# Logistic Regression\n\n- Linear regression: $y=\\mathbf{w}^T\\mathbf{x}+b$ \n- Logistic regression: $z=\\mathbf{w}^T\\mathbf{x}+b$ \nThe output $z$ from the linear regression equation is then passed through a sigmoid function i.e. $y=\\sigma(z)$, where $\\sigma(z)=\\frac{1}{1+e^{-z}}$ \n<img width=\"300px\" src=\"data:image/png;base64,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\"/> \n<br><br>\n- A graph of the sigmoid/logistic function is shown below; this function has a non-linear mapping. \n<img width=\"350px\" src=\"data:image/png;base64,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\"/>",
"_____no_output_____"
],
[
"## Implementation in Python",
"_____no_output_____"
],
[
"### Sigmoid Function\n\nWe will try to implement the sigmoid function using ```np.exp```. You may refer to the formula above.\n\n```python\nExpected output:\narray([0.26894142, 0.31479902, 0.36457644, 0.41742979, 0.47225076,\n 0.52774924, 0.58257021, 0.63542356, 0.68520098, 0.73105858])\n```\n \n> Replace all instance of ```None``` below with your own code.",
"_____no_output_____"
]
],
[
[
"def sigmoid(z):\n return None ## Insert code here\n\nsigmoid(np.linspace(-1, 1, 10))",
"_____no_output_____"
]
],
[
[
"You can try plotting the sigmoid curve that we have been talking about too!\n\n> Replace all instance of ```None``` below with your own code.",
"_____no_output_____"
]
],
[
[
"x = np.linspace(-10, 10, 100)\ny = sigmoid(None) ## Insert code here\n\nplt.figure(figsize=(6, 4))\nplt.plot(x, y)\nplt.plot(np.linspace(-10, 10, 100), np.ones(100)*0.5)\nplt.plot(np.zeros(100), np.linspace(0, 1, 100))\nplt.legend(['Sigmoid curve', 'Decision boundary'])\nplt.title('Sigmoid Function');",
"_____no_output_____"
]
],
[
[
"### Hypothesis and Loss Function\n\nThe hypothesis is $h_\\theta(x)=\\sigma(\\sum_{i=1}^m\\theta_ix_i)=\\sigma(\\mathbf{\\theta}^T\\mathbf{x})$. [Note: $\\mathbf{\\theta}$ is a vector of weights]\n\n\n*How did this happen? $z=\\mathbf{w}^T\\mathbf{x}+b=\\mathbf{\\theta}^T\\mathbf{x}$*\n\n<br>\n\nThe loss function mentioned in the slides is listed below:\n\\\\[J(\\theta)=-\\frac{1}{m}\\sum_{i=1}^m\\bigg(y^{(i)}\\log{(h_\\theta(x^{(i)}))}+(1-y^{(i)})\\log{(1-h_\\theta(x^{(i)}))}\\bigg)\\\\]\n\n<br>\n\nWe will observe how incorrect predictions will cause a large loss. Again, we use the power of vectorization to speed up our calculations. \nWe will also concatenate a vector of 1's to serve as our bias term.\n\n> There are some revisions to the code block below. \n> Replace all instance of ```None``` below with your own code.",
"_____no_output_____"
]
],
[
[
"x = np.array([[1, 2, 3, 4, 5], ## Modify this line of code\n [-5, 0, 3, -10, -1]]) ## Modify this line of code\ny = np.array([[1, 0, 0, 1, 0]])\n\nx = None ## New line of code\n\n# Initialise the weights with random values\nnp.random.seed(0)\ntheta = np.random.randn(None, 1) ## Modify this line of code\n\ndef print_loss(iter=0, verbose=True):\n\n hypothesis = None ## Insert code here\n loss_contribution = -(None * np.log(None) + None)\n \n if verbose:\n print('\\033[36mh(x)\\ty Loss (Individual Contribution)\\033[0m')\n for i in range(5):\n print(f'{hypothesis[0][i].round(3)}\\t{y[0][i]} {loss_contribution[0][i].round(3)}')\n print('')\n\n print(f'\\033[94mLoss (Iteration #{iter})\\033[0m = {None.round(3)}') ## Insert code here\n\nprint_loss()",
"_____no_output_____"
]
],
[
[
"### Gradient Descent\n\nNow, we update the weights using the gradient descent algorithm. The formula is given below:\n\n\\\\[\\theta_j\\leftarrow\\theta_j-\\alpha\\frac{1}{m}\\sum_{i=1}^m\\bigg(h_\\theta(x^{(i)})-y^{(i)}\\bigg)x^{(i)}_j\\\\]\n\n> *Try re-running the code block above and changing the learning rate (alpha) yourself!*",
"_____no_output_____"
]
],
[
[
"alpha = 0.05\n\ntheta = theta - alpha * np.expand_dims((np.mean(sigmoid(x)-y, axis=0) @ x.T), -1)\nprint('Iteration #1:')\nprint_loss(1)\nprint('\\n---------------------------------')\n\nfor i in range(2, 11):\n theta = theta - alpha * np.expand_dims((np.mean(sigmoid(x)-y, axis=0) @ x.T), -1) ## Modify this line of code\n print_loss(i, verbose=False)",
"_____no_output_____"
]
],
[
[
"As you can see, the loss decreases with each iteration. When it converges to a minima, the weights can then be used to classify from new input data.\n\nLet's have a look at the current weights.",
"_____no_output_____"
]
],
[
[
"theta",
"_____no_output_____"
]
],
[
[
"We can fix the prediction class at 0 to help visualise the decision boundary with $0=\\theta_0+\\theta_1x_0+\\theta_2x_1$. \nThis can be rearranged to:\n\\\\[x_1=\\frac{-\\theta_0-\\theta_1x_0}{\\theta_2}\\\\]\n\nLet's visualise this decision boundary!",
"_____no_output_____"
]
],
[
[
"from matplotlib.colors import ListedColormap\n\npredictors = x[[1,2], :]\nclasses = y\nweights = np.squeeze(theta)\n\nplt.figure(figsize=(5, 4))\nplt.scatter(predictors[0,:], predictors[1,:], marker='o', c=classes, cmap=ListedColormap(['Red', 'Green']))\nplt.xlabel('$x_0$')\nplt.ylabel('$x_1$')\n\nx0 = np.linspace(-2, 6, 100)\nplt.plot(x0, (0-weights[0]-weights[1]*x0)/weights[2])\nplt.legend(['Decision boundary', 'Data points']);",
"_____no_output_____"
]
],
[
[
"Note: The same decision boundary appears when we use a prediction class of 1. \n\nAll these math might seem a bit daunting. Fortunately, logistic regression is already implemented in ```scikit-learn```. Hence, we will use this for convenience in future examples.",
"_____no_output_____"
],
[
"## Binary Classification\n\nThis data science problem is to predict whether a customer purchases an item. Firstly, import the dataset. \nWe will be utilising ```sklearn.linear_model.LogisticRegression```. You can read more about this classifier from the documentation [here](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html).",
"_____no_output_____"
]
],
[
[
"import pandas as pd\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport seaborn as sns\n%matplotlib inline\n\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.preprocessing import StandardScaler\nfrom sklearn.linear_model import LogisticRegression\n\n# Explore data\ndf_log1 = pd.read_csv('purchase.csv')\ndf_log1.head()",
"_____no_output_____"
]
],
[
[
"### Exploratory Data Analysis (EDA)\nThere is only a little class imbalance.\n",
"_____no_output_____"
]
],
[
[
"plt.figure(figsize=(4, 3))\nsns.countplot(x='Purchased', data=df_log1, palette='Blues_d');",
"_____no_output_____"
]
],
[
[
"### Model Fitting and Predictions\n\nExtract the $x$ and $y$ data, then perform train-test split. After that, apply standardization to the training data before fitting the logistic regression classifier. (Click [here](https://www.analyticsvidhya.com/blog/2020/04/feature-scaling-machine-learning-normalization-standardization/) for a detailed explanation of why standardization is used)\n\n*Reflection Question:* Why is male/female not converted to 0/1 to be used in predictions?\n\n> Replace all instance of ```None``` below with your own code.",
"_____no_output_____"
]
],
[
[
"# Extract x and y data\nx = df_log1.iloc[:,[2,3]].values\ny = df_log1.iloc[:, 4].values\n\n# Train-test split\nx_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.2, random_state=0)\n\n# Feature Scaling\nsc = StandardScaler()\nx_train = sc.fit_transform(None) ## Insert code here\nx_test = sc.transform(None) ## Insert code here\n\n# Fitting logistic regression to the training set\nClassifier = None (random_state=0) ## Insert code here\nClassifier.fit(None, None) ## Insert code here\n\n# Predicting the results\nprint(\"Train accuracy score : \", Classifier.score(x_train, y_train))\nprint(\"Test accuracy score : \", Classifier.score(x_test, y_test))",
"_____no_output_____"
]
],
[
[
"We can see below that the model actually predicts the probabilities that the set of inputs is a 0 or 1. \nComparison is then done to determine the actual prediction. \n\n*The code below demonstrates the last 10 prediction values by the model.*",
"_____no_output_____"
]
],
[
[
"probs = Classifier.predict_proba(x_test)\nprint('\\033[36mIndex\\tP(0)\\tP(1)\\tPred\\033[0m')\nfor i in range(len(probs)-10, len(probs)):\n print(f'{i}\\t{probs[i,0].round(3)}\\t{probs[i,1].round(3)}\\t{int(probs[i,1].round(0))}')",
"_____no_output_____"
]
],
[
[
"The predictions can also be obtained directly by calling the ```predict()``` method. \nYou can compare the last 10 values below with the results above.",
"_____no_output_____"
]
],
[
[
"Classifier.predict(x_test)",
"_____no_output_____"
]
],
[
[
"The two plots below show the decision boundary and the actual labels of the data points for **both train and test set**. \nFrom the plots, it can be seen that the 2D space is divided into two regions - 0 and 1.",
"_____no_output_____"
]
],
[
[
"fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(16, 5))\ncmap = ['red', 'green']\n\ndef plot_log_reg(X_set, y_set, axes):\n\n X1, X2 = np.meshgrid(np.arange(start = X_set[:, 0].min() - 1, stop = X_set[:, 0].max() + 1, step = 0.01),\n np.arange(start = X_set[:, 1].min() - 1, stop = X_set[:, 0].max() + 1, step = 0.01))\n\n ax[axes].contourf(X1, X2, Classifier.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape), alpha = 0.75)\n\n for i, j in enumerate(np.unique(y_set)):\n ax[axes].scatter(X_set[y_set==j, 0], X_set[y_set==j, 1], c=cmap[i], label = j) \n\n ax[axes].set_xlim(X1.min(), X1.max())\n ax[axes].set_ylim(X1.min(), X1.max())\n ax[axes].set_title(f'Logistic Regression ({\"Train\" if not axes else \"Test\"} set)')\n ax[axes].set_xlabel('Age')\n ax[axes].set_ylabel('Estimated Salary')\n ax[axes].legend();\n\nplot_log_reg(x_train, y_train, 0)\nplot_log_reg(x_test, y_test, 1)",
"_____no_output_____"
]
],
[
[
"## Multiclass Classification\n\nFrom the documentation, ```sklearn.linear_model.LogisticRegression``` implements the one-vs-rest (OvR) scheme or multinomial for multiclass cases.\n\nThis data science problem is to predict the species of flower based on its features. Firstly, import the [dataset](https://github.com/mwaskom/seaborn-data/blob/master/iris.csv). \n<img width=\"800px\" src=\"data:image/png;base64,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\"/>",
"_____no_output_____"
]
],
[
[
"df_log2 = pd.read_csv('iris.csv')\ndf_log2.head()",
"_____no_output_____"
]
],
[
[
"Pre-process the data as usual; EDA won't be shown in the example here but it should be done to gain insight from the dataset.\n\n> Replace all instance of ```None``` below with your own code.",
"_____no_output_____"
]
],
[
[
"# Extract x and y data\nx = df_log2.iloc[:, range(None)].values ## Insert code here\ny = df_log2.iloc[:, None].values ## Insert code here\n\n# Train-test split\nx_train, x_test, y_train, y_test = None (x, y, test_size=0.2, random_state=1) ## Insert code here\n\n# Feature Scaling\nsc = StandardScaler()\nx_train = sc.fit_transform(x_train)\nx_test = sc.transform(x_test)",
"_____no_output_____"
]
],
[
[
"We will try out both OvR and Multinomial in the code blocks below.\n\n> Replace all instance of ```None``` below with your own code.",
"_____no_output_____"
]
],
[
[
"# Fitting logistic regression to the training set\nClassifier = LogisticRegression(random_state=0, multi_class='ovr')\nClassifier.fit(None, None) ## Insert code here\n\n# Predicting the results\nprint('\\033[94mOne-vs-rest (OvR) scheme\\033[0m')\nprint('- Train accuracy score :', Classifier.score(x_train, y_train))\nprint('- Test accuracy score :', Classifier.score(x_test, y_test), end='\\n\\n')\n\n# Sample of predictions\nprobs = Classifier.predict_proba(None) ## Insert code here\nprint(f'\\033[36mIndex\\tP({Classifier.classes_[0]})\\tP({Classifier.classes_[1]})\\tP({Classifier.classes_[2]})\\tPrediction\\033[0m')\nfor i in range(10):\n print(f'{i}\\t{probs[i,0].round(3)}\\t\\t{probs[i,1].round(3)}\\t\\t{probs[i,2].round(3)}\\t\\t{Classifier.predict(None)[i]}') ## Insert code here",
"_____no_output_____"
],
[
"# Fitting logistic regression to the training set\nClassifier = LogisticRegression(random_state=0, multi_class=None) ## Insert code here\nClassifier.fit(x_train, y_train)\n\n# Predicting the results\nprint('\\033[94mMultinomial\\033[0m')\nprint('- Train accuracy score :', Classifier.score(x_train, y_train))\nprint('- Test accuracy score :', Classifier.score(x_test, y_test), end='\\n\\n')\n\n# Sample of predictions\nprobs = Classifier.predict_proba(x_test) ## Insert code here\nprint(f'\\033[36mIndex\\tP({Classifier.classes_[0]})\\tP({Classifier.classes_[1]})\\tP({Classifier.classes_[2]})\\tPrediction\\033[0m')\nfor i in range(10):\n print(f'{i}\\t{probs[i,0].round(3)}\\t\\t{probs[i,1].round(3)}\\t\\t{probs[i,2].round(3)}\\t\\t{Classifier.predict(x_test)[i]}')",
"_____no_output_____"
]
],
[
[
"Check out the coefficients of the features used to make the classification (multinomial method).",
"_____no_output_____"
]
],
[
[
"Classifier.coef_",
"_____no_output_____"
]
],
[
[
"===== End of Logistic Regression Part =====\n\n<br><hr>",
"_____no_output_____"
],
[
"# Thank you for listening!\n\nThat is all, we have come to the end of our workshop! There's also a way to implement binary and multiclass classification in neural networks. They make use of the same *sigmoid* and *softmax* function that you have just learned as activation functions for the neurons, so don't forget them! \n\n",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
]
] |
cb90c48af97eaf8fd1f5e7e4f79ab91ea6efed22
| 288,979 |
ipynb
|
Jupyter Notebook
|
2018-08-18_ch_Radial_Base_Functions.ipynb
|
caheredia/Radial_Base_Functions
|
7b7344935f3b01d80d15a4ef6f2b05f6f77b8aa8
|
[
"MIT"
] | null | null | null |
2018-08-18_ch_Radial_Base_Functions.ipynb
|
caheredia/Radial_Base_Functions
|
7b7344935f3b01d80d15a4ef6f2b05f6f77b8aa8
|
[
"MIT"
] | null | null | null |
2018-08-18_ch_Radial_Base_Functions.ipynb
|
caheredia/Radial_Base_Functions
|
7b7344935f3b01d80d15a4ef6f2b05f6f77b8aa8
|
[
"MIT"
] | null | null | null | 183.945894 | 82,630 | 0.838047 |
[
[
[
"import pandas as pd\nimport numpy as np \nimport matplotlib as plt \nimport seaborn as sns\n%matplotlib inline",
"_____no_output_____"
]
],
[
[
"# Build dataframe with data for plotting\nhttp://koaning.io/radial-basis-functions.html\n\n$\\phi_{i} (x) = exp\\left( \\frac{-1}{2 \\alpha} (x-m_i)^2 \\right)$",
"_____no_output_____"
]
],
[
[
"def rbf(x, alpha, m):\n return np.exp(-1/(2*alpha)*(x-m)**2)",
"_____no_output_____"
],
[
"# Generate data with rbf\nx = np.arange(-2, 5, 0.01)\ndf=pd.DataFrame(data=x, columns=['x']).set_index('x')\n## make 5 radial functions, with different m \nfor i in range(5):\n df[str(i)] = rbf(x,.25,i)\ndf.rename_axis('m', axis='columns', inplace = True)\ndf.head()",
"_____no_output_____"
],
[
"df.plot();",
"_____no_output_____"
]
],
[
[
"# Example: pattern learning with rbf and linear regression\nAssuming $\\alpha=1$, then \n\n$\\phi_{i} (x) = exp\\left( \\frac{-1}{2} (x-m_i)^2 \\right)$",
"_____no_output_____"
],
[
"## Generate monthly data",
"_____no_output_____"
]
],
[
[
"# Generate rbf\nx = np.arange(0, 12, 0.01)\ndf=pd.DataFrame(data=x, columns=['x'])\ndf['y'] = np.sin(x) + 2*np.cos(x/2) + np.random.normal(loc=0.0, scale=.2, size=len(df))\ndf.head()",
"_____no_output_____"
],
[
"df.plot(kind='scatter', x='x', y='y');",
"_____no_output_____"
]
],
[
[
"## With $\\alpha =1$ add a radial function for each month ",
"_____no_output_____"
]
],
[
[
"for i in range(1,13):\n df[str(i)] = rbf(df['x'],1,i)",
"_____no_output_____"
],
[
"df.head()",
"_____no_output_____"
],
[
"df.set_index('x').plot(figsize=(12,6));",
"_____no_output_____"
]
],
[
[
"## Plot two different fits: one with floor and the other with rbf",
"_____no_output_____"
]
],
[
[
"dfz =pd.DataFrame()\nfor i in range(1,13):\n dfz[i]=np.floor(df['x'])",
"_____no_output_____"
]
],
[
[
"# Forecasting bridge data with rbf linear regression ",
"_____no_output_____"
]
],
[
[
"# lin reg with z model \nX = dfz\nY=df['y']\nmodel = sm.OLS( Y, X ).fit()\ndf['model_z'] = model.predict(X)\nmodel.summary()",
"/Users/cristian/anaconda/lib/python3.6/site-packages/statsmodels/regression/linear_model.py:1471: RuntimeWarning: divide by zero encountered in double_scalars\n return np.sqrt(eigvals[0]/eigvals[-1])\n"
],
[
"df.columns",
"_____no_output_____"
],
[
"#lin reg with rbf\nX = df[['1', '2', '3', '4', '5', '6', '7', '8', '9', '10', '11', '12']]\nY=df['y']\nmodel = sm.OLS( Y, X ).fit()\ndf['model_rbf'] = model.predict(X)\nmodel.summary()",
"_____no_output_____"
],
[
"df[['model_z','x', 'model_rbf','y']].plot(x='x', alpha=.5);",
"_____no_output_____"
]
],
[
[
"# Testing linear regression methods and results",
"_____no_output_____"
]
],
[
[
"import statsmodels.formula.api as sm",
"_____no_output_____"
],
[
"x = np.arange(-1, 1.1, .1)\ndf = pd.DataFrame(data=x, columns=['x'])\ndf['intercept'] =np.ones(len(df))\ndf['y'] = 2*df['x']+3+ np.random.normal(loc=0.0, scale=.2, size=len(df)) \ndf.head()",
"_____no_output_____"
],
[
"df.plot(kind='scatter', x='x',y='y');",
"_____no_output_____"
],
[
"df.plot( x='x', y=['y','model']);",
"/Users/cristian/anaconda/lib/python3.6/site-packages/pandas/plotting/_core.py:1714: UserWarning: Pandas doesn't allow columns to be created via a new attribute name - see https://pandas.pydata.org/pandas-docs/stable/indexing.html#attribute-access\n series.name = label\n"
]
],
[
[
"## add more independent variables",
"_____no_output_____"
]
],
[
[
"x = np.arange(-1, 1.1, .1)\ndf = pd.DataFrame(data=x, columns=['x'])\ndf['intercept'] =np.ones(len(df))\ndf['cos'] = 2*np.cos(df['x'])\ndf['y'] = 2*df['x']+3 + df['cos'] + np.random.normal(loc=0.0, scale=.2, size=len(df))\ndf.head()",
"_____no_output_____"
],
[
"df.plot(kind='scatter', x='x',y='y');",
"_____no_output_____"
],
[
"X = df[['x', 'intercept', 'cos']]\nY = df['y']\nmodel = sm.OLS( Y, X ).fit()\ndf['model'] = model.predict(X)\nmodel.summary()",
"_____no_output_____"
],
[
"df[ ['x', 'y','model' ] ].plot( x=\"x\");",
"_____no_output_____"
]
],
[
[
"## add non-contributing column ",
"_____no_output_____"
]
],
[
[
"x = np.arange(-1, 1.1, .1)\ndf = pd.DataFrame(data=x, columns=['x'])\ndf['intercept'] =np.ones(len(df))\ndf['cos'] = 2*np.cos(df['x'])\ndf['sin'] = 2*np.sin(df['x'])\ndf['y'] = 2*df['x']+3 + df['cos'] + np.random.normal(loc=0.0, scale=.2, size=len(df))\ndf.head()",
"_____no_output_____"
],
[
"X = df[['x', 'intercept', 'cos', 'sin']]\nY = df['y']\nmodel = sm.OLS( Y, X ).fit()\ndf['model'] = model.predict(X)\nmodel.summary()",
"_____no_output_____"
],
[
"df[ ['x', 'y','model' ] ].plot( x=\"x\");",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
]
] |
cb90c59a8d43b1e43d4cbfb46817bac42fbbb5ab
| 80,582 |
ipynb
|
Jupyter Notebook
|
Seals/Notebooks/TimelineComparison.ipynb
|
radical-experiments/iceberg_escience
|
e5c230a23395a71a4adf554730ea3d77f923166c
|
[
"MIT"
] | 1 |
2019-05-24T02:19:29.000Z
|
2019-05-24T02:19:29.000Z
|
Seals/Notebooks/TimelineComparison.ipynb
|
radical-experiments/iceberg_seals
|
e5c230a23395a71a4adf554730ea3d77f923166c
|
[
"MIT"
] | null | null | null |
Seals/Notebooks/TimelineComparison.ipynb
|
radical-experiments/iceberg_seals
|
e5c230a23395a71a4adf554730ea3d77f923166c
|
[
"MIT"
] | null | null | null | 252.60815 | 34,924 | 0.898451 |
[
[
[
"%matplotlib inline\nimport numpy as np\nfrom matplotlib import pyplot as plt\nfrom matplotlib import cm\nimport pandas as pd\nimport matplotlib as mpl\nmpl.rcParams['text.usetex'] = True\nmpl.rcParams['text.latex.unicode'] = True\n\nblues = cm.get_cmap(plt.get_cmap('Blues'))\ngreens = cm.get_cmap(plt.get_cmap('Greens'))\nreds = cm.get_cmap(plt.get_cmap('Reds'))\noranges = cm.get_cmap(plt.get_cmap('Oranges'))\npurples = cm.get_cmap(plt.get_cmap('Purples'))\ngreys = cm.get_cmap(plt.get_cmap('Greys'))\n\nfrom IPython.core.display import display, HTML\ndisplay(HTML(\"<style>.container { width:100% !important; }</style>\"))\nimport warnings\nwarnings.filterwarnings('ignore')",
"_____no_output_____"
],
[
"des2a = pd.read_csv('../Data/design2a.csv')\ndes2b = pd.read_csv('../Data/design2b.csv')\n",
"_____no_output_____"
],
[
"fig, axis = plt.subplots(nrows=2,ncols=4,figsize=(11,5),sharex='row',sharey='row')\nx1 = np.arange(1)\nx2 = np.arange(1)\n_ = axis[0,0].bar(x1,des2a['CpuMeanTime'][0],yerr=des2a['CpuStdTime'][0], color=blues(150))\n_ = axis[0,0].bar(x2+1,des2a['GpuMeanTime'][0],yerr=des2a['GpuStdTime'][0], color=reds(150))\n_ = axis[0,0].set_xticks([0,1])\n_ = axis[0,0].grid('on', linestyle=':', linewidth=2)\n_ = axis[0,0].set_xticklabels(['CPUs', 'GPUs'], fontsize=22)\n\n_ = axis[0,1].bar(x1,des2a['CpuMeanTime'][1],yerr=des2a['CpuStdTime'][1], color=blues(150))\n_ = axis[0,1].bar(x2+1,des2a['GpuMeanTime'][1],yerr=des2a['GpuStdTime'][1], color=reds(150))\n_ = axis[0,1].set_xticks([0,1])\n_ = axis[0,1].grid('on', linestyle=':', linewidth=2)\n_ = axis[0,1].set_xticklabels(['CPUs', 'GPUs'], fontsize=22)\n\n_ = axis[0,2].bar(x1,des2a['CpuMeanTime'][2],yerr=des2a['CpuStdTime'][2], color=blues(150))\n_ = axis[0,2].bar(x2+1,des2a['GpuMeanTime'][2],yerr=des2a['GpuStdTime'][2], color=reds(150))\n_ = axis[0,2].set_xticks([0,1])\n_ = axis[0,2].grid('on', linestyle=':', linewidth=2)\n_ = axis[0,2].set_xticklabels(['CPUs', 'GPUs'], fontsize=22)\n\n_ = axis[0,3].bar(x1,des2a['CpuMeanTime'][3],yerr=des2a['CpuStdTime'][3], color=blues(150))\n_ = axis[0,3].bar(x2+1,des2a['GpuMeanTime'][3],yerr=des2a['GpuStdTime'][3], color=reds(150))\n_ = axis[0,3].set_xticks([0,1])\n_ = axis[0,3].grid('on', linestyle=':', linewidth=2)\n_ = axis[0,3].set_xticklabels(['CPUs', 'GPUs'], fontsize=22)\n_ = axis[0,0].set_yticklabels(axis[0,0].get_yticks().astype('int').tolist(),fontsize=22)\n\n_ = axis[1,0].hist(eval(des2a['Images'][0]),bins=50)\n_ = axis[1,1].hist(eval(des2a['Images'][1]),bins=50)\n_ = axis[1,2].hist(eval(des2a['Images'][2]),bins=50)\n_ = axis[1,3].hist(eval(des2a['Images'][3]),bins=50)\n_ = axis[0,0].set_title('Node 1',fontsize=24)\n_ = axis[0,1].set_title('Node 2',fontsize=24)\n_ = axis[0,2].set_title('Node 3',fontsize=24)\n_ = axis[0,3].set_title('Node 4',fontsize=24)\n_ = axis[1,0].set_xlabel('Image Size\\nin MBs', fontsize=24)\n_ = axis[1,1].set_xlabel('Image Size\\nin MBs', fontsize=24)\n_ = axis[1,2].set_xlabel('Image Size\\nin MBs', fontsize=24)\n_ = axis[1,3].set_xlabel('Image Size\\nin MBs', fontsize=24)\n_ = axis[0,0].set_ylabel('Time', fontsize=24)\n_ = axis[1,0].set_yticklabels(axis[1,0].get_yticks().astype('int').tolist(),fontsize=22)\n_ = axis[1,0].set_xticklabels(axis[1,0].get_xticks().astype('int').tolist(),fontsize=22)\n_ = axis[1,1].set_xticklabels(axis[1,1].get_xticks().astype('int').tolist(),fontsize=22)\n_ = axis[1,2].set_xticklabels(axis[1,2].get_xticks().astype('int').tolist(),fontsize=22)\n_ = axis[1,3].set_xticklabels(axis[1,3].get_xticks().astype('int').tolist(),fontsize=22)\n\n_ = axis[1,0].grid('on', linestyle=':', linewidth=2)\n_ = axis[1,1].grid('on', linestyle=':', linewidth=2)\n_ = axis[1,2].grid('on', linestyle=':', linewidth=2)\n_ = axis[1,3].grid('on', linestyle=':', linewidth=2)\n_ = axis[1,0].set_ylabel('Number of\\nImages', fontsize=24)\n_ = axis[0,0].set_ylabel('Time', fontsize=24)\n_ = axis[1,0].set_yticklabels(axis[1,0].get_yticks().astype('int').tolist(),fontsize=22)\n#fig.savefig('design2_timelines.pdf',dpi=800,bbox_inches='tight')",
"_____no_output_____"
],
[
"level0 = axis[0,0].get_ylim()\nlevel1 = axis[1,0].get_ylim()",
"_____no_output_____"
],
[
"fig, axis = plt.subplots(nrows=2,ncols=4,figsize=(11,5),sharex='row',sharey='row')\n\nx1 = np.arange(1)\nx2 = np.arange(1)\n_ = axis[0,0].set_ylim(level0)\n_ = axis[1,0].set_ylim(level1)\n_ = axis[0,0].bar(x1,des2b['CpuMeanTime'][0],yerr=des2b['CpuStdTime'][0], color=blues(150))\n_ = axis[0,0].bar(x2+1,des2b['GpuMeanTime'][0],yerr=des2b['GpuStdTime'][0], color=reds(150))\n_ = axis[0,0].set_xticks([0,1])\n_ = axis[0,0].grid('on', linestyle=':', linewidth=2)\n_ = axis[0,0].set_xticklabels(['CPUs', 'GPUs'], fontsize=22)\n\n_ = axis[0,1].bar(x1,des2b['CpuMeanTime'][1],yerr=des2b['CpuStdTime'][1], color=blues(150))\n_ = axis[0,1].bar(x2+1,des2b['GpuMeanTime'][1],yerr=des2b['GpuStdTime'][1], color=reds(150))\n_ = axis[0,1].set_xticks([0,1])\n_ = axis[0,1].grid('on', linestyle=':', linewidth=2)\n_ = axis[0,1].set_xticklabels(['CPUs', 'GPUs'], fontsize=22)\n\n_ = axis[0,2].bar(x1,des2b['CpuMeanTime'][2],yerr=des2b['CpuStdTime'][2], color=blues(150))\n_ = axis[0,2].bar(x2+1,des2b['GpuMeanTime'][2],yerr=des2b['GpuStdTime'][2], color=reds(150))\n_ = axis[0,2].set_xticks([0,1])\n_ = axis[0,2].grid('on', linestyle=':', linewidth=2)\n_ = axis[0,2].set_xticklabels(['CPUs', 'GPUs'], fontsize=22)\n\n_ = axis[0,3].bar(x1,des2b['CpuMeanTime'][3],yerr=des2b['CpuStdTime'][3], color=blues(150))\n_ = axis[0,3].bar(x2+1,des2b['GpuMeanTime'][3],yerr=des2b['GpuStdTime'][3], color=reds(150))\n_ = axis[0,3].set_xticks([0,1])\n_ = axis[0,3].grid(axis='both', linestyle=':', linewidth=2)\n_ = axis[0,3].set_xticklabels(['CPUs', 'GPUs'], fontsize=22)\n_ = axis[0,0].set_yticklabels(axis[0,0].get_yticks().astype('int').tolist(),fontsize=22)\n\n_ = axis[1,0].grid('on', linestyle=':', linewidth=2)\n_ = axis[1,1].grid('on', linestyle=':', linewidth=2)\n_ = axis[1,2].grid('on', linestyle=':', linewidth=2)\n_ = axis[1,3].grid('on', linestyle=':', linewidth=2)\n_ = axis[1,0].hist(eval(des2b['Images'][0]),bins=50)\n_ = axis[1,1].hist(eval(des2b['Images'][1]),bins=50)\n_ = axis[1,2].hist(eval(des2b['Images'][2]),bins=50)\n_ = axis[1,3].hist(eval(des2b['Images'][3]),bins=50)\n_ = axis[0,0].set_title('Node 1',fontsize=24)\n_ = axis[0,1].set_title('Node 2',fontsize=24)\n_ = axis[0,2].set_title('Node 3',fontsize=24)\n_ = axis[0,3].set_title('Node 4',fontsize=24)\n_ = axis[1,0].set_xlabel('Image Size\\nin MBs', fontsize=24)\n_ = axis[1,1].set_xlabel('Image Size\\nin MBs', fontsize=24)\n_ = axis[1,2].set_xlabel('Image Size\\nin MBs', fontsize=24)\n_ = axis[1,3].set_xlabel('Image Size\\nin MBs', fontsize=24)\n_ = axis[1,0].set_ylabel('Number of\\nImages', fontsize=24)\n_ = axis[0,0].set_ylabel('Time', fontsize=24)\n_ = axis[1,0].set_yticklabels(axis[1,0].get_yticks().astype('int').tolist(),fontsize=22)\n_ = axis[1,0].set_xticklabels(axis[1,0].get_xticks().astype('int').tolist(),fontsize=22)\n_ = axis[1,1].set_xticklabels(axis[1,1].get_xticks().astype('int').tolist(),fontsize=22)\n_ = axis[1,2].set_xticklabels(axis[1,2].get_xticks().astype('int').tolist(),fontsize=22)\n_ = axis[1,3].set_xticklabels(axis[1,3].get_xticks().astype('int').tolist(),fontsize=22)\n#fig.savefig('design2a_timelines.pdf',dpi=800,bbox_inches='tight')",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code"
]
] |
cb90d2ece3032349c195ec6b61c1a9f570384d3f
| 6,165 |
ipynb
|
Jupyter Notebook
|
Audio Visual/Problems/HandGestureDetection_OpenCV_Video.ipynb
|
pavi-ninjaac/ai-seed
|
2fa0b19d1b5d902d2510eefa4eb1b8be26610cd8
|
[
"Apache-2.0"
] | 53 |
2021-08-28T07:41:49.000Z
|
2022-03-09T02:20:17.000Z
|
Audio Visual/Problems/HandGestureDetection_OpenCV_Video.ipynb
|
pavi-ninjaac/ai-seed
|
2fa0b19d1b5d902d2510eefa4eb1b8be26610cd8
|
[
"Apache-2.0"
] | 142 |
2021-07-27T07:23:10.000Z
|
2021-08-25T14:57:24.000Z
|
Audio Visual/Problems/HandGestureDetection_OpenCV_Video.ipynb
|
pavi-ninjaac/ai-seed
|
2fa0b19d1b5d902d2510eefa4eb1b8be26610cd8
|
[
"Apache-2.0"
] | 38 |
2021-07-27T04:54:08.000Z
|
2021-08-23T02:27:20.000Z
| 31.136364 | 172 | 0.487916 |
[
[
[
"# HandGestureDetection using OpenCV",
"_____no_output_____"
],
[
"This code template is for Hand Gesture detection in a video using OpenCV Library.",
"_____no_output_____"
],
[
"### Required Packages",
"_____no_output_____"
]
],
[
[
"!pip install opencv-python\n!pip install mediapipe",
"_____no_output_____"
],
[
"import cv2\nimport mediapipe as mp\nimport time",
"_____no_output_____"
]
],
[
[
"### Hand Detection\nFor detecting hands in the image, we use the detectMultiScale() function.\n\nIt detects objects of different sizes in the input image. The detected objects are returned as a list of rectangles which can be then plotted around the hands.\n\n####Tuning Parameters:\n**image** - Matrix of the type CV_8U containing an image where objects are detected.\n\n**objects** - Vector of rectangles where each rectangle contains the detected object, the rectangles may be partially outside the original image.\n\n**scaleFactor** - Parameter specifying how much the image size is reduced at each image scale.\n\n**minNeighbors** - Parameter specifying how many neighbors each candidate rectangle should have to retain it.",
"_____no_output_____"
]
],
[
[
"class handDetector():\n def __init__(self, mode = False, maxHands = 2, detectionCon = 0.5, trackCon = 0.5):\n self.mode = mode\n self.maxHands = maxHands\n self.detectionCon = detectionCon\n self.trackCon = trackCon\n\n self.mpHands = mp.solutions.hands\n self.hands = self.mpHands.Hands(self.mode, self.maxHands, self.detectionCon, self.trackCon)\n self.mpDraw = mp.solutions.drawing_utils\n \n def findHands(self,img, draw = True):\n imgRGB = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n self.results = self.hands.process(imgRGB)\n # print(results.multi_hand_landmarks)\n\n if self.results.multi_hand_landmarks:\n for handLms in self.results.multi_hand_landmarks:\n if draw:\n self.mpDraw.draw_landmarks(img, handLms, self.mpHands.HAND_CONNECTIONS)\n return img\n\n def findPosition(self, img, handNo = 0, draw = True):\n\n lmlist = []\n if self.results.multi_hand_landmarks:\n myHand = self.results.multi_hand_landmarks[handNo]\n for id, lm in enumerate(myHand.landmark):\n h, w, c = img.shape\n cx, cy = int(lm.x * w), int(lm.y * h)\n lmlist.append([id, cx, cy])\n if draw:\n cv2.circle(img, (cx, cy), 3, (255, 0, 255), cv2.FILLED)\n return lmlist",
"_____no_output_____"
]
],
[
[
"To run the handDetector(), save this file with .py extension and allow your webcam to take a video. The co-ordinates will be outputted at the terminal.",
"_____no_output_____"
]
],
[
[
"pTime = 0\ncTime = 0\ncap = cv2.VideoCapture(0)\ndetector = handDetector()\n\nwhile True:\n success, img = cap.read()\n img = detector.findHands(img)\n lmlist = detector.findPosition(img)\n if len(lmlist) != 0:\n print(lmlist[4])\n\n cTime = time.time()\n fps = 1 / (cTime - pTime)\n pTime = cTime\n\n cv2.putText(img, str(int(fps)), (10, 70), cv2.FONT_HERSHEY_PLAIN, 3, (255, 0, 255), 3)\n\n cv2.imshow(\"Image\", img)\n cv2.waitKey(1)",
"_____no_output_____"
]
],
[
[
"#### Creator: Ayush Gupta , Github: [Profile](https://github.com/guptayush179)",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
]
] |
cb90dd3748b93b66d75d8d928b8d8565bb50858c
| 34,056 |
ipynb
|
Jupyter Notebook
|
notebooks/_01-tb-convert-wide-format-car-data-to-long-format.ipynb
|
timothyb0912/check-yourself
|
bbca2afcd63c068401913de2dd98f64569373c97
|
[
"MIT"
] | 7 |
2018-12-30T04:31:41.000Z
|
2021-08-31T17:07:28.000Z
|
notebooks/_01-tb-convert-wide-format-car-data-to-long-format.ipynb
|
timothyb0912/check-yourself
|
bbca2afcd63c068401913de2dd98f64569373c97
|
[
"MIT"
] | 3 |
2020-03-09T16:47:37.000Z
|
2021-04-25T15:55:13.000Z
|
notebooks/_01-tb-convert-wide-format-car-data-to-long-format.ipynb
|
timothyb0912/check-yourself
|
bbca2afcd63c068401913de2dd98f64569373c97
|
[
"MIT"
] | 2 |
2019-12-27T15:45:40.000Z
|
2020-07-29T20:31:57.000Z
| 34.715596 | 708 | 0.336974 |
[
[
[
"The purpose of this notebook is to convert the wide-format car data to long-format. The car data comes from the mlogit package. The data description is reproduced below. Note the data originally comes from McFadden and Train (2000).\n\n#### Description\n- Cross-Sectional Dataset\n- Number of Observations: 4,654\n- Unit of Observation: Individual\n- Country: United States\n\n#### Format\nA dataframe containing :\n- choice: choice of a vehicule amoung 6 propositions\n- college: college education?\n- hsg2: size of household greater than 2?\n- coml5: commulte lower than 5 miles a day?\n- typez: body type, one of regcar (regular car), sportuv (sport utility vehicule), sportcar, stwagon (station wagon), truck, van, for each proposition z from 1 to 6\n- fuelz: fuel for proposition z, one of gasoline, methanol, cng (compressed natural gas), electric. pricez price of vehicule divided by the logarithme of income\n- rangez: hundreds of miles vehicule can travel between refuelings/rechargings\n- accz: acceleration, tens of seconds required to reach 30 mph from stop\n- speedz: highest attainable speed in hundreds of mph\n- pollutionz: tailpipe emissions as fraction of those for new gas vehicule\n- sizez: 0 for a mini, 1 for a subcompact, 2 for a compact and 3 for a mid–size or large vehicule\n- spacez: fraction of luggage space in comparable new gas vehicule\n- costz: cost per mile of travel(tens of cents). Either cost of home recharging for electric vehicule or the cost of station refueling otherwise\n- stationz: fraction of stations that can refuel/recharge vehicule\n\n#### Source\nMcFadden, Daniel and Kenneth Train (2000) “Mixed MNL models for discrete response”, Journal of Applied Econometrics, 15(5), 447–470.\n\n\nJournal of Applied Econometrics data archive : http://jae.wiley.com/jae/",
"_____no_output_____"
]
],
[
[
"import pandas as pd\nimport numpy as np\nimport pylogit as pl",
"_____no_output_____"
]
],
[
[
"# Load the Car data",
"_____no_output_____"
]
],
[
[
"wide_car = pd.read_csv(\"../data/raw/car_wide_format.csv\")\nwide_car.head().T",
"_____no_output_____"
]
],
[
[
"# Convert the Car dataset to long-format",
"_____no_output_____"
]
],
[
[
"# Look at the columns of the car data\nprint(wide_car.columns.tolist())",
"['choice', 'college', 'hsg2', 'coml5', 'type1', 'type2', 'type3', 'type4', 'type5', 'type6', 'fuel1', 'fuel2', 'fuel3', 'fuel4', 'fuel5', 'fuel6', 'price1', 'price2', 'price3', 'price4', 'price5', 'price6', 'range1', 'range2', 'range3', 'range4', 'range5', 'range6', 'acc1', 'acc2', 'acc3', 'acc4', 'acc5', 'acc6', 'speed1', 'speed2', 'speed3', 'speed4', 'speed5', 'speed6', 'pollution1', 'pollution2', 'pollution3', 'pollution4', 'pollution5', 'pollution6', 'size1', 'size2', 'size3', 'size4', 'size5', 'size6', 'space1', 'space2', 'space3', 'space4', 'space5', 'space6', 'cost1', 'cost2', 'cost3', 'cost4', 'cost5', 'cost6', 'station1', 'station2', 'station3', 'station4', 'station5', 'station6']\n"
],
[
"# Create the list of individual specific variables\nind_variables = wide_car.columns.tolist()[1:4]\n\n# Specify the variables that vary across individuals and some or all alternatives\n# The keys are the column names that will be used in the long format dataframe.\n# The values are dictionaries whose key-value pairs are the alternative id and\n# the column name of the corresponding column that encodes that variable for\n# the given alternative. Examples below.\nnew_name_to_old_base = {'body_type': 'type{}',\n 'fuel_type': 'fuel{}',\n 'price_over_log_income': 'price{}',\n 'range': 'range{}',\n 'acceleration': 'acc{}',\n 'top_speed': 'speed{}',\n 'pollution': 'pollution{}',\n 'vehicle_size': 'size{}',\n 'luggage_space': 'space{}',\n 'cents_per_mile': 'cost{}',\n 'station_availability': 'station{}'}\n\nalt_varying_variables =\\\n {k: dict([(x, v.format(x)) for x in range(1, 7)])\n for k, v in list(new_name_to_old_base.items())}\n\n# Specify the availability variables\n# Note that the keys of the dictionary are the alternative id's.\n# The values are the columns denoting the availability for the\n# given mode in the dataset.\navailability_variables =\\\n {x: 'avail_{}'.format(x) for x in range(1, 7)}\nfor col in availability_variables.values():\n wide_car[col] = 1\n\n##########\n# Determine the columns for: alternative ids, the observation ids and the choice\n##########\n# The 'custom_alt_id' is the name of a column to be created in the long-format data\n# It will identify the alternative associated with each row.\ncustom_alt_id = \"alt_id\"\n\n# Create a custom id column that ignores the fact that this is a \n# panel/repeated-observations dataset. Note the +1 ensures the id's start at one.\nobs_id_column = \"obs_id\"\nwide_car[obs_id_column] =\\\n np.arange(1, wide_car.shape[0] + 1, dtype=int)\n\n\n# Create a variable recording the choice column\nchoice_column = \"choice\"\n# Store the original choice column in a new variable\nwide_car['orig_choices'] = wide_car['choice'].values\n# Alter the original choice column\nchoice_str_to_value = {'choice{}'.format(x): x for x in range(1, 7)}\nwide_car[choice_column] =\\\n wide_car[choice_column].map(choice_str_to_value)\n\n# Convert the wide-format data to long format\nlong_car =\\\n pl.convert_wide_to_long(wide_data=wide_car,\n ind_vars=ind_variables,\n alt_specific_vars=alt_varying_variables,\n availability_vars=availability_variables,\n obs_id_col=obs_id_column,\n choice_col=choice_column,\n new_alt_id_name=custom_alt_id)\n\nlong_car.head().T",
"/Users/timothyb0912/anaconda/lib/python2.7/site-packages/pylogit-0.1.2-py2.7.egg/pylogit/choice_tools.py:421: UserWarning: Note, there are 78 variables in wide_data but the inputs ind_vars, alt_specific_vars, and subset_specific_vars only account for 77 variables.\n"
],
[
"# Save the long-format data\nlong_car.to_csv(\"../data/interim/car_long_format.csv\",\n index=False)",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
]
] |
cb90df497dcce6163c458021e03076ceb8f09e2d
| 7,485 |
ipynb
|
Jupyter Notebook
|
6. Getting Started with Numpy/1. Introduction_to_Numpy.ipynb
|
AshishJangra27/Data-Science-Live-Course-GeeksForGeeks
|
4fefa9c855dd515a974ee4c0d9a41886e3c0c1f8
|
[
"Apache-2.0"
] | 2 |
2021-09-11T15:44:23.000Z
|
2021-09-25T17:49:26.000Z
|
6. Getting Started with Numpy/1. Introduction_to_Numpy.ipynb
|
AshishJangra27/Data-Science-Live-Course-GeeksForGeeks
|
4fefa9c855dd515a974ee4c0d9a41886e3c0c1f8
|
[
"Apache-2.0"
] | null | null | null |
6. Getting Started with Numpy/1. Introduction_to_Numpy.ipynb
|
AshishJangra27/Data-Science-Live-Course-GeeksForGeeks
|
4fefa9c855dd515a974ee4c0d9a41886e3c0c1f8
|
[
"Apache-2.0"
] | null | null | null | 21.695652 | 103 | 0.368337 |
[
[
[
"!pip install numpy",
"Requirement already satisfied: numpy in /usr/local/lib/python3.7/dist-packages (1.19.5)\n"
],
[
"import numpy as np",
"_____no_output_____"
],
[
"lst = [[1,2,3],\n [1,2,3],\n [1,2,3]]\n ",
"_____no_output_____"
],
[
"type(lst)",
"_____no_output_____"
],
[
"arr = np.array(lst)",
"_____no_output_____"
],
[
"arr",
"_____no_output_____"
],
[
"type(arr)",
"_____no_output_____"
],
[
"arr.ndim",
"_____no_output_____"
],
[
"arr.shape",
"_____no_output_____"
],
[
"arr.size",
"_____no_output_____"
],
[
"arr.dtype",
"_____no_output_____"
],
[
"zeros = np.zeros((5,3))",
"_____no_output_____"
],
[
"zeros",
"_____no_output_____"
],
[
"ones = np.ones((6,3))",
"_____no_output_____"
],
[
"ones",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb90e0071e7978baaa7351342fb2b593bf75bbb8
| 87,507 |
ipynb
|
Jupyter Notebook
|
nb/fred-infl-unem-fed.ipynb
|
maidenlane/five
|
bf14dd37b0f14d6998893c2b0478275a0fc55a82
|
[
"BSD-3-Clause"
] | 1 |
2020-04-24T05:29:26.000Z
|
2020-04-24T05:29:26.000Z
|
fecon235-master/nb/fred-infl-unem-fed.ipynb
|
maidenlane/five
|
bf14dd37b0f14d6998893c2b0478275a0fc55a82
|
[
"BSD-3-Clause"
] | null | null | null |
fecon235-master/nb/fred-infl-unem-fed.ipynb
|
maidenlane/five
|
bf14dd37b0f14d6998893c2b0478275a0fc55a82
|
[
"BSD-3-Clause"
] | 1 |
2020-04-24T05:34:06.000Z
|
2020-04-24T05:34:06.000Z
| 108.434944 | 35,234 | 0.846949 |
[
[
[
"# Score for the Fed's dual mandate\n\nThe U.S. Congress established three key objectives for monetary policy \nin the Federal Reserve Act: *Maximum employment, stable prices*, and \nmoderate long-term interest rates. The first two objectives are \nsometimes referred to as the Federal Reserve's **dual mandate**. \n\nHere we examine unemployment and inflation data to construct \na time-series which gives a numerical score to the \nFed's performance on the dual mandate. \n(This notebook could be extended to studies of \nthe **Phillips curve**, see Appendix 1).\n\nThe key is to find comparable units to measure performance \nand a suitable scalar measure to show deviation \nfrom the dual mandate. \nOur visualization features *time-sequential* scatter plots using color *heat* map.\n\nShort URL: https://git.io/phillips",
"_____no_output_____"
],
[
"*Dependencies:*\n\n- fecon235 repository https://github.com/rsvp/fecon235\n- Python: matplotlib, pandas\n \n*CHANGE LOG*\n\n 2016-11-14 Fix #2 by v5 and PREAMBLE-p6.16.0428 upgrades. \n Switch from fecon to fecon235 for main import module. \n Minor edits given additional year of data.\n 2015-12-15 Switch to yi_0sys dependencies. Phillips curve.\n 2015-11-18 First version.",
"_____no_output_____"
]
],
[
[
"from fecon235.fecon235 import *",
"_____no_output_____"
],
[
"# PREAMBLE-p6.16.0428 :: Settings and system details\nfrom __future__ import absolute_import, print_function\nsystem.specs()\npwd = system.getpwd() # present working directory as variable.\nprint(\" :: $pwd:\", pwd)\n# If a module is modified, automatically reload it:\n%load_ext autoreload\n%autoreload 2\n# Use 0 to disable this feature.\n\n# Notebook DISPLAY options:\n# Represent pandas DataFrames as text; not HTML representation:\nimport pandas as pd\npd.set_option( 'display.notebook_repr_html', False )\nfrom IPython.display import HTML # useful for snippets\n# e.g. HTML('<iframe src=http://en.mobile.wikipedia.org/?useformat=mobile width=700 height=350></iframe>')\nfrom IPython.display import Image \n# e.g. Image(filename='holt-winters-equations.png', embed=True) # url= also works\nfrom IPython.display import YouTubeVideo\n# e.g. YouTubeVideo('1j_HxD4iLn8', start='43', width=600, height=400)\nfrom IPython.core import page\nget_ipython().set_hook('show_in_pager', page.as_hook(page.display_page), 0)\n# Or equivalently in config file: \"InteractiveShell.display_page = True\", \n# which will display results in secondary notebook pager frame in a cell.\n\n# Generate PLOTS inside notebook, \"inline\" generates static png:\n%matplotlib inline \n# \"notebook\" argument allows interactive zoom and resize.",
" :: Python 2.7.11\n :: IPython 4.2.0\n :: jupyter_core 4.1.0\n :: notebook 4.1.0\n :: matplotlib 1.5.1\n :: numpy 1.10.4\n :: pandas 0.18.0\n :: pandas_datareader 0.2.1\n :: Repository: fecon235 v5.16.1107 develop\n :: Timestamp: 2016-11-15, 20:53:09 UTC\n :: $pwd: /media/yaya/virt15h/virt/dbx/Dropbox/ipy/fecon235/nb\n"
]
],
[
[
"## Comparable unit for comparison\n\nA 1% change in inflation may have different economic significance \nthan a 1% change in unemployment. \nWe retrieve historical data, then let one standard deviation \nrepresent one unit for scoring purposes. \nNote that the past scores will thus be represented *ex-post*, \nwhich is to say, influenced by recent data.\n\nThe dual mandate is not required to be explicitly \nstated numerically by the Federal Reserve, \nso we assign explicit target values \nfor unemployment and inflation \n(those frequently mentioned in Congressional testimonies \nand used as benchmarks by the market). \nThese targets can be finely reset below \nso that the Fed performance can be re-evaluated.\n\nSidenote: \"The natural rate of unemployment \n[**NAIRU**, *non-accelerating inflation rate of unemployment*] \nis the rate of unemployment \narising from all sources except fluctuations in aggregate demand. \nEstimates of potential GDP are based on the long-term natural rate. \nThe short-term natural rate is used to gauge the amount of current \nand projected slack in labor markets.\" \nSee https://research.stlouisfed.org/fred2/series/NROU \nand Appendix 1 for details.",
"_____no_output_____"
]
],
[
[
"# Set DUAL MANDATE, assumed throughout this notebook:\nunem_target = 5.0\ninfl_target = 2.0\n\n# The Fed varies the targets over time, \n# sometimes only implicitly. So for example,\n# there may be disagreement among Board members\n# regarding NAIRU -- but we set it to what\n# seems to be the prevalent market assumption.",
"_____no_output_____"
]
],
[
[
"## Unemployment rate",
"_____no_output_____"
]
],
[
[
"unem = get( m4unemp )\n# m4 implies monthly frequency.",
"_____no_output_____"
],
[
"# # Starts 1948-01-01, uncomment to view:\n# stats( unem )",
"_____no_output_____"
],
[
"# Standard deviation for unemployment rate:\nunem_std = unem.std()\nunem_std",
"_____no_output_____"
],
[
"# Uncomment to plot raw unemployment rate:\n# plot( unem )",
"_____no_output_____"
],
[
"# Score unemployment as standard deviations from target: \nunem_score = todf( (unem - unem_target) / unem_std )",
"_____no_output_____"
]
],
[
[
"## Inflation",
"_____no_output_____"
]
],
[
[
"# Use synthetic inflation\n# which averages CPI and PCE for both headline and core versions:\ninfl_level = get( m4infl )",
"_____no_output_____"
],
[
"# Get the YoY inflation rate:\ninfl = todf(pcent( infl_level, 12 ))",
"_____no_output_____"
],
[
"# # Starts 1960-01-01, uncomment to view:\n# stats(infl)",
"_____no_output_____"
],
[
"infl_std = infl.std()\ninfl_std",
"_____no_output_____"
],
[
"# # Uncomment to plot inflation rate:\n# plot( infl )",
"_____no_output_____"
],
[
"# Score inflation as standard deviations from target:\ninfl_score = todf( (infl - infl_target) / infl_std )",
"_____no_output_____"
]
],
[
[
"## Expressing duality as complex number\n\nWe encode each joint score for unemployment and inflation into a single complex number. \nLet *u* be the unemployment score and *i* the inflation score. \n(Note: we follow the Python/engineering convention by letting **j** \nbe the imaginary number $\\sqrt -1$.) \nSo let *z* be our dual encoding as follows:\n\n$ z = u + i \\mathbf{j} $\n\n(In the history of mathematics, this was the precursor \nto the idea of a *vector*.)",
"_____no_output_____"
]
],
[
[
"# Let's start constructing our 4-column dataframe:\nscores = paste( [unem_score, infl_score, infl_score, infl_score ] )\n# Third and fouth columns are dummy placeholders to be replaced later.",
"_____no_output_____"
],
[
"# Give names to the scores columns:\nscores.columns = ['z_unem', 'z_infl', 'z', 'z_norm']",
"_____no_output_____"
],
[
"# Fill in THIRD column z as complex number per our discussion:\nscores.z = scores.z_unem + (scores.z_infl * 1j)\n# The imaginary number in itself in Python is represented as 1j, \n# since j may be a variable elsewhere.",
"_____no_output_____"
]
],
[
[
"## Computing the Fed score\n\nEach dual score can be interpreted as a vector in the \ncomplex plane. Its component parts, real for unemployment \nand imaginary for inflation, measure deviation from \nrespective targets in units expressed as standard deviations.\n\n*Our **key idea** is to use the length of this vector (from \nthe origin, 0+0**j**, representing the dual mandate) as \nthe Fed score* = |z|.\n\nPython, which natively handles complex numbers, can \ncompute the *norm* of such a vector using abs(z).\n\nLater we shall visualize the trajectory of the component \nparts using a color heat map.",
"_____no_output_____"
]
],
[
[
"# Finally fill-in the FOURTH placeholder column:\nscores.z_norm = abs( scores.z ) ",
"_____no_output_____"
],
[
"# Tail end of recent scores:\ntail( scores )\n# ... nicely revealing the data structure:",
"_____no_output_____"
]
],
[
[
"## Visualizing Fed scores\n\nz_norm is expressed in standard deviation units, thus \nit truly represents deviation from the dual mandate \non a Gaussian scale.",
"_____no_output_____"
]
],
[
[
"# Define descriptive dataframe from our mathematical construct:\nfed_score = todf( scores.z_norm )",
"_____no_output_____"
],
[
"# FED SCORE\nplot( fed_score )",
"_____no_output_____"
]
],
[
[
"*We can say that a score greater than 2 is definitively cause for concern. \nFor example, between 1974 and 1984, there are two peaks \nextending into 4 standard deviations mainly due to high inflation. \nFed score during the Great Recession hit 3 mainly due to \nsevere unemployment.*",
"_____no_output_____"
]
],
[
[
"stats( fed_score )",
" Y\ncount 681.000000\nmean 1.306697\nstd 1.026325\nmin 0.038818\n25% 0.469673\n50% 0.989438\n75% 1.793291\nmax 4.224767\n\n :: Index on min:\nY 2001-09-01\ndtype: datetime64[ns]\n\n :: Index on max:\nY 1980-06-01\ndtype: datetime64[ns]\n\n :: Head:\n Y\nT \n1960-01-01 0.158333\n1960-02-01 0.128790\n1960-03-01 0.254965\n :: Tail:\n Y\nT \n2016-07-01 0.261664\n2016-08-01 0.201106\n2016-09-01 0.138719\n\n :: Correlation matrix:\n Y\nY 1.0\n"
]
],
[
[
"## Remarks on fed_score\n\nOur fed_score *handles both positive and negative deviations from the Fed's dual mandate*, \nmoreover, it handles them jointly using a historical fair measuring unit: \nthe standard deviation. This avoids using ad-hoc weights to balance \nthe importance between unemployment and inflation.\n\nThe **fed_score is always a positve real number (since it is a norm) \nwhich is zero if and only if the Fed has achieved targets** \n(which we have explicitly specified). \n*That score can be interpreted as the number of standard deviations \naway from the dual mandate.*\n\nOur **fed_score** can also be simply interpreted \nas an economic **crisis level** indicator \n(much like a *n-alarm* fire) when Fed monetary \npolicy becomes crucial for the US economy.\n\nSince 1960, ex-post fed_score averages around 1.31 where \nthe mid-50% percentile range is approximately [0.47, 1.79].\nBut keep in mind that this computation relies on our \ncurrent fixed targets, so fed_score is most \nuseful in accessing recent performance \nfrom the perspective of historical variance. \nThis means that the fed_score for a particular \npoint in time may change as new incoming data arrives.\n\n2015-12-15 notes: The current fed_score is 0.46 \ngravitating towards zero as the Fed is preparing \nits first rate hike in a decade.\n\n2016-11-14 notes: The current fed_score is 0.14 \nas the Fed is expected to announce \nthe second rate hike since the Great Recession. \nThe labor market has exhibited sustained improvement.",
"_____no_output_____"
],
[
"## CONCLUSION: visualize fed_score components over time\n\nWhen FOMC makes its releases, the market becomes \nfixated on the so-called blue dot chart \nof expected future interest rates by its members. \nThe next scatter plot helps us to understand \nthe motivations and constraints behind \nthe Federal Reserve's monetary policy.\n\nWe see that the key components of the U.S. economy, \nunemployment and inflation, came back \n\"full circle\" to target from 2005 to 2016, \nor more accurately \"full figure eight\" \nwith *major* (3 standard) deviations seen in the unemployment component.",
"_____no_output_____"
]
],
[
[
"# Scatter plot of recent data using color heat map:\nscatter( scores['2005':], col=[0, 1] )\n# (Ignore FutureWarning from matplotlib/collections.py:590\n# regarding elementwise comparison due to upstream numpy.)",
"_____no_output_____"
]
],
[
[
"In this scatter plot: z_unem is shown along the x-axis, \nz_infl along the y-axis, \nsuch that the color *heat* map depicts chronological movement \nfrom *blue to green to red*. \n\nThe coordinate (0, 0) represents our Fed dual mandate, \nso it is easy to see deviations from target. \nGeometrically a point's distance from the origin is what \nwe have computed as fed_score (= z_norm, in the complex plane).\n\n- - - -",
"_____no_output_____"
],
[
"## Appendix 1: Phillips curve\n\nThe Phillips curve purports to explain the relationship between \ninflation and unemployment, however, as it turns out the \nrelationship is one of mere correlation during certain periods \nin a given country. It is too simplistic to assert that \ndecreased unemployment (i.e. increased levels of employment) \nwill cause with the inflation rate to increase.\n\n\"Most economists no longer use the Phillips curve in its original \nform because it was shown to be too simplistic. This can be seen \nin a cursory analysis of US inflation and unemployment data from \n1953-92. There is no single curve that will fit the data, but \nthere are three rough aggregations—1955–71, 1974–84, and 1985–92 — \neach of which shows a general, downwards slope, but at three \nvery different levels with the shifts occurring abruptly. \nThe data for 1953-54 and 1972-73 do not group easily, \nand a more formal analysis posits up to five groups/curves \nover the period.\n\nModern versions distinguish between short-run and long-run \neffects on unemployment. The \"short-run Phillips curve\" is also called \nthe \"expectations-augmented Phillips curve,\" since it shifts up when \ninflationary expectations rise. In the long run, this implies \nthat monetary policy cannot affect unemployment, \nwhich adjusts back to its \"natural rate\", also called the \"NAIRU\" or \n\"long-run Phillips curve\". However, this long-run \"neutrality\" of \nmonetary policy does allow for short run fluctuations and the ability \nof the monetary authority to temporarily decrease unemployment \nby increasing permanent inflation, and vice versa.\n\nIn many recent stochastic general equilibrium models, with sticky prices, \nthere is a positive relation between the rate of inflation and \nthe level of demand, and therefore a negative relation between \nthe rate of inflation and the rate of unemployment. This relationship is \noften called the \"New Keynesian Phillips curve.\" Like the \nexpectations-augmented Phillips curve, the New Keynesian Phillips curve \nimplies that increased inflation can lower unemployment temporarily, \nbut cannot lower it permanently.\"\n\nFor more details, see https://en.wikipedia.org/wiki/Phillips_curve\n\nCuriously, according to Edmund Phelps who won the 2006 Nobel Prize in Economics,\nthe **long-run Phillips Curve is *vertical* **\nsuch that the rate of inflation has no effect on unemployment at its NAIRU. \nThe name \"NAIRU\" arises because with actual unemployment below it, \ninflation accelerates, while with unemployment above it, \ninflation decelerates. With the actual rate equal to it, \ninflation is stable, neither accelerating nor decelerating.",
"_____no_output_____"
]
],
[
[
"Image(url=\"https://upload.wikimedia.org/wikipedia/commons/e/e3/NAIRU-SR-and-LR.svg\", embed=False)",
"_____no_output_____"
]
],
[
[
"We know from our data that economic reality is far more \ncomplicated than the ideal diagram above.\n\nFor example, in the late 1990s, the unemployment rate \nfell below 4%, much lower than almost all estimates of the NAIRU. \nBut inflation stayed very moderate rather than accelerating.\n\nSince our z data captures the normalized version of \nunemployment vs inflation rates, it is easy \nto visualize the actual paths using our scatter plots.",
"_____no_output_____"
]
],
[
[
"# Uncomment for # Scatter plot of selected data using color heat map:\n# scatter( scores['1995':'2000'], col=[0, 1] )",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb9103614ee990e55ebf65f46768593b075c129f
| 9,483 |
ipynb
|
Jupyter Notebook
|
tests/README.ipynb
|
nbabulkov/textworld-challenge
|
feb84d39c42207b285c2e5ed88a46e85c7c65eaa
|
[
"MIT"
] | null | null | null |
tests/README.ipynb
|
nbabulkov/textworld-challenge
|
feb84d39c42207b285c2e5ed88a46e85c7c65eaa
|
[
"MIT"
] | null | null | null |
tests/README.ipynb
|
nbabulkov/textworld-challenge
|
feb84d39c42207b285c2e5ed88a46e85c7c65eaa
|
[
"MIT"
] | null | null | null | 34.234657 | 256 | 0.47833 |
[
[
[
"from glob import glob\nfrom os.path import join as pjoin\n\nimport gym\nimport textworld.gym\nfrom textworld import EnvInfos\n\nGAMES_PATH = \"sample_games\" # This assumes `sample_games.zip` was first unzipped.\ngamefiles = glob(pjoin(GAMES_PATH, \"*.ulx\"))\nprint(\"Found {} games.\".format(len(gamefiles)))",
"Found 10 games.\n"
]
],
[
[
"## Playing a game ",
"_____no_output_____"
]
],
[
[
"gamefile = gamefiles[0] # Pick a game.\n\nrequested_infos = EnvInfos(description=True, inventory=True, extras=[\"recipe\", \"walkthrough\"])\nenv_id = textworld.gym.register_games([gamefile], requested_infos)\n\nagent = textworld.agents.HumanAgent()\n\nenv = gym.make(env_id)\nobs, infos = env.reset()\n\n# Since we asked for more infos, let's print them.\nprint(\"Walkthrough:\", \". \".join(infos[\"extra.walkthrough\"]))\nprint(infos[\"extra.recipe\"])\n\nenv.render() # Print the initial observation.\n\nscore = 0\ndone = False\nwhile not done:\n command = input('> ')\n ob, score, done, infos = env.step(command)\n env.render()\n ",
"/home/nbabulkov/Code/TextWorld/venv/lib/python3.7/site-packages/gym/envs/registration.py:14: PkgResourcesDeprecationWarning: Parameters to load are deprecated. Call .resolve and .require separately.\n result = entry_point.load(False)\n"
]
],
[
[
"## Visualizing a `TextWorld.Game` object",
"_____no_output_____"
]
],
[
[
"import textworld\ngamefile = gamefiles[0] # Pick a game.\ngame = textworld.Game.load(gamefile.replace(\".ulx\", \".json\"))\ntextworld.render.visualize(game)",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb910c62e8ffd9130f4687d59b4ae50a22339173
| 313,598 |
ipynb
|
Jupyter Notebook
|
Assignment1/ECE285-Assignment-1.ipynb
|
yansun1996/ECE285-MachineLearning-ImageProcessing
|
efca42b55f8d2490e19f26f74dc01774f051cf43
|
[
"MIT"
] | null | null | null |
Assignment1/ECE285-Assignment-1.ipynb
|
yansun1996/ECE285-MachineLearning-ImageProcessing
|
efca42b55f8d2490e19f26f74dc01774f051cf43
|
[
"MIT"
] | null | null | null |
Assignment1/ECE285-Assignment-1.ipynb
|
yansun1996/ECE285-MachineLearning-ImageProcessing
|
efca42b55f8d2490e19f26f74dc01774f051cf43
|
[
"MIT"
] | 7 |
2019-04-25T19:17:29.000Z
|
2020-03-31T01:01:46.000Z
| 225.447879 | 180,296 | 0.921913 |
[
[
[
"# 1 Getting started – Python, Platform and Jupyter",
"_____no_output_____"
]
],
[
[
"print(\"HelloWorld!\")",
"HelloWorld!\n"
]
],
[
[
"# 2 Numpy",
"_____no_output_____"
],
[
"### 2.1 Arrays",
"_____no_output_____"
],
[
"##### 1. Run the following:",
"_____no_output_____"
]
],
[
[
"import numpy as np\na = np.array([1, 2, 3])\nprint(type(a))\nprint(a.shape)\nprint(a[0], a[1], a[2])\na[0] = 5",
"<class 'numpy.ndarray'>\n(3,)\n1 2 3\n"
],
[
"print(a[0], a[1], a[2])",
"5 2 3\n"
]
],
[
[
"What are the rank, shape of a, and the current values of a[0], a[1], a[2]?",
"_____no_output_____"
],
[
"Answer:\n 1. The rank of a is 1.\n 2. The shape of a is (3,).\n 3. a[0], a[1], a[2] = 5, 2, 3",
"_____no_output_____"
],
[
"##### 2. Run the following:",
"_____no_output_____"
]
],
[
[
"b = np.array([[1, 2, 3], [4, 5, 6]])",
"_____no_output_____"
],
[
"print(b[0, 0], b[0, 1], b[1, 0])",
"1 2 4\n"
]
],
[
[
"What are the rank, shape of b, and the current values of b[0, 0], b[0, 1], b[1, 0]?\n",
"_____no_output_____"
],
[
"Answer:\n 1. The rank of a is 2.\n 2. The shape of a is (2,3).\n 3. b[0, 0], b[0, 1], b[1, 0] = 1, 2, 4",
"_____no_output_____"
],
[
"##### 3. Numpy also provides many functions to create arrays. Assign the correct comment to each of the following instructions:",
"_____no_output_____"
]
],
[
[
"a = np.zeros((2,2)) # an array of all zeros\nb = np.ones((1,2)) # an array of all ones\nc = np.full((2,2), 7) # a constant array\nd = np.eye(2) # a 2x2 identity matrix\ne = np.random.random((2,2)) # an array filled with random values",
"_____no_output_____"
]
],
[
[
"### 2.2 Array Indexing",
"_____no_output_____"
],
[
"##### 4. Run the following",
"_____no_output_____"
]
],
[
[
"a = np.array([[1,2,3,4], [5,6,7,8], [9,10,11,12]])\nb = a[:2, 1:3]",
"_____no_output_____"
]
],
[
[
"The shape of a and b are:",
"_____no_output_____"
]
],
[
[
"print(a.shape)\nprint(b.shape)",
"(3, 4)\n(2, 2)\n"
]
],
[
[
"The values in b are:",
"_____no_output_____"
]
],
[
[
"b",
"_____no_output_____"
]
],
[
[
"##### 5. A slice of an array is a view into the same data. Follow the comments in the following snippet",
"_____no_output_____"
]
],
[
[
"print(a[0, 1]) # Prints \"2\"\nb[0, 0] = 77\nprint(a[0, 1])",
"2\n77\n"
]
],
[
[
"Last line prints 77, modifying b also modifies the original array.",
"_____no_output_____"
],
[
"##### 6. You can also mix integer indexing with slice indexing. However, doing so will yield an array of lower rank than the original array. Note that this is quite different from the way that MATLAB handles array slicing. Create the following rank 2 array with shape (3, 4):\n",
"_____no_output_____"
]
],
[
[
"a = np.array([[1,2,3,4], [5,6,7,8], [9,10,11,12]])",
"_____no_output_____"
],
[
"row_r1 = a[1, :] # Rank 1 view of the second row of a\nrow_r2 = a[1:2, :] # Rank 2 view of the second row of a\nprint(row_r1, row_r1.shape) # Prints \"[5 6 7 8] (4,)\"\nprint(row_r2, row_r2.shape) # Prints \"[[5 6 7 8]] (1, 4)\"\ncol_r1 = a[:, 1]\ncol_r2 = a[:, 1:2]",
"[5 6 7 8] (4,)\n[[5 6 7 8]] (1, 4)\n"
]
],
[
[
"What are the values and shapes of col r1 and col r2?",
"_____no_output_____"
]
],
[
[
"print(col_r1, col_r1.shape)\nprint(col_r2, col_r2.shape)",
"[ 2 6 10] (3,)\n[[ 2]\n [ 6]\n [10]] (3, 1)\n"
]
],
[
[
"##### 7. Run the following:",
"_____no_output_____"
]
],
[
[
"a = np.array([[1,2], [3, 4], [5, 6]])\nprint(a[[0, 1, 2], [0, 1, 0]])\nprint(np.array([a[0, 0], a[1, 1], a[2, 0]]))",
"[1 4 5]\n[1 4 5]\n"
]
],
[
[
"They are equivalent.",
"_____no_output_____"
],
[
"##### 8. When using integer array indexing, you can duplicate several times the same element from the source array. Compare the following instructions:",
"_____no_output_____"
]
],
[
[
"b = a[[0, 0], [1, 1]]\nc = np.array([a[0, 1], a[0, 1]])",
"_____no_output_____"
],
[
"print(b)\nprint(c)",
"[2 2]\n[2 2]\n"
]
],
[
[
"They are equivalent.",
"_____no_output_____"
],
[
"##### 9. One useful trick with integer array indexing is selecting or mutating one element from each row of a matrix. Run the following code",
"_____no_output_____"
]
],
[
[
"a = np.array([[1,2,3], [4,5,6], [7,8,9], [10, 11, 12]])\nb = np.array([0, 2, 0, 1])\nprint(a[np.arange(4), b])\na[np.arange(4), b] += 10\nprint(a)",
"[ 1 6 7 11]\n[[11 2 3]\n [ 4 5 16]\n [17 8 9]\n [10 21 12]]\n"
]
],
[
[
"Observation:\n\nnp.arange(4) = [0, 1, 2, 3] and b = [0, 2, 0, 1]\nSo a[np.arange(4), b]) will be [a[0,0], a[1,2], a[2,0], a[3,1]].\nThat is [1, 6, 7, 11]\n\nModifying this diretly via a[np.arange(4), b] += 10 will also modify the original values in a",
"_____no_output_____"
],
[
"##### 10. Run the following",
"_____no_output_____"
]
],
[
[
"a = np.array([[1,2], [3, 4], [5, 6]])\nbool_idx = (a > 2)",
"_____no_output_____"
]
],
[
[
"What is the result stored in bool idx?\nRun the following:\n\nWhat are the values printed? What is their shape?",
"_____no_output_____"
]
],
[
[
"print(a[bool_idx])\nprint(a[a > 2])\nprint(a[bool_idx].shape)\nprint(a[a > 2].shape)",
"[3 4 5 6]\n[3 4 5 6]\n(4,)\n(4,)\n"
]
],
[
[
"Their shapes are the same, (4,)",
"_____no_output_____"
],
[
"### 2.3 Datatypes",
"_____no_output_____"
],
[
"##### 11. Here is an example\n\nWhat are the datatypes of x, y, z?",
"_____no_output_____"
]
],
[
[
"x = np.array([1, 2]) # Let numpy choose the datatype\nprint(x.dtype)\ny = np.array([1.0, 2.0]) # Let numpy choose the datatype\nprint(y.dtype)\nz = np.array([1, 2], dtype=np.int32) # Force a particular datatype\nprint(z.dtype)",
"int64\nfloat64\nint32\n"
]
],
[
[
"### 2.4 Array math",
"_____no_output_____"
],
[
"##### 12. Give the output for each print statement in the following code snippet",
"_____no_output_____"
]
],
[
[
"x = np.array([[1,2],[3,4]], dtype=np.float64)\ny = np.array([[5,6],[7,8]], dtype=np.float64)\n# Elementwise sum; both produce the array\nprint(x + y)\nprint(np.add(x, y))\n# Elementwise difference; both produce the array\nprint(x - y)\nprint(np.subtract(x, y))\n# Elementwise product; both produce the array\nprint(x * y)\nprint(np.multiply(x, y))\n# Elementwise division; both produce the array\nprint(x / y)\nprint(np.divide(x, y))\n# Elementwise square root; produces the array\nprint(np.sqrt(x))",
"[[ 6. 8.]\n [10. 12.]]\n[[ 6. 8.]\n [10. 12.]]\n[[-4. -4.]\n [-4. -4.]]\n[[-4. -4.]\n [-4. -4.]]\n[[ 5. 12.]\n [21. 32.]]\n[[ 5. 12.]\n [21. 32.]]\n[[0.2 0.33333333]\n [0.42857143 0.5 ]]\n[[0.2 0.33333333]\n [0.42857143 0.5 ]]\n[[1. 1.41421356]\n [1.73205081 2. ]]\n"
]
],
[
[
"##### 13. What are the mathematical operations performed by the last two instructions?",
"_____no_output_____"
]
],
[
[
"v = np.array([9,10])\nw = np.array([11, 12])\nprint(v.dot(w))\nprint(np.dot(v, w))",
"219\n219\n"
]
],
[
[
"It performs the inner product of vectors. 9\\*11 + 10\\*12 = 219",
"_____no_output_____"
],
[
"##### 14. What are the mathematical operations performed in the snippet below?",
"_____no_output_____"
]
],
[
[
"x = np.array([[1,2],[3,4]])\nprint(x.dot(v))\nprint(np.dot(x, v))",
"[29 67]\n[29 67]\n"
]
],
[
[
"It performs the matrix product.",
"_____no_output_____"
],
[
"##### 15. Write a code to compute the product of the matrices x with the following matrix y\n",
"_____no_output_____"
]
],
[
[
"y = np.array([[5,6,7],[7,8,9]])\nprint(np.dot(x,y))",
"[[19 22 25]\n [43 50 57]]\n"
],
[
"print(x.shape)\nprint(y.shape)",
"(2, 2)\n(2, 3)\n"
],
[
"print(np.dot(x,y))",
"[[19 22 25]\n [43 50 57]]\n"
],
[
"print(np.dot(y.T,x))",
"[[26 38]\n [30 44]\n [34 50]]\n"
],
[
"x = np.array([[1,2],[3,4]])\nprint(np.sum(x)) # Compute sum of all elements; prints \"10\"\nprint(np.sum(x, axis=0)) # Compute sum of each column; prints \"[4 6]\"\nprint(np.sum(x, axis=1)) # Compute sum of each row; prints \"[3 7]\"",
"10\n[4 6]\n[3 7]\n"
]
],
[
[
"##### 16. In the following",
"_____no_output_____"
]
],
[
[
"x = np.array([[1,2,3], [3,4,5]])\nprint(x)\nprint(x.T)",
"[[1 2 3]\n [3 4 5]]\n[[1 3]\n [2 4]\n [3 5]]\n"
]
],
[
[
"What is the transpose of x? How is it different from x?",
"_____no_output_____"
],
[
"Exchange the element based on the symmetry for diagonal.",
"_____no_output_____"
],
[
"### 2.5 Broadcasting",
"_____no_output_____"
],
[
"##### 17. Complete the following code in order to add the vector v to each row of a matrix x:",
"_____no_output_____"
]
],
[
[
"x = np.array([[1,2,3], [4,5,6], [7,8,9], [10, 11, 12]])\nv = np.array([1,0,1])\ny = np.zeros(x.shape)\nfor i in range(x.shape[0]):\n for j in range(x.shape[1]):\n y[i,j] = x[i, j] + v[j]\nprint(y)",
"[[ 2. 2. 4.]\n [ 5. 5. 7.]\n [ 8. 8. 10.]\n [11. 11. 13.]]\n"
]
],
[
[
"##### 18. The previous solution works well; however when the matrix x is very large, computing an explicit loop in Python could be slow. An alternative is to use tile. Run the following and interpret the result.",
"_____no_output_____"
]
],
[
[
"vv = np.tile(v, (4, 1))\nprint(vv)",
"[[1 0 1]\n [1 0 1]\n [1 0 1]\n [1 0 1]]\n"
]
],
[
[
"Next, complete the code to compute y from x and vv without explicit loops.",
"_____no_output_____"
]
],
[
[
"x = np.array([[1,2,3], [4,5,6], [7,8,9], [10, 11, 12]])\nv = np.array([1,0,1])\nvv = np.tile(v, (4, 1))\ny = x + vv\nprint(y)",
"[[ 2 2 4]\n [ 5 5 7]\n [ 8 8 10]\n [11 11 13]]\n"
]
],
[
[
"##### 19. Numpy broadcasting allows us to perform this computation without actually creating multiple copies of v. This is simply obtained as follows\n",
"_____no_output_____"
]
],
[
[
"x = np.array([[1,2,3], [4,5,6], [7,8,9], [10, 11, 12]])\nv = np.array([1, 0, 1])\ny = x + v # Add v to each row of x using broadcasting\nprint(y)",
"[[ 2 2 4]\n [ 5 5 7]\n [ 8 8 10]\n [11 11 13]]\n"
]
],
[
[
"Compute the outer product of v and w using broadcasting",
"_____no_output_____"
]
],
[
[
"v = np.array([1,2,3]) # v has shape (3,)\nw = np.array([4,5]) # w has shape (2,)",
"_____no_output_____"
],
[
"print(np.reshape(v, (3, 1)) * w)",
"[[ 4 5]\n [ 8 10]\n [12 15]]\n"
]
],
[
[
"##### 20. Add v to each row of x using broadcasting",
"_____no_output_____"
]
],
[
[
"x = np.array([[1,2,3], [4,5,6]])",
"_____no_output_____"
],
[
"print(x + v)",
"[[2 4 6]\n [5 7 9]]\n"
]
],
[
[
"##### 21. Add w to each column of x using broadcasting.",
"_____no_output_____"
]
],
[
[
"print(x + np.reshape(w, (2, 1)))",
"[[ 5 6 7]\n [ 9 10 11]]\n"
]
],
[
[
"##### 22. Numpy treats scalars as arrays of shape (). Write a code to multiply x with 2.",
"_____no_output_____"
]
],
[
[
"print(x * 2)",
"[[ 2 4 6]\n [ 8 10 12]]\n"
]
],
[
[
"### 2.6 Numpy Documentation",
"_____no_output_____"
],
[
"# 3 Matplotlib",
"_____no_output_____"
],
[
"### 3.1 Plotting",
"_____no_output_____"
],
[
"##### An important function in matplotlib is plot, which allows you to plot 2D graphs. Here is a simple example:",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport matplotlib.pyplot as plt\n# Compute the x and y coordinates for points on a sine curve\nx = np.arange(0, 3 * np.pi, 0.1)\ny = np.sin(x)\n# Plot the points using matplotlib\nplt.plot(x, y)\nplt.show() # You must call plt.show() to make graphics appear.",
"_____no_output_____"
]
],
[
[
"##### 24. Modify the above example by adding these extra instructions before plt.show()",
"_____no_output_____"
]
],
[
[
"# Compute the x and y coordinates for points on a sine curve\nx = np.arange(0, 3 * np.pi, 0.1)\ny = np.sin(x)\n# Plot the points using matplotlib\nplt.plot(x, y)\n\nplt.xlim(0, 3 * np.pi)\nplt.xlabel('x')\nplt.ylabel('y')\nplt.title('Sinusoidal curve(s)')\nplt.legend(['Sine'])\n\nplt.show() # You must call plt.show() to make graphics appear.",
"_____no_output_____"
]
],
[
[
"##### 25. Write a code to plot both sine and cosine curves in the same plot with proper legend.",
"_____no_output_____"
]
],
[
[
"# Compute the x and y coordinates for points on a sine curve\nx = np.arange(0, 3 * np.pi, 0.1)\ny1 = np.sin(x)\ny2 = np.cos(x)\n# Plot the points using matplotlib\nplt.plot(x, y1)\nplt.plot(x, y2)\n\nplt.xlim(0, 3 * np.pi)\nplt.xlabel('x')\nplt.ylabel('y')\nplt.title('Sin(x) and Cos(x) curve(s)')\nplt.legend(['Sine', 'Cosine'])\n\nplt.show() # You must call plt.show() to make graphics appear.",
"_____no_output_____"
]
],
[
[
"### 3.2 Subplots",
"_____no_output_____"
]
],
[
[
"x = np.arange(0, 3 * np.pi, 0.1)\ny_sin = np.sin(x)\n# Set up a subplot grid that has height 2 and width 1,\n# and set the first such subplot as active.\nplt.subplot(2, 1, 1)\n# Make the first plot\nplt.plot(x, y_sin)\nplt.xlim(0, 3 * np.pi)\nplt.title('Sine')\n\nplt.show()",
"_____no_output_____"
]
],
[
[
"##### 26. Complete the above code to create a second subplot in the same grid representing the cosine function.",
"_____no_output_____"
]
],
[
[
"x = np.arange(0, 3 * np.pi, 0.1)\ny_sin = np.sin(x)\ny_cos = np.cos(x)\n\n# Set up a subplot grid that has height 2 and width 1,\n# and set the first such subplot as active.\nplt.subplot(2, 1, 1)\n# Make the first plot\nplt.plot(x, y_sin)\nplt.xlim(0, 3 * np.pi)\nplt.title('Sine')\n\nplt.subplot(2, 1, 2)\n# Make the second plot\nplt.plot(x, y_cos)\nplt.xlim(0, 3 * np.pi)\nplt.title('Cosine')\n\nplt.show()",
"_____no_output_____"
]
],
[
[
"### 3.3 Images",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport imageio\nimport matplotlib.pyplot as plt\nimg = imageio.imread('assets/cat.jpg')\nimg_tinted = img * [1, 0.95, 0.9]\n# Show the original image\nplt.subplot(1, 2, 1)\nplt.imshow(img)\n# Show the tinted image\nplt.subplot(1, 2, 2)\n# A slight gotcha with imshow is that it might give strange results\n# if presented with data that is not uint8. To work around this, we\n# explicitly cast the image to uint8 before displaying it.\nplt.imshow(np.uint8(img_tinted))\nplt.show()",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb911b0599d043a12558240d8c1acd9d2f5794e4
| 28,985 |
ipynb
|
Jupyter Notebook
|
Modulo2/Clase7_RendimientoRiesgoPortafoliosII.ipynb
|
duarteandres/porinvo2020
|
93c1d90653382b29a5cf1e5d60b591d8400013b2
|
[
"MIT"
] | null | null | null |
Modulo2/Clase7_RendimientoRiesgoPortafoliosII.ipynb
|
duarteandres/porinvo2020
|
93c1d90653382b29a5cf1e5d60b591d8400013b2
|
[
"MIT"
] | null | null | null |
Modulo2/Clase7_RendimientoRiesgoPortafoliosII.ipynb
|
duarteandres/porinvo2020
|
93c1d90653382b29a5cf1e5d60b591d8400013b2
|
[
"MIT"
] | null | null | null | 27.34434 | 291 | 0.460618 |
[
[
[
"# ¿Cómo medir rendimiento y riesgo en un portafolio? II\n\n<img style=\"float: right; margin: 0px 0px 15px 15px;\" src=\"http://www.picpedia.org/clipboard/images/stock-portfolio.jpg\" width=\"600px\" height=\"400px\" />\n\n> La clase pasada y la presente, están dedicadas a obtener medidas de rendimiento y riesgo en un portafolio.\n\n> Vimos que podemos obtener los rendimientos de un portafolio mediante la relación $r_p=\\sum_{i=1}^{n}w_ir_i$, y una vez teniendo los rendimientos del portafolio, lo podemos tratar como un activo individual.\n\n> Por otra parte, vimos que si conocemos los rendimientos esperados de cada activo que conforma el portafolio $E[r_i]$, podemos calcular el rendimiento esperado del portafolio como el promedio ponderado de los rendimientos esperados de los activos $E[r_p]=\\sum_{i=1}^{n}w_iE[r_i]$.\n\n> Sin embargo, vimos que esto no es válido para la medida de riesgo (desviación estándar). Es decir, la varianza (o volatilidad, o desviación estándar) no es el promedio ponderado de las varianzas individuales. Anticipamos que esto es clave en el concepto de **diversificación**.",
"_____no_output_____"
],
[
"**Objetivos:**\n- Medir el riesgo en un portafolio a partir del riesgo de cada uno de los activos que lo conforman.\n\n*Referencias*\n- Notas del curso \"Portfolio Selection and Risk Management\", Rice University, disponible en Coursera.\n___",
"_____no_output_____"
],
[
"## 1. Midiendo el riesgo en un portafolio\n\n### 1.1. Volatilidad de un portafolio\n\nRetomamos el ejemplo qur veníamos trabajando la clase pasada...",
"_____no_output_____"
],
[
"**Ejemplo.** Supongamos que tenemos inversión en activos de Toyota, Walmart y Pfizer. Tenemos cuatro posibles estados económicos:",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport pandas as pd",
"_____no_output_____"
],
[
"# Creamos tabla\ntabla = pd.DataFrame(columns=['Prob', 'Toyota', 'Walmart', 'Pfizer'], \n index=['Expansion', 'Normal', 'Recesion', 'Depresion'])\ntabla.index.name = 'Estado'\ntabla['Prob']=np.array([0.1, 0.4, 0.3, 0.2])\ntabla['Toyota']=np.array([0.06, 0.075, 0.02, -0.03])\ntabla['Walmart']=np.array([0.045, 0.055, 0.04, -0.01])\ntabla['Pfizer']=np.array([0.025, -0.005, 0.01, 0.13])\n\ntabla.round(4)",
"_____no_output_____"
],
[
"## Rendimientos esperados\n# Toyota\nErT = (tabla['Prob']*tabla['Toyota']).sum()\n# Walmart\nErW = (tabla['Prob']*tabla['Walmart']).sum()\n# Pfizer\nErP = (tabla['Prob']*tabla['Pfizer']).sum()\n# Mostrar\nErT, ErW, ErP",
"_____no_output_____"
],
[
"## Volatilidad\n# Toyota\nsT = (tabla['Prob'] * (tabla['Toyota'] - ErT)**2).sum()**0.5\n# Walmart\nsW = (tabla['Prob'] * (tabla['Walmart'] - ErW)**2).sum()**0.5\n# Pfizer\nsP = (tabla['Prob'] * (tabla['Pfizer'] - ErP)**2).sum()**0.5\n# Mostrar\nsT, sW, sP",
"_____no_output_____"
],
[
"# Portafolio 0.5Toyota+0.5Pfizer\ntabla['PortTP'] = 0.5 * tabla['Toyota'] + 0.5 * tabla['Pfizer']\ntabla",
"_____no_output_____"
],
[
"# Rendimiento portafolio (como activo individual)\nErTP = (tabla['Prob']*tabla['PortTP']).sum()\n# Rendimiento portafolio (como suma ponderada de rendimientos individuales)\nErTP2 = 0.5 * ErT + 0.5 * ErP\nErTP, ErTP2",
"_____no_output_____"
],
[
"# Volatilidad del portafolio\nsTP = (tabla['Prob'] * (tabla['PortTP'] - ErTP)**2).sum()**0.5\nsTP",
"_____no_output_____"
],
[
"# Notar que sTP < 0.5 * sT + 0.5 * sP\n# la volatilidad del portafolio siempre es menor\n# a la suma ponderada de las volatilidades individuales\nsTP, 0.5 * sT + 0.5 * sP, sTP < 0.5 * sT + 0.5 * sP",
"_____no_output_____"
]
],
[
[
"**Actividad.** Encontrar la volatilidad del portafolio formado $0.5$ Toyota y $0.5$ Walmart.",
"_____no_output_____"
]
],
[
[
"# Encontrar los rendimientos del portafolio en cada estado de la economía\ntabla['ProbTW'] = 0.5 * tabla['Toyota'] + 0.5 * tabla['Walmart']\ntabla",
"_____no_output_____"
],
[
"# Encontrar el rendimiento esperado del portafolio\nErTW = (tabla['Prob'] * tabla['ProbTW']).sum()\nErTW",
"_____no_output_____"
],
[
"# Encontrar la volatilidad de Toyota, Walmart y el portafolio\nsTW = (tabla['Prob']*(tabla['ProbTW']-ErTW)**2).sum()**0.5\nsTW",
"_____no_output_____"
],
[
"# Notar que sTW < 0.5 * sT + 0.5 * sW\n# la volatilidad del portafolio siempre es menor\n# a la suma ponderada de las volatilidades individuales\nsTW, 0.5 * sT + 0.5 * sW, sTW < 0.5 * sT + 0.5 * sW",
"_____no_output_____"
]
],
[
[
"### 1.2. Midiendo el co-movimiento entre instrumentos\n\n- Una vez más, concluimos que la volatilidad (varianza) **NO** es el promedio ponderado de las varianzas individales.\n\n- Por el contrario, la varianza de los rendimientos de un portafolio está afectada por el movimiento relativo de un activo individual respecto a otro.\n\n- Por tanto, necesitamos definir las medidas de **covarianza** y **correlación**, que nos permiten evaluar las fluctuaciones relativas entre los activos.",
"_____no_output_____"
],
[
"#### Covarianza:\n\nEs una medida el movimiento relativo entre dos instrumentos.\n\nMatemáticamente, si tenemos dos activos $A_1$ y $A_2$ cuyos rendimientos son $r_1$ y $r_2$, respectivamente, entonces la covarianza de los rendimientos de los activos es\n\n$$\\text{cov}(r_1,r_2)=\\sigma_{12}=\\sum_{j=1}^{m}p_j(r_{1j}-E[r_1])(r_{2j}-E[r_2]).$$\n\n$$\\text{cov}(r_2,r_1)=\\sigma_{21}=\\sum_{j=1}^{m}p_j(r_{2j}-E[r_2])(r_{1j}-E[r_1]) = \\sigma_{12}.$$",
"_____no_output_____"
],
[
"Podemos notar fácilmente que la covarianza de los rendimientos de un activo con los rendimientos del mismo activo corresponde a la varianza\n\n$$\\text{cov}(r_1,r_1)=\\sigma_{11}=\\sum_{j=1}^{m} p_j(r_{1j}-E[r_1])(r_{1j}-E[r_1])=\\sigma_1^2=\\text{var}(r_1).$$",
"_____no_output_____"
],
[
"**Ejemplo.** Calcular la covarianza entre los rendimientos de Toyota y Pfizer.",
"_____no_output_____"
]
],
[
[
"# Mostrar tabla\ntabla",
"_____no_output_____"
],
[
"# Calcular la covarianza\ncovTP = (tabla['Prob'] * (tabla['Toyota'] - ErT) * (tabla['Pfizer'] - ErP)).sum()\ncovTP",
"_____no_output_____"
]
],
[
[
"**Actividad.** Calcular la covarianza entre los rendimientos de Toyota y Walmart.",
"_____no_output_____"
]
],
[
[
"# Calcular la covarianza\ncovTW = (tabla['Prob'] * (tabla['Toyota'] - ErT) * (tabla['Walmart'] - ErW)).sum()\ncovTW",
"_____no_output_____"
]
],
[
[
"¿Qué nos dice este número?\n- El signo nos dice las direcciones relativas entre los rendimientos de cada activo. Por ejemplo, la covarianza entre los rendimientos de Toyota y Pfizer es negativa... ver los rendimientos.\n- La magnitud de la covarianza no nos dice mucho acerca de la fuerza con que se relacionan o no estos rendimientos.",
"_____no_output_____"
],
[
"**Correlación:**\n\nUn posible problema de la covarianza es que la magnitud de esta medida no nos dice mucho acerca de la fuerza de los movimientos relativos. La *correlación* es una medida normalizada del movimiento relativo entre los rendimientos de dos activos.\n\nMatemáticamente,\n\n$$\\text{corr}(r_1,r_2)=\\rho_{12}=\\rho_{21}=\\frac{\\sigma_{12}}{\\sigma_1\\sigma_{2}}.$$",
"_____no_output_____"
],
[
"Propiedades:\n\n- Podemos notar fácilmente que la correlación de los rendimientos de un activo con los rendimientos del mismo activo es $1$: $$\\text{corr}(r_1,r_1)=\\rho_{11}=\\frac{\\sigma_{11}}{\\sigma_1\\sigma_1}=\\frac{\\sigma_{1}^2}{\\sigma_1\\sigma_1}=1.$$\n- El signo de la correlación y la covarianza es el mismo.\n- La correlación satisface: $$-1\\leq\\rho_{12}\\leq 1.$$",
"_____no_output_____"
],
[
"**Ejemplo.** Calcular la correlación entre los rendimientos de Toyota y Pfizer.",
"_____no_output_____"
]
],
[
[
"rTP = covTP / (sT * sP)\nrTP",
"_____no_output_____"
],
[
"sTP, 0.5 * sT + 0.5 * sP",
"_____no_output_____"
]
],
[
[
"**Actividad.** Calcular la correlación entre los rendimientos de Toyota y Walmart.",
"_____no_output_____"
]
],
[
[
"rTW = covTW / (sT * sW)\nrTW",
"_____no_output_____"
],
[
"sTW, 0.5 * sT + 0.5 * sW",
"_____no_output_____"
]
],
[
[
"**Conclusión.**\n- Es una medida normalizada de la fluctuación relativa de los rendimientos de dos activos.\n- En los ejemplos que vimos, sería conveniente invertir en el portafolio de Toyota y Pfizer puesto que su correlación es negativa, y esto impactaría positivamente en la diversificación del riesgo.\n\n___",
"_____no_output_____"
],
[
"## 2. Uniendo todo...",
"_____no_output_____"
],
[
"- Entonces, vimos mediante ejemplos que el riesgo en un portafolio se ve afectado significativamente por como los rendimientos de los activos se mueven relativamente.\n- Este movimiento relativo lo medimos mediante la covarianza o la correlación.\n- Si se mueven de una manera que no están perfectamente correlacionados ($\\rho<1$), entonces el riesgo del portafolio siempre será menor que el promedio ponderado de los riesgos individuales.",
"_____no_output_____"
],
[
"<img style=\"float: left; margin: 0px 0px 15px 15px;\" src=\"https://www.publicdomainpictures.net/pictures/20000/velka/happy-child.jpg\" width=\"300px\" height=\"200px\" />\n\n## Ésta es la razón por la que combinar activos en un portafolio permite diversificar el riesgo...",
"_____no_output_____"
],
[
"Entonces, ¿cómo podemos incorporar esta medida en el cálculo de la varianza del portafolio? \n- <font color=blue> Ver en el tablero...</font>\n- ¿Cómo sería para dos activos?",
"_____no_output_____"
],
[
"**Ejemplo.** Calcular por fórmula para el portafolio de Toyota y Pfizer. Comparar.",
"_____no_output_____"
]
],
[
[
"wT = 0.5\nwP = 0.5",
"_____no_output_____"
],
[
"sTP2 = (wT**2 * sT**2 + 2 * wT * wP * covTP + wP**2 * sP**2)**0.5",
"_____no_output_____"
],
[
"sTP, sTP2",
"_____no_output_____"
]
],
[
[
"**Actividad.** Calcular por fórmula para el portafolio de Toyota y Walmart. Comparar.",
"_____no_output_____"
],
[
"## 2.1. <font color=blue> Ver en el tablero...</font>\n### Matriz de varianza covarianza.\n### Matriz de correlación.",
"_____no_output_____"
],
[
"# Anuncios parroquiales\n\n## 1. Recordar quiz la siguiente clase. Temas: Clases 6 y 7.\n## 2. Tarea: revisar archivo \"Tarea4_MidiendoRendimientoRiesgo\" en clase. ",
"_____no_output_____"
],
[
"<script>\n $(document).ready(function(){\n $('div.prompt').hide();\n $('div.back-to-top').hide();\n $('nav#menubar').hide();\n $('.breadcrumb').hide();\n $('.hidden-print').hide();\n });\n</script>\n\n<footer id=\"attribution\" style=\"float:right; color:#808080; background:#fff;\">\nCreated with Jupyter by Esteban Jiménez Rodríguez.\n</footer>",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
]
] |
cb913810bec9c7e17bbdf82b5526ad101709488b
| 33,989 |
ipynb
|
Jupyter Notebook
|
part2 - splitting and preprocessing/splitting_and_preprocessing.ipynb
|
azsom/Supervised-Learning
|
42280c1c3af237f7a61a6f542ff54ce8641fb10d
|
[
"MIT"
] | 3 |
2021-01-31T03:51:49.000Z
|
2021-09-29T13:27:02.000Z
|
part2 - splitting and preprocessing/splitting_and_preprocessing.ipynb
|
azsom/Supervised-Learning
|
42280c1c3af237f7a61a6f542ff54ce8641fb10d
|
[
"MIT"
] | null | null | null |
part2 - splitting and preprocessing/splitting_and_preprocessing.ipynb
|
azsom/Supervised-Learning
|
42280c1c3af237f7a61a6f542ff54ce8641fb10d
|
[
"MIT"
] | 6 |
2021-01-17T11:06:33.000Z
|
2022-02-16T00:26:07.000Z
| 34.896304 | 225 | 0.610374 |
[
[
[
"# <center>Welcome to Supervised Learning</center>\n## <center>Part 2: How to prepare your data for supervised machine learning</center>\n## <center>Instructor: Andras Zsom</center>\n### <center>https://github.com/azsom/Supervised-Learning<center>",
"_____no_output_____"
],
[
"## The topic of the course series: supervised Machine Learning (ML)\n- how to build an ML pipeline from beginning to deployment\n- we assume you already performed data cleaning\n- this is the first course out of 6 courses\n - Part 1: Introduction to machine learning and the bias-variance tradeoff\n - **Part 2: How to prepare your data for supervised machine learning**\n - Part 3: Evaluation metrics in supervised machine learning\n - Part 4: SVMs, Random Forests, XGBoost\n - Part 5: Missing data in supervised ML\n - Part 6: Interpretability\n- you can complete the courses in sequence or complete individual courses based on your interest",
"_____no_output_____"
],
[
"### Structured data\n| X|feature_1|feature_2|...|feature_j|...|feature_m|<font color='red'>Y</font>|\n|-|:-:|:-:|:-:|:-:|:-:|:-:|:-:|\n|__data_point_1__|x_11|x_12|...|x_1j|...|x_1m|__<font color='red'>y_1</font>__|\n|__data_point_2__|x_21|x_22|...|x_2j|...|x_2m|__<font color='red'>y_2</font>__|\n|__...__|...|...|...|...|...|...|__<font color='red'>...</font>__|\n|__data_point_i__|x_i1|x_i2|...|x_ij|...|x_im|__<font color='red'>y_i</font>__|\n|__...__|...|...|...|...|...|...|__<font color='red'>...</font>__|\n|__data_point_n__|x_n1|x_n2|...|x_nj|...|x_nm|__<font color='red'>y_n</font>__|\n\nWe focus on the feature matrix (X) in this course.",
"_____no_output_____"
],
[
"### Learning objectives of this course\n\nBy the end of the course, you will be able to\n- describe why data splitting is necessary in machine learning\n- summarize the properties of IID data\n- list examples of non-IID datasets\n- apply IID splitting techniques\n- apply non-IID splitting techniques\n- identify when a custom splitting strategy is necessary\n- describe the two motivating concepts behind preprocessing\n- apply various preprocessors to categorical and continuous features\n- perform preprocessing with a sklearn pipeline and ColumnTransformer\n",
"_____no_output_____"
],
[
"# Module 1: Split IID data\n### Learning objectives of this module:\n- describe why data splitting is necessary in machine learning\n- summarize the properties of IID data\n- apply IID splitting techniques",
"_____no_output_____"
],
[
"## Why do we split the data?\n- we want to find the best hyper-parameters of our ML algorithms\n - fit models to training data\n - evaluate each model on validation set\n - we find hyper-parameter values that optimize the validation score\n- we want to know how the model will perform on previously unseen data - the generalization error\n - apply our final model on the test set\n \n### We need to split the data into three parts!",
"_____no_output_____"
],
[
"## Ask yourself these questions!\n- What is the intended use of the model? What is it supposed to do/predict?\n- What data/info do you have available at the time of prediction?\n- Your split must mimic the intended use of the model only then will you accurately estimate how well the model will perform on previously unseen points (generalization error).\n- two examples:\n - if you want to predict the outcome of a new patient's visit to the ER:\n - your test score must be based on patients not included in training and validation\n - your validation score must be based on patients not included in training\n - points of one patient should not be distributed over multiple sets because your generalization error will be off\n - predict stocks price\n - it is a time series data\n - if you predict the stocks price at a certain time in development, make sure that you only use information predating that time",
"_____no_output_____"
],
[
"## How should we split the data into train/validation/test?\n\n- data is **Independent and Identically Distributed** (iid)\n - all samples stem from the same generative process and the generative process is assumed to have no memory of past generated samples\n - identify cats and dogs on images\n - predict the house price\n - predict if someone's salary is above or below 50k\n- examples of not iid data:\n - data generated by time-dependent processes\n - data has group structure (samples collected from e.g., different subjects, experiments, measurement devices)",
"_____no_output_____"
],
[
"## Splitting strategies for iid data: basic approach\n- 60% train, 20% validation, 20% test for small datasets\n- 98% train, 1% validation, 1% test for large datasets\n - if you have 1 million points, you still have 10000 points in validation and test which is plenty to assess model performance\n",
"_____no_output_____"
],
[
"### Let's work with the adult data!\n\nhttps://archive.ics.uci.edu/ml/datasets/adult",
"_____no_output_____"
]
],
[
[
"import pandas as pd\nfrom sklearn.model_selection import train_test_split \n\ndf = pd.read_csv('data/adult_data.csv')\n\n# let's separate the feature matrix X, and target variable y\ny = df['gross-income'] # remember, we want to predict who earns more than 50k or less than 50k\nX = df.loc[:, df.columns != 'gross-income'] # all other columns are features\nprint(y)\nprint(X.head())\n",
"_____no_output_____"
],
[
"help(train_test_split)",
"_____no_output_____"
],
[
"random_state = 42\n\n# first split to separate out the training set\nX_train, X_other, y_train, y_other = train_test_split(X,y,train_size = 0.6,random_state=random_state)\nprint('training set:',X_train.shape, y_train.shape) # 60% of points are in train\nprint(X_other.shape, y_other.shape) # 40% of points are in other\nprint(X_train.head())\n\n# second split to separate out the validation and test sets\nX_val, X_test, y_val, y_test = train_test_split(X_other,y_other,train_size = 0.5,random_state=random_state)\nprint('validation set:',X_val.shape, y_val.shape) # 20% of points are in validation\nprint('test set:',X_test.shape, y_test.shape) # 20% of points are in test\nprint(X_val.head())\nprint(X_test.head())",
"_____no_output_____"
]
],
[
[
"## Randomness due to splitting\n- the model performance, validation and test scores will change depending on which points are in train, val, test\n - inherent randomness or uncertainty of the ML pipeline\n- change the random state a couple of times and repeat the whole ML pipeline to assess how much the random splitting affects your test score\n - you would expect a similar uncertainty when the model is deployed",
"_____no_output_____"
],
[
"## Splitting strategies for iid data: k-fold splitting\n\n<center><img src=\"figures/grid_search_cross_validation.png\" width=\"600\"></center>\n",
"_____no_output_____"
]
],
[
[
"from sklearn.model_selection import KFold\nhelp(KFold)",
"_____no_output_____"
],
[
"random_state = 42\n\n# first split to separate out the test set\nX_other, X_test, y_other, y_test = train_test_split(X,y,test_size = 0.2,random_state=random_state)\nprint(X_other.shape,y_other.shape)\nprint('test set:',X_test.shape,y_test.shape)\n\n# do KFold split on other\nkf = KFold(n_splits=5,shuffle=True,random_state=random_state)\nfor train_index, val_index in kf.split(X_other,y_other):\n X_train = X_other.iloc[train_index]\n y_train = y_other.iloc[train_index]\n X_val = X_other.iloc[val_index]\n y_val = y_other.iloc[val_index]\n print(' training set:',X_train.shape, y_train.shape) \n print(' validation set:',X_val.shape, y_val.shape) \n # the validation set contains different points in each iteration\n print(X_val[['age','workclass','education']].head())\n ",
"_____no_output_____"
]
],
[
[
"## How many splits should I create?\n- tough question, 3-5 is most common\n- if you do $n$ splits, $n$ models will be trained, so the larger the $n$, the most computationally intensive it will be to train the models\n- KFold is usually better suited for small datasets\n- KFold is good to estimate uncertainty due to random splitting of train and val, but it is not perfect\n - the test set remains the same",
"_____no_output_____"
],
[
"### Why shuffling iid data is important?\n- by default, data is not shuffled by Kfold which can introduce errors!\n<center><img src=\"figures/kfold.png\" width=\"600\"></center>\n",
"_____no_output_____"
],
[
"## Imbalanced data\n- imbalanced data: only a small fraction of the points are in one of the classes, usually ~5% or less but there is no hard limit here\n- examples:\n - people visit a bank's website. do they sign up for a new credit card?\n - most customers just browse and leave the page\n - usually 1% or less of the customers get a credit card (class 1), the rest leaves the page without signing up (class 0).\n - fraud detection\n - only a tiny fraction of credit card payments are fraudulent\n - rare disease diagnosis\n- the issue with imbalanced data:\n - if you apply train_test_split or KFold, you might not have class 1 points in one of your sets by chance\n - this is what we need to fix",
"_____no_output_____"
],
[
"## Solution: stratified splits",
"_____no_output_____"
]
],
[
[
"random_state = 42\n\nX_train, X_other, y_train, y_other = train_test_split(X,y,train_size = 0.6,random_state=random_state)\nX_val, X_test, y_val, y_test = train_test_split(X_other,y_other,train_size = 0.5,random_state=random_state)\n\nprint('**balance without stratification:**')\n# a variation on the order of 1% which would be too much for imbalanced data!\nprint(y_train.value_counts(normalize=True))\nprint(y_val.value_counts(normalize=True))\nprint(y_test.value_counts(normalize=True))\n\nX_train, X_other, y_train, y_other = train_test_split(X,y,train_size = 0.6,stratify=y,random_state=random_state)\nX_val, X_test, y_val, y_test = train_test_split(X_other,y_other,train_size = 0.5,stratify=y_other,random_state=random_state)\nprint('**balance with stratification:**')\n# very little variation (in the 4th decimal point only) which is important if the problem is imbalanced\nprint(y_train.value_counts(normalize=True))\nprint(y_val.value_counts(normalize=True))\nprint(y_test.value_counts(normalize=True))",
"_____no_output_____"
]
],
[
[
"## Stratified folds\n<center><img src=\"figures/stratified_kfold.png\" width=\"600\"></center>\n",
"_____no_output_____"
]
],
[
[
"from sklearn.model_selection import StratifiedKFold\nhelp(StratifiedKFold)",
"_____no_output_____"
],
[
"# what we did before: variance in balance on the order of 1%\nrandom_state = 42\n\nX_other, X_test, y_other, y_test = train_test_split(X,y,test_size = 0.2,random_state=random_state)\nprint('test balance:',y_test.value_counts(normalize=True))\n\n# do KFold split on other\nkf = KFold(n_splits=5,shuffle=True,random_state=random_state)\nfor train_index, val_index in kf.split(X_other,y_other):\n X_train = X_other.iloc[train_index]\n y_train = y_other.iloc[train_index]\n X_val = X_other.iloc[val_index]\n y_val = y_other.iloc[val_index]\n print('train balance:')\n print(y_train.value_counts(normalize=True))\n print('val balance:')\n print(y_val.value_counts(normalize=True))",
"_____no_output_____"
],
[
"# stratified K Fold: variation in balance is very small (4th decimal point)\nrandom_state = 42\n\n# stratified train-test split\nX_other, X_test, y_other, y_test = train_test_split(X,y,test_size = 0.2,stratify=y,random_state=random_state)\nprint('test balance:',y_test.value_counts(normalize=True))\n\n# do StratifiedKFold split on other\nkf = StratifiedKFold(n_splits=5,shuffle=True,random_state=random_state)\nfor train_index, val_index in kf.split(X_other,y_other):\n X_train = X_other.iloc[train_index]\n y_train = y_other.iloc[train_index]\n X_val = X_other.iloc[val_index]\n y_val = y_other.iloc[val_index]\n print('train balance:')\n print(y_train.value_counts(normalize=True))\n print('val balance:')\n print(y_val.value_counts(normalize=True))",
"_____no_output_____"
]
],
[
[
"# Module 2: Split non-IID data\n### Learning objectives of this module:\n- list examples of non-IID datasets\n- apply non-IID splitting techniques\n- identify when a custom splitting strategy is necessary",
"_____no_output_____"
],
[
"## Examples of non-iid data\n- if there is any sort of time or group structure in your data, it is likely non-iid\n - group structure:\n - each point is someone's visit to the ER and some people visited the ER multiple times\n - each point is a customer's visit to website and customers tend to return regularly\n - time structure\n - each point is the stocks price at a given time\n - eahc point is a person's health or activity status\n ",
"_____no_output_____"
],
[
"## Group-based split: GroupShuffleSplit\n<center><img src=\"figures/groupshufflesplit.png\" width=\"600\"></center>\n",
"_____no_output_____"
]
],
[
[
"import numpy as np\nfrom sklearn.model_selection import GroupShuffleSplit\nX = np.ones(shape=(8, 2))\ny = np.ones(shape=(8, 1))\ngroups = np.array([1, 1, 2, 2, 2, 3, 3, 3])\n\ngss = GroupShuffleSplit(n_splits=10, train_size=.8, random_state=42)\n\nfor train_idx, test_idx in gss.split(X, y, groups):\n print(\"TRAIN:\", train_idx, \"TEST:\", test_idx)\n",
"_____no_output_____"
]
],
[
[
"## Group-based split: GroupKFold\n<center><img src=\"figures/groupkfold.png\" width=\"600\"></center>\n",
"_____no_output_____"
]
],
[
[
"from sklearn.model_selection import GroupKFold\n\ngroup_kfold = GroupKFold(n_splits=3)\n\nfor train_index, test_index in group_kfold.split(X, y, groups):\n print(\"TRAIN:\", train_index, \"TEST:\", test_index)\n",
"_____no_output_____"
],
[
"help(GroupKFold)",
"_____no_output_____"
]
],
[
[
"## Data leakage in time series data is similar!\n- do NOT use information in validation or test which will not be available once your model is deployed\n - don't use future information!\n \n<center><img src=\"figures/timeseriessplit.png\" width=\"600\"></center>\n",
"_____no_output_____"
]
],
[
[
"import numpy as np\nfrom sklearn.model_selection import TimeSeriesSplit\nX = np.array([[1, 2], [3, 4], [1, 2], [3, 4], [1, 2], [3, 4]])\ny = np.array([1, 2, 3, 4, 5, 6])\ntscv = TimeSeriesSplit()\nfor train_index, test_index in tscv.split(X):\n print(\"TRAIN:\", train_index, \"TEST:\", test_index)\n X_train, X_test = X[train_index], X[test_index]\n y_train, y_test = y[train_index], y[test_index]\n",
"_____no_output_____"
]
],
[
[
"## When should you develop your own splitting function?\n- there are certain splitting strategies sklearn can't handle at the moment\n - time series data with group structure is one example\n - if you want certain groups to be in certain sets\n - group structure in classification where all points in a group belong to a certain class\n - you might want a roughly equal number of groups of each class to be in each set \n- check out the [model selection](https://scikit-learn.org/stable/modules/classes.html#module-sklearn.model_selection) part of sklearn\n - if the splitting stragey you want to follow is not there, implement your own function\n",
"_____no_output_____"
],
[
"# Module 3: Preprocess continuous and categorical features\n### Learning objectives of this module:\n- describe the two motivating concepts behind preprocessing\n- apply various preprocessors to categorical and continuous features\n- perform preprocessing with a sklearn pipeline and ColumnTransformer",
"_____no_output_____"
],
[
"### Data almost never comes in a format that's directly usable in ML\n- ML works with numerical data but some columns are text (e.g., home country, educational level, gender, race)\n - some ML algorithms accept (and prefer) a non-numerical feature matrix (like [CatBoost](https://catboost.ai/) ) but that's not standard\n - sklearn throws an error message if the feature matrix contains non-numerical elements\n- the order of magnitude of numerical features can vary greatly which is not good for most ML algorithms (e.g., salary in USD, age in years, time spent on the site in sec)\n - many ML algorithms are distance-based and they perform better and converge faster if the features are standardized (features have a mean of 0 and the same standard deviation, usually 1)\n - Lasso and Ridge regression because of the penalty term, K Nearest Neightbors, SVM, linear models if you want to use the coefficients to measure feature importance (more on this in part 6), neural networks\n - tree-based methods don't require standardization \n - check out part 1 to learn more about linear and logistic regression, Lasso and Ridge\n - check out part 4 to learn more about SVMs, tree-based methods, and K Nearest Neighbors",
"_____no_output_____"
],
[
"### scikit-learn transformers to the rescue!\n\nPreprocessing is done with various transformers. All transformes have three methods:\n- **fit** method: estimates parameters necessary to do the transformation,\n- **transform** method: transforms the data based on the estimated parameters,\n- **fit_transform** method: both steps are performed at once, this can be faster than doing the steps separately.",
"_____no_output_____"
],
[
"### Transformers we cover \n- **OrdinalEncoder** - converts categorical features into an integer array\n- **OneHotEncoder** - converts categorical features into dummy arrays\n- **StandardScaler** - standardizes continuous features by removing the mean and scaling to unit variance",
"_____no_output_____"
],
[
"## Ordered categorical data: OrdinalEncoder",
"_____no_output_____"
],
[
"Let's assume we have a categorical feature and training and test sets\n\nThe cateogies can be ordered or ranked\n\nE.g., educational level in the adult dataset",
"_____no_output_____"
]
],
[
[
"import pandas as pd\n\ntrain_edu = {'educational level':['Bachelors','Masters','Bachelors','Doctorate','HS-grad','Masters']} \ntest_edu = {'educational level':['HS-grad','Masters','Masters','College','Bachelors']}\n\nX_train = pd.DataFrame(train_edu)\nX_test = pd.DataFrame(test_edu)",
"_____no_output_____"
],
[
"from sklearn.preprocessing import OrdinalEncoder\nhelp(OrdinalEncoder)",
"_____no_output_____"
],
[
"# initialize the encoder\ncats = ['HS-grad','Bachelors','Masters','Doctorate']\n\nenc = OrdinalEncoder(categories = [cats]) # The ordered list of \n# categories need to be provided. By default, the categories are alphabetically ordered!\n\n# fit the training data\nenc.fit(X_train)\n# print the categories - not really important because we manually gave the ordered list of categories\nprint(enc.categories_)\n# transform X_train. We could have used enc.fit_transform(X_train) to combine fit and transform\nX_train_oe = enc.transform(X_train)\nprint(X_train_oe)\n# transform X_test\nX_test_oe = enc.transform(X_test) # OrdinalEncoder always throws an error message if \n # it encounters an unknown category in test\nprint(X_test_oe)",
"_____no_output_____"
]
],
[
[
"## Unordered categorical data: one-hot encoder",
"_____no_output_____"
],
[
"some categories cannot be ordered. e.g., workclass, relationship status\n\nfirst feature: gender (male, female, unknown)\n\nsecond feature: browser used \n\nthese categories cannot be ordered",
"_____no_output_____"
]
],
[
[
"train = {'gender':['Male','Female','Unknown','Male','Female','Female'],\\\n 'browser':['Safari','Safari','Internet Explorer','Chrome','Chrome','Internet Explorer']}\ntest = {'gender':['Female','Male','Unknown','Female'],'browser':['Chrome','Firefox','Internet Explorer','Safari']}\n\nX_train = pd.DataFrame(train)\nX_test = pd.DataFrame(test)",
"_____no_output_____"
],
[
"# How do we convert this to numerical features?\nfrom sklearn.preprocessing import OneHotEncoder\n\nhelp(OneHotEncoder)",
"_____no_output_____"
],
[
"# initialize the encoder\nenc = OneHotEncoder(sparse=False) # by default, OneHotEncoder returns a sparse matrix. sparse=False returns a 2D array\n# fit the training data\nenc.fit(X_train)\nprint('categories:',enc.categories_)\nprint('feature names:',enc.get_feature_names())\n# transform X_train\nX_train_ohe = enc.transform(X_train)\n#print(X_train_ohe)\n# do all of this in one step\nX_train_ohe = enc.fit_transform(X_train)\nprint(X_train_ohe)\n\n# transform X_test\nX_test_ohe = enc.transform(X_test)\nprint('X_test transformed')\nprint(X_test_ohe)",
"_____no_output_____"
]
],
[
[
"## Continuous features: StandardScaler",
"_____no_output_____"
]
],
[
[
"train = {'salary':[50_000,75_000,40_000,1_000_000,30_000,250_000,35_000,45_000]}\ntest = {'salary':[25_000,55_000,1_500_000,60_000]}\n\nX_train = pd.DataFrame(train)\nX_test = pd.DataFrame(test)",
"_____no_output_____"
],
[
"from sklearn.preprocessing import StandardScaler\nhelp(StandardScaler)",
"_____no_output_____"
],
[
"scaler = StandardScaler()\nprint(scaler.fit_transform(X_train))\nprint(scaler.transform(X_test))",
"_____no_output_____"
]
],
[
[
"## How and when to do preprocessing in the ML pipeline?\n- **SPLIT YOUR DATA FIRST!**\n- **APPLY TRANSFORMER.FIT ONLY ON YOUR TRAINING DATA!** Then transform the validation and test sets.\n- One of the most common mistake practitioners make is leaking statistics!\n - fit_transform is applied to the whole dataset, then the data is split into train/validation/test\n - this is wrong because the test set statistics impacts how the training and validation sets are transformed\n - but the test set must be separated from train and val, and val must be separated from train\n - or fit_transform is applied to the train, then fit_transform is applied to the validation set, and fit_transform is applied to the test set\n - this is wrong because the relative position of the points change\n<center><img src=\"figures/no_separate_scaling.png\" width=\"1200\"></center>\n",
"_____no_output_____"
],
[
"## Scikit-learn's pipelines\n\n- Preprocessing and model training (not the splitting) can be chained together into a scikit-learn pipeline which consists of transformers and one final estimator which is usually your classifier or regression model.\n- It neatly combines the preprocessing steps and it helps to avoid leaking statistics.\n\nhttps://scikit-learn.org/stable/auto_examples/compose/plot_column_transformer_mixed_types.html\n",
"_____no_output_____"
]
],
[
[
"import pandas as pd\nimport numpy as np\n\nfrom sklearn.compose import ColumnTransformer\nfrom sklearn.pipeline import Pipeline\nfrom sklearn.preprocessing import StandardScaler, OneHotEncoder, OrdinalEncoder, MinMaxScaler\nfrom sklearn.model_selection import train_test_split\n\nnp.random.seed(0)\n\ndf = pd.read_csv('data/adult_data.csv')\n\n# let's separate the feature matrix X, and target variable y\ny = df['gross-income'] # remember, we want to predict who earns more than 50k or less than 50k\nX = df.loc[:, df.columns != 'gross-income'] # all other columns are features\n\nrandom_state = 42\n\n# first split to separate out the training set\nX_train, X_other, y_train, y_other = train_test_split(X,y,train_size = 0.6,random_state=random_state)\n\n# second split to separate out the validation and test sets\nX_val, X_test, y_val, y_test = train_test_split(X_other,y_other,train_size = 0.5,random_state=random_state)\n",
"_____no_output_____"
],
[
"# collect which encoder to use on each feature\n# needs to be done manually\nordinal_ftrs = ['education'] \nordinal_cats = [[' Preschool',' 1st-4th',' 5th-6th',' 7th-8th',' 9th',' 10th',' 11th',' 12th',' HS-grad',\\\n ' Some-college',' Assoc-voc',' Assoc-acdm',' Bachelors',' Masters',' Prof-school',' Doctorate']]\nonehot_ftrs = ['workclass','marital-status','occupation','relationship','race','sex','native-country']\nstd_ftrs = ['capital-gain','capital-loss','age','hours-per-week']\n\n# collect all the encoders\npreprocessor = ColumnTransformer(\n transformers=[\n ('ord', OrdinalEncoder(categories = ordinal_cats), ordinal_ftrs),\n ('onehot', OneHotEncoder(sparse=False,handle_unknown='ignore'), onehot_ftrs),\n ('std', StandardScaler(), std_ftrs)])\n\n# for now we only preprocess, later on we will add other steps here\n# note the final scaler which is a standard scaler\n# the ordinal and one hot encoded features do not have a mean of 0 and an std of 1\n# the final scaler standardizes those features\nclf = Pipeline(steps=[('preprocessor', preprocessor),('final scaler',StandardScaler())]) \n\nX_train_prep = clf.fit_transform(X_train)\nX_val_prep = clf.transform(X_val)\nX_test_prep = clf.transform(X_test)\n\nprint(X_train.shape)\nprint(X_train_prep.shape)\n\nprint(np.mean(X_train_prep,axis=0))\nprint(np.std(X_train_prep,axis=0))\nprint(np.mean(X_val_prep,axis=0))\nprint(np.std(X_val_prep,axis=0))\nprint(np.mean(X_test_prep,axis=0))\nprint(np.std(X_test_prep,axis=0))\n",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
]
] |
cb9138bed7c12c94df2d7baf732314f0f0b15ea0
| 24,250 |
ipynb
|
Jupyter Notebook
|
jupyter_french/topic05_ensembles_random_forests/topic5_part1_bagging-fr_def.ipynb
|
salman394/AI-ml--course
|
2ed3a1382614dd00184e5179026623714ccc9e8c
|
[
"Unlicense"
] | null | null | null |
jupyter_french/topic05_ensembles_random_forests/topic5_part1_bagging-fr_def.ipynb
|
salman394/AI-ml--course
|
2ed3a1382614dd00184e5179026623714ccc9e8c
|
[
"Unlicense"
] | null | null | null |
jupyter_french/topic05_ensembles_random_forests/topic5_part1_bagging-fr_def.ipynb
|
salman394/AI-ml--course
|
2ed3a1382614dd00184e5179026623714ccc9e8c
|
[
"Unlicense"
] | null | null | null | 64.323607 | 871 | 0.670598 |
[
[
[
"<center>\n<img src=\"https://github.com/Yorko/mlcourse.ai/blob/master/img/ods_stickers.jpg?raw=true\" />\n \n<br>\n\n<div style=\"font-weight:700;font-size:25px\"> [mlcourse.ai](https://mlcourse.ai) - Open Machine Learning Course </div>\n\n<br>\n\nAuteurs: [Vitaliy Radchenko](https://www.linkedin.com/in/vitaliyradchenk0/) et [Yury Kashnitsky](https://yorko.github.io). Traduit et édité par [Christina Butsko](https://www.linkedin.com/in/christinabutsko/), [Egor Polusmak](https://www.linkedin.com/in/egor-polusmak/), [Anastasia Manokhina](https://www.linkedin.com/in/anastasiamanokhina/), [Anna Shirshova](http://linkedin.com/in/anna-shirshova-b908458b), [Yuanyuan Pao](https://www.linkedin.com/in/yuanyuanpao/) et [Ousmane Cissé](https://github.com/oussou-dev). Ce matériel est soumis aux termes et conditions de la licence [Creative Commons CC BY-NC-SA 4.0](https://creativecommons.org/licenses/by-nc-sa/4.0/). L'utilisation gratuite est autorisée à des fins non commerciales.",
"_____no_output_____"
],
[
"# <center> Thème 5. Ensembles et Forêts aléatoires (random forest) </center>\n<center> <div style=\"font-weight:700;font-size:25px\"> Partie 1. Bagging (Boostrap Aggregating) </div> </center>\n\n",
"_____no_output_____"
],
[
"<h1>Sommaire<span class=\"tocSkip\"></span></h1>\n<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Outline\" data-toc-modified-id=\"Outline-1\">Outline</a></span></li><li><span><a href=\"#1.-Ensembles\" data-toc-modified-id=\"1.-Ensembles-2\">1. Ensembles</a></span></li><li><span><a href=\"#2.-Bootstrapping\" data-toc-modified-id=\"2.-Bootstrapping-3\">2. Bootstrapping</a></span></li><li><span><a href=\"#3.-Bagging\" data-toc-modified-id=\"3.-Bagging-4\">3. Bagging</a></span></li><li><span><a href=\"#4.-L'erreur-Out-of-bag\" data-toc-modified-id=\"4.-L'erreur-Out-of-bag-5\">4. L'erreur Out-of-bag</a></span></li><li><span><a href=\"#5.-Mission-d'entraînement\" data-toc-modified-id=\"5.-Mission-d'entraînement-6\">5. Mission d'entraînement</a></span></li><li><span><a href=\"#6.-Ressources-utiles\" data-toc-modified-id=\"6.-Ressources-utiles-7\">6. Ressources utiles</a></span></li></ul></div>",
"_____no_output_____"
],
[
"Dans des articles précédents, vous avez exploré différents algorithmes de classification ainsi que des techniques pouvant être utilisées pour valider et évaluer correctement la qualité de vos modèles.\n\nSupposons maintenant que vous avez choisi le meilleur modèle possible pour un problème particulier et que vous vous efforcez d’améliorer encore sa précision. Dans ce cas, vous devez appliquer des techniques d’apprentissage automatique plus avancées, appelées collectivement **_ensembles_**.\n\nUn *ensemble* est une collection d’éléments qui contribuent collectivement à un tout. Un exemple familier est un ensemble musical, qui associe les sons de plusieurs instruments de musique pour créer une harmonie, ou des ensembles architecturaux, qui sont un ensemble de bâtiments conçus comme une unité. Dans les ensembles, le résultat (global) harmonieux est plus important que la performance d'une partie individuelle.",
"_____no_output_____"
],
[
"## 1. Ensembles\n\nLe théorème [Condorcet's jury theorem](https://en.wikipedia.org/wiki/Condorcet%27s_jury_theorem) (1784) traite d'un ensemble dans un certain sens. Il indique que, si chaque membre du jury rend un jugement indépendant et que la probabilité d'une décision correcte par chaque juré est supérieure à 0,5, la probabilité d'une décision correcte par l'ensemble du jury augmente avec le nombre total de jurés et tend à un. D'autre part, si la probabilité d'avoir raison est inférieure à 0,5 pour chaque juré, la probabilité d'une décision correcte par l'ensemble du jury diminue avec le nombre de jurés et tend vers zéro.\n\nÉcrivons une expression analytique pour ce théorème:\n\n- $\\large N$ est le nombre total de jurés;\n- $\\large m$ est un nombre minimal de jurés qui formeraient une majorité, c'est-à-dire $\\large m = floor(N/2) + 1$;\n- $\\large {N \\choose i}$ est le nombre de combinaisons $\\large i$ d'un ensemble contenant des éléments $\\large N$.\n- $\\large p$ est la probabilité que le juré prenne la décision correcte;\n- $\\large \\mu$ est la probabilité que le jury entier prenne la bonne décision.\n\nEnsuite:\n\n$$ \\large \\mu = \\sum_{i=m}^{N}{N\\choose i}p^i(1-p)^{N-i} $$\n\nOn peut voir que si $\\large p > 0.5$, alors $\\large \\mu > p$. De plus, si $\\large N \\rightarrow \\infty $, alors $\\large \\mu \\rightarrow 1$.\n\nRegardons un autre exemple d'ensembles : une observation connue sous le nom de [ Wisdom of the crowd](https://en.wikipedia.org/wiki/Wisdom_of_the_crowd). <img src=\"https://github.com/Yorko/mlcourse.ai/blob/master/img/bull.png?raw=true\" align=\"right\" width=15% height=15%> En 1906, [Francis Galton](https://en.wikipedia.org/wiki/Francis_ Galton) s'est rendu à une foire à Plymouth où il a assisté à un concours pour les agriculteurs. 800 participants ont essayé d'estimer le poids d'un taureau abattu. Le poids réel du taureau était de 1198 livres. Bien qu'aucun des fermiers ne puisse deviner le poids exact de l'animal, la moyenne de leurs prédictions était de 1197 livres.\n\nUne idée similaire de réduction des erreurs a été adoptée dans le domaine de l'apprentissage automatique.",
"_____no_output_____"
],
[
"## 2. Bootstrapping\n\n*Bagging* (également connu sous le nom de [Bootstrap aggregation](https://en.wikipedia.org/wiki/Bootstrap_aggregating)) est l'une des premières et des plus fondamentales des techniques d'ensemble. Il a été proposé par [Leo Breiman](https://en.wikipedia.org/fr/wiki/Leo_Breiman) en 1994. Le Bagging repose sur la méthode statistique de [bootstrapping](https://en.wikipedia.org/wiki/Bootstrapping_%28statistics%29), ce qui permet d'évaluer de nombreuses statistiques de modèles complexes.\n\nLa méthode de Bootstrap se présente ainsi. Soit un échantillon $\\large X$ de taille $\\large N$. Nous pouvons créer un nouvel échantillon à partir de l'échantillon d'origine en tirant des éléments $\\large N$ de celui-ci de manière aléatoire et uniforme, avec remplacement. En d'autres termes, nous sélectionnons un élément aléatoire dans l'échantillon d'origine de taille $\\large N$ et procédons de la sorte $\\large N$ fois. Tous les éléments sont également susceptibles d'être sélectionnés, ainsi chaque élément est tiré avec la probabilité égale à $\\large \\frac{1}{N}$.\n\nDisons que nous tirons les boules d'un sac une à la fois. A chaque étape, la boule sélectionnée est remise dans le sac de sorte que la sélection suivante soit faite de manière égale, c'est-à-dire à partir du même nombre de boules $\\large N$. Notez que, comme nous remettons les boules dans le sac, il peut y avoir des doublons dans le nouvel échantillon. Appelons ce nouvel exemple $\\large X_1$.\n\nEn répétant cette procédure $\\large M$ fois, nous créons $\\large M$ * échantillons bootstrap * $\\large X_1, \\dots, X_M$. Au final, nous disposons d’un nombre suffisant d’échantillons et pouvons calculer diverses statistiques de la distribution initiale.\n\n\n\nPour notre exemple, nous utiliserons le jeu de données `telecom_churn` bien connu. Précédemment, lorsque nous avions discuté de l'importance des caractéristiques, nous avions constaté que l'une des caractéristiques les plus importantes de cet ensemble de données était le nombre d'appels au service clientèle. Visualisons les données et regardons la distribution de cette caractéristique.",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport pandas as pd\nimport seaborn as sns\n\nsns.set()\n%matplotlib inline\nfrom matplotlib import pyplot as plt\n\ntelecom_data = pd.read_csv(\"../../data/telecom_churn.csv\")\n\ntelecom_data.loc[telecom_data[\"Churn\"] == False, \"Customer service calls\"].hist(\n label=\"Loyal\"\n)\ntelecom_data.loc[telecom_data[\"Churn\"] == True, \"Customer service calls\"].hist(\n label=\"Churn\"\n)\nplt.xlabel(\"Number of calls\")\nplt.ylabel(\"Density\")\nplt.legend();",
"_____no_output_____"
]
],
[
[
"On dirait que les clients fidèles font moins d’appels au service clientèle que ceux qui finissent par partir. À présent, il peut être judicieux d’estimer le nombre moyen d’appels au service clientèle dans chaque groupe. Comme notre ensemble de données est petit, nous ne pourrions pas obtenir une bonne estimation en calculant simplement la moyenne de l'échantillon initial. Nous ferons mieux d'appliquer la méthode bootstrap. Générons 1000 nouveaux échantillons bootstrap à partir de notre population d'origine et produisons une estimation par intervalle de la moyenne.",
"_____no_output_____"
]
],
[
[
"def get_bootstrap_samples(data, n_samples):\n \"\"\"Generate bootstrap samples using the bootstrap method.\"\"\"\n indices = np.random.randint(0, len(data), (n_samples, len(data)))\n samples = data[indices]\n return samples\n\n\ndef stat_intervals(stat, alpha):\n \"\"\"Produce an interval estimate.\"\"\"\n boundaries = np.percentile(stat, [100 * alpha / 2.0, 100 * (1 - alpha / 2.0)])\n return boundaries\n\n\n# on sauvegarde les données sur les clients fidèles et les anciens clients pour fractionner l'ensemble de données\nloyal_calls = telecom_data.loc[\n telecom_data[\"Churn\"] == False, \"Customer service calls\"\n].values\nchurn_calls = telecom_data.loc[\n telecom_data[\"Churn\"] == True, \"Customer service calls\"\n].values\n\n# nous fixonx random.seed() à 0 pour conserver la reproductibité des résultats\nnp.random.seed(0)\n\n# Générer les échantillons à l'aide du bootstrapping et calculer la moyenne pour chacun d'entre eux\nloyal_mean_scores = [\n np.mean(sample) for sample in get_bootstrap_samples(loyal_calls, 1000)\n]\nchurn_mean_scores = [\n np.mean(sample) for sample in get_bootstrap_samples(churn_calls, 1000)\n]\n\n# afficher les estimations d'intervalle résultantes\nprint(\n \"Service calls from loyal: mean interval\", stat_intervals(loyal_mean_scores, 0.05)\n)\nprint(\n \"Service calls from churn: mean interval\", stat_intervals(churn_mean_scores, 0.05)\n)",
"_____no_output_____"
]
],
[
[
"Pour l'interprétation des intervalles de confiance, vous pouvez lire cette [note](https://www.graphpad.com/guides/prism/7/statistics/stat_more_about_confidence_interval.htm?toc=0&printWindow) concise ou tout cours sur les statistiques. Il n'est pas exact de dire qu'un intervalle de confiance contient 95% des valeurs. Notez que l'intervalle pour les clients fidèles est plus étroit, ce qui est raisonnable car ils effectuent moins d'appels (0, 1 ou 2) par rapport aux clients désabonnés qui appellent jusqu'à en avoir marre et décident de changer de fournisseur.",
"_____no_output_____"
],
[
"## 3. Bagging\n\nMaintenant que vous avez compris l’idée du Bootstraping, nous pouvons passer au *Bagging*.\n\nSupposons que nous ayons un jeu de données d'entraînement $\\large X$. À l’aide du Bootstraping , nous générons des exemples $\\large X_1, \\dots, X_M$. Maintenant, pour chaque échantillon bootstrap, nous entraînons son propre classifieur $\\large a_i(x)$. Le classifieur final fera la moyenne des sorties de tous ces classifieurs individuels. En cas de classification, cette technique correspond au vote:\n$$\\large a(x) = \\frac{1}{M}\\sum_{i = 1}^M a_i(x).$$\n\nL'image ci-dessous illustre cet algorithme:\n<img src=\"https://github.com/Yorko/mlcourse.ai/blob/master/img/bagging.png?raw=true\" alt=\"image\"/>",
"_____no_output_____"
],
[
"Considérons un problème de régression avec les algorithmes de base $\\large b_1(x), \\dots , b_n(x)$. Supposons qu'il existe une fonction cible idéale de vraies réponses $\\large y(x)$ définie pour toutes les entrées et que la distribution $\\large p(x)$ est définie. Nous pouvons ensuite exprimer l'erreur pour chaque fonction de régression comme suit :\n\n$$\\large \\varepsilon_i(x) = b_i(x) - y(x), \\quad i = 1, \\dots, n$$\n\nEt la valeur attendue de l'erreur quadratique moyenne :\n\n$$\\large \\E_x\\left[\\left(b_i(x) - y(x)\\right)^{2}\\right] = \\E_x\\left[\\varepsilon_i^{2}(x)\\right].$$\n\nEnsuite, l’erreur moyenne sur toutes les fonctions de régression se présentera comme suit :\n$$ \\large \\E_1 = \\frac{1}{n} \\E_x\\left[ \\sum_i^n \\varepsilon_i^{2}(x)\\right]$$\n\nNous supposerons que les erreurs sont non biaisées et non corrélées, c'est-à-dire :\n\n$$\\large \\begin{array}{rcl} \\E_x\\left[\\varepsilon_i(x)\\right] &=& 0, \\\\\n\\E_x\\left[\\varepsilon_i(x)\\varepsilon_j(x)\\right] &=& 0, \\quad i \\neq j. \\end{array}$$\n\nConstruisons maintenant une nouvelle fonction de régression qui calculera la moyenne des valeurs des fonctions individuelles :\n\n$$\\large a(x) = \\frac{1}{n}\\sum_{i=1}^{n}b_i(x)$$\n\nTrouvons son erreur quadratique moyenne :\n\n$$\\large \\begin{array}{rcl}\\E_n &=& \\E_x\\left[\\frac{1}{n}\\sum_{i=1}^{n}b_i(x)-y(x)\\right]^2 \\\\\n&=& \\E_x\\left[\\frac{1}{n}\\sum_{i=1}^{n}\\varepsilon_i\\right]^2 \\\\\n&=& \\frac{1}{n^2}\\E_x\\left[\\sum_{i=1}^{n}\\varepsilon_i^2(x) + \\sum_{i \\neq j}\\varepsilon_i(x)\\varepsilon_j(x)\\right] \\\\\n&=& \\frac{1}{n}\\E_1\\end{array}$$\n\nAinsi, en faisant la moyenne des réponses individuelles, nous avons réduit l’erreur quadratique moyenne par un facteur de $\\large n$.\n\nDans notre leçon précédente, rappelons les composants qui constituent l'erreur totale hors échantillon : \n\n$$\\large \\begin{array}{rcl} \n\\Err\\left(\\vec{x}\\right) &=& \\E\\left[\\left(y - \\hat{f}\\left(\\vec{x}\\right)\\right)^2\\right] \\\\\n&=& \\sigma^2 + f^2 + \\Var\\left(\\hat{f}\\right) + \\E\\left[\\hat{f}\\right]^2 - 2f\\E\\left[\\hat{f}\\right] \\\\\n&=& \\left(f - \\E\\left[\\hat{f}\\right]\\right)^2 + \\Var\\left(\\hat{f}\\right) + \\sigma^2 \\\\\n&=& \\Bias\\left(\\hat{f}\\right)^2 + \\Var\\left(\\hat{f}\\right) + \\sigma^2\n\\end{array}$$",
"_____no_output_____"
],
[
"Le Bagging réduit la variance d'un classifieur en diminuant la différence d'erreur lorsque nous entraînons le modèle sur différents jeux de données. En d'autres termes, le Bagging empêche le sur-apprentissage. L'efficacité du Bagging vient du fait que les modèles individuels sont très différents en raison des données d'entraînement différentes et que leurs erreurs s'annulent pendant le vote. De plus, les valeurs aberrantes sont probablement omises dans certains exemples de bootstrap d'entraînement.\n\nLa bibliothèque `scikit-learn` prend en charge le Bagging avec les méta-estimateurs `BaggingRegressor` et `BaggingClassifier`. Vous pouvez utiliser la plupart des algorithmes comme base.\n\nExaminons comment fonctionne le Bagging et comparons-le avec un arbre de décision. Pour cela, nous allons utiliser un exemple tiré de [la documentation de sklearn](http://scikit-learn.org/stable/auto_examples/ensemble/plot_bias_variance.html#sphx-glr-auto-examples-ensemble-plot-bias-variance-py).\n\n\n\nL'erreur pour l'arbre de décision:\n$$ \\large 0.0255 \\, (\\Err) = 0.0003 \\, (\\Bias^2) + 0.0152 \\, (\\Var) + 0.0098 \\, (\\sigma^2) $$\n\nL'erreur lors de l'utilisation du Bagging:\n$$ \\large 0.0196 \\, (\\Err) = 0.0004 \\, (\\Bias^2) + 0.0092 \\, (\\Var) + 0.0098 \\, (\\sigma^2) $$\n\nComme vous pouvez le voir sur le graphique ci-dessus, la variance de l'erreur est beaucoup plus faible pour le Bagging. Rappelez-vous que nous l'avons déjà prouvé théoriquement.\n\nLe Bagging est efficace sur de petits ensembles de données. L'abandon d'une petite partie des données d'apprentissage conduit à la construction de classifieurs de base très différents. Si vous avez un grand jeu de données, vous générerez des échantillons de bootstrap de taille beaucoup plus petite.\n\nL'exemple ci-dessus ne sera probablement pas applicable à un travail réel. En effet, nous avons fortement présumé que nos erreurs individuelles ne sont pas corrélées. Le plus souvent, c’est beaucoup trop optimiste pour des applications réelles. Lorsque cette hypothèse est fausse, la réduction d'erreur ne sera pas aussi significative. Dans les cours suivants, nous aborderons certaines méthodes d'ensemble plus sophistiquées, qui permettent des prévisions plus précises sur des problèmes du monde réel.",
"_____no_output_____"
],
[
"## 4. L'erreur Out-of-bag\n\nPour la suite, dans le cas d'une forêt aléatoire, il n'est pas nécessaire d'utiliser des échantillons de validation croisée ou un échantilon de données de test (hold-out samples) pour obtenir une estimation non biaisée de l'erreur. Pourquoi ? Parce que, dans les techniques d'ensemble, l'estimation de l'erreur se fait en interne.\n \nLes arbres aléatoires sont construits à l'aide de différents échantillons bootstrap de l'ensemble de données d'origine. Environ 37% des entrées sont exclues d'un échantillon bootstrap particulier et ne sont pas utilisées dans la construction de l'arbre $\\large k$-th.\n\nC'est facile à prouver. Supposons qu'il y a des exemples $\\large \\ell$ dans notre jeu de données. A chaque étape, chaque point de données a une probabilité égale de se retrouver dans un échantillon bootstrap avec remplacement, probabilité $\\large\\frac{1}{\\ell}.$ La probabilité qu’il n’existe pas d’échantillon bootstrap contenant un élément particulier de l'ensemble de données (c’est-à-dire qu’il a été omis $\\large \\ell$ fois) est égale à $\\large (1 - \\frac{1}{\\ell})^\\ell$. Lorsque $\\large \\ell \\rightarrow +\\infty$, il devient égal à la [seconde limite remarquable](https://en.wikipedia.org/wiki/List_of_limits) $\\large \\frac{1}{e}$. Ensuite, la probabilité de choisir un exemple spécifique est $\\large \\approx 1 - \\frac{1}{e} \\approx 63\\%$.\n\nVisualisons le fonctionnement de l'estimation de l'erreurs Out-of-Bag (ou OOBE) :\n\n\n\nLa partie supérieure de la figure ci-dessus représente notre jeu de données d'origine. Nous l'avons divisé en ensembles d'entraînement (à gauche) et de test (à droite). Dans l'image de gauche, nous dessinons une grille qui divise parfaitement notre jeu de données en fonction des classes. Maintenant, nous utilisons la même grille pour estimer la part des réponses correctes sur notre ensemble de tests. Nous pouvons voir que notre classifieur a donné des réponses incorrectes dans les 4 cas qui n'ont pas été utilisés pendant l'entraînement (à gauche). Par conséquent, la précision de notre classifieur est $\\large \\frac{11}{15}*100\\% = 73.33\\%$.\n\nEn résumé, chaque algorithme de base est formé sur $\\large \\approx 63\\%$ des exemples originaux. Il peut être validé sur le $\\large \\approx 37\\%$ restant. L'estimation Out-of-Bag n'est rien de plus que l'estimation moyenne des algorithmes de base sur ces $\\large \\approx 37\\%$ d'entrées laissées en dehors de l'entraînement.",
"_____no_output_____"
],
[
"## 5. Mission d'entraînement\nVous pouvez vous entraîner avec [cette mission](https://www.kaggle.com/kashnitsky/a5-demo-logit-and-rf-for-credit-scoring) dans laquelle vous travaillerez avec la régression logistique et le Random Forest dans un tâche de scoring de crédit. Avec une [solution](https://www.kaggle.com/kashnitsky/a5-demo-logit-and-rf-for-credit-scoring-sol).\n\n## 6. Ressources utiles\n- Cours principal [site](https://mlcourse.ai), [repo github](https://github.com/Yorko/mlcourse.ai) et [chaîne YouTube](https://www.youtube.com/watch?v=QKTuw4PNOsU&list=PLVlY_7IJCMJeRfZ68eVfEcu-UcN9BbwiX)\n- mlcourse.ai [video](https://www.youtube.com/watch?v=neXJL-AqI_c) sur le Random Forest\n- Medium [\"story\"](https://medium.com/open-machine-learning-course/open-machine-learning-course-topic-5-ensembles-of-algorithms-and-random-forest-8e05246cbba7) based on this notebook\n- Course materials as a [Kaggle Dataset](https://www.kaggle.com/kashnitsky/mlcourse)\n- Si vous lisez le Russse : un [article](https://habrahabr.ru/company/ods/blog/324402/) sur Habrahabr avec ~ le même material. Et une [video](https://youtu.be/G0DmuuFeC30) on YouTube\n- Chapter 15 of the book “[Elements of Statistical Learning](https://statweb.stanford.edu/~tibs/ElemStatLearn/)” by Jerome H. Friedman, Robert Tibshirani, and Trevor Hastie.\n- More about practical applications of random forests and other algorithms can be found in the [official documentation](http://scikit-learn.org/stable/modules/ensemble.html) of `scikit-learn`.\n- For a more in-depth discussion of variance and decorrelation of random forests, see the [original paper](https://www.stat.berkeley.edu/~breiman/randomforest2001.pdf).",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
]
] |
cb915f98b44cb7f2b0b10eb09661bd6f806e6ccd
| 104,214 |
ipynb
|
Jupyter Notebook
|
4/Q2/Q2.ipynb
|
ellietoulabi/DataMiningCourseHomeworks
|
fd7a76f9e96939c7f81e09bd73e50fa74e75b78f
|
[
"MIT"
] | 2 |
2021-12-07T18:47:43.000Z
|
2022-03-03T12:38:42.000Z
|
4/Q2/Q2.ipynb
|
ellietoulabi/DataMiningCourseHomeworks
|
fd7a76f9e96939c7f81e09bd73e50fa74e75b78f
|
[
"MIT"
] | null | null | null |
4/Q2/Q2.ipynb
|
ellietoulabi/DataMiningCourseHomeworks
|
fd7a76f9e96939c7f81e09bd73e50fa74e75b78f
|
[
"MIT"
] | null | null | null | 84.589286 | 32,892 | 0.748047 |
[
[
[
"## 2. Random Forest",
"_____no_output_____"
],
[
"### a)",
"_____no_output_____"
]
],
[
[
"import pandas as pd",
"_____no_output_____"
],
[
"headers = [\"Number of times pregnant\",\n \"Plasma glucose concentration a 2 hours in an oral glucose tolerance test\",\n \"Diastolic blood pressure (mm Hg)\",\n \"Triceps skinfold thickness (mm)\",\n \"2-Hour serum insulin (mu U/ml)\",\n \"Body mass index (weight in kg/(height in m)^2)\",\n \"Diabetes pedigree function\",\n \"Age (years)\",\n \"Class variable(0 or 1)\"]\n\nraw_df = pd.read_csv('Diabetes.csv', names= headers)\nprint(raw_df.shape)\nraw_df.head(6)",
"(768, 9)\n"
]
],
[
[
"<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n\n<p style =\" direction:rtl;text-align:right;\">\nابتدا یک لیست از اسامی هدر ها ایجاد کرده و سپس فایل csv را به کمک ان فراخوانی می کنیم \n</p>\n\n<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n",
"_____no_output_____"
],
[
"### b)",
"_____no_output_____"
],
[
"#### Handling Missing Values",
"_____no_output_____"
]
],
[
[
"raw_df.isnull().sum()",
"_____no_output_____"
],
[
"raw_df['Class variable(0 or 1)'].value_counts()",
"_____no_output_____"
]
],
[
[
"<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n\n<p style =\" direction:rtl;text-align:right;\">\n همانطور که مشخص است داده missing در دیتاست وجود ندارد.\n همینطور با توجه به محموع تعداد هرکدام از مقادیر یکتا که برابر است به تعداد کل رکوردها می توان فهمید که در ستون کلاس null نداریم.\n این تحلیل را برای سایر ستون ها هم می توان انجام داد.\n</p>\n\n<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n",
"_____no_output_____"
],
[
"#### Normalizing Numerical Columns and Encoding Categorical Ones",
"_____no_output_____"
]
],
[
[
"raw_df.dtypes",
"_____no_output_____"
]
],
[
[
"<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n\n<p style =\" direction:rtl;text-align:right;\">\nمتغیر کتگوریکال نداریم پس نیازی به انکود نیست\n</p>\n\n<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n",
"_____no_output_____"
]
],
[
[
"from sklearn.preprocessing import normalize",
"_____no_output_____"
],
[
"numericals = pd.DataFrame(raw_df.loc[:, raw_df.columns != 'Class variable(0 or 1)'])\ncategoricals = pd.DataFrame(raw_df['Class variable(0 or 1)'])\n\n\n#remove outlires\nprint(\"Before Removing Outliers: \",raw_df.shape)\nQ1 = raw_df.loc[:, raw_df.columns != 'Class variable(0 or 1)'].quantile(0.25)\nQ3 = raw_df.loc[:, raw_df.columns != 'Class variable(0 or 1)'].quantile(0.75)\nIQR = Q3 - Q1 \nmask = ~((raw_df < (Q1 - 1.5 * IQR)) | (raw_df > (Q3 + 1.5 * IQR))).any(axis=1)\nprint(\"#Outliers = \",raw_df[~mask].dropna().shape[0])\nprint(\"#Not outliers = \",raw_df.shape[0]-raw_df[~mask].dropna().shape[0])\n\nraw_df= raw_df[mask]\nprint(\"After Removing Outliers: \",raw_df.shape)\nraw_df.head()\n",
"Before Removing Outliers: (639, 9)\n#Outliers = 48\n#Not outliers = 591\nAfter Removing Outliers: (591, 9)\n"
],
[
"numericals = pd.DataFrame(raw_df.loc[:, raw_df.columns != 'Class variable(0 or 1)'])\ncategoricals = pd.DataFrame(raw_df['Class variable(0 or 1)'])\n\n#normalize\nraw_df.loc[:, raw_df.columns != 'Class variable(0 or 1)'] = normalize(numericals, norm='l2')\n\ndf.loc[:,:]=raw_df\ndf.head()",
"_____no_output_____"
]
],
[
[
"<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n\n<p style =\" direction:rtl;text-align:right;\">\nبا جزئیاتی کاملا مشابه سوال قبل حذف داده های پرت و نرمال سازی انجام شده است.</p>\n\n<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n",
"_____no_output_____"
],
[
"### c)",
"_____no_output_____"
]
],
[
[
"Y = df['Class variable(0 or 1)']\nX = df.drop('Class variable(0 or 1)', axis=1)",
"_____no_output_____"
]
],
[
[
"### d)",
"_____no_output_____"
]
],
[
[
"from sklearn.model_selection import train_test_split",
"_____no_output_____"
],
[
"X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.2)",
"_____no_output_____"
],
[
"print('Distribution of defferent classes in training data:(%) ')\ny = pd.DataFrame(y_train)\n((y.groupby('Class variable(0 or 1)')['Class variable(0 or 1)'].count())/y_train.shape[0])*100",
"Distribution of defferent classes in training data:(%) \n"
],
[
"print('Distribution of defferent classes in test data:(%) ')\ny = pd.DataFrame(y_test)\n((y.groupby('Class variable(0 or 1)')['Class variable(0 or 1)'].count())/y_test.shape[0])*100",
"Distribution of defferent classes in test data:(%) \n"
]
],
[
[
"<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n\n<p style =\" direction:rtl;text-align:right;\">\nبه کمک تابع train_test_split داده ها را به دو بخش آموزشی و تست با نسبت خواسته شده تقسیم می کنیم.سپس درصد وجود هر کلاس را در داده های آموزش و تست محاسبه می کنیم.\n همانطور که مشحص است هر کدام از این دسته ها درصدهای نزدیک به همی در دو نوع داده دارند\n</p>\n\n<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n",
"_____no_output_____"
],
[
"### e)",
"_____no_output_____"
]
],
[
[
"from sklearn.ensemble import RandomForestClassifier",
"_____no_output_____"
],
[
"myclassifier = RandomForestClassifier(max_depth=3, criterion='entropy').fit(X_train, y_train)",
"_____no_output_____"
]
],
[
[
"<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n\n<p style =\" direction:rtl;text-align:right;\">\nکلاسیفایر جنگل تصادفی را با پارامترهای گفته شده در سوال می سازیم .\n سپس داده ها را برای آموزش به مدل دادیم.\n\n</p>\n\n<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n",
"_____no_output_____"
],
[
"### f)",
"_____no_output_____"
]
],
[
[
"from sklearn.metrics import classification_report, confusion_matrix, accuracy_score",
"_____no_output_____"
],
[
"y_predicted = myclassifier.predict(X_test)\nprint(\"[+] confusion matrix\\n\")\nprint(confusion_matrix(y_test, y_predicted))\nprint(\"\\n[+] classification report\\n\")\nprint(classification_report(y_test, y_predicted))",
"[+] confusion matrix\n\n[[72 12]\n [22 13]]\n\n[+] classification report\n\n precision recall f1-score support\n\n 0 0.77 0.86 0.81 84\n 1 0.52 0.37 0.43 35\n\n accuracy 0.71 119\n macro avg 0.64 0.61 0.62 119\nweighted avg 0.69 0.71 0.70 119\n\n"
],
[
"print(\"Mean Accuracy = \",myclassifier.score(X_test,y_test,y_predicted))",
"Mean Accuracy = 0.52\n"
],
[
"print(\"Accuracy Score = \" ,accuracy_score(y_test, y_predicted))",
"Accuracy Score = 0.7142857142857143\n"
]
],
[
[
"<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n\n<p style =\" direction:rtl;text-align:right;\">\nتوضیح مفصل confusion matrix را در بخش های بعد ذکر خواهم کرد.\n اما در این مرحله خطای mse و دقت را به صورت فوق گزارش کرده ام که دقت حوالی 70 است.\n</p>\n\n<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n",
"_____no_output_____"
],
[
"### g)",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport matplotlib.pyplot as plt\nfrom matplotlib.legend_handler import HandlerLine2D",
"_____no_output_____"
],
[
"# ranging from 1 to 32\n\nmax_depths = np.linspace(1, 32, 32, endpoint=True)\ntrain_results = []\ntest_results = []\n\nfor max_depth in max_depths:\n rf = RandomForestClassifier(max_depth=max_depth, criterion='entropy')\n rf.fit(X_train, y_train)\n train_pred = rf.predict(X_train)\n ac = accuracy_score(y_train, train_pred)\n train_results.append(ac)\n y_pred = rf.predict(X_test)\n acc = accuracy_score(y_test, y_pred)\n test_results.append(acc)\n \nfig, axes = plt.subplots(1,1,figsize=(9,9)) \ntrains, = plt.plot(max_depths, train_results, 'b', label='Train Accuracy')\ntests, = plt.plot(max_depths, test_results, 'r', label='Test Accuracy')\nplt.legend(handler_map={trains: HandlerLine2D(numpoints=2)})\nplt.ylabel('Acccuracy Score')\nplt.xlabel('max_depth')\nplt.show()",
"_____no_output_____"
],
[
"from sklearn.metrics import roc_curve, auc\n\nmax_depths = np.linspace(1, 32, 32, endpoint=True)\ntrain_results = []\ntest_results = []\nfor max_depth in max_depths:\n dt = RandomForestClassifier(max_depth=max_depth, criterion='entropy')\n dt.fit(X_train, y_train)\n train_pred = dt.predict(X_train)\n false_positive_rate, true_positive_rate, thresholds = roc_curve(y_train, train_pred)\n roc_auc = auc(false_positive_rate, true_positive_rate)\n train_results.append(roc_auc)\n y_pred = dt.predict(X_test)\n false_positive_rate, true_positive_rate, thresholds = roc_curve(y_test, y_pred)\n roc_auc = auc(false_positive_rate, true_positive_rate)\n test_results.append(roc_auc)\n \nfig, axes = plt.subplots(1,1,figsize=(9,9)) \ntrains, = plt.plot(max_depths, train_results, 'b', label='Train Accuracy')\ntests, = plt.plot(max_depths, test_results, 'r', label='Test Accuracy')\nplt.legend(handler_map={trains: HandlerLine2D(numpoints=2)})\nplt.ylabel('Acccuracy Score')\nplt.xlabel('max_depth')\nplt.show()",
"_____no_output_____"
],
[
"myclassifier = RandomForestClassifier(max_depth=5, criterion='entropy').fit(X_train, y_train)",
"_____no_output_____"
],
[
"y_predicted = myclassifier.predict(X_test)\nprint(\"[+] confusion matrix\\n\")\nprint(confusion_matrix(y_test, y_predicted))\nprint(\"\\n[+] classification report\\n\")\nprint(classification_report(y_test, y_predicted))\nprint(\"Mean Accuracy = \",myclassifier.score(X_test,y_test,y_predicted))\nprint(\"Accuracy Score = \" ,accuracy_score(y_test, y_predicted))",
"[+] confusion matrix\n\n[[80 4]\n [26 9]]\n\n[+] classification report\n\n precision recall f1-score support\n\n 0 0.75 0.95 0.84 84\n 1 0.69 0.26 0.37 35\n\n accuracy 0.75 119\n macro avg 0.72 0.60 0.61 119\nweighted avg 0.74 0.75 0.70 119\n\nMean Accuracy = 0.6923076923076923\nAccuracy Score = 0.7478991596638656\n"
]
],
[
[
"<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n\n<p style =\" direction:rtl;text-align:right;\">\n این پارامتر بیشینه عمق درخت را نمایش می دهد. اگر ست نشود مدل تا جایی ادامه می دهد تا به نودهای تماما خالص برسد یا برگها کمتر ازmin_samples_split باشند\nهمانطور که در نمودار هایی که برای دقت در دو بخش train و test رسم کرده ایم مشخص است بعد از مدتی ( حوالی 10 )\nمدل ما با خطر overfit شدن مواجه می شود.\nیعنی برای داده های آموزش نتایج بسیار خوبی حاصل می شود اما قادر به پیش بینی داده های جدید مخواهد بود.\nبه همین علت باید این پارامتر را به صورتی مقداررهی کنیم که نه این اتفاق رخ بدهد نه اینکه مدل به علت مقدار کم آن به خوبی آموزش نبیند.\nپس این مقدار را برابر 5 قرارداده ام که قبل از بیش برازش است و دقت تست خوبی هم دارد.\n</p>\n\n<hr style = \"border-top: 3px solid #000000 ; border-radius: 3px;\">\n",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
]
] |
cb9166a423d0dec44280fd6533489c41219342c2
| 20,766 |
ipynb
|
Jupyter Notebook
|
notebooks/chierici_practical_part2.ipynb
|
mchierici/embo2019popgen
|
1738a13d88cc37b43816b1f04e49890f84422858
|
[
"MIT"
] | null | null | null |
notebooks/chierici_practical_part2.ipynb
|
mchierici/embo2019popgen
|
1738a13d88cc37b43816b1f04e49890f84422858
|
[
"MIT"
] | null | null | null |
notebooks/chierici_practical_part2.ipynb
|
mchierici/embo2019popgen
|
1738a13d88cc37b43816b1f04e49890f84422858
|
[
"MIT"
] | null | null | null | 27.762032 | 431 | 0.594144 |
[
[
[
"<div>\n <h1 style=\"text-align: center;\">Machine learning from scratch - Part II</h1>\n <h2 style=\"text-align: center;\">EMBO practical course on population genomics 2019 @ Procida, Italy</h2>\n<div>\n\n---\n\n### Authors: Marco Chierici & Margherita Francescatto\n### _FBK/MPBA_\n\n---",
"_____no_output_____"
],
[
"**Recap.** We are using a subset of the SEQC neuroblastoma data set [Zhang et al, Genome Biology, 2015] consisting of 272 samples (136 training, 136 test). The data was preprocessed a bit to facilitate the progress of the tutorial.",
"_____no_output_____"
],
[
"We start by loading the modules we need to process the data.",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport pylab as pl ## for plotting\nimport pandas as pd ## for reading text files and manipulating data frames\nfrom sklearn import neighbors ## kNN classifier\nfrom sklearn import svm ## SVM classifier\nfrom sklearn.ensemble import RandomForestClassifier ## RF classifier\nfrom sklearn.model_selection import cross_val_score ## needed to train in CV\n%matplotlib inline\nnp.random.seed(42) ## set random seed just in case",
"_____no_output_____"
]
],
[
[
"Define files to read:",
"_____no_output_____"
]
],
[
[
"## for convenience, define the data directory as a variable\nDATA_DIR = \"../data/\" #\"/path/to/your/data\"",
"_____no_output_____"
],
[
"DATA_TR = DATA_DIR + \"MAV-G_272_tr.txt.gz\"\nDATA_TS = DATA_DIR + \"MAV-G_272_ts.txt.gz\"\nLABS_TR = DATA_DIR + \"labels_tr.txt\"\nLABS_TS = DATA_DIR + \"labels_ts.txt\"",
"_____no_output_____"
]
],
[
[
"Read in the files as _pandas dataframes_ (they are conceptually like R data frames):",
"_____no_output_____"
]
],
[
[
"data_tr = pd.read_csv(DATA_TR, sep = \"\\t\")\ndata_ts = pd.read_csv(DATA_TS, sep = \"\\t\")",
"_____no_output_____"
]
],
[
[
"Since we already looked at the data in the first part of the dataset, we move directly to the juicy stuff.",
"_____no_output_____"
],
[
"We drop the first column from the train and test expression sets, since it's just the sample IDs...",
"_____no_output_____"
]
],
[
[
"data_tr = data_tr.drop('sampleID',axis =1)\ndata_ts = data_ts.drop('sampleID',axis =1)",
"_____no_output_____"
]
],
[
[
"...and store the data into Numpy arrays.",
"_____no_output_____"
]
],
[
[
"x_tr = data_tr.values\nx_ts = data_ts.values",
"_____no_output_____"
]
],
[
[
"Now we read in the files containing labels and select the column with the CLASS target to do our first round of analyses.",
"_____no_output_____"
]
],
[
[
"labs_tr = pd.read_csv(LABS_TR, sep = \"\\t\")\nlabs_ts = pd.read_csv(LABS_TS, sep = \"\\t\")\nclass_lab_tr = labs_tr[['CLASS']]\nclass_lab_ts = labs_ts[['CLASS']]\ny_tr = class_lab_tr.values.ravel()\ny_ts = class_lab_ts.values.ravel()",
"_____no_output_____"
]
],
[
[
"In the previous part of the tutorial, we focused on the k-NN classifiers. In the previous lecture, however, we explored theoretical aspects related to two other broadly used classifiers: Support Vector Machines (SVMs) and Random Forests (RFs). In this second part of tutorial, the first thing we want to do is assessing how these two alternative classification methods perform on our neuroblastoma dataset.",
"_____no_output_____"
],
[
"We start with SVM. We first rescale the data, import the relevant model and create an instance of the SVM classifier.",
"_____no_output_____"
]
],
[
[
"from sklearn.preprocessing import MinMaxScaler\n## first you need to create a \"scaler\" object\nscaler = MinMaxScaler(feature_range=(-1,1))\n## then you actually scale data by fitting the scaler object on the data\nscaler.fit(x_tr)\nx_tr = scaler.transform(x_tr)\nx_ts = scaler.transform(x_ts)",
"_____no_output_____"
],
[
"## import support vector classifier (SVC) and create an instance\nfrom sklearn.svm import SVC\nsvc = SVC(random_state=42, verbose=1, kernel='linear')",
"_____no_output_____"
]
],
[
[
"Note that the specification _kernel = 'linear'_ implies that a linear kernel will be used. If you remember from the lecture, this means that a linear function is used to define the decision boundaries of the classifier. Alternatives include _‘poly’_ and _‘rbf’_ for polynomial or gaussian kernels respectively. Scikit-learn offers an alternative implementation of linear SVMs. You can find more details in Scikit User Guide. ",
"_____no_output_____"
],
[
"As previously done with the k-NN classifier, we fit the SVM and get the predictions for the test data.",
"_____no_output_____"
]
],
[
[
"## fit the model and get the predictions\nsvc.fit(x_tr, y_tr)\nclass_pred_ts = svc.predict(x_ts)",
"_____no_output_____"
]
],
[
[
"Now we give a look at the classification metrics introduced in the first part of the tutorial. to access the functions, we need to load the metrics module.",
"_____no_output_____"
]
],
[
[
"from sklearn import metrics\nprint('MCC = ', metrics.matthews_corrcoef(class_lab_ts, class_pred_ts))\nprint('ACC = ', metrics.accuracy_score(class_lab_ts, class_pred_ts))\nprint('SENS = ', metrics.recall_score(class_lab_ts, class_pred_ts))",
"_____no_output_____"
]
],
[
[
"We can also give a look at the classification report.",
"_____no_output_____"
]
],
[
[
"print(metrics.classification_report(class_lab_ts, class_pred_ts))",
"_____no_output_____"
]
],
[
[
"Exercise: **one-shot Random Forest classification**. _Hint:_ the RF classifier is implemented in the Scikit learn class RandomForestClassifier, from _sklearn.ensemble_ module.",
"_____no_output_____"
]
],
[
[
"## space for exercise\n\nfrom sklearn.ensemble import RandomForestClassifier as RFC\nclf = RFC(n_estimators=500)\nclf.fit(x_tr, y_tr)\ny_pred_rfc = clf.predict(x_ts)\nprint('MCC = ', metrics.matthews_corrcoef(class_lab_ts, y_pred_rfc))\nprint('ACC = ', metrics.accuracy_score(class_lab_ts, y_pred_rfc))\nprint('SENS = ', metrics.recall_score(class_lab_ts, y_pred_rfc))\nprint(metrics.classification_report(class_lab_ts, y_pred_rfc))",
"_____no_output_____"
]
],
[
[
"## Parameter tuning",
"_____no_output_____"
],
[
"As mentioned in the lecture, Scikit learn offers a very useful and flexible tool for parameter tuning called _GridSearchCV_. While the tool is very sophisticated and efficient, it is useful to at least try an example _by hand_ to understand what is happening in the background.\n\nFor this example we use a linear SVM and try to tune the C parameter. You might remember from the lectures that the paramenter C essentially controls how much we want to avoid misclassifying each training example. Large values of C result in smaller margins, i.e. closer fitting to the training data. As mentioned in the classes, the drawback is over-fitting, resulting in poor generalization.",
"_____no_output_____"
]
],
[
[
"## define the sequence of C values we want to use in the search of the best one\nC_list = [0.000001, 0.00001, 0.0001, 0.001, 0.01, 0.1]\nfor C in C_list:\n print('C = ', C)\n svc = svm.SVC(kernel='linear', C=C)\n svc.fit(x_tr, class_lab_tr.values.ravel())\n class_pred_ts = svc.predict(x_ts)\n print('MCC = ', metrics.matthews_corrcoef(class_lab_ts, class_pred_ts))\n print('ACC = ', metrics.accuracy_score(class_lab_ts, class_pred_ts))\n print('SENS = ', metrics.recall_score(class_lab_ts, class_pred_ts), \"\\n\")",
"_____no_output_____"
]
],
[
[
"From C = 1e-03 the classification performance reaches a plateau. C = 1e-04 yields the highest MCC on the test set: when tuning the parameters we would consider this as the best choice for the problem.",
"_____no_output_____"
],
[
"**Exercise:** as you already saw in the lectures, there are many parameters that can be tuned, also when considering only one simple classifier. For example, if you consider SVM with 'rbf' kernel, you could check performance changes with different values of C **and** gamma, for example using two nested loops.",
"_____no_output_____"
]
],
[
[
"## space for exercise",
"_____no_output_____"
]
],
[
[
"As we mentioned, Scikit offers fully automated parameter tuning engine. We illustrate its power with a simple example on our data. We use GridSearchCV to search through a grid of C and gamma parameter options for SVM with 'rbf' kernel. In order to do this we define a small function svc_param_selection that does the work for us.",
"_____no_output_____"
]
],
[
[
"from sklearn.model_selection import GridSearchCV\n\ndef svc_param_selection(X, y, nfolds):\n Cs = [0.001, 0.01, 0.1, 1, 10]\n gammas = [0.001, 0.01, 0.1, 1, 'auto']\n param_grid = {'C': Cs, 'gamma' : gammas}\n grid_search = GridSearchCV(svm.SVC(kernel='rbf'), param_grid, cv=nfolds, n_jobs=4)\n grid_search.fit(X, y)\n grid_search.best_params_\n return grid_search.best_params_\n\nsvc_param_selection(x_tr, y_tr, 5)",
"_____no_output_____"
]
],
[
[
"## Feature ranking",
"_____no_output_____"
],
[
"As mentioned in the lecture, random forests have a built in tool for feature ranking",
"_____no_output_____"
]
],
[
[
"# Build a forest and compute the feature importances\nrf = RandomForestClassifier(n_estimators=250)\nrf.fit(x_tr, y_tr)",
"_____no_output_____"
]
],
[
[
"For the sake of completeness make the predictions and check the classification performance.",
"_____no_output_____"
]
],
[
[
"class_pred_ts = rf.predict(x_ts)\nprint('MCC = ', metrics.matthews_corrcoef(class_lab_ts, class_pred_ts))\nprint('ACC = ', metrics.accuracy_score(class_lab_ts, class_pred_ts))\nprint('SENS = ', metrics.recall_score(class_lab_ts, class_pred_ts))",
"_____no_output_____"
]
],
[
[
"Now extract the feature importances and display the first 10:",
"_____no_output_____"
]
],
[
[
"importances = rf.feature_importances_\nindices = np.argsort(importances)[::-1]\n\n# Print the feature ranking\nprint(\"Feature ranking (top 10 features):\")\nfor f in range(10):\n print(\"%d. feature %d (%f)\" % (f + 1, indices[f], importances[indices[f]]))",
"_____no_output_____"
]
],
[
[
"Would be nice to know to which genes they actually correspond. If you remember the gene names are the column names of the pandas dataframe containing the training/test data.",
"_____no_output_____"
]
],
[
[
"columnsNamesArr = data_tr.columns.values\nfor i in range(10):\n print(columnsNamesArr[indices[i]])",
"_____no_output_____"
]
],
[
[
"## Extra exercises",
"_____no_output_____"
],
[
"The classification task considered so far (CLASS) is quite easy, since the classes reflect extreme disease outcomes (favorable vs unfavorable).\n\nA more interesting task could be the prediction of Event-Free Survival (EFS). To do this, an extended version of the dataset is provided in the `/data/marco` directory:",
"_____no_output_____"
]
],
[
[
"DATA_TR = DATA_DIR + \"MAV-G_498_tr.txt.gz\"\nDATA_TS = DATA_DIR + \"MAV-G_498_ts.txt.gz\"\nLABS_TR = DATA_DIR + \"labels_full_tr.txt\"\nLABS_TS = DATA_DIR + \"labels_full_ts.txt\"",
"_____no_output_____"
]
],
[
[
"Read the data in and prepare the `x_tr`, `x_ts`, `y_tr`, `y_ts` Numpy arrays, as before, using \"EFS\" as target variable.",
"_____no_output_____"
],
[
"Recalling concepts from the first practical, perform an explorative PCA analyisis, plotting the first two components.",
"_____no_output_____"
],
[
"Train a kNN classifier in one-shot mode: fit the model on the training set and predict the labels on the test set. Compute performance metrics using the provided true labels of the test set.",
"_____no_output_____"
],
[
"Experiment with different classifier(s) and/or different parameters.",
"_____no_output_____"
],
[
"Try tuning the parameters (e.g. using GridSearchCV) and find the optimal parameter set.",
"_____no_output_____"
],
[
"Using the optimal parameters, run a (iterated) cross-validation on the training set; compute the average cross-validation metrics.",
"_____no_output_____"
],
[
"Using the optimal parameters, train a model on the whole training set and predict the labels of the test set. Compute the metrics and compare them with the average cross-validation metrics. What do you expect? Use the trained model to rank the features and inspect the top ones.",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
]
] |
cb9171410898a6c3f1ee9ae2d279d615e25acbb0
| 2,128 |
ipynb
|
Jupyter Notebook
|
doc/docstrings/set_context.ipynb
|
hercules261188/seaborn
|
f0b2537d21d1122e8fead3576d15619ceaf8e422
|
[
"BSD-3-Clause"
] | 8,852 |
2015-01-02T15:44:51.000Z
|
2022-03-31T10:52:24.000Z
|
doc/docstrings/set_context.ipynb
|
hercules261188/seaborn
|
f0b2537d21d1122e8fead3576d15619ceaf8e422
|
[
"BSD-3-Clause"
] | 2,330 |
2015-01-02T00:23:18.000Z
|
2022-03-31T12:36:51.000Z
|
doc/docstrings/set_context.ipynb
|
hercules261188/seaborn
|
f0b2537d21d1122e8fead3576d15619ceaf8e422
|
[
"BSD-3-Clause"
] | 2,044 |
2015-01-02T18:31:12.000Z
|
2022-03-31T10:38:55.000Z
| 20.266667 | 84 | 0.530545 |
[
[
[
"import seaborn as sns",
"_____no_output_____"
]
],
[
[
"Call the function with the name of a context to set the default for all plots:",
"_____no_output_____"
]
],
[
[
"sns.set_context(\"notebook\")\nsns.lineplot(x=[0, 1, 2], y=[1, 3, 2])",
"_____no_output_____"
]
],
[
[
"You can independently scale the font elements relative to the current context:",
"_____no_output_____"
]
],
[
[
"sns.set_context(\"notebook\", font_scale=1.25)\nsns.lineplot(x=[0, 1, 2], y=[1, 3, 2])",
"_____no_output_____"
]
],
[
[
"It is also possible to override some of the parameters with specific values:",
"_____no_output_____"
]
],
[
[
"sns.set_context(\"notebook\", rc={\"lines.linewidth\": 3})\nsns.lineplot(x=[0, 1, 2], y=[1, 3, 2])",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb9175d6d04d3d371418f0f75b01d78bd3248610
| 184,405 |
ipynb
|
Jupyter Notebook
|
write_to_vtk/frozen_in_electron_assumption_does_not_work.ipynb
|
jensv/relative_canonical_helicity_tools
|
51dc408b5d547c3122b77b0058eab32a3fca2d4c
|
[
"MIT"
] | null | null | null |
write_to_vtk/frozen_in_electron_assumption_does_not_work.ipynb
|
jensv/relative_canonical_helicity_tools
|
51dc408b5d547c3122b77b0058eab32a3fca2d4c
|
[
"MIT"
] | null | null | null |
write_to_vtk/frozen_in_electron_assumption_does_not_work.ipynb
|
jensv/relative_canonical_helicity_tools
|
51dc408b5d547c3122b77b0058eab32a3fca2d4c
|
[
"MIT"
] | null | null | null | 158.832903 | 18,976 | 0.852282 |
[
[
[
"import numpy as np\nfrom scipy.spatial import Delaunay\nfrom scipy.interpolate import LinearNDInterpolator\nfrom scipy.constants import mu_0\nfrom scipy.constants import elementary_charge as q_e\nfrom scipy.constants import proton_mass as m_i\nfrom astropy.convolution import convolve, convolve_fft\nfrom scipy.signal import fftconvolve\nfrom scipy.interpolate import SmoothBivariateSpline\nimport write_canonical_flux_tube_quantities as wcf\nreload(wcf)\n\nfrom datetime import date\nfrom datetime import datetime\nimport visit_writer\n\nimport structured_3d_vtk as struc_3d\nreload(struc_3d)\n\n%matplotlib inline\nimport matplotlib.pyplot as plt\n\nimport seaborn as sns\nsns.set_style('white')\n\nimport os\n\nimport ion_current_to_mach_number as ic_to_mach\nreload(ic_to_mach)\n\nimport read_from_sql\n\nfrom mpl_toolkits.mplot3d import Axes3D",
"_____no_output_____"
]
],
[
[
"# Read idl files and generate inteprolators",
"_____no_output_____"
]
],
[
[
"bx_all_planes = wcf.save_idl_quantity_to_unstructured_grids('bx', 'B_x', 'now',\n x_min=-0.032, x_max=0.028,\n y_min=-0.022, y_max=0.032, \n z_min=0.249, z_max=0.416)\nby_all_planes = wcf.save_idl_quantity_to_unstructured_grids('by', 'B_y', 'now',\n x_min=-0.032, x_max=0.028,\n y_min=-0.022, y_max=0.032, \n z_min=0.249, z_max=0.416)\nbz_all_planes = wcf.save_idl_quantity_to_unstructured_grids('bz', 'B_z', 'now',\n x_min=-0.032, x_max=0.028,\n y_min=-0.022, y_max=0.032, \n z_min=0.249, z_max=0.416)\nte_all_planes = wcf.save_idl_quantity_to_unstructured_grids('te', 'T_e', 'now', \n x_min=-0.026, x_max=0.028,\n y_min=-0.03, y_max=0.028, \n z_min=0.249, z_max=0.416,\n bounds=(1e-3, 1e3))\nn_all_planes = wcf.save_idl_quantity_to_unstructured_grids('n', 'n', 'now',\n x_min=-0.026, x_max=0.028,\n y_min=-0.03, y_max=0.028, \n z_min=0.249, z_max=0.416,\n bounds=(1e3, 1e22))\nn_three_planes = wcf.remove_plane(0.302, n_all_planes)\nte_three_planes = wcf.remove_plane(0.302, te_all_planes)\n\nn_all_planes = n_three_planes\nte_all_planes = te_three_planes\n\nbx_triangulation, bx_interpolators = wcf.give_delaunay_and_interpolator(bx_all_planes)\nby_triangulation, by_interpolators = wcf.give_delaunay_and_interpolator(by_all_planes)\nbz_triangulation, bz_interpolators = wcf.give_delaunay_and_interpolator(bz_all_planes)\nte_triangulation, te_interpolators = wcf.give_delaunay_and_interpolator(te_all_planes)\nn_triangulation, n_interpolators = wcf.give_delaunay_and_interpolator(n_all_planes)",
"_____no_output_____"
]
],
[
[
"# Read Mach measurements from database and build interpolators",
"_____no_output_____"
]
],
[
[
"timesteps = 250\n\ndatabase = '/home/jensv/rsx/jens_analysis/shots_database/source/shots.db'\ntable = 'Shots'\nz_direction_1, z_direction_2 = 0, 180\ny_direction_1, y_direction_2 = 90, 270\nangle_signs = {0: 1,\n 180: -1,\n 90: -1,\n 0: 1}\nmin_spectral_density = 1.6e-8\n\ncondition_z_0416 = (\"campaigns = 'mach_probe_plane_campaign_1'\"\n \" AND fiducial_pre_crowbar_gyration_spectral_density > \"\n + str(min_spectral_density) +\n \" AND mach_signals_exist = 1\"\n \" AND (mach_orientation = \" + str(z_direction_1) +\n \" OR mach_orientation = \" + str(z_direction_2) + \")\")\n\ncondition_y_0416 = (\"campaigns = 'mach_probe_plane_campaign_1'\"\n \" AND fiducial_pre_crowbar_gyration_spectral_density > \"\n + str(min_spectral_density) +\n \" AND mach_signals_exist = 1\"\n \" AND (mach_orientation = \" + str(y_direction_1) +\n \" OR mach_orientation = \" + str(y_direction_2) + \")\")\n\ncursor, connection = read_from_sql.cursor_with_rows(condition_z_0416,\n database,\n table)\nz_0416_shots = cursor.fetchall()\ncursor.close()\nconnection.close()\n\ncursor, connection = read_from_sql.cursor_with_rows(condition_y_0416,\n database,\n table)\ny_0416_shots = cursor.fetchall()\ncursor.close()\nconnection.close()\n\ncondition_z_302 = (\"campaigns = 'mach_probe_plane_campaign_2'\"\n \" AND fiducial_pre_crowbar_gyration_spectral_density > \"\n + str(min_spectral_density) +\n \" AND mach_signals_exist = 1\"\n \" AND (mach_orientation = \" + str(z_direction_1) +\n \" OR mach_orientation = \" + str(z_direction_2) + \")\")\n\ncursor, connection = read_from_sql.cursor_with_rows(condition_z_302,\n database,\n table)\nz_0302_shots = cursor.fetchall()\ncursor.close()\nconnection.close()\n\nmach_z_0416_measurements = ic_to_mach.run_mach_analysis(z_0416_shots,\n timesteps, \n angle_signs) \nmach_y_0416_measurements = ic_to_mach.run_mach_analysis(y_0416_shots,\n timesteps, \n angle_signs)\nmach_z_0302_measurements = ic_to_mach.run_mach_analysis(z_0302_shots,\n timesteps, \n angle_signs)\n\nmach_z_0416_measurements['delays'] = np.arange(timesteps)\nmach_y_0416_measurements['delays'] = np.arange(timesteps)\nmach_z_0302_measurements['delays'] = np.arange(timesteps)\n\nmach_z_0416_measurements = struc_3d.average_duplicate_points(mach_z_0416_measurements)\nmach_y_0416_measurements = struc_3d.average_duplicate_points(mach_y_0416_measurements)\nmach_z_0302_measurements = struc_3d.average_duplicate_points(mach_z_0302_measurements)\n\nmach_y_measurements = {0.416: mach_y_0416_measurements}\nmach_z_measurements = {0.302: mach_z_0302_measurements,\n 0.416: mach_z_0416_measurements}\n\nmach_y_all_planes = wcf.save_quantity_to_unstructured_grids(mach_y_measurements, \n 'Mach_y', 'Mach_y', '2016-07-26',\n planes=[0.416],\n x_min=-0.052, x_max=0.052,\n y_min=-0.022, y_max=0.032, \n z_min=0.249, z_max=0.416,\n bounds=(-10, 10))\n\nmach_z_all_planes = wcf.save_quantity_to_unstructured_grids(mach_z_measurements, \n 'Mach_z', 'Mach_z', '2016-07-26',\n planes=[0.302, 0.416],\n x_min=-0.032, x_max=0.032,\n y_min=-0.022, y_max=0.032, \n z_min=0.249, z_max=0.416,\n bounds=(-10, 10))\nmach_y_all_planes = wcf.remove_nan_points(mach_y_all_planes)\nmach_z_all_planes = wcf.remove_nan_points(mach_z_all_planes)\n\nmach_y_triangulation, mach_y_interpolators = wcf.give_delaunay_and_interpolator(mach_y_all_planes)\nmach_z_triangulation, mach_z_interpolators = wcf.give_delaunay_and_interpolator(mach_z_all_planes)",
"_____no_output_____"
]
],
[
[
"# Just work on timestep 0",
"_____no_output_____"
]
],
[
[
"time_point = 0",
"_____no_output_____"
]
],
[
[
"# Interpolate Temperature and density",
"_____no_output_____"
]
],
[
[
"(x_min, x_max, \n y_min, y_max,\n z_min, z_max) = wcf.joint_mach_bdot_tp_extent()\n\nspatial_increment = 0.001\nmesh = np.meshgrid(np.linspace(x_min, x_max, np.ceil((x_max-x_min)/spatial_increment)),\n np.linspace(y_min, y_max, np.ceil((y_max-y_min)/spatial_increment)),\n np.linspace(z_min, z_max, np.ceil((z_max-z_min)/spatial_increment)))\n\nmesh_wo_edges = wcf.remove_edges_mesh([np.array(mesh[0]), \n np.array(mesh[1]), \n np.array(mesh[2])])\nones = np.ones(mesh_wo_edges[0].shape)\n\n\nte_interpolator = te_interpolators[time_point]\nn_interpolator = n_interpolators[time_point]\n \ntemperature = wcf.scalar_on_mesh(te_interpolator, mesh_wo_edges)\ndensity = wcf.scalar_on_mesh(n_interpolator, mesh_wo_edges)\nb_field, b_field_norm = wcf.b_field_on_mesh([bx_interpolators[time_point], \n by_interpolators[time_point],\n bz_interpolators[time_point]], \n mesh_wo_edges, \n bias=2e-2)",
"_____no_output_____"
],
[
"temperature_plane_normalized = temperature / np.nanmax(np.nanmax(temperature, 0), 0)[None, None, :]",
"_____no_output_____"
]
],
[
[
"# Determine ion_velocity_term1",
"_____no_output_____"
]
],
[
[
"alpha = 1\nfilter_width = 15\n\n(x_min, x_max, \n y_min, y_max,\n z_min, z_max) = wcf.joint_mach_bdot_tp_extent()\n\nspatial_increment = 0.001\nmesh = np.meshgrid(np.linspace(x_min, x_max, np.ceil((x_max-x_min)/spatial_increment)),\n np.linspace(y_min, y_max, np.ceil((y_max-y_min)/spatial_increment)),\n np.linspace(z_min, z_max, np.ceil((z_max-z_min)/spatial_increment)))\n\nmesh_wo_edges = wcf.remove_edges_mesh([np.array(mesh[0]), \n np.array(mesh[1]), \n np.array(mesh[2])])\nones = np.ones(mesh_wo_edges[0].shape)\n\ntime_point = 200\n\nbx_interpolator = bx_interpolators[time_point]\nby_interpolator = by_interpolators[time_point]\nbz_interpolator = bz_interpolators[time_point]\nte_interpolator = te_interpolators[time_point]\nn_interpolator = n_interpolators[time_point]\n\nbx_derivative = wcf.triangulate_derivatives(mesh, bx_triangulation, bx_interpolator, \n increment=0.0000001)\nbx_derivative = wcf.remove_edges_derivative_meshes(bx_derivative)\nby_derivative = wcf.triangulate_derivatives(mesh, by_triangulation, by_interpolator, \n increment=0.0000001)\nby_derivative = wcf.remove_edges_derivative_meshes(by_derivative)\nbz_derivative = wcf.triangulate_derivatives(mesh, bz_triangulation, bz_interpolator, \n increment=0.0000001)\nbz_derivative = wcf.remove_edges_derivative_meshes(bz_derivative)\n\n\ncurrent = wcf.current_on_mesh([bx_derivative, \n by_derivative, \n bz_derivative])\nb_field, b_field_norm = wcf.b_field_on_mesh([bx_interpolator, \n by_interpolator,\n bz_interpolator], \n mesh_wo_edges, \n bias=2e-2)\n\ntemperature = wcf.scalar_on_mesh(te_interpolator, mesh_wo_edges)\ndensity = wcf.scalar_on_mesh(n_interpolator, mesh_wo_edges)\n\ncurrent = np.asarray(current)\ndensity = np.asarray(density)\nb_field_norm = np.asarray(b_field_norm)\n\ndensity = wcf.boxcar_filter_quantity_mesh(density, filter_width)\n\nfor direction in xrange(len(current)):\n current[direction] = wcf.boxcar_filter_quantity_mesh(current[direction], filter_width)\n\ndensity_constant = 1e18*np.ones(density.shape)\n\nion_velocity_term_1 = wcf.calc_ion_velocity_term_1(current, density, q_e)\nion_velocity_term_1_constant_density = wcf.calc_ion_velocity_term_1(current, density_constant, q_e)\nion_velocity_term_2 = wcf.calc_ion_velocity_term_2(b_field_norm, alpha)\n\nion_vorticity_term_1 = wcf.calc_ion_vorticity_term_1(current, density, q_e, mesh_wo_edges)\nion_vorticity_term_1_constant_density = wcf.calc_ion_vorticity_term_1(current, density_constant, q_e, mesh_wo_edges)\nion_vorticity_term_2 = wcf.calc_ion_vorticity_term_2(b_field_norm, alpha, mesh_wo_edges)\n\nmesh_wo_edges = wcf.remove_edges_mesh([np.array(mesh[0]), \n np.array(mesh[1]), \n np.array(mesh[2])])",
"_____no_output_____"
]
],
[
[
"# Determine measured ion_velocity u_i",
"_____no_output_____"
]
],
[
[
"(x_min, x_max, \n y_min, y_max,\n z_min, z_max) = wcf.joint_mach_bdot_tp_extent()\n\nspatial_increment = 0.001\nmesh = np.meshgrid(np.linspace(x_min, x_max, np.ceil((x_max-x_min)/spatial_increment)),\n np.linspace(y_min, y_max, np.ceil((y_max-y_min)/spatial_increment)),\n np.linspace(z_min, z_max, np.ceil((z_max-z_min)/spatial_increment)))\n\nmach_y_interpolator = mach_y_interpolators[0]\nmach_z_interpolator = mach_z_interpolators[0]\nte_interpolator = te_interpolators[0]\n \nmach_y = wcf.scalar_on_mesh(mach_y_interpolator, mesh[:2])\nmach_z = wcf.scalar_on_mesh(mach_z_interpolator, mesh)\nte = wcf.scalar_on_mesh(te_interpolator, mesh)\n\nu_i_y = np.sqrt(te*q_e/m_i)*mach_y\nu_i_z = np.sqrt(te*q_e/m_i)*mach_z\n\nu_i_y = np.reshape(u_i_y, mesh[0].shape)\nu_i_z = np.reshape(u_i_z, mesh[0].shape)\n\nu_i_y = wcf.remove_edges_scalar_quantity_meshes(u_i_y)\nu_i_z = wcf.remove_edges_scalar_quantity_meshes(u_i_z)",
"_____no_output_____"
],
[
"np.nanmax(u_i_y)",
"_____no_output_____"
],
[
"np.sum(np.isnan(u_i_z[:,:,-1]))",
"_____no_output_____"
],
[
"alpha_from_y = (u_i_y - ion_velocity_term_1[1])/ b_field_norm[1]\nalpha_from_z = (u_i_z - ion_velocity_term_1[2])/ b_field_norm[2]\n\nalpha_from_y_flattened_z04 = alpha_from_y[:,:,-1].ravel()\nalpha_from_z_flattened_z04 = alpha_from_z[:,:,-1].ravel()\npoints_x = mesh_wo_edges[0][:, :, -1].ravel()\npoints_y = mesh_wo_edges[1][:, :, -1].ravel()\npoints_z_z04 = mesh[2][0,0,-1]*np.ones(points_x.shape)\nstd_z04 = np.expand_dims(np.zeros(points_x.shape), 0)\n\ndata_y_z04 = {'a_out': np.expand_dims(alpha_from_y_flattened_z04, 0),\n 'x_out': points_x,\n 'y_out': points_y,\n 'z_out': points_z_z04,\n 'std': std_z04,\n 'delays': np.asarray([0])\n }\ndata_z_z04 = {'a_out': np.expand_dims(alpha_from_z_flattened_z04, 0),\n 'x_out': points_x,\n 'y_out': points_y,\n 'z_out': points_z_z04,\n 'std': std_z04,\n 'delays': np.asarray([0])\n }\n\ndata_y_z04 = wcf.remove_nan_points(data_y_z04)\ndata_z_z04 = wcf.remove_nan_points(data_z_z04)\n\npositions_to_remove = np.where(np.abs(data_y_z04['a_out'][0]) > np.abs(np.nanmean(data_y_z04['a_out'][0])*10000))\n\ndata_y_z04 = wcf.remove_positions(data_y_z04, positions_to_remove)\n\nalpha_z_z04_triangulation, alpha_z_z04_interpolators = wcf.give_delaunay_and_interpolator(data_z_z04)\nalpha_y_z04_triangulation, alpha_y_z04_interpolators = wcf.give_delaunay_and_interpolator(data_y_z04)",
"_____no_output_____"
],
[
"alpha_z_test = wcf.scalar_on_mesh(alpha_z_z04_interpolators[0], [mesh_wo_edges[0], mesh_wo_edges[1]])",
"_____no_output_____"
],
[
"alpha_z_test[:,:,0]",
"_____no_output_____"
],
[
"alpha_y_test = wcf.scalar_on_mesh(alpha_y_z04_interpolators[0], [mesh_wo_edges[0], mesh_wo_edges[1]])",
"_____no_output_____"
],
[
"alpha_y_test[:,:,0]",
"_____no_output_____"
]
],
[
[
"# Compare contour plots of ion velocities",
"_____no_output_____"
]
],
[
[
"plt.contourf(mesh_wo_edges[0][:,:,0], mesh_wo_edges[1][:,:,0],\n u_i_z[:,:,-1])\nplt.colorbar()\nplt.show()",
"_____no_output_____"
],
[
"plt.contourf(mesh_wo_edges[0][:,:,0], mesh_wo_edges[1][:,:,0],\n b_field_norm[2][:, :, -1]*alpha_test[:,:,-1]\n + ion_velocity_term_1[2][:, :, -1])\nplt.colorbar()\nplt.show()",
"_____no_output_____"
]
],
[
[
"# Now look at y using alpha_z interpolation",
"_____no_output_____"
]
],
[
[
"plt.contourf(mesh_wo_edges[0][:,:,0], mesh_wo_edges[1][:,:,0],\n u_i_y[:,:,-1])\nplt.colorbar()\nplt.show()",
"_____no_output_____"
],
[
"plt.contourf(mesh_wo_edges[0][:,:,0], mesh_wo_edges[1][:,:,0],\n b_field_norm[1][:, :, -1]*alpha_test[:,:,-1]\n + ion_velocity_term_1[1][:, :, -1])\nplt.colorbar()\nplt.show()",
"_____no_output_____"
]
],
[
[
"# Now look at u_y using alpha_y interpolation",
"_____no_output_____"
]
],
[
[
"plt.contourf(mesh_wo_edges[0][:,:,0], mesh_wo_edges[1][:,:,0],\n u_i_y[:,:,-1])\nplt.colorbar()\nplt.show()",
"_____no_output_____"
],
[
"plt.contourf(mesh_wo_edges[0][:,:,0], mesh_wo_edges[1][:,:,0],\n b_field_norm[1][:, :, -1]*alpha_y_test[:,:,-1]\n + ion_velocity_term_1[1][:, :, -1])\nplt.colorbar()\nplt.show()",
"_____no_output_____"
],
[
"alpha_y_test_smooth = wcf.boxcar_filter_quantity_mesh(alpha_y_test, 15)",
"_____no_output_____"
],
[
"plt.contourf(mesh_wo_edges[0][:,:,0], mesh_wo_edges[1][:,:,0],\n b_field_norm[0][:, :, -1]*alpha_y_test[:,:,-1]\n + ion_velocity_term_1[0][:, :, -1], vmax=1e5)\n\nplt.show()",
"_____no_output_____"
],
[
"plt.contourf(mesh_wo_edges[0][:,:,0], mesh_wo_edges[1][:,:,0],\n b_field_norm[0][:, :, -1]*alpha_y_test_smooth[:,:,-1]\n + ion_velocity_term_1[0][:, :, -1])\nplt.colorbar()\nplt.show()",
"_____no_output_____"
],
[
"np.max(ion_velocity_term_1[0][:, :, -1])",
"_____no_output_____"
],
[
"np.max(ion_velocity_term_1[1][:, :, -1])",
"_____no_output_____"
],
[
"np.max(b_field_norm[1][:, :, -1])",
"_____no_output_____"
],
[
"np.max(b_field_norm[0][:, :, -1])",
"_____no_output_____"
],
[
"np.nanmean(ion_velocity_term_1[0][:, :, -1] / ion_velocity_term_1[1][:, :, -1])",
"_____no_output_____"
],
[
"np.nanmax(b_field_norm[0][:, :, -1]*alpha_y_test[:,:,-1] + ion_velocity_term_1[0][:, :, -1])",
"_____no_output_____"
],
[
"np.where((b_field_norm[0][:, :, -1]*alpha_y_test[:,:,-1] + ion_velocity_term_1[0][:, :, -1]) > 1.2e8)",
"_____no_output_____"
],
[
"(b_field_norm[0][:, :, -1]*alpha_y_test[:,:,-1] + ion_velocity_term_1[0][:, :, -1])[13,1]",
"_____no_output_____"
],
[
"b_field_norm[0][13, 1, -1]",
"_____no_output_____"
],
[
"alpha_y_test[13, 1,-1]",
"_____no_output_____"
],
[
"ion_velocity_term_1[0][13,1,-1]",
"_____no_output_____"
],
[
"ion_velocity_term_1[1][:, :, -1]",
"_____no_output_____"
],
[
"alpha_from_y_flattened_z04 = alpha_from_y[:,:,-1].ravel()\nalpha_from_z_flattened_z04 = alpha_from_z[:,:,-1].ravel()\npoints_x = mesh_wo_edges[0][:, :, -1].ravel()\npoints_y = mesh_wo_edges[1][:, :, -1].ravel()\npoints_z_z04 = mesh[2][0,0,-1]*np.ones(points_x.shape)\nstd_z04 = np.expand_dims(np.zeros(points_x.shape), 0)\n\ndata_y_z04 = {'a_out': np.expand_dims(alpha_from_y_flattened_z04, 0),\n 'x_out': points_x,\n 'y_out': points_y,\n 'z_out': points_z_z04,\n 'std': std_z04\n }",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb917d8bba60ed852467ea7d047fdf0cd68c2e44
| 792,539 |
ipynb
|
Jupyter Notebook
|
basic_examples_compact.ipynb
|
cod3licious/simec
|
45392475debc37883a79eddb0e90f2669ce0a44e
|
[
"MIT"
] | 28 |
2016-11-05T16:05:04.000Z
|
2022-03-05T15:47:45.000Z
|
basic_examples_compact.ipynb
|
cod3licious/simec
|
45392475debc37883a79eddb0e90f2669ce0a44e
|
[
"MIT"
] | 4 |
2018-04-28T03:10:50.000Z
|
2020-06-26T13:59:52.000Z
|
basic_examples_compact.ipynb
|
cod3licious/simec
|
45392475debc37883a79eddb0e90f2669ce0a44e
|
[
"MIT"
] | 5 |
2017-02-10T11:54:37.000Z
|
2022-02-17T13:45:45.000Z
| 1,303.518092 | 186,092 | 0.947161 |
[
[
[
"# Similarity Encoders with Keras\n## using the model definition from `simec.py`",
"_____no_output_____"
]
],
[
[
"from __future__ import unicode_literals, division, print_function, absolute_import\nfrom builtins import range\nimport numpy as np\nnp.random.seed(28)\nimport matplotlib.pyplot as plt\nfrom sklearn.manifold import Isomap\nfrom sklearn.decomposition import KernelPCA\nfrom sklearn.preprocessing import StandardScaler\nfrom sklearn.datasets import fetch_mldata, fetch_20newsgroups\n\nimport tensorflow as tf\ntf.set_random_seed(28)\nimport keras\nfrom keras.models import Sequential, Model\nfrom keras.layers import Input, Dense, Activation\n\n# https://github.com/cod3licious/nlputils\nfrom nlputils.features import FeatureTransform, features2mat\n\nfrom simec import SimilarityEncoder\nfrom utils import center_K, check_embed_match, check_similarity_match\nfrom utils_plotting import plot_mnist, plot_20news\n\n%matplotlib inline\n%load_ext autoreload\n%autoreload 2",
"/home/franzi/anaconda2/envs/python36/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n from ._conv import register_converters as _register_converters\nUsing TensorFlow backend.\n"
]
],
[
[
"### MNIST with Linear Kernel",
"_____no_output_____"
]
],
[
[
"# load digits\nmnist = fetch_mldata('MNIST original', data_home='data')\nX = mnist.data/255. # normalize to 0-1\ny = np.array(mnist.target, dtype=int)\n# subsample 10000 random data points\nnp.random.seed(42)\nn_samples = 10000\nn_test = 2000\nrnd_idx = np.random.permutation(X.shape[0])[:n_samples]\nX_test, y_test = X[rnd_idx[:n_test],:], y[rnd_idx[:n_test]]\nX, y = X[rnd_idx[n_test:],:], y[rnd_idx[n_test:]]\nss = StandardScaler(with_std=False)\nX = ss.fit_transform(X)\nX_test = ss.transform(X_test)\nn_train, n_features = X.shape",
"_____no_output_____"
],
[
"# centered linear kernel matrix\nK_lin = center_K(np.dot(X, X.T))",
"_____no_output_____"
],
[
"# linear kPCA\nkpca = KernelPCA(n_components=2, kernel='linear')\nX_embed = kpca.fit_transform(X)\nX_embed_test = kpca.transform(X_test)\nplot_mnist(X_embed, y, X_embed_test, y_test, title='MNIST - linear Kernel PCA')\nprint(\"error similarity match: msqe: %.10f ; r^2: %.10f ; rho: %.10f\" % check_similarity_match(X_embed, K_lin))",
"error similarity match: msqe: 51.4531662441 ; r^2: 0.4438507851 ; rho: 0.6341602248\n"
],
[
"# on how many target similarities you want to train - faster and works equally well than training on all\nn_targets = 1000 # K_lin.shape[1]\n# initialize the model\nsimec = SimilarityEncoder(X.shape[1], 2, n_targets, s_ll_reg=0.5, S_ll=K_lin[:n_targets,:n_targets])\n# train the model to get an embedding with which the target similarities\n# can be linearly approximated\nsimec.fit(X, K_lin[:,:n_targets], epochs=25)\n# get the embeddings\nX_embeds = simec.transform(X)\nX_embed_tests = simec.transform(X_test)\nplot_mnist(X_embeds, y, X_embed_tests, y_test, title='MNIST - SimEc (lin. kernel, linear)')\n# correlation with the embedding produced by the spectral method should be high\nprint(\"correlation with lin kPCA : %f\" % check_embed_match(X_embed, X_embeds)[1])\nprint(\"correlation with lin kPCA (test): %f\" % check_embed_match(X_embed_test, X_embed_tests)[1])\n# similarity match error should be similar to the one from kpca\nprint(\"error similarity match: msqe: %.10f ; r^2: %.10f ; rho: %.10f\" % check_similarity_match(X_embeds, K_lin))",
"Epoch 1/25\n8000/8000 [==============================] - 1s 155us/step - loss: 138.9871\nEpoch 2/25\n8000/8000 [==============================] - 1s 93us/step - loss: 134.7487\nEpoch 3/25\n8000/8000 [==============================] - 1s 88us/step - loss: 128.0694\nEpoch 4/25\n8000/8000 [==============================] - 1s 93us/step - loss: 121.4029\nEpoch 5/25\n8000/8000 [==============================] - 1s 93us/step - loss: 116.2597\nEpoch 6/25\n8000/8000 [==============================] - 1s 95us/step - loss: 112.6398\nEpoch 7/25\n8000/8000 [==============================] - 1s 93us/step - loss: 109.3631\nEpoch 8/25\n8000/8000 [==============================] - 1s 87us/step - loss: 105.8619\nEpoch 9/25\n8000/8000 [==============================] - 1s 89us/step - loss: 102.6419\nEpoch 10/25\n8000/8000 [==============================] - 1s 94us/step - loss: 100.4048\nEpoch 11/25\n8000/8000 [==============================] - 1s 92us/step - loss: 99.2771\nEpoch 12/25\n8000/8000 [==============================] - 1s 92us/step - loss: 98.8235\nEpoch 13/25\n8000/8000 [==============================] - 1s 90us/step - loss: 98.6205\nEpoch 14/25\n8000/8000 [==============================] - 1s 93us/step - loss: 98.5062\nEpoch 15/25\n8000/8000 [==============================] - 1s 90us/step - loss: 98.4333\nEpoch 16/25\n8000/8000 [==============================] - 1s 90us/step - loss: 98.3796\nEpoch 17/25\n8000/8000 [==============================] - 1s 89us/step - loss: 98.3371\nEpoch 18/25\n8000/8000 [==============================] - 1s 90us/step - loss: 98.3028\nEpoch 19/25\n8000/8000 [==============================] - 1s 96us/step - loss: 98.2718\nEpoch 20/25\n8000/8000 [==============================] - 1s 93us/step - loss: 98.2441\nEpoch 21/25\n8000/8000 [==============================] - 1s 88us/step - loss: 98.2178\nEpoch 22/25\n8000/8000 [==============================] - 1s 90us/step - loss: 98.1937\nEpoch 23/25\n8000/8000 [==============================] - 1s 93us/step - loss: 98.1695\nEpoch 24/25\n8000/8000 [==============================] - 1s 88us/step - loss: 98.1459\nEpoch 25/25\n8000/8000 [==============================] - 1s 90us/step - loss: 98.1206\ncorrelation with lin kPCA : 0.989194\ncorrelation with lin kPCA (test): 0.989490\nerror similarity match: msqe: 57769.3758877199 ; r^2: 0.4418963431 ; rho: 0.6326533868\n"
]
],
[
[
"### Non-linear MNIST embedding with isomap",
"_____no_output_____"
]
],
[
[
"# isomap\nisomap = Isomap(n_neighbors=10, n_components=2)\nX_embed = isomap.fit_transform(X)\nX_embed_test = isomap.transform(X_test)\nplot_mnist(X_embed, y, X_embed_test, y_test, title='MNIST - isomap')",
"_____no_output_____"
],
[
"# non-linear SimEc to approximate isomap solution\nK_geod = center_K(-0.5*(isomap.dist_matrix_**2))\nn_targets = 1000\n# initialize the model\nsimec = SimilarityEncoder(X.shape[1], 2, n_targets, hidden_layers=[(20, 'tanh')], s_ll_reg=0.5, \n S_ll=K_geod[:n_targets,:n_targets], opt=keras.optimizers.Adamax(lr=0.01))\n# train the model to get an embedding with which the target similarities\n# can be linearly approximated\nsimec.fit(X, K_geod[:,:n_targets], epochs=25)\n# get the embeddings\nX_embeds = simec.transform(X)\nX_embed_tests = simec.transform(X_test)\nplot_mnist(X_embeds, y, X_embed_tests, y_test, title='MNIST - SimEc (isomap, 1 h.l.)')\nprint(\"correlation with isomap : %f\" % check_embed_match(X_embed, X_embeds)[1])\nprint(\"correlation with isomap (test): %f\" % check_embed_match(X_embed_test, X_embed_tests)[1])",
"Epoch 1/25\n8000/8000 [==============================] - 1s 144us/step - loss: 51178.2788\nEpoch 2/25\n8000/8000 [==============================] - 1s 123us/step - loss: 46771.8850\nEpoch 3/25\n8000/8000 [==============================] - 1s 124us/step - loss: 41210.9276\nEpoch 4/25\n8000/8000 [==============================] - 1s 122us/step - loss: 38588.5507\nEpoch 5/25\n8000/8000 [==============================] - 1s 122us/step - loss: 37866.1624\nEpoch 6/25\n8000/8000 [==============================] - 1s 121us/step - loss: 37444.9817\nEpoch 7/25\n8000/8000 [==============================] - 1s 123us/step - loss: 37145.0221\nEpoch 8/25\n8000/8000 [==============================] - 1s 122us/step - loss: 36876.5603\nEpoch 9/25\n8000/8000 [==============================] - 1s 122us/step - loss: 36615.4620\nEpoch 10/25\n8000/8000 [==============================] - 1s 121us/step - loss: 36335.6757\nEpoch 11/25\n8000/8000 [==============================] - 1s 123us/step - loss: 36026.0781\nEpoch 12/25\n8000/8000 [==============================] - 1s 134us/step - loss: 35676.8645\nEpoch 13/25\n8000/8000 [==============================] - 1s 122us/step - loss: 35261.1747\nEpoch 14/25\n8000/8000 [==============================] - 1s 136us/step - loss: 34778.4479\nEpoch 15/25\n8000/8000 [==============================] - 1s 129us/step - loss: 34220.4407\nEpoch 16/25\n8000/8000 [==============================] - 1s 124us/step - loss: 33603.7760\nEpoch 17/25\n8000/8000 [==============================] - 1s 136us/step - loss: 32931.2144\nEpoch 18/25\n8000/8000 [==============================] - 1s 121us/step - loss: 32257.5660\nEpoch 19/25\n8000/8000 [==============================] - 1s 125us/step - loss: 31622.0433\nEpoch 20/25\n8000/8000 [==============================] - 1s 131us/step - loss: 31084.5188\nEpoch 21/25\n8000/8000 [==============================] - 1s 135us/step - loss: 30660.2642\nEpoch 22/25\n8000/8000 [==============================] - 1s 125us/step - loss: 30352.2243\nEpoch 23/25\n8000/8000 [==============================] - 1s 125us/step - loss: 30165.2209\nEpoch 24/25\n8000/8000 [==============================] - 1s 125us/step - loss: 30061.4422\nEpoch 25/25\n8000/8000 [==============================] - 1s 125us/step - loss: 29994.4935\ncorrelation with isomap : 0.904213\ncorrelation with isomap (test): 0.809611\n"
]
],
[
[
"## 20newsgroups embedding",
"_____no_output_____"
]
],
[
[
"## load the data and transform it into a tf-idf representation\ncategories = [\n \"comp.graphics\",\n \"rec.autos\",\n \"rec.sport.baseball\",\n \"sci.med\",\n \"sci.space\",\n \"soc.religion.christian\",\n \"talk.politics.guns\"\n]\nnewsgroups_train = fetch_20newsgroups(subset='train', remove=(\n 'headers', 'footers', 'quotes'), data_home='data', categories=categories, random_state=42)\nnewsgroups_test = fetch_20newsgroups(subset='test', remove=(\n 'headers', 'footers', 'quotes'), data_home='data', categories=categories, random_state=42)\n# store in dicts (if the text contains more than 3 words)\ntextdict = {i: t for i, t in enumerate(newsgroups_train.data) if len(t.split()) > 3}\ntextdict.update({i: t for i, t in enumerate(newsgroups_test.data, len(newsgroups_train.data)) if len(t.split()) > 3})\ntrain_ids = [i for i in range(len(newsgroups_train.data)) if i in textdict]\ntest_ids = [i for i in range(len(newsgroups_train.data), len(textdict)) if i in textdict]\nprint(\"%i training and %i test samples\" % (len(train_ids), len(test_ids)))\n# transform into tf-idf features\nft = FeatureTransform(norm='max', weight=True, renorm='max')\ndocfeats = ft.texts2features(textdict, fit_ids=train_ids)\n# organize in feature matrix\nX, featurenames = features2mat(docfeats, train_ids)\nX_test, _ = features2mat(docfeats, test_ids, featurenames)\nprint(\"%i features\" % len(featurenames))\ntargets = np.hstack([newsgroups_train.target,newsgroups_test.target])\ny = targets[train_ids]\ny_test = targets[test_ids]\ntarget_names = newsgroups_train.target_names\nn_targets = 1000",
"3959 training and 2359 test samples\n45813 features\n"
],
[
"# linear kPCA\nkpca = KernelPCA(n_components=2, kernel='linear')\nX_embed = kpca.fit_transform(X)\nX_embed_test = kpca.transform(X_test)\nplot_20news(X_embed, y, target_names, X_embed_test, y_test,\n title='20newsgroups - linear Kernel PCA', legend=True)\n# compute linear kernel and center\nK_lin = center_K(X.dot(X.T).A)\nK_lin_test = center_K(X_test.dot(X_test.T).A)\nprint(\"similarity approximation : msqe: %.10f ; r^2: %.10f ; rho: %.10f\" % check_similarity_match(X_embed, K_lin))\nprint(\"similarity approximation (test): msqe: %.10f ; r^2: %.10f ; rho: %.10f\" % check_similarity_match(X_embed_test, K_lin_test))",
"similarity approximation : msqe: 0.0331868908 ; r^2: 0.2263374840 ; rho: 0.6310964767\nsimilarity approximation (test): msqe: 0.0421617582 ; r^2: 0.1775899837 ; rho: 0.6165146584\n"
],
[
"# project to 2d with linear similarity encoder\n# careful: our input is sparse!!!\nsimec = SimilarityEncoder(X.shape[1], 2, n_targets, sparse_inputs=True, opt=keras.optimizers.SGD(lr=50.))\n# train the model to get an embedding with which the target similarities\n# can be linearly approximated\nsimec.fit(X, K_lin[:,:n_targets], epochs=25)\n# get the embeddings\nX_embeds = simec.transform(X)\nX_embed_tests = simec.transform(X_test)\nplot_20news(X_embeds, y, target_names, X_embed_tests, y_test,\n title='20 newsgroups - SimEc (lin. kernel, linear)', legend=True)\nprint(\"correlation with lin kPCA : %f\" % check_embed_match(X_embed, X_embeds)[1])\nprint(\"correlation with lin kPCA (test): %f\" % check_embed_match(X_embed_test, X_embed_tests)[1])\nprint(\"similarity approximation : msqe: %.10f ; r^2: %.10f ; rho: %.10f\" % check_similarity_match(X_embeds, K_lin))\nprint(\"similarity approximation (test): msqe: %.10f ; r^2: %.10f ; rho: %.10f\" % check_similarity_match(X_embed_tests, K_lin_test))",
"Epoch 1/25\n3959/3959 [==============================] - 1s 140us/step - loss: 0.0366\nEpoch 2/25\n3959/3959 [==============================] - 0s 104us/step - loss: 0.0318\nEpoch 3/25\n3959/3959 [==============================] - 0s 102us/step - loss: 0.0315\nEpoch 4/25\n3959/3959 [==============================] - 0s 104us/step - loss: 0.0311\nEpoch 5/25\n3959/3959 [==============================] - 0s 110us/step - loss: 0.0311\nEpoch 6/25\n3959/3959 [==============================] - 0s 109us/step - loss: 0.0311\nEpoch 7/25\n3959/3959 [==============================] - 0s 103us/step - loss: 0.0310\nEpoch 8/25\n3959/3959 [==============================] - 0s 106us/step - loss: 0.0311\nEpoch 9/25\n3959/3959 [==============================] - 0s 101us/step - loss: 0.0310\nEpoch 10/25\n3959/3959 [==============================] - 0s 106us/step - loss: 0.0310\nEpoch 11/25\n3959/3959 [==============================] - 0s 102us/step - loss: 0.0310\nEpoch 12/25\n3959/3959 [==============================] - 0s 111us/step - loss: 0.0311\nEpoch 13/25\n3959/3959 [==============================] - 0s 100us/step - loss: 0.0310\nEpoch 14/25\n3959/3959 [==============================] - 0s 102us/step - loss: 0.0310\nEpoch 15/25\n3959/3959 [==============================] - 0s 98us/step - loss: 0.0310\nEpoch 16/25\n3959/3959 [==============================] - 0s 100us/step - loss: 0.0310\nEpoch 17/25\n3959/3959 [==============================] - 0s 97us/step - loss: 0.0310\nEpoch 18/25\n3959/3959 [==============================] - 0s 98us/step - loss: 0.0310\nEpoch 19/25\n3959/3959 [==============================] - 0s 104us/step - loss: 0.0310\nEpoch 20/25\n3959/3959 [==============================] - 0s 98us/step - loss: 0.0310\nEpoch 21/25\n3959/3959 [==============================] - 0s 100us/step - loss: 0.0310\nEpoch 22/25\n3959/3959 [==============================] - 0s 103us/step - loss: 0.0311\nEpoch 23/25\n3959/3959 [==============================] - 0s 103us/step - loss: 0.0310\nEpoch 24/25\n3959/3959 [==============================] - 0s 106us/step - loss: 0.0310\nEpoch 25/25\n3959/3959 [==============================] - 0s 116us/step - loss: 0.0310\ncorrelation with lin kPCA : 0.939694\ncorrelation with lin kPCA (test): 0.957125\nsimilarity approximation : msqe: 0.1682884082 ; r^2: 0.2108829888 ; rho: 0.6228658243\nsimilarity approximation (test): msqe: 0.1391057260 ; r^2: 0.1664079620 ; rho: 0.6075047460\n"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
]
] |
cb91951a20111eed815b0ef682cd3f90173120d8
| 139,144 |
ipynb
|
Jupyter Notebook
|
ml_files/cluster_analyze-v6.ipynb
|
cbarros7/capstone_project_holberton
|
62d765e154d76d2ec8b9ff6cb271d7484d262129
|
[
"MIT"
] | null | null | null |
ml_files/cluster_analyze-v6.ipynb
|
cbarros7/capstone_project_holberton
|
62d765e154d76d2ec8b9ff6cb271d7484d262129
|
[
"MIT"
] | null | null | null |
ml_files/cluster_analyze-v6.ipynb
|
cbarros7/capstone_project_holberton
|
62d765e154d76d2ec8b9ff6cb271d7484d262129
|
[
"MIT"
] | null | null | null | 72.170124 | 12,952 | 0.754341 |
[
[
[
"# Analyzing data from clusters FAMD-v6",
"_____no_output_____"
]
],
[
[
"import pandas as pd\nimport numpy as np\nimport json\nimport datetime as dt\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nimport statsmodels.api as sm\nimport statsmodels.formula.api as smf\nimport statsmodels.stats.multicomp as multi\nimport scipy.stats\n# Configuración de pandas \npd.set_option('display.max_rows', 500)\npd.set_option('display.max_columns', 500)\npd.set_option('display.width', 1000)",
"_____no_output_____"
]
],
[
[
"## Loading data",
"_____no_output_____"
]
],
[
[
"path_file = 'data/clusterizacion_v6.csv'\ndata = pd.read_csv(path_file)\ndata.head()",
"_____no_output_____"
],
[
"data.isnull().any()",
"_____no_output_____"
]
],
[
[
"## Data analysis",
"_____no_output_____"
],
[
"This section is to analyze clusters behaviour with two cathegorization of clients",
"_____no_output_____"
],
[
"### Cathegorical data",
"_____no_output_____"
]
],
[
[
"sns.countplot(x='cluster', data = data)\nplt.suptitle('Frecuency of observation by cluster')",
"_____no_output_____"
],
[
"data['cluster'].value_counts()",
"_____no_output_____"
],
[
"data['arrears_days'].describe()",
"_____no_output_____"
]
],
[
[
"We can see that there are more customers on the cluster_1 than the cluster_0",
"_____no_output_____"
],
[
"### Analyzing relationship on other variables with two clusters",
"_____no_output_____"
]
],
[
[
"#----------------------------------------------------------------",
"_____no_output_____"
]
],
[
[
"#### Cathegorical = cluster_id_2 vs Quantitative = 'Column name'",
"_____no_output_____"
]
],
[
[
"data.info()",
"<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 577 entries, 0 to 576\nData columns (total 32 columns):\n # Column Non-Null Count Dtype \n--- ------ -------------- ----- \n 0 Unnamed: 0 577 non-null int64 \n 1 client_id 577 non-null int64 \n 2 loan_id 577 non-null int64 \n 3 state 577 non-null object \n 4 arrears_days 577 non-null int64 \n 5 total_paid 577 non-null float64\n 6 percentage 577 non-null float64\n 7 Monto Acumulado 577 non-null int64 \n 8 Uso de los recursos 577 non-null object \n 9 Plazo 577 non-null object \n 10 Sector 577 non-null object \n 11 Ingresos 577 non-null object \n 12 Ubicación 577 non-null object \n 13 Estrato Mínimo 577 non-null int64 \n 14 Procesos judiciales 577 non-null object \n 15 Alertas 577 non-null object \n 16 Score Bureau Empresa 577 non-null float64\n 17 Huellas de Consulta 577 non-null float64\n 18 Tiempo en el negocio 577 non-null int64 \n 19 Website empresa 577 non-null object \n 20 Instagram empresa 577 non-null object \n 21 LinkedIn empresa 577 non-null object \n 22 LinkedIn empresarios 577 non-null object \n 23 Edad empresarios 577 non-null int64 \n 24 Activador 577 non-null object \n 25 Número de accionistas 577 non-null float64\n 26 Impacto 577 non-null object \n 27 Acceso previso a la banca 577 non-null object \n 28 # Empleados 577 non-null int64 \n 29 Mujeres empresarias 577 non-null object \n 30 Mujeres en cargos directivos 577 non-null int64 \n 31 cluster 577 non-null int64 \ndtypes: float64(5), int64(11), object(16)\nmemory usage: 144.4+ KB\n"
],
[
"sns.catplot(x='cluster', y='arrears_days', kind='bar', data=data)\nplt.suptitle('Cluster vs arrears_days')",
"_____no_output_____"
]
],
[
[
"On this graph we can see that cluster 1 has less arrears_days than cluster 0",
"_____no_output_____"
]
],
[
[
"sns.catplot(x='cluster', y='Monto Acumulado', kind='bar', data=data)\nplt.suptitle('Cluster vs Monto Acumulado')",
"_____no_output_____"
]
],
[
[
"Cluster 1 has less 'Monto Acumulado' than cluster 0",
"_____no_output_____"
]
],
[
[
"sns.catplot(x='cluster', y='Score Bureau Empresa', kind='bar', data=data)\nplt.suptitle('Cluster vs Score Bureau Empresa')",
"_____no_output_____"
]
],
[
[
"Score Bureau between the two clusters are relatively similar",
"_____no_output_____"
]
],
[
[
"sns.catplot(x='cluster', y='Huellas de Consulta', kind='bar', data=data)\nplt.suptitle('Cluster vs Huellas de Consulta')",
"_____no_output_____"
]
],
[
[
"Cluster 1 has less 'Huellas de consulta' than cluster 0",
"_____no_output_____"
]
],
[
[
"sns.catplot(x='cluster', y='Número de accionistas', kind='bar', data=data)\nplt.suptitle('Cluster vs Número de accionistas')",
"_____no_output_____"
]
],
[
[
"Cluster 0 has more 'Accionistas' than cluster 1",
"_____no_output_____"
]
],
[
[
"sns.catplot(x='cluster', y='# Empleados', kind='bar', data=data)\nplt.suptitle('Cluster vs # Empleados')",
"_____no_output_____"
]
],
[
[
"Cluster 1 has more 'Empleados' than cluster 0",
"_____no_output_____"
]
],
[
[
"sns.catplot(x='cluster', y='Mujeres en cargos directivos', kind='bar', data=data)\nplt.suptitle('Cluster vs Mujeres en cargos directivos')",
"_____no_output_____"
]
],
[
[
"Cluster 1 has more 'Mujeres en cargos directivos' than cluster 0",
"_____no_output_____"
]
],
[
[
"#--------------------------------------------------------",
"_____no_output_____"
]
],
[
[
"#### Cathegorical = cluster_id_2 vs Cathegorical = 'Column name'",
"_____no_output_____"
]
],
[
[
"data.info()",
"<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 577 entries, 0 to 576\nData columns (total 32 columns):\n # Column Non-Null Count Dtype \n--- ------ -------------- ----- \n 0 Unnamed: 0 577 non-null int64 \n 1 client_id 577 non-null int64 \n 2 loan_id 577 non-null int64 \n 3 state 577 non-null object \n 4 arrears_days 577 non-null int64 \n 5 total_paid 577 non-null float64\n 6 percentage 577 non-null float64\n 7 Monto Acumulado 577 non-null int64 \n 8 Uso de los recursos 577 non-null object \n 9 Plazo 577 non-null object \n 10 Sector 577 non-null object \n 11 Ingresos 577 non-null object \n 12 Ubicación 577 non-null object \n 13 Estrato Mínimo 577 non-null int64 \n 14 Procesos judiciales 577 non-null object \n 15 Alertas 577 non-null object \n 16 Score Bureau Empresa 577 non-null float64\n 17 Huellas de Consulta 577 non-null float64\n 18 Tiempo en el negocio 577 non-null int64 \n 19 Website empresa 577 non-null object \n 20 Instagram empresa 577 non-null object \n 21 LinkedIn empresa 577 non-null object \n 22 LinkedIn empresarios 577 non-null object \n 23 Edad empresarios 577 non-null int64 \n 24 Activador 577 non-null object \n 25 Número de accionistas 577 non-null float64\n 26 Impacto 577 non-null object \n 27 Acceso previso a la banca 577 non-null object \n 28 # Empleados 577 non-null int64 \n 29 Mujeres empresarias 577 non-null object \n 30 Mujeres en cargos directivos 577 non-null int64 \n 31 cluster 577 non-null int64 \ndtypes: float64(5), int64(11), object(16)\nmemory usage: 144.4+ KB\n"
],
[
"data.groupby('cluster')['state'].value_counts()/len(data)",
"_____no_output_____"
]
],
[
[
"Cluster 0 has more clients on state PAID than cluster 1, but at the same time has more clients on Late than cluster 1",
"_____no_output_____"
]
],
[
[
"data.groupby('cluster')['Uso de los recursos'].value_counts()/len(data)",
"_____no_output_____"
]
],
[
[
"We can see the distribution between the two clusters",
"_____no_output_____"
]
],
[
[
"data.groupby('cluster')['Plazo'].value_counts()/len(data)",
"_____no_output_____"
],
[
"# Cluster 0 has more clients on 'Plazo' on 13 to 24 months. On the other hand we have cluster 1 with 'Plazo' less than 12 months",
"_____no_output_____"
],
[
"data.groupby('cluster')['Sector'].value_counts()/len(data)",
"_____no_output_____"
],
[
"# Cluster 0 has a ranking of (Servicios, Comercio and Industria) - Cluster 1 has a ranking of (Servicios, Industria y Comercio)",
"_____no_output_____"
],
[
"data.groupby('cluster')['Ingresos'].value_counts()/len(data)",
"_____no_output_____"
],
[
"data.groupby('cluster')['Acceso previso a la banca'].value_counts()/len(data)",
"_____no_output_____"
],
[
"data.groupby('cluster')['Mujeres empresarias'].value_counts()/len(data)",
"_____no_output_____"
],
[
"data.groupby('cluster')['Activador'].value_counts()/len(data)",
"_____no_output_____"
],
[
"data.groupby('cluster')['Website empresa'].value_counts()/len(data)",
"_____no_output_____"
],
[
"data.groupby('cluster')['Estrato Mínimo'].value_counts()/len(data)",
"_____no_output_____"
],
[
"data.groupby('cluster')['Ubicación'].value_counts()/len(data)",
"_____no_output_____"
],
[
"data.groupby('cluster')['Procesos judiciales'].value_counts()/len(data)",
"_____no_output_____"
],
[
"data.groupby('cluster')['Instagram empresa'].value_counts()/len(data)",
"_____no_output_____"
],
[
"data.groupby('cluster')['Impacto'].value_counts()/len(data)",
"_____no_output_____"
],
[
"#end",
"_____no_output_____"
],
[
"#--------------------------------------------------------------------",
"_____no_output_____"
]
],
[
[
"### Multi-linear regression",
"_____no_output_____"
]
],
[
[
"X = data[['Huellas de Consulta', 'Score Bureau Empresa']]\nY = data['arrears_days']\nX = sm.add_constant(X) # adding a constant\n \nmodel = sm.OLS(Y, X).fit()\npredictions = model.predict(X) \n \nprint_model = model.summary()\nprint(print_model)",
" OLS Regression Results \n==============================================================================\nDep. Variable: arrears_days R-squared: 0.079\nModel: OLS Adj. R-squared: 0.076\nMethod: Least Squares F-statistic: 24.68\nDate: Wed, 28 Oct 2020 Prob (F-statistic): 5.21e-11\nTime: 11:29:58 Log-Likelihood: -3606.5\nNo. Observations: 577 AIC: 7219.\nDf Residuals: 574 BIC: 7232.\nDf Model: 2 \nCovariance Type: nonrobust \n========================================================================================\n coef std err t P>|t| [0.025 0.975]\n----------------------------------------------------------------------------------------\nconst 123.8634 19.777 6.263 0.000 85.019 162.708\nHuellas de Consulta 4.0733 0.680 5.990 0.000 2.738 5.409\nScore Bureau Empresa -0.1020 0.023 -4.359 0.000 -0.148 -0.056\n==============================================================================\nOmnibus: 371.639 Durbin-Watson: 1.474\nProb(Omnibus): 0.000 Jarque-Bera (JB): 3412.454\nSkew: 2.820 Prob(JB): 0.00\nKurtosis: 13.494 Cond. No. 3.18e+03\n==============================================================================\n\nWarnings:\n[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n[2] The condition number is large, 3.18e+03. This might indicate that there are\nstrong multicollinearity or other numerical problems.\n"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb9199118cf71f0cc9a6bff6eadbc4d3120d7210
| 3,471 |
ipynb
|
Jupyter Notebook
|
devices/lighting/save configs to file.ipynb
|
naren-m/home_assistant
|
a20d64b79d97cbb2c9bae197c0900515ed8bfbfc
|
[
"MIT"
] | null | null | null |
devices/lighting/save configs to file.ipynb
|
naren-m/home_assistant
|
a20d64b79d97cbb2c9bae197c0900515ed8bfbfc
|
[
"MIT"
] | null | null | null |
devices/lighting/save configs to file.ipynb
|
naren-m/home_assistant
|
a20d64b79d97cbb2c9bae197c0900515ed8bfbfc
|
[
"MIT"
] | null | null | null | 26.496183 | 638 | 0.50533 |
[
[
[
"import hue\n\nh = hue.Hue()",
"_____no_output_____"
],
[
"h.toggle(2)",
"(u'Living room hanging', 'is on. Turning it off') {}\n"
],
[
"lightGroups = h.getAllGroups()\nlightsInfo = h.getAllLights()",
"_____no_output_____"
],
[
"import json\nwith open('tmp.txt', 'w') as filename:\n json.dump(lightsInfo, filename, indent=4)\n\nprint lightGroups",
"{u'1': {u'name': u'Bedroom', u'lights': [u'3', u'1'], u'state': {u'any_on': False, u'all_on': False}, u'action': {u'on': False, u'hue': 34076, u'colormode': u'xy', u'effect': u'none', u'alert': u'none', u'xy': [0.3144, 0.3301], u'bri': 254, u'ct': 153, u'sat': 251}, u'type': u'Room', u'class': u'Bedroom'}, u'2': {u'name': u'Living room', u'lights': [u'4', u'5', u'2'], u'state': {u'any_on': False, u'all_on': False}, u'action': {u'on': False, u'hue': 34076, u'colormode': u'xy', u'effect': u'none', u'alert': u'none', u'xy': [0.3144, 0.3301], u'bri': 254, u'ct': 153, u'sat': 251}, u'type': u'Room', u'class': u'Living room'}}\n"
],
[
"for k, v in lightGroups.iteritems():\n print k, v",
"1 {u'name': u'Bedroom', u'lights': [u'3', u'1'], u'state': {u'any_on': False, u'all_on': False}, u'action': {u'on': False, u'hue': 34076, u'colormode': u'xy', u'effect': u'none', u'alert': u'none', u'xy': [0.3144, 0.3301], u'bri': 254, u'ct': 153, u'sat': 251}, u'type': u'Room', u'class': u'Bedroom'}\n2 {u'name': u'Living room', u'lights': [u'4', u'5', u'2'], u'state': {u'any_on': False, u'all_on': False}, u'action': {u'on': False, u'hue': 34076, u'colormode': u'xy', u'effect': u'none', u'alert': u'none', u'xy': [0.3144, 0.3301], u'bri': 254, u'ct': 153, u'sat': 251}, u'type': u'Room', u'class': u'Living room'}\n"
],
[
"from collections import MutableMapping\nclass LightGroup(MutableMapping):\n def __init__(self):\n pass\n \n def \n ",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb91b78ae7596ca72483c1cd48190c08c573aa32
| 46,546 |
ipynb
|
Jupyter Notebook
|
Practical Examples/10 - Dogs vs Cats/features_extraction.ipynb
|
IgorMeloS/Computer-Vision-Training
|
cc458d17ee0ff880ce6ccc47179d667bd55f4bd1
|
[
"Apache-2.0"
] | null | null | null |
Practical Examples/10 - Dogs vs Cats/features_extraction.ipynb
|
IgorMeloS/Computer-Vision-Training
|
cc458d17ee0ff880ce6ccc47179d667bd55f4bd1
|
[
"Apache-2.0"
] | null | null | null |
Practical Examples/10 - Dogs vs Cats/features_extraction.ipynb
|
IgorMeloS/Computer-Vision-Training
|
cc458d17ee0ff880ce6ccc47179d667bd55f4bd1
|
[
"Apache-2.0"
] | null | null | null | 74.354633 | 112 | 0.684183 |
[
[
[
"# Feature Extraction on Dogs vs Cats dataset using a pre-trained ResNet50",
"_____no_output_____"
],
[
"## Importing Libraries\n",
"_____no_output_____"
]
],
[
[
"from config import dogs_vs_cats_config as config\nfrom tensorflow.keras.applications import ResNet50\nfrom tensorflow.keras.applications.resnet50 import preprocess_input\nfrom tensorflow.keras.applications import imagenet_utils\nfrom tensorflow.keras.preprocessing.image import img_to_array\nfrom tensorflow.keras.preprocessing.image import load_img\nfrom sklearn.preprocessing import LabelEncoder\nfrom compvis.io import HDF5DatasetWriter\nfrom imutils import paths\nimport numpy as np\nimport progressbar\nimport random\nimport os",
"_____no_output_____"
]
],
[
[
"## Setting the dataset",
"_____no_output_____"
],
[
"**Loading the image paths**",
"_____no_output_____"
]
],
[
[
"imagesPath = list(paths.list_images(config.IMAGES_PATH))\nrandom.shuffle(imagesPath)",
"_____no_output_____"
]
],
[
[
"**Encoding the labels**",
"_____no_output_____"
]
],
[
[
"labels = [p.split(os.path.sep)[-1].split(\".\")[0] for p in imagesPath]\nle = LabelEncoder()\nlabels = le.fit_transform(labels)",
"_____no_output_____"
],
[
"bs = 16",
"_____no_output_____"
]
],
[
[
"## Feature Extraction",
"_____no_output_____"
]
],
[
[
"model = ResNet50(weights = \"imagenet\", include_top = False)",
"_____no_output_____"
],
[
"model.summary()",
"Model: \"resnet50\"\n__________________________________________________________________________________________________\nLayer (type) Output Shape Param # Connected to \n==================================================================================================\ninput_1 (InputLayer) [(None, None, None, 0 \n__________________________________________________________________________________________________\nconv1_pad (ZeroPadding2D) (None, None, None, 3 0 input_1[0][0] \n__________________________________________________________________________________________________\nconv1_conv (Conv2D) (None, None, None, 6 9472 conv1_pad[0][0] \n__________________________________________________________________________________________________\nconv1_bn (BatchNormalization) (None, None, None, 6 256 conv1_conv[0][0] \n__________________________________________________________________________________________________\nconv1_relu (Activation) (None, None, None, 6 0 conv1_bn[0][0] \n__________________________________________________________________________________________________\npool1_pad (ZeroPadding2D) (None, None, None, 6 0 conv1_relu[0][0] \n__________________________________________________________________________________________________\npool1_pool (MaxPooling2D) (None, None, None, 6 0 pool1_pad[0][0] \n__________________________________________________________________________________________________\nconv2_block1_1_conv (Conv2D) (None, None, None, 6 4160 pool1_pool[0][0] \n__________________________________________________________________________________________________\nconv2_block1_1_bn (BatchNormali (None, None, None, 6 256 conv2_block1_1_conv[0][0] \n__________________________________________________________________________________________________\nconv2_block1_1_relu (Activation (None, None, None, 6 0 conv2_block1_1_bn[0][0] \n__________________________________________________________________________________________________\nconv2_block1_2_conv (Conv2D) (None, None, None, 6 36928 conv2_block1_1_relu[0][0] \n__________________________________________________________________________________________________\nconv2_block1_2_bn (BatchNormali (None, None, None, 6 256 conv2_block1_2_conv[0][0] \n__________________________________________________________________________________________________\nconv2_block1_2_relu (Activation (None, None, None, 6 0 conv2_block1_2_bn[0][0] \n__________________________________________________________________________________________________\nconv2_block1_0_conv (Conv2D) (None, None, None, 2 16640 pool1_pool[0][0] \n__________________________________________________________________________________________________\nconv2_block1_3_conv (Conv2D) (None, None, None, 2 16640 conv2_block1_2_relu[0][0] \n__________________________________________________________________________________________________\nconv2_block1_0_bn (BatchNormali (None, None, None, 2 1024 conv2_block1_0_conv[0][0] \n__________________________________________________________________________________________________\nconv2_block1_3_bn (BatchNormali (None, None, None, 2 1024 conv2_block1_3_conv[0][0] \n__________________________________________________________________________________________________\nconv2_block1_add (Add) (None, None, None, 2 0 conv2_block1_0_bn[0][0] \n conv2_block1_3_bn[0][0] \n__________________________________________________________________________________________________\nconv2_block1_out (Activation) (None, None, None, 2 0 conv2_block1_add[0][0] \n__________________________________________________________________________________________________\nconv2_block2_1_conv (Conv2D) (None, None, None, 6 16448 conv2_block1_out[0][0] \n__________________________________________________________________________________________________\nconv2_block2_1_bn (BatchNormali (None, None, None, 6 256 conv2_block2_1_conv[0][0] \n__________________________________________________________________________________________________\nconv2_block2_1_relu (Activation (None, None, None, 6 0 conv2_block2_1_bn[0][0] \n__________________________________________________________________________________________________\nconv2_block2_2_conv (Conv2D) (None, None, None, 6 36928 conv2_block2_1_relu[0][0] \n__________________________________________________________________________________________________\nconv2_block2_2_bn (BatchNormali (None, None, None, 6 256 conv2_block2_2_conv[0][0] \n__________________________________________________________________________________________________\nconv2_block2_2_relu (Activation (None, None, None, 6 0 conv2_block2_2_bn[0][0] \n__________________________________________________________________________________________________\nconv2_block2_3_conv (Conv2D) (None, None, None, 2 16640 conv2_block2_2_relu[0][0] \n__________________________________________________________________________________________________\nconv2_block2_3_bn (BatchNormali (None, None, None, 2 1024 conv2_block2_3_conv[0][0] \n__________________________________________________________________________________________________\nconv2_block2_add (Add) (None, None, None, 2 0 conv2_block1_out[0][0] \n conv2_block2_3_bn[0][0] \n__________________________________________________________________________________________________\nconv2_block2_out (Activation) (None, None, None, 2 0 conv2_block2_add[0][0] \n__________________________________________________________________________________________________\nconv2_block3_1_conv (Conv2D) (None, None, None, 6 16448 conv2_block2_out[0][0] \n__________________________________________________________________________________________________\nconv2_block3_1_bn (BatchNormali (None, None, None, 6 256 conv2_block3_1_conv[0][0] \n__________________________________________________________________________________________________\nconv2_block3_1_relu (Activation (None, None, None, 6 0 conv2_block3_1_bn[0][0] \n__________________________________________________________________________________________________\nconv2_block3_2_conv (Conv2D) (None, None, None, 6 36928 conv2_block3_1_relu[0][0] \n__________________________________________________________________________________________________\nconv2_block3_2_bn (BatchNormali (None, None, None, 6 256 conv2_block3_2_conv[0][0] \n__________________________________________________________________________________________________\nconv2_block3_2_relu (Activation (None, None, None, 6 0 conv2_block3_2_bn[0][0] \n__________________________________________________________________________________________________\nconv2_block3_3_conv (Conv2D) (None, None, None, 2 16640 conv2_block3_2_relu[0][0] \n__________________________________________________________________________________________________\nconv2_block3_3_bn (BatchNormali (None, None, None, 2 1024 conv2_block3_3_conv[0][0] \n__________________________________________________________________________________________________\nconv2_block3_add (Add) (None, None, None, 2 0 conv2_block2_out[0][0] \n conv2_block3_3_bn[0][0] \n__________________________________________________________________________________________________\nconv2_block3_out (Activation) (None, None, None, 2 0 conv2_block3_add[0][0] \n__________________________________________________________________________________________________\nconv3_block1_1_conv (Conv2D) (None, None, None, 1 32896 conv2_block3_out[0][0] \n__________________________________________________________________________________________________\nconv3_block1_1_bn (BatchNormali (None, None, None, 1 512 conv3_block1_1_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block1_1_relu (Activation (None, None, None, 1 0 conv3_block1_1_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block1_2_conv (Conv2D) (None, None, None, 1 147584 conv3_block1_1_relu[0][0] \n__________________________________________________________________________________________________\nconv3_block1_2_bn (BatchNormali (None, None, None, 1 512 conv3_block1_2_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block1_2_relu (Activation (None, None, None, 1 0 conv3_block1_2_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block1_0_conv (Conv2D) (None, None, None, 5 131584 conv2_block3_out[0][0] \n__________________________________________________________________________________________________\nconv3_block1_3_conv (Conv2D) (None, None, None, 5 66048 conv3_block1_2_relu[0][0] \n__________________________________________________________________________________________________\nconv3_block1_0_bn (BatchNormali (None, None, None, 5 2048 conv3_block1_0_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block1_3_bn (BatchNormali (None, None, None, 5 2048 conv3_block1_3_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block1_add (Add) (None, None, None, 5 0 conv3_block1_0_bn[0][0] \n conv3_block1_3_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block1_out (Activation) (None, None, None, 5 0 conv3_block1_add[0][0] \n__________________________________________________________________________________________________\nconv3_block2_1_conv (Conv2D) (None, None, None, 1 65664 conv3_block1_out[0][0] \n__________________________________________________________________________________________________\nconv3_block2_1_bn (BatchNormali (None, None, None, 1 512 conv3_block2_1_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block2_1_relu (Activation (None, None, None, 1 0 conv3_block2_1_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block2_2_conv (Conv2D) (None, None, None, 1 147584 conv3_block2_1_relu[0][0] \n__________________________________________________________________________________________________\nconv3_block2_2_bn (BatchNormali (None, None, None, 1 512 conv3_block2_2_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block2_2_relu (Activation (None, None, None, 1 0 conv3_block2_2_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block2_3_conv (Conv2D) (None, None, None, 5 66048 conv3_block2_2_relu[0][0] \n__________________________________________________________________________________________________\nconv3_block2_3_bn (BatchNormali (None, None, None, 5 2048 conv3_block2_3_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block2_add (Add) (None, None, None, 5 0 conv3_block1_out[0][0] \n conv3_block2_3_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block2_out (Activation) (None, None, None, 5 0 conv3_block2_add[0][0] \n__________________________________________________________________________________________________\nconv3_block3_1_conv (Conv2D) (None, None, None, 1 65664 conv3_block2_out[0][0] \n__________________________________________________________________________________________________\nconv3_block3_1_bn (BatchNormali (None, None, None, 1 512 conv3_block3_1_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block3_1_relu (Activation (None, None, None, 1 0 conv3_block3_1_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block3_2_conv (Conv2D) (None, None, None, 1 147584 conv3_block3_1_relu[0][0] \n__________________________________________________________________________________________________\nconv3_block3_2_bn (BatchNormali (None, None, None, 1 512 conv3_block3_2_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block3_2_relu (Activation (None, None, None, 1 0 conv3_block3_2_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block3_3_conv (Conv2D) (None, None, None, 5 66048 conv3_block3_2_relu[0][0] \n__________________________________________________________________________________________________\nconv3_block3_3_bn (BatchNormali (None, None, None, 5 2048 conv3_block3_3_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block3_add (Add) (None, None, None, 5 0 conv3_block2_out[0][0] \n conv3_block3_3_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block3_out (Activation) (None, None, None, 5 0 conv3_block3_add[0][0] \n__________________________________________________________________________________________________\nconv3_block4_1_conv (Conv2D) (None, None, None, 1 65664 conv3_block3_out[0][0] \n__________________________________________________________________________________________________\nconv3_block4_1_bn (BatchNormali (None, None, None, 1 512 conv3_block4_1_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block4_1_relu (Activation (None, None, None, 1 0 conv3_block4_1_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block4_2_conv (Conv2D) (None, None, None, 1 147584 conv3_block4_1_relu[0][0] \n__________________________________________________________________________________________________\nconv3_block4_2_bn (BatchNormali (None, None, None, 1 512 conv3_block4_2_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block4_2_relu (Activation (None, None, None, 1 0 conv3_block4_2_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block4_3_conv (Conv2D) (None, None, None, 5 66048 conv3_block4_2_relu[0][0] \n__________________________________________________________________________________________________\nconv3_block4_3_bn (BatchNormali (None, None, None, 5 2048 conv3_block4_3_conv[0][0] \n__________________________________________________________________________________________________\nconv3_block4_add (Add) (None, None, None, 5 0 conv3_block3_out[0][0] \n conv3_block4_3_bn[0][0] \n__________________________________________________________________________________________________\nconv3_block4_out (Activation) (None, None, None, 5 0 conv3_block4_add[0][0] \n__________________________________________________________________________________________________\nconv4_block1_1_conv (Conv2D) (None, None, None, 2 131328 conv3_block4_out[0][0] \n__________________________________________________________________________________________________\nconv4_block1_1_bn (BatchNormali (None, None, None, 2 1024 conv4_block1_1_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block1_1_relu (Activation (None, None, None, 2 0 conv4_block1_1_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block1_2_conv (Conv2D) (None, None, None, 2 590080 conv4_block1_1_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block1_2_bn (BatchNormali (None, None, None, 2 1024 conv4_block1_2_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block1_2_relu (Activation (None, None, None, 2 0 conv4_block1_2_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block1_0_conv (Conv2D) (None, None, None, 1 525312 conv3_block4_out[0][0] \n__________________________________________________________________________________________________\nconv4_block1_3_conv (Conv2D) (None, None, None, 1 263168 conv4_block1_2_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block1_0_bn (BatchNormali (None, None, None, 1 4096 conv4_block1_0_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block1_3_bn (BatchNormali (None, None, None, 1 4096 conv4_block1_3_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block1_add (Add) (None, None, None, 1 0 conv4_block1_0_bn[0][0] \n conv4_block1_3_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block1_out (Activation) (None, None, None, 1 0 conv4_block1_add[0][0] \n__________________________________________________________________________________________________\nconv4_block2_1_conv (Conv2D) (None, None, None, 2 262400 conv4_block1_out[0][0] \n__________________________________________________________________________________________________\nconv4_block2_1_bn (BatchNormali (None, None, None, 2 1024 conv4_block2_1_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block2_1_relu (Activation (None, None, None, 2 0 conv4_block2_1_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block2_2_conv (Conv2D) (None, None, None, 2 590080 conv4_block2_1_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block2_2_bn (BatchNormali (None, None, None, 2 1024 conv4_block2_2_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block2_2_relu (Activation (None, None, None, 2 0 conv4_block2_2_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block2_3_conv (Conv2D) (None, None, None, 1 263168 conv4_block2_2_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block2_3_bn (BatchNormali (None, None, None, 1 4096 conv4_block2_3_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block2_add (Add) (None, None, None, 1 0 conv4_block1_out[0][0] \n conv4_block2_3_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block2_out (Activation) (None, None, None, 1 0 conv4_block2_add[0][0] \n__________________________________________________________________________________________________\nconv4_block3_1_conv (Conv2D) (None, None, None, 2 262400 conv4_block2_out[0][0] \n__________________________________________________________________________________________________\nconv4_block3_1_bn (BatchNormali (None, None, None, 2 1024 conv4_block3_1_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block3_1_relu (Activation (None, None, None, 2 0 conv4_block3_1_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block3_2_conv (Conv2D) (None, None, None, 2 590080 conv4_block3_1_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block3_2_bn (BatchNormali (None, None, None, 2 1024 conv4_block3_2_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block3_2_relu (Activation (None, None, None, 2 0 conv4_block3_2_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block3_3_conv (Conv2D) (None, None, None, 1 263168 conv4_block3_2_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block3_3_bn (BatchNormali (None, None, None, 1 4096 conv4_block3_3_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block3_add (Add) (None, None, None, 1 0 conv4_block2_out[0][0] \n conv4_block3_3_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block3_out (Activation) (None, None, None, 1 0 conv4_block3_add[0][0] \n__________________________________________________________________________________________________\nconv4_block4_1_conv (Conv2D) (None, None, None, 2 262400 conv4_block3_out[0][0] \n__________________________________________________________________________________________________\nconv4_block4_1_bn (BatchNormali (None, None, None, 2 1024 conv4_block4_1_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block4_1_relu (Activation (None, None, None, 2 0 conv4_block4_1_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block4_2_conv (Conv2D) (None, None, None, 2 590080 conv4_block4_1_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block4_2_bn (BatchNormali (None, None, None, 2 1024 conv4_block4_2_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block4_2_relu (Activation (None, None, None, 2 0 conv4_block4_2_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block4_3_conv (Conv2D) (None, None, None, 1 263168 conv4_block4_2_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block4_3_bn (BatchNormali (None, None, None, 1 4096 conv4_block4_3_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block4_add (Add) (None, None, None, 1 0 conv4_block3_out[0][0] \n conv4_block4_3_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block4_out (Activation) (None, None, None, 1 0 conv4_block4_add[0][0] \n__________________________________________________________________________________________________\nconv4_block5_1_conv (Conv2D) (None, None, None, 2 262400 conv4_block4_out[0][0] \n__________________________________________________________________________________________________\nconv4_block5_1_bn (BatchNormali (None, None, None, 2 1024 conv4_block5_1_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block5_1_relu (Activation (None, None, None, 2 0 conv4_block5_1_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block5_2_conv (Conv2D) (None, None, None, 2 590080 conv4_block5_1_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block5_2_bn (BatchNormali (None, None, None, 2 1024 conv4_block5_2_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block5_2_relu (Activation (None, None, None, 2 0 conv4_block5_2_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block5_3_conv (Conv2D) (None, None, None, 1 263168 conv4_block5_2_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block5_3_bn (BatchNormali (None, None, None, 1 4096 conv4_block5_3_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block5_add (Add) (None, None, None, 1 0 conv4_block4_out[0][0] \n conv4_block5_3_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block5_out (Activation) (None, None, None, 1 0 conv4_block5_add[0][0] \n__________________________________________________________________________________________________\nconv4_block6_1_conv (Conv2D) (None, None, None, 2 262400 conv4_block5_out[0][0] \n__________________________________________________________________________________________________\nconv4_block6_1_bn (BatchNormali (None, None, None, 2 1024 conv4_block6_1_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block6_1_relu (Activation (None, None, None, 2 0 conv4_block6_1_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block6_2_conv (Conv2D) (None, None, None, 2 590080 conv4_block6_1_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block6_2_bn (BatchNormali (None, None, None, 2 1024 conv4_block6_2_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block6_2_relu (Activation (None, None, None, 2 0 conv4_block6_2_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block6_3_conv (Conv2D) (None, None, None, 1 263168 conv4_block6_2_relu[0][0] \n__________________________________________________________________________________________________\nconv4_block6_3_bn (BatchNormali (None, None, None, 1 4096 conv4_block6_3_conv[0][0] \n__________________________________________________________________________________________________\nconv4_block6_add (Add) (None, None, None, 1 0 conv4_block5_out[0][0] \n conv4_block6_3_bn[0][0] \n__________________________________________________________________________________________________\nconv4_block6_out (Activation) (None, None, None, 1 0 conv4_block6_add[0][0] \n__________________________________________________________________________________________________\nconv5_block1_1_conv (Conv2D) (None, None, None, 5 524800 conv4_block6_out[0][0] \n__________________________________________________________________________________________________\nconv5_block1_1_bn (BatchNormali (None, None, None, 5 2048 conv5_block1_1_conv[0][0] \n__________________________________________________________________________________________________\nconv5_block1_1_relu (Activation (None, None, None, 5 0 conv5_block1_1_bn[0][0] \n__________________________________________________________________________________________________\nconv5_block1_2_conv (Conv2D) (None, None, None, 5 2359808 conv5_block1_1_relu[0][0] \n__________________________________________________________________________________________________\nconv5_block1_2_bn (BatchNormali (None, None, None, 5 2048 conv5_block1_2_conv[0][0] \n__________________________________________________________________________________________________\nconv5_block1_2_relu (Activation (None, None, None, 5 0 conv5_block1_2_bn[0][0] \n__________________________________________________________________________________________________\nconv5_block1_0_conv (Conv2D) (None, None, None, 2 2099200 conv4_block6_out[0][0] \n__________________________________________________________________________________________________\nconv5_block1_3_conv (Conv2D) (None, None, None, 2 1050624 conv5_block1_2_relu[0][0] \n__________________________________________________________________________________________________\nconv5_block1_0_bn (BatchNormali (None, None, None, 2 8192 conv5_block1_0_conv[0][0] \n__________________________________________________________________________________________________\nconv5_block1_3_bn (BatchNormali (None, None, None, 2 8192 conv5_block1_3_conv[0][0] \n__________________________________________________________________________________________________\nconv5_block1_add (Add) (None, None, None, 2 0 conv5_block1_0_bn[0][0] \n conv5_block1_3_bn[0][0] \n__________________________________________________________________________________________________\nconv5_block1_out (Activation) (None, None, None, 2 0 conv5_block1_add[0][0] \n__________________________________________________________________________________________________\nconv5_block2_1_conv (Conv2D) (None, None, None, 5 1049088 conv5_block1_out[0][0] \n__________________________________________________________________________________________________\nconv5_block2_1_bn (BatchNormali (None, None, None, 5 2048 conv5_block2_1_conv[0][0] \n__________________________________________________________________________________________________\nconv5_block2_1_relu (Activation (None, None, None, 5 0 conv5_block2_1_bn[0][0] \n__________________________________________________________________________________________________\nconv5_block2_2_conv (Conv2D) (None, None, None, 5 2359808 conv5_block2_1_relu[0][0] \n__________________________________________________________________________________________________\nconv5_block2_2_bn (BatchNormali (None, None, None, 5 2048 conv5_block2_2_conv[0][0] \n__________________________________________________________________________________________________\nconv5_block2_2_relu (Activation (None, None, None, 5 0 conv5_block2_2_bn[0][0] \n__________________________________________________________________________________________________\nconv5_block2_3_conv (Conv2D) (None, None, None, 2 1050624 conv5_block2_2_relu[0][0] \n__________________________________________________________________________________________________\nconv5_block2_3_bn (BatchNormali (None, None, None, 2 8192 conv5_block2_3_conv[0][0] \n__________________________________________________________________________________________________\nconv5_block2_add (Add) (None, None, None, 2 0 conv5_block1_out[0][0] \n conv5_block2_3_bn[0][0] \n__________________________________________________________________________________________________\nconv5_block2_out (Activation) (None, None, None, 2 0 conv5_block2_add[0][0] \n__________________________________________________________________________________________________\nconv5_block3_1_conv (Conv2D) (None, None, None, 5 1049088 conv5_block2_out[0][0] \n__________________________________________________________________________________________________\nconv5_block3_1_bn (BatchNormali (None, None, None, 5 2048 conv5_block3_1_conv[0][0] \n__________________________________________________________________________________________________\nconv5_block3_1_relu (Activation (None, None, None, 5 0 conv5_block3_1_bn[0][0] \n__________________________________________________________________________________________________\nconv5_block3_2_conv (Conv2D) (None, None, None, 5 2359808 conv5_block3_1_relu[0][0] \n__________________________________________________________________________________________________\nconv5_block3_2_bn (BatchNormali (None, None, None, 5 2048 conv5_block3_2_conv[0][0] \n__________________________________________________________________________________________________\nconv5_block3_2_relu (Activation (None, None, None, 5 0 conv5_block3_2_bn[0][0] \n__________________________________________________________________________________________________\nconv5_block3_3_conv (Conv2D) (None, None, None, 2 1050624 conv5_block3_2_relu[0][0] \n__________________________________________________________________________________________________\nconv5_block3_3_bn (BatchNormali (None, None, None, 2 8192 conv5_block3_3_conv[0][0] \n__________________________________________________________________________________________________\nconv5_block3_add (Add) (None, None, None, 2 0 conv5_block2_out[0][0] \n conv5_block3_3_bn[0][0] \n__________________________________________________________________________________________________\nconv5_block3_out (Activation) (None, None, None, 2 0 conv5_block3_add[0][0] \n==================================================================================================\nTotal params: 23,587,712\nTrainable params: 23,534,592\nNon-trainable params: 53,120\n__________________________________________________________________________________________________\n"
]
],
[
[
"**Creating the dataset to store the features**",
"_____no_output_____"
]
],
[
[
"dataset = HDF5DatasetWriter((len(imagesPath), 100352), config.FEATURES, dataKey=\"features\", bufSize=500)",
"The supplied 'outputPath' already exist \nDo you want overwrite (be sure)? Enter yes or no: yes\n"
],
[
"dataset.storeClassLabels(le.classes_)",
"_____no_output_____"
]
],
[
[
"**Extracting features on the batch**",
"_____no_output_____"
]
],
[
[
"widgets = [\"Extracting Features: \", progressbar.Percentage(), \" \",\n progressbar.Bar(), \" \", progressbar.ETA()]\npbar = progressbar.ProgressBar(maxval=len(imagesPath),\n widgets=widgets).start()\n\nfor i in np.arange(0, len(imagesPath), bs):\n batch_paths = imagesPath[i : i + bs]\n batch_labels = labels[i: i + bs]\n batch_images = []\n \n for (j, imagePath) in enumerate(batch_paths):\n \n image = load_img(imagePath, target_size=(224, 224))\n image = img_to_array(image)\n \n image = np.expand_dims(image, axis = 0)\n image = preprocess_input(image)\n \n batch_images.append(image)\n batch_images = np.vstack(batch_images)\n features = model.predict(batch_images, batch_size=bs)\n features = features.reshape((features.shape[0], 7*7*2048))\n dataset.add(features, batch_labels)\n pbar.update(i)\ndataset.close()\npbar.finish()",
"Extracting Features: 93% |################################## | ETA: 0:00:22\r"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb91be63414a16970c06b88ad0af7097203aa0e8
| 27,758 |
ipynb
|
Jupyter Notebook
|
workshops/cdnextcon_2017/Walk-Through.ipynb
|
modi975/tensorflow-workshop
|
55679d7e8819da6a974768ce2fdba358b71726b2
|
[
"Apache-2.0"
] | 1 |
2022-03-07T03:38:53.000Z
|
2022-03-07T03:38:53.000Z
|
workshops/cdnextcon_2017/Walk-Through.ipynb
|
modi975/tensorflow-workshop
|
55679d7e8819da6a974768ce2fdba358b71726b2
|
[
"Apache-2.0"
] | null | null | null |
workshops/cdnextcon_2017/Walk-Through.ipynb
|
modi975/tensorflow-workshop
|
55679d7e8819da6a974768ce2fdba358b71726b2
|
[
"Apache-2.0"
] | null | null | null | 25.396157 | 268 | 0.532495 |
[
[
[
"\n** ----- IMPORTANT ------ ** \nThe code presented here assumes that you're running TensorFlow v1.3.0 or higher, this was not released yet so the easiet way to run this is update your TensorFlow version to TensorFlow's master. \n\n\nTo do that go [here](https://github.com/tensorflow/tensorflow#installation) and then execute: \n`pip install --ignore-installed --upgrade <URL for the right binary for your machine>`. \n\nFor example, considering a Linux CPU-only running python2: \n`pip install --upgrade https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=cpu-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.1-cp27-none-linux_x86_64.whl`\n\n## Here is walk-through to help getting started with tensorflow\n\n\n1) Simple Linear Regression with low-level TensorFlow \n2) Simple Linear Regression with a canned estimator \n3) Playing with real data: linear regressor and DNN \n4) Building a custom estimator to classify handwritten digits (MNIST)\n\n### [What's next?](https://goo.gl/hZaLPA)",
"_____no_output_____"
],
[
"## Dependencies",
"_____no_output_____"
]
],
[
[
"from __future__ import absolute_import\nfrom __future__ import division\nfrom __future__ import print_function\n\nimport collections\n\n# tensorflow\nimport tensorflow as tf\nprint('Expected TensorFlow version is v1.3.0 or higher')\nprint('Your TensorFlow version:', tf.__version__)\n\n# data manipulation\nimport numpy as np\nimport pandas as pd\n\n# visualization\nimport matplotlib\nimport matplotlib.pyplot as plt\n%matplotlib inline\nmatplotlib.rcParams['figure.figsize'] = [12,8]",
"_____no_output_____"
]
],
[
[
"## 1) Simple Linear Regression with low-level TensorFlow",
"_____no_output_____"
],
[
"### Generating data\n\nThis function creates a noisy dataset that's roughly linear, according to the equation y = mx + b + noise.\n\nNotice that the expected value for m is 0.1 and for b is 0.3. This is the values we expect the model to predict.",
"_____no_output_____"
]
],
[
[
"def make_noisy_data(m=0.1, b=0.3, n=100):\n x = np.random.randn(n)\n noise = np.random.normal(scale=0.01, size=len(x))\n y = m * x + b + noise\n return x, y",
"_____no_output_____"
]
],
[
[
"Create training data",
"_____no_output_____"
]
],
[
[
"x_train, y_train = make_noisy_data()",
"_____no_output_____"
]
],
[
[
"Plot the training data",
"_____no_output_____"
]
],
[
[
"plt.plot(x_train, y_train, 'b.')",
"_____no_output_____"
]
],
[
[
"### The Model",
"_____no_output_____"
]
],
[
[
"# input and output\nx = tf.placeholder(shape=[None], dtype=tf.float32, name='x')\ny_label = tf.placeholder(shape=[None], dtype=tf.float32, name='y_label')\n\n# variables\nW = tf.Variable(tf.random_normal([1], name=\"W\")) # weight\nb = tf.Variable(tf.random_normal([1], name=\"b\")) # bias\n\n# actual model\ny = W * x + b",
"_____no_output_____"
]
],
[
[
"### The Loss and Optimizer\n\nDefine a loss function (here, squared error) and an optimizer (here, gradient descent).",
"_____no_output_____"
]
],
[
[
"loss = tf.reduce_mean(tf.square(y - y_label))\noptimizer = tf.train.GradientDescentOptimizer(learning_rate=0.1)\ntrain = optimizer.minimize(loss)",
"_____no_output_____"
]
],
[
[
"### The Training Loop and generating predictions",
"_____no_output_____"
]
],
[
[
"init = tf.global_variables_initializer()\n\nwith tf.Session() as sess:\n sess.run(init) # initialize variables\n for i in range(100): # train for 100 steps\n sess.run(train, feed_dict={x: x_train, y_label:y_train})\n\n x_plot = np.linspace(-3, 3, 101) # return evenly spaced numbers over a specified interval\n # using the trained model to predict values for the training data\n y_plot = sess.run(y, feed_dict={x: x_plot})\n\n # saving final weight and bias\n final_W = sess.run(W)\n final_b = sess.run(b)",
"_____no_output_____"
]
],
[
[
"### Visualizing predictions",
"_____no_output_____"
]
],
[
[
"plt.scatter(x_train, y_train)\nplt.plot(x_plot, y_plot, 'g')",
"_____no_output_____"
]
],
[
[
"### What is the final weight and bias?",
"_____no_output_____"
]
],
[
[
"print('W:', final_W, 'expected: 0.1')\nprint('b:', final_b, 'expected: 0.3')",
"_____no_output_____"
]
],
[
[
"## 2) Simple Linear Regression with a canned estimator ",
"_____no_output_____"
],
[
"### Input Pipeline",
"_____no_output_____"
]
],
[
[
"x_dict = {'x': x_train}\ntrain_input = tf.estimator.inputs.numpy_input_fn(x_dict, y_train,\n shuffle=True,\n num_epochs=None) # repeat forever",
"_____no_output_____"
]
],
[
[
"### Describe input feature usage",
"_____no_output_____"
]
],
[
[
"features = [tf.feature_column.numeric_column('x')] # because x is a real number",
"_____no_output_____"
]
],
[
[
"### Build and train the model",
"_____no_output_____"
]
],
[
[
"estimator = tf.estimator.LinearRegressor(features)\nestimator.train(train_input, steps = 1000)",
"_____no_output_____"
]
],
[
[
"### Generating and visualizing predictions",
"_____no_output_____"
]
],
[
[
"x_test_dict = {'x': np.linspace(-5, 5, 11)}\ndata_source = tf.estimator.inputs.numpy_input_fn(x_test_dict, shuffle=False)\n\npredictions = list(estimator.predict(data_source))\npreds = [p['predictions'][0] for p in predictions]\n\nfor y in predictions:\n print(y['predictions'])",
"_____no_output_____"
],
[
"plt.scatter(x_train, y_train)\nplt.plot(x_test_dict['x'], preds, 'g')",
"_____no_output_____"
]
],
[
[
"## 3) Playing with real data: linear regressor and DNN ",
"_____no_output_____"
],
[
"### Get the data\n\nThe Adult dataset is from the Census bureau and the task is to predict whether a given adult makes more than $50,000 a year based attributes such as education, hours of work per week, etc.\n\nBut the code here presented can be easilly aplicable to any csv dataset that fits in memory.\n\nMore about the data [here](https://archive.ics.uci.edu/ml/machine-learning-databases/adult/old.adult.names)",
"_____no_output_____"
]
],
[
[
"census_train_url = 'https://archive.ics.uci.edu/ml/machine-learning-databases/adult/adult.data'\ncensus_train_path = tf.contrib.keras.utils.get_file('census.train', census_train_url)\n\ncensus_test_url = 'https://archive.ics.uci.edu/ml/machine-learning-databases/adult/adult.test'\ncensus_test_path = tf.contrib.keras.utils.get_file('census.test', census_test_url)",
"_____no_output_____"
]
],
[
[
"### Load the data",
"_____no_output_____"
]
],
[
[
"column_names = [\n 'age', 'workclass', 'fnlwgt', 'education', 'education-num',\n 'marital-status', 'occupation', 'relationship', 'race', 'sex',\n 'capital-gain', 'capital-loss', 'hours-per-week', 'native-country',\n 'income'\n]\n\ncensus_train = pd.read_csv(census_train_path, index_col=False, names=column_names) \ncensus_test = pd.read_csv(census_train_path, index_col=False, names=column_names) \n\ncensus_train_label = census_train.pop('income') == \" >50K\" \ncensus_test_label = census_test.pop('income') == \" >50K\"",
"_____no_output_____"
],
[
"census_train.head(10)",
"_____no_output_____"
],
[
"census_train_label[:20]",
"_____no_output_____"
]
],
[
[
"### Input pipeline",
"_____no_output_____"
]
],
[
[
"train_input = tf.estimator.inputs.pandas_input_fn(\n census_train, \n census_train_label,\n shuffle=True, \n batch_size = 32, # process 32 examples at a time\n num_epochs=None,\n)",
"_____no_output_____"
],
[
"test_input = tf.estimator.inputs.pandas_input_fn(\n census_test, \n census_test_label, \n shuffle=True, \n num_epochs=1)",
"_____no_output_____"
],
[
"features, labels = train_input()\nfeatures",
"_____no_output_____"
]
],
[
[
"### Feature description",
"_____no_output_____"
]
],
[
[
"features = [\n tf.feature_column.numeric_column('hours-per-week'),\n tf.feature_column.bucketized_column(tf.feature_column.numeric_column('education-num'), list(range(25))),\n tf.feature_column.categorical_column_with_vocabulary_list('sex', ['male','female']),\n tf.feature_column.categorical_column_with_hash_bucket('native-country', 1000),\n]",
"_____no_output_____"
],
[
"estimator = tf.estimator.LinearClassifier(features, model_dir='census/linear',n_classes=2)",
"_____no_output_____"
],
[
"estimator.train(train_input, steps=5000)",
"_____no_output_____"
]
],
[
[
"### Evaluate the model",
"_____no_output_____"
]
],
[
[
"estimator.evaluate(test_input)",
"_____no_output_____"
]
],
[
[
"## DNN model",
"_____no_output_____"
],
[
"### Update input pre-processing",
"_____no_output_____"
]
],
[
[
"features = [\n tf.feature_column.numeric_column('education-num'),\n tf.feature_column.numeric_column('hours-per-week'),\n tf.feature_column.numeric_column('age'),\n tf.feature_column.indicator_column(\n tf.feature_column.categorical_column_with_vocabulary_list('sex',['male','female'])),\n tf.feature_column.embedding_column( # now using embedding!\n tf.feature_column.categorical_column_with_hash_bucket('native-country', 1000), 10)\n]",
"_____no_output_____"
],
[
"estimator = tf.estimator.DNNClassifier(hidden_units=[20,20], \n feature_columns=features, \n n_classes=2, \n model_dir='census/dnn')",
"_____no_output_____"
],
[
"estimator.train(train_input, steps=5000)",
"_____no_output_____"
],
[
"estimator.evaluate(test_input)",
"_____no_output_____"
]
],
[
[
"## Custom Input Pipeline using Datasets API",
"_____no_output_____"
],
[
"### Read the data",
"_____no_output_____"
]
],
[
[
"def census_input_fn(path):\n def input_fn(): \n dataset = (\n tf.contrib.data.TextLineDataset(path)\n .map(csv_decoder)\n .shuffle(buffer_size=100)\n .batch(32)\n .repeat())\n\n columns = dataset.make_one_shot_iterator().get_next()\n\n income = tf.equal(columns.pop('income'),\" >50K\") \n\n return columns, income\n \n return input_fn",
"_____no_output_____"
],
[
"csv_defaults = collections.OrderedDict([\n ('age',[0]),\n ('workclass',['']),\n ('fnlwgt',[0]),\n ('education',['']),\n ('education-num',[0]),\n ('marital-status',['']),\n ('occupation',['']),\n ('relationship',['']),\n ('race',['']),\n ('sex',['']),\n ('capital-gain',[0]),\n ('capital-loss',[0]),\n ('hours-per-week',[0]),\n ('native-country',['']),\n ('income',['']),\n])",
"_____no_output_____"
],
[
"def csv_decoder(line):\n parsed = tf.decode_csv(line, csv_defaults.values())\n return dict(zip(csv_defaults.keys(), parsed))\n ",
"_____no_output_____"
]
],
[
[
"### Try the input function",
"_____no_output_____"
]
],
[
[
"tf.reset_default_graph()\ncensus_input = census_input_fn(census_train_path)\ntraining_batch = census_input()",
"_____no_output_____"
],
[
"with tf.Session() as sess:\n features, high_income = sess.run(training_batch)",
"_____no_output_____"
],
[
"print(features['education'])",
"_____no_output_____"
],
[
"print(features['age'])",
"_____no_output_____"
],
[
"print(high_income)",
"_____no_output_____"
]
],
[
[
"## 4) Building a custom estimator to classify handwritten digits (MNIST)\n\n\nImage from: http://rodrigob.github.io/are_we_there_yet/build/images/mnist.png?1363085077",
"_____no_output_____"
]
],
[
[
"train,test = tf.contrib.keras.datasets.mnist.load_data()\nx_train,y_train = train \nx_test,y_test = test\n\nmnist_train_input = tf.estimator.inputs.numpy_input_fn({'x':np.array(x_train, dtype=np.float32)},\n np.array(y_train,dtype=np.int32),\n shuffle=True,\n num_epochs=None)\n\nmnist_test_input = tf.estimator.inputs.numpy_input_fn({'x':np.array(x_test, dtype=np.float32)},\n np.array(y_test,dtype=np.int32),\n shuffle=True,\n num_epochs=1)\n",
"_____no_output_____"
]
],
[
[
"### tf.estimator.LinearClassifier",
"_____no_output_____"
]
],
[
[
"estimator = tf.estimator.LinearClassifier([tf.feature_column.numeric_column('x',shape=784)], \n n_classes=10,\n model_dir=\"mnist/linear\")\nestimator.train(mnist_train_input, steps = 10000)",
"_____no_output_____"
],
[
"estimator.evaluate(mnist_test_input)",
"_____no_output_____"
]
],
[
[
"### Examine the results with [TensorBoard](http://0.0.0.0:6006)\n$> tensorboard --logdir mnnist/DNN",
"_____no_output_____"
]
],
[
[
"estimator = tf.estimator.DNNClassifier(hidden_units=[256],\n feature_columns=[tf.feature_column.numeric_column('x',shape=784)], \n n_classes=10,\n model_dir=\"mnist/DNN\")\nestimator.train(mnist_train_input, steps = 10000)",
"_____no_output_____"
],
[
"estimator.evaluate(mnist_test_input)",
"_____no_output_____"
],
[
"# Parameters\nBATCH_SIZE = 128\nSTEPS = 10000",
"_____no_output_____"
]
],
[
[
"## A Custom Model",
"_____no_output_____"
]
],
[
[
"def build_cnn(input_layer, mode):\n with tf.name_scope(\"conv1\"): \n conv1 = tf.layers.conv2d(inputs=input_layer,filters=32, kernel_size=[5, 5],\n padding='same', activation=tf.nn.relu)\n\n with tf.name_scope(\"pool1\"): \n pool1 = tf.layers.max_pooling2d(inputs=conv1, pool_size=[2, 2], strides=2)\n\n with tf.name_scope(\"conv2\"): \n conv2 = tf.layers.conv2d(inputs=pool1,filters=64, kernel_size=[5, 5],\n padding='same', activation=tf.nn.relu)\n\n with tf.name_scope(\"pool2\"): \n pool2 = tf.layers.max_pooling2d(inputs=conv2, pool_size=[2, 2], strides=2)\n\n with tf.name_scope(\"dense\"): \n pool2_flat = tf.reshape(pool2, [-1, 7 * 7 * 64])\n dense = tf.layers.dense(inputs=pool2_flat, units=1024, activation=tf.nn.relu)\n\n with tf.name_scope(\"dropout\"): \n is_training_mode = mode == tf.estimator.ModeKeys.TRAIN\n dropout = tf.layers.dropout(inputs=dense, rate=0.4, training=is_training_mode)\n\n logits = tf.layers.dense(inputs=dropout, units=10)\n\n return logits\n",
"_____no_output_____"
],
[
"def model_fn(features, labels, mode):\n # Describing the model\n input_layer = tf.reshape(features['x'], [-1, 28, 28, 1])\n \n tf.summary.image('mnist_input',input_layer)\n \n logits = build_cnn(input_layer, mode)\n \n # Generate Predictions\n classes = tf.argmax(input=logits, axis=1)\n predictions = {\n 'classes': classes,\n 'probabilities': tf.nn.softmax(logits, name='softmax_tensor')\n }\n\n if mode == tf.estimator.ModeKeys.PREDICT:\n # Return an EstimatorSpec object\n return tf.estimator.EstimatorSpec(mode=mode, predictions=predictions)\n\n with tf.name_scope('loss'):\n loss = tf.nn.sparse_softmax_cross_entropy_with_logits(labels=labels, logits=logits)\n \n loss = tf.reduce_sum(loss)\n tf.summary.scalar('loss', loss)\n \n with tf.name_scope('accuracy'):\n accuracy = tf.cast(tf.equal(tf.cast(classes,tf.int32),labels),tf.float32)\n accuracy = tf.reduce_mean(accuracy)\n tf.summary.scalar('accuracy', accuracy)\n\n # Configure the Training Op (for TRAIN mode)\n if mode == tf.estimator.ModeKeys.TRAIN:\n train_op = tf.contrib.layers.optimize_loss(\n loss=loss,\n global_step=tf.train.get_global_step(),\n learning_rate=1e-4,\n optimizer='Adam')\n\n return tf.estimator.EstimatorSpec(mode=mode, predictions=predictions,\n loss=loss, train_op=train_op)\n\n # Configure the accuracy metric for evaluation\n eval_metric_ops = {\n 'accuracy': tf.metrics.accuracy(\n classes,\n input=labels)\n }\n\n return tf.estimator.EstimatorSpec(mode=mode, predictions=predictions,\n loss=loss, eval_metric_ops=eval_metric_ops)\n",
"_____no_output_____"
]
],
[
[
"## Runs estimator",
"_____no_output_____"
]
],
[
[
"# create estimator\nrun_config = tf.contrib.learn.RunConfig(model_dir='mnist/CNN')\nestimator = tf.estimator.Estimator(model_fn=model_fn, config=run_config)\n\n# train for 10000 steps\nestimator.train(input_fn=mnist_train_input, steps=10000)\n\n# evaluate\nestimator.evaluate(input_fn=mnist_test_input)\n\n# predict\npreds = estimator.predict(input_fn=test_input_fn)",
"_____no_output_____"
]
],
[
[
"## Distributed tensorflow: using experiments",
"_____no_output_____"
]
],
[
[
"# Run an experiment\nfrom tensorflow.contrib.learn.python.learn import learn_runner\n\n# Enable TensorFlow logs\ntf.logging.set_verbosity(tf.logging.INFO)",
"_____no_output_____"
],
[
"# create experiment\ndef experiment_fn(run_config, hparams):\n # create estimator\n estimator = tf.estimator.Estimator(model_fn=model_fn,\n config=run_config)\n return tf.contrib.learn.Experiment(\n estimator,\n train_input_fn=train_input_fn,\n eval_input_fn=test_input_fn,\n train_steps=STEPS\n )\n\n# run experiment\nlearn_runner.run(experiment_fn,\n run_config=run_config)",
"_____no_output_____"
]
],
[
[
"### Examine the results with [TensorBoard](http://0.0.0.0:6006)\n$> tensorboard --logdir mnist/CNN",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
]
] |
cb91c072143bf31d080186d3046984145b63a619
| 8,629 |
ipynb
|
Jupyter Notebook
|
notebooks/astropy lomb-scargle.ipynb
|
sejones22/usrp20
|
5a3f6771f0927e63921c35281096508d8f893c72
|
[
"MIT"
] | null | null | null |
notebooks/astropy lomb-scargle.ipynb
|
sejones22/usrp20
|
5a3f6771f0927e63921c35281096508d8f893c72
|
[
"MIT"
] | null | null | null |
notebooks/astropy lomb-scargle.ipynb
|
sejones22/usrp20
|
5a3f6771f0927e63921c35281096508d8f893c72
|
[
"MIT"
] | null | null | null | 110.628205 | 7,172 | 0.888052 |
[
[
[
"import numpy as np\nrand = np.random.RandomState(42)\nt = 100 * rand.rand(100)\ny = np.sin(2 * np.pi * t) + 0.1 * rand.randn(100)",
"_____no_output_____"
],
[
"from astropy.timeseries import LombScargle\nfrequency, power = LombScargle(t, y).autopower()",
"_____no_output_____"
],
[
"import matplotlib.pyplot as plt \nplt.plot(1/frequency, power); ",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code"
]
] |
cb91c32fc9f27e1dfc23e8e622e1b0a71a6ed615
| 54,836 |
ipynb
|
Jupyter Notebook
|
examples/toggle-switch example.ipynb
|
ayush9pandey/sbmlReduce
|
0612bbb1def5c1aa5f8d3ca125998d0fefa46e53
|
[
"BSD-3-Clause"
] | 3 |
2020-08-10T20:21:07.000Z
|
2021-09-21T19:47:41.000Z
|
examples/toggle-switch example.ipynb
|
ayush9pandey/sbmlReduce
|
0612bbb1def5c1aa5f8d3ca125998d0fefa46e53
|
[
"BSD-3-Clause"
] | 13 |
2020-10-14T21:46:17.000Z
|
2021-09-20T22:18:32.000Z
|
examples/toggle-switch example.ipynb
|
ayush9pandey/AutoReduce
|
0612bbb1def5c1aa5f8d3ca125998d0fefa46e53
|
[
"BSD-3-Clause"
] | 1 |
2020-07-21T23:19:47.000Z
|
2020-07-21T23:19:47.000Z
| 180.976898 | 15,896 | 0.888559 |
[
[
[
"from autoreduce import *\nimport numpy as np\nfrom sympy import symbols",
"_____no_output_____"
],
[
"# Post conservation law and other approximations phenomenological model at the RNA level\nn = 4 # Number of states \nnouts = 2 # Number of outputs\n\n# Inputs by user \nx_init = np.zeros(n)\nn = 4 # Number of states \ntimepoints_ode = np.linspace(0, 100, 100)\nC = [[0, 0, 1, 0], [0, 0, 0, 1]]\nnstates_tol = 3\nerror_tol = 0.3\n# System dynamics symbolically\n\n# params = [100, 50, 10, 5, 5, 0.02, 0.02, 0.01, 0.01]\n# params = [1, 1, 5, 0.1, 0.2, 1, 1, 100, 100] # Parameter set for which reduction doesn't work\n# K,b_t,b_l,d_t,d_l,del_t,del_l,beta_t,beta_l = params\n\nx0 = symbols('x0')\nx1 = symbols('x1')\nx2 = symbols('x2')\nx3 = symbols('x3')\nx = [x0, x1, x2, x3]\n\nK = symbols('K')\nb_t = symbols('b_t')\nb_l = symbols('b_l')\nd_t = symbols('d_t')\nd_l = symbols('d_l')\ndel_t = symbols('del_t')\ndel_l = symbols('del_l')\nbeta_t = symbols('beta_t')\nbeta_l = symbols('beta_l')\nparams = [K,b_t,b_l,d_t,d_l,del_t,del_l,beta_t,beta_l]\nf0 = K * b_t**2/(b_t**2 + x[3]**2) - d_t * x[0]\nf1 = K * b_l**2/(b_l**2 + x[2]**2) - d_l * x[1]\nf2 = beta_t * x[0] - del_t * x[2]\nf3 = beta_l * x[1] - del_l * x[3]\nf = [f0,f1,f2,f3]\n# parameter values\nparams_values = [100, 50, 10, 5, 5, 0.02, 0.02, 0.01, 0.01]\nsys = System(x, f, params = params, params_values = params_values, C = C, x_init = x_init)",
"_____no_output_____"
],
[
"from autoreduce.utils import get_ODE\nsys_ode = get_ODE(sys, timepoints_ode)\nsol = sys_ode.solve_system().T\ntry:\n import matplotlib.pyplot as plt\n plt.plot(timepoints_ode, np.transpose(np.array(C)@sol))\n plt.xlabel('Time')\n plt.ylabel('[Outputs]')\n plt.show()\nexcept:\n print('Plotting libraries missing.')",
"_____no_output_____"
],
[
"from autoreduce.utils import get_SSM\ntimepoints_ssm = np.linspace(0,100,100)\nsys_ssm = get_SSM(sys, timepoints_ssm)\nSs = sys_ssm.compute_SSM() # len(timepoints) x len(params) x len(states)\nout_Ss = []\nfor i in range(len(params)):\n out_Ss.append((np.array(C)@(Ss[:,i,:].T)))\nout_Ss = np.reshape(np.array(out_Ss), (len(timepoints_ssm), len(params), nouts))",
"SSM Progress: |██████████████████████████████████████████████████| 100.0% Complete\n"
],
[
"try:\n import seaborn as sn\n import matplotlib.pyplot as plt\n for j in range(nouts):\n sn.heatmap(out_Ss[:,:,j].T)\n plt.xlabel('Time')\n plt.ylabel('Parameters')\n plt.title('Sensitivity of output[{0}] with respect to all parameters'.format(j))\n plt.show()\nexcept:\n print('Plotting libraries missing.')",
"_____no_output_____"
],
[
"from autoreduce.utils import get_reducible\ntimepoints_ssm = np.linspace(0,100,10)\ntimepoints_ode = np.linspace(0, 100, 100)\nsys_reduce = get_reducible(sys, timepoints_ode, timepoints_ssm)\nresults = sys_reduce.reduce_simple()",
"Successful time-scale separation solution obtained with states: [x2, x3]!\nSSM Progress: |██████████████████████████████████████████████████| 100.0% Complete\nSSM Progress: |██████████████████████████████████████████████████| 100.0% Complete\n"
],
[
"list(results.keys())[0].f[1]",
"_____no_output_____"
],
[
"reduced_system, collapsed_system = sys_reduce.solve_timescale_separation([x0,x1], fast_states = [x3, x2])",
"Successful time-scale separation solution obtained with states: [x0, x1]!\n"
],
[
"reduced_system.f[1]",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb91c404ef8af127862bf6b3468b2dbb19fd6f54
| 116,952 |
ipynb
|
Jupyter Notebook
|
Data_visualization_practice.ipynb
|
oubush/my_cheat_sheets
|
16e09e3d87be8dc6b555ec1d3620419a026127b8
|
[
"MIT"
] | null | null | null |
Data_visualization_practice.ipynb
|
oubush/my_cheat_sheets
|
16e09e3d87be8dc6b555ec1d3620419a026127b8
|
[
"MIT"
] | null | null | null |
Data_visualization_practice.ipynb
|
oubush/my_cheat_sheets
|
16e09e3d87be8dc6b555ec1d3620419a026127b8
|
[
"MIT"
] | null | null | null | 338.991304 | 15,132 | 0.935991 |
[
[
[
"# Matplotlib",
"_____no_output_____"
],
[
"Курс \"Программирование на Python\" (Stepik.org)",
"_____no_output_____"
]
],
[
[
"from pylab import *",
"_____no_output_____"
],
[
"x = linspace(0, 5, 10)\ny = x ** 2\nprint(x)\nprint(y)",
"[0. 0.55555556 1.11111111 1.66666667 2.22222222 2.77777778\n 3.33333333 3.88888889 4.44444444 5. ]\n[ 0. 0.30864198 1.2345679 2.77777778 4.9382716 7.71604938\n 11.11111111 15.12345679 19.75308642 25. ]\n"
],
[
"figure()\nplot(x,y,'r')\nxlabel('x')\nylabel('y')\ntitle('title')\nshow()",
"_____no_output_____"
],
[
"%matplotlib inline",
"_____no_output_____"
],
[
"figure()\nplot(x,y,'r')\nxlabel('x')\nylabel('y')\ntitle('title')\nshow()",
"_____no_output_____"
],
[
"fig = plt.figure()\n\naxes = fig.add_axes([0.1, 0.1, 0.8, 0.8]) # left, bottom, width, height (range 0 to 1)\n\naxes.plot(x,y,'r')\n\naxes.set_xlabel('x')\naxes.set_ylabel('y')\naxes.set_title('title');",
"_____no_output_____"
],
[
"fig = plt.figure()\n\naxes1 = fig.add_axes([0.1, 0.1, 0.8, 0.8]) # left, bottom, width, height (range 0 to 1)\naxes2 = fig.add_axes([0.2, 0.5, 0.4, 0.3]) # left, bottom, width, height (range 0 to 1)\n\naxes1.plot(x,y,'r')\naxes1.set_xlabel('x')\naxes1.set_ylabel('y')\naxes1.set_title('title')\n\naxes2.plot(y,x,'g')\naxes2.set_xlabel('y')\naxes2.set_ylabel('x')\naxes2.set_title('insert title');",
"_____no_output_____"
],
[
"fig, axes = plt.subplots(nrows = 1, ncols = 2)\n\nfor ax, color in zip(axes, ['r', 'g']):\n ax.plot(x,y,color)\n ax.set_xlabel('x')\n ax.set_ylabel('y')\n ax.set_title('title')",
"_____no_output_____"
],
[
"fig, axes = plt.subplots(nrows = 1, ncols = 2)\n\nfor ax, color in zip(axes, ['r', 'g']):\n ax.plot(x,y,color)\n ax.set_xlabel('x')\n ax.set_ylabel('y')\n ax.set_title('title')\nfig.tight_layout()",
"_____no_output_____"
],
[
"fig, axes = plt.subplots(figsize=(12,3))\n\naxes.plot(x,y,'b')\naxes.set_xlabel('x')\naxes.set_ylabel('y')\naxes.set_title('title');",
"_____no_output_____"
],
[
"fig, ax = plt.subplots()\n\nax.plot(x,x**2,label=\"y = x**2\")\nax.plot(x,x**3,label=\"y = x**3\")\nax.legend(loc=2) # upper left corner\nax.set_xlabel('x')\nax.set_ylabel('y')\nax.set_title('title');",
"_____no_output_____"
],
[
"from numpy import random\nn = random.randn(100000)\nfig, ax = plt.subplots(1,2,figsize=(12,4))\n\nax[0].hist(n)\nax[0].set_title('hist')\nax[0].set_xlim((min(n), max(n)))\n\nax[1].hist(n, cumulative=True, bins=50)\nax[1].set_title('cumulative hist')\nax[1].set_xlim((min(n), max(n)));",
"_____no_output_____"
]
]
] |
[
"markdown",
"code"
] |
[
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb91c527adbcdb234bac9a707f3d22aae416b27e
| 2,530 |
ipynb
|
Jupyter Notebook
|
examples/diffusion_decay.ipynb
|
bjodah/chemreac
|
dbe38a10cf6b88e66192bcc998721b61aabbd9dc
|
[
"BSD-2-Clause"
] | 14 |
2015-03-11T21:46:15.000Z
|
2020-06-06T16:01:38.000Z
|
examples/diffusion_decay.ipynb
|
bjodah/chemreac
|
dbe38a10cf6b88e66192bcc998721b61aabbd9dc
|
[
"BSD-2-Clause"
] | 20 |
2015-01-21T16:11:36.000Z
|
2020-01-06T10:30:46.000Z
|
examples/diffusion_decay.ipynb
|
chemreac/chemreac
|
dbe38a10cf6b88e66192bcc998721b61aabbd9dc
|
[
"BSD-2-Clause"
] | 3 |
2015-08-13T12:06:17.000Z
|
2021-12-17T01:12:20.000Z
| 21.810345 | 126 | 0.541107 |
[
[
[
"import numpy as np\nimport matplotlib.pyplot as plt\nfrom chempy import ReactionSystem, Substance\nfrom chemreac import ReactionDiffusion\nfrom chemreac.integrate import run\nfrom chemreac.util.plotting import plot_C_vs_x\n%matplotlib notebook",
"_____no_output_____"
],
[
"rsys = ReactionSystem.from_string(\"\"\"\nA -> B; 3\n-> A; Radiolytic(2)\n\"\"\", substance_factory=Substance, checks=())\nrsys",
"_____no_output_____"
],
[
"N = 128\nfields = [[0]*N]\nfields[0][0] = 1\nrd = ReactionDiffusion.from_ReactionSystem(rsys, N=N, fields=fields, D=rsys.as_per_substance_array({'A': 4e-9, 'B': 0}))",
"_____no_output_____"
],
[
"tout = np.logspace(-12, 3)\nintegr = run(rd, C0=[1e-99]*N*2, tout=tout)",
"_____no_output_____"
],
[
"plot_C_vs_x(rd, tout, integr.Cout, rd.substance_names, -1)",
"_____no_output_____"
],
[
"from ipywidgets import interact\nfrom chemreac.util.anim import animate_C_vs_x",
"_____no_output_____"
],
[
"fig, ax = plt.subplots(1, 1, figsize=(5, 2))\nax.set_xscale('log')\nax.set_yscale('log')\nax.set_ylim([1e-50, 1e4])\nupdate = animate_C_vs_x(integr, ax)\ninteract(update, ti=(0, integr.tout.size-1))",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb91d1199f88bbf72f420c36f6b0d69f2179bbea
| 32,094 |
ipynb
|
Jupyter Notebook
|
Function Operation on List.ipynb
|
teamindelible/Getting-Started-with-Python
|
e9c34cfc774c8210b909dbf9f887158e7789071a
|
[
"MIT"
] | null | null | null |
Function Operation on List.ipynb
|
teamindelible/Getting-Started-with-Python
|
e9c34cfc774c8210b909dbf9f887158e7789071a
|
[
"MIT"
] | null | null | null |
Function Operation on List.ipynb
|
teamindelible/Getting-Started-with-Python
|
e9c34cfc774c8210b909dbf9f887158e7789071a
|
[
"MIT"
] | null | null | null | 27.221374 | 138 | 0.48439 |
[
[
[
"# Define a smallest_number function that accepts a list of numbers.\n# It should return the smallest value in the list.\n\ndef smallest_number(numbers):\n numbers = list(numbers)\n smallest = numbers[0]\n for number in numbers:\n if number < smallest:\n smallest = number\n return smallest\n\nsmallest_number((1, 2, 3)) #=> 1\n# smallest_number([3, 2, 1]) => 1\n# smallest_number([4, 5, 4]) => 4\n# smallest_number([-3, -2, -1]) => -3",
"_____no_output_____"
],
[
"# Define a sum_of_lengths function that accepts a list of strings.\n# The function should return the sum of the string lengths.\n\ndef sum_of_lengths(list_of_strings):\n sums = 0\n for string in list_of_strings:\n sums += len(string)\n return sums \n\nsum_of_lengths([\"Hello\", \"Bob\"]) #=> 8\n#sum_of_lengths([\"Nonsense\"]) #=> 8",
"_____no_output_____"
],
[
"# Define a product function that accepts a list of numbers.\n# The function should return the product of the numbers.\n# The list will always have at least one value\n#\ndef product(numbers):\n products = 1\n for number in numbers:\n products *= number\n return products\n\n\n#product([1, 2, 3]) #=> 6\nproduct([4, 5, 6, 7]) #=> 840\n#product([10]) #=> 10",
"_____no_output_____"
],
[
"# Define a function concatenate that accepts a list of strings. \n# The function should return a concatenated string which consists of all list elements whose length is greater than 2 characters.\n\ndef concatenate(lists):\n concat = \"\"\n for string in lists:\n if len(string) > 2:\n concat += string\n return concat\n\n#concatenate([\"abc\", \"def\", \"ghi\"]) # => \"abcdefghi\"\nconcatenate([\"abc\", \"de\", \"fgh\", \"ic\"]) #=> \"abcfgh\"\n# concatenate([\"ab\", \"cd\", \"ef\", \"gh\"]) => \"\"",
"_____no_output_____"
],
[
"# Write a function super_sum that accepts a list of strings. \n# The function should sum the index positions of the first occurence of the letter “s” in each word. \n# Not every word is guaranteed to have an “s”.\n# Don’t use the sum keyword as it’s a built-in keyword.\n#\n\ndef super_sum(string_list):\n index = 0\n for string in string_list:\n index += string.find('s')\n return index\n \n# super_sum([]) => 0\n# super_sum([\"mustache\"]) => 2\n#super_sum([\"mustache\", \"greatest\"]) #=> 8\n#super_sum([\"mustache\", \"pessimist\"]) #=> 4\nsuper_sum([\"mustache\", \"greatest\", \"almost\"]) #=> 12",
"_____no_output_____"
],
[
"# Define an in_list function that accepts a list of strings and a separate string.\n# Return the index where the string exists in the list. If the string does not exist, return -1.\n# Do NOT use the find or index methods.\n \ndef in_list(values, target): \n for idx, value in enumerate(values):\n if target == value:\n return idx\n return -1\n\nstrings = [\"enchanted\", \"sparks fly\", \"long live\"]\nin_list(strings, \"enchanted\") #==> 0\n# in_list(strings, \"sparks fly\") ==> 1\n# in_list(strings, \"fifteen\") ==> -1\n# in_list(strings, \"love story\") ==> -1",
"_____no_output_____"
],
[
"in_list(strings, \"fifteen\") # ==> -1",
"_____no_output_____"
],
[
"# Define a sum_of_values_and_indices function that accepts a list of numbers. \n# It should return the sum of all of the elements along with their index values.\n#\ndef sum_of_values_and_indices(numbers):\n total = 0\n for idx, num in enumerate(numbers):\n total += (idx + num)\n return total\n\nsum_of_values_and_indices([1, 2, 3]) #=> (1 + 0) + (2 + 1) + (3 + 2) => 9\n# sum_of_values_and_indices([0, 0, 0, 0]) => 6\n# sum_of_values_and_indices([]) => 0",
"_____no_output_____"
],
[
"# Declare a function same_index_values that accepts two lists.\n# The function should return a list of the index positions in which the two lists have equal elements\n#\ndef same_index_values(list_1, list_2):\n position = []\n for idx, i in enumerate(list_1):\n if i == list_2[idx]:\n position.append(idx)\n return position\n\n#same_values([1, 2, 3], [3, 2, 1]) => [1]\nsame_index_values([\"a\", \"b\", \"c\", \"d\"], [\"c\", \"b\", \"a\", \"d\"]) #=> [1, 3]",
"_____no_output_____"
],
[
"# Declare a sum_from function that accepts two numbers as arguments.\n# The second number will always be greater than the first number.\n# The function should return the sum of all numbers from the first number to the second number (inclusive).\n#\ndef sum_from(number_1, number_2):\n summation = 0\n for number in range(number_1, number_2 + 1):\n summation += number\n return summation\n\n# sum_from(1, 2) # 1 + 2 => 3\nsum_from(1, 5) # 1 + 2 + 3 + 4 + 5 => 15\n# sum_from(3, 8) # 3 + 4 + 5 + 6 + 7 + 8 => 33\n# sum_from(9, 12) # 9 + 10 + 11 + 12 => 42",
"_____no_output_____"
],
[
"# Declare a length_match function that accepts a list of strings and an integer.\n# It should return a count of the number of strings whose length is equal to the number.\n#\ndef length_match(strings, interger):\n count = 0\n for idx, string in enumerate(strings):\n if len(string) == interger:\n count += 1\n return count\n \nlength_match([\"cat\", \"dog\", \"kangaroo\", \"mouse\"], 3) #=> 2\n# length_match([\"cat\", \"dog\", \"kangaroo\", \"mouse\"], 5)) => 1\n# length_match([\"cat\", \"dog\", \"kangaroo\", \"mouse\"], 4)) => 0\n# length_match([], 5)) => 0",
"_____no_output_____"
],
[
"# Given the tuple below, destructure the three values and\n# assign them to position, city and salary variables\n# Do NOT use index positions (i.e. job_opening[1])\njob_opening = (\"Software Engineer\", \"New York City\", 100000)\n\nposition, city, salary = job_opening\n\n# Given the tuple below, \n# - destructure the first value and assign it to a street variable\n# - destructure the last value and assign it to a zip_code variable\n# - destructure the middle two values into a list and assign it to a city_and_state variable\naddress = (\"35 Elm Street\", \"San Francisco\", \"CA\", \"94107\")\n\nstreet, *city_and_state, zip_code = address",
"_____no_output_____"
],
[
"# Declare a sum_of_evens_and_odds function that accepts a tuple of numbers.\n# It should return a tuple with two numeric values:\n# -- the sum of the even numbers\n# -- the sum of the odd numbers.\n\ndef sum_of_evens_and_odds(numbers):\n sum_odd = 0\n sum_even = 0\n sums = 0, 0\n #sums = 0\n for num in numbers:\n if num%2 == 1:\n sum_odd += num\n else:\n sum_even += num\n sums = sum_even, sum_odd\n return sums\n\nsum_of_evens_and_odds((1, 2, 3, 4)) #=> (6, 4)\n#sum_of_evens_and_odds((1, 3, 5)) #=> (0, 9)\n#sum_of_evens_and_odds((2, 4, 6)) #=> (12, 0)",
"_____no_output_____"
],
[
"# this function returns the smallest and the highest value in a list\n\ndef s_and_h(numbers):\n smallest = numbers[0]\n highest = numbers[0]\n smallest_and_highest = 0, 0\n for number in numbers:\n if number < smallest:\n smallest = number\n for number in numbers:\n if number > highest:\n highest = number\n smallest_and_highest = (smallest, highest)\n return smallest_and_highest",
"_____no_output_____"
],
[
"s_and_h([9,1,2,3,0,4,5,6])",
"_____no_output_____"
],
[
"# Declare a function right_words that accepts a list of words and a number.\n# Return a new list with the words that have a length equal to the number.\n# Do NOT use list comprehension.\n\ndef right_words(words, number):\n new_list = []\n for word in words:\n if len(word) == number:\n new_word = new_list.append(word)\n return new_list\n\n#right_words(['cat', 'dog', 'bean', 'ace', 'heart'], 3)\nright_words(['cat', 'dog', 'bean', 'ace', 'heart'], 5) #=> ['heart']",
"_____no_output_____"
],
[
"# Define an only_evens function that accepts a list of numbers. \n# It should return a new list consisting of only the even numbers from the original list.\n\ndef only_evens(numbers):\n even_numbers = []\n for number in numbers:\n if number % 2 == 0:\n even_numbers.append(number)\n return even_numbers\n\nonly_evens([4, 8, 15, 16, 23, 42]) #=> [4, 8, 16, 42]\n# only_evens([1, 3, 5]) => []\n# only_evens([]) => []",
"_____no_output_____"
],
[
"# Define a long_strings function that accepts a list of strings. \n# It should return a new list consisting of only the strings that have 5 characters or more.\n\ndef long_strings(strings):\n five_plus_strings = []\n for string in strings:\n if len(string) >= 5:\n five_plus_strings.append(string)\n return five_plus_strings\n\nlong_strings([\"Hello\", \"Goodbye\", \"Sam\"]) #=> [\"Hello\", \"Goodbye\"]\n# long_strings([\"Ace\", \"Cat\", \"Job\"] => []\n# long_strings([] => []\n",
"_____no_output_____"
],
[
"# Write a factors function that accepts a positive whole number\n# It should return a list of all of the number's factors in ascending order\n# HINT: Could the range function be helpful here? Or maybe a while loop?\n\ndef factors(pwn):\n factor_s = []\n for number in range(1, pwn + 1):\n if pwn % number == 0:\n factor_s.append(number)\n return factor_s\n\n# factors(1) => [1]\n# factors(2) => [1, 2]\n# factors(10) => [1, 2, 5, 10]\nfactors(64) #=> [1, 2, 4, 8, 16, 32, 64]",
"_____no_output_____"
],
[
"# Declare a delete_all function that accepts a list of strings and a target string\n# Remove all occurrences of the target string from the list and return it\n\ndef delete_all(strings, target):\n new_list = []\n for string in strings:\n if string != target:\n new_list.append(string) \n return new_list\n\n#delete_all([1, 3, 5], 3) #=> [1, 5]\ndelete_all([5, 3, 5], 5) #=> [3]\n#delete_all([4, 4, 4], 4) #=> []\n#delete_all([4, 4, 4], 6) #=> [4, 4, 4]",
"_____no_output_____"
],
[
"# Declare a push_or_pop function that accepts a list of numbers\n# Build up and return a new list by iterating over the list of numbers\n# If a number is greater than 5, add it to the end of the new list\n# If a number is less than or equal to 5, remove the last element added to the new list\n# Assume the order of numbers in the argument will never require removing from an empty list\n\ndef push_or_pop(numbers):\n new_list = []\n for number in numbers:\n if number > 5:\n new_list.append(number)\n else:\n new_list.pop(-1)\n return new_list\n\n# push_or_pop([10]) => [10]\npush_or_pop([10, 4]) #=> []\n#push_or_pop([10, 20, 30]) #=> [10, 20, 30]\n#push_or_pop([10, 20, 3, 30, 4, 50, 60]) #=>[10, 50, 60]\npush_or_pop([10, 20, 2, 30]) #=> [10, 30]",
"_____no_output_____"
],
[
"bf = [\"egg\", \"fish\", \"meat\"]\nlc = [\"water\", \"beans\", \"soup\"]\ndn = [\"fat\", \"thin\", \"short\"]\n\nprint(list(zip(bf, lc, dn)))\n\nfor b, l, d in zip(bf, lc, dn):\n print(f\"My meal today were {b}, {l} and {d}.\")",
"[('egg', 'water', 'fat'), ('fish', 'beans', 'thin'), ('meat', 'soup', 'short')]\nMy meal today were egg, water and fat.\nMy meal today were fish, beans and thin.\nMy meal today were meat, soup and short.\n"
],
[
"food = [\n [\"egg\", \"fish\", \"meat\"], \n [\"water\", \"beans\", \"soup\"], \n [\"fat\", \"thin\", \"short\"]\n]\n\nprint(len(food)) ##length of list\nprint(food[0]) ##accessing list in food\nprint(food[1])\nprint(food[2])\nprint(len(food[0]))\n\nprint(food[1][1]) ##accessing elements in list of lists",
"3\n['egg', 'fish', 'meat']\n['water', 'beans', 'soup']\n['fat', 'thin', 'short']\n3\nbeans\n"
],
[
"food = [\n [\"egg\", \"fish\", \"meat\"], \n [\"water\", \"beans\", \"soup\"], \n [\"fat\", \"thin\", \"short\"]\n]\n\nall_food = []\nfor f in food:\n for i in f:\n all_food.append(i)\n \nprint(all_food)",
"['egg', 'fish', 'meat', 'water', 'beans', 'soup', 'fat', 'thin', 'short']\n"
],
[
"# Define a nested_sum function that accepts a list of lists of numbers\n# The function should return the sum of the values\n# The list may contain empty lists\n#\ndef nested_sum(numbers):\n sums = 0\n for lists in numbers:\n for number in lists:\n sums = sums + number\n return sums\n\n#nested_sum([[1, 2, 3], [4, 5]]) #=> # 15\n# nested_sum([[1, 2, 3], [], [], [4], [5]]) => # 15\nnested_sum([[]]) #=> 0",
"_____no_output_____"
],
[
"# Define a fancy_concatenate function that accepts a list of lists of strings\n# The function should return a concatenated string\n# The strings in a list should only be concatenated if the length of the list is 3\n#\ndef fancy_concatenate(strings):\n concat = \"\"\n for lists in strings:\n for string in lists:\n if len(lists) == 3:\n concat += string\n return concat \n\n\n \n# fancy_concatenate([[\"A, \"B\", \"C\"]]) => \"ABC\"\n#fancy_concatenate([[\"A\", \"B\", \"C\"], [\"D\", \"E\", \"F\"]]) #=> \"ABCDEF\"\nfancy_concatenate([[\"A\", \"B\", \"C\"], [\"D\", \"E\", \"F\", \"G\"]]) #=> \"ABC\"\n# fancy_concatenate([[\"A\", \"B\", \"C\"], [\"D\", \"E\"]]) => \"ABC\"\n# fancy_concatenate([[\"A\", \"B\"], [\"C\", \"D\"]]) => \"\"",
"_____no_output_____"
],
[
"donuts = [\"Boston Cream\",\n \"Jelly\", \n \"Vanilla Cream\", \n \"Chcocolate Cream\"]\n\nprint([donut for donut in donuts if \"Cream\" in donut]) \nprint([donut.split(\" \")[0] for donut in donuts if \"Cream\" in donut]) ",
"['Boston Cream', 'Vanilla Cream', 'Chcocolate Cream']\n['Boston', 'Vanilla', 'Chcocolate']\n"
],
[
"# Uncomment the commented lines of code bellow and complete the list comprehension logic\n\n# The floats variable should store the floating point values for each string in the values list.\nvalues = [\"3.14\", \"9.99\", \"567.324\", \"5.678\"]\nfloats = [float(value) for value in values]\n\n# The letters variable should store a list of 5 strings. \n# Each of the strings should be a character from name concatenated together 3 times.\n# i.e. ['BBB', 'ooo', 'rrr', 'iii', 'sss']\nname = \"Boris\"\nletters = [letter*3 for letter in name]\n\n# The 'lengths' list should store a list with the lengths\n# of each string in the 'elements' list\nelements = [\"Hydrogen\", \"Helium\", \"Lithium\", \"Boron\", \"Carbon\"]\nlengths = [len(element) for element in elements]",
"_____no_output_____"
],
[
"# Declare a function destroy_elements that accepts two lists.\n# It should return a list of all elements from the first list that are NOT contained in the second list.\n# Use list comprehension in your solution.\n#\ndef destroy_elements(list_1, list_2):\n #return [element for element in list_1 if element not in list_2] #list comprehension\n new_list = []\n for element in list_1:\n if element not in list_2:\n new_list.append(element)\n return new_list\n\n#destroy_elements([1, 2, 3], [1, 2]) #=> [3]\n#destroy_elements([1, 2, 3], [1, 2, 3]) #=> []\ndestroy_elements([1, 2, 3], [4, 5]) #=> [1, 2, 3]",
"_____no_output_____"
],
[
"def delete_all(strings, target):\n return list(filter(lambda x: x != target, strings))\n\ndelete_all([1, 3, 5], 3) #=> [1, 5]\n#delete_all([5, 3, 5], 5) #=> [3]\n#delete_all([4, 4, 4], 4) #=> []\n#delete_all([4, 4, 4], 6) #=> [4, 4, 4]",
"_____no_output_____"
],
[
"def delete_all(strings, target):\n return [d for d in strings if d != target]\n\n# delete_all([1, 3, 5], 3) => [1, 5]\n#delete_all([5, 3, 5], 5) #=> [3]\ndelete_all([4, 4, 4], 4) #=> []\n# delete_all([4, 4, 4], 6) => [4, 4, 4]",
"_____no_output_____"
],
[
"# Define an encrypt_message function that accepts a string.\n# The input string will consist of only alphabetic characters.\n# The function should return a string where all characters have been moved\n# \"up\" two spots in the alphabet. For example, \"a\" will become \"c\".\n\ndef encrypt_message(string):\n alphabet = \"abcdefghijklmnopqrstuvwxyz\"\n new_string = \"\"\n for letter in string:\n moved = (alphabet.index(letter) + 2) % 26\n new_string += alphabet[moved]\n return new_string\n\n#encrypt_message(\"abc\")\nencrypt_message(\"man\")\n#encrypt_message(\"xyz\")",
"_____no_output_____"
],
[
"# Define a cleanup function that accepts a list of strings.\n# The function should return the strings joined together by a space.\n# There's one BIG problem -- some of the strings are empty or only consist of spaces!\n# These should NOT be included in the final string\n#\ndef cleanup(strings):\n result = []\n for string in strings:\n string = string.strip()\n if not string:\n continue\n result.append(string)\n return \" \".join(result)\n\n#cleanup([\"cat\", \"er\", \"pillar\"]) #=> \"cat er pillar\"\ncleanup([\"cat\", \" \", \"er\", \"\", \"pillar\"]) #=> \"cat er pillar\"\n#cleanup([\"\", \"\", \" \"]) #=> \"\"",
"_____no_output_____"
],
[
"# Define a word_lengths function that accepts a string. \n# It should return a list with the lengths of each word.\n#\ndef word_lengths(string):\n word_length = []\n string_list = string.split(\" \")\n print(string_list)\n for word in string_list:\n word_length.append(len(word))\n return word_length\n \n#word_lengths(\"Mary Poppins was a nanny\") #=> [4, 7, 3, 1, 5]\nword_lengths(\"Somebody stole my donut\") #=> [8, 5, 2, 5]",
"['Somebody', 'stole', 'my', 'donut']\n"
],
[
"# Declare a function right_words that accepts a list of words and a number.\n# Return a new list with the words that have a length equal to the number.\n# Do NOT use list comprehension.\n#\ndef right_words(words, number):\n new_list = []\n for word in words:\n if len(word) == number:\n new_list.append(word)\n return new_list\n \nright_words(['cat', 'dog', 'bean', 'ace', 'heart'], 3) #=> ['cat', 'dog', 'ace']\n# right_words(['cat', 'dog', 'bean', 'ace', 'heart'], 5) => ['heart']\n# right_words([], 4) => []",
"_____no_output_____"
],
[
"# Declare an only_odds function.\n# It should return a list with only the odd numbers.\n# Do NOT use list comprehension.\n#\ndef only_odds(numbers):\n new_list = []\n for number in numbers:\n if number%2 == 1:\n new_list.append(number)\n return new_list\n\n#only_odds([1, 3, 5, 6, 7, 8]) #=> [1, 3, 5, 7]\nonly_odds([2, 4, 6, 8]) #=> []",
"_____no_output_____"
],
[
"# Declare a count_of_a function that accepts a list of strings.\n# It should return a list with counts of how many “a” characters appear per string.\n# Do NOT use list comprehension.\n#\ndef count_of_a(words):\n return (list(map(lambda lists:lists.count(\"a\"), words)))\n\n#words = [\"alligator\", \"aardvark\", \"albatross\"] #=> [2, 3, 2]\n#count_of_a(list3)\ncount_of_a([\"alligator\", \"aardvark\", \"albatross\"]) \n#count_of_a([\"plywood\"]) => [0]\n# count_of_a([]) => []",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb91d8d645fc67aa3f16a4caeec4df59f1cc8af3
| 10,252 |
ipynb
|
Jupyter Notebook
|
Notebooks/General.ipynb
|
radical-experiments/campaign_manager
|
337660cf07a97933b9b516d6612353bd3f6592a8
|
[
"MIT"
] | null | null | null |
Notebooks/General.ipynb
|
radical-experiments/campaign_manager
|
337660cf07a97933b9b516d6612353bd3f6592a8
|
[
"MIT"
] | null | null | null |
Notebooks/General.ipynb
|
radical-experiments/campaign_manager
|
337660cf07a97933b9b516d6612353bd3f6592a8
|
[
"MIT"
] | null | null | null | 28.960452 | 96 | 0.496879 |
[
[
[
"from matplotlib import pyplot as plt\nfrom matplotlib import cm\nimport pandas as pd\nfrom pprint import pprint\nfrom random import randint, random, gauss, uniform \nimport numpy as np\n#import matplotlib as mpl\n#mpl.rcParams['text.usetex'] = True\n#mpl.rcParams['text.latex.unicode'] = True\n\nblues = cm.get_cmap(plt.get_cmap('Blues'))\ngreens = cm.get_cmap(plt.get_cmap('Greens'))\nreds = cm.get_cmap(plt.get_cmap('Reds'))\noranges = cm.get_cmap(plt.get_cmap('Oranges'))\npurples = cm.get_cmap(plt.get_cmap('Purples'))\ngreys = cm.get_cmap(plt.get_cmap('Greys'))\nset1 = cm.get_cmap(plt.get_cmap('Set1'))\n\ndef tableau20(color):\n # Use coordinated colors. These are the \"Tableau 20\" colors as\n # RGB. Each pair is strong/light. For a theory of color\n tableau20 = [(31 , 119, 180), (174, 199, 232), # blue [ 0,1 ]\n (255, 127, 14 ), (255, 187, 120), # orange [ 2,3 ]\n (44 , 160, 44 ), (152, 223, 138), # green [ 4,5 ]\n (214, 39 , 40 ), (255, 152, 150), # red [ 6,7 ]\n (148, 103, 189), (197, 176, 213), # purple [ 8,9 ]\n (140, 86 , 75 ), (196, 156, 148), # brown [10,11]\n (227, 119, 194), (247, 182, 210), # pink [12,13]\n (188, 189, 34 ), (219, 219, 141), # yellow [14,15]\n (23 , 190, 207), (158, 218, 229), # cyan [16,17]\n (65 , 68 , 81 ), (96 , 99 , 106), # gray [18,19]\n (127, 127, 127), (143, 135, 130), # gray [20,21]\n (165, 172, 175), (199, 199, 199), # gray [22,23]\n (207, 207, 207)] # gray [24]\n # Scale the RGB values to the [0, 1] range, which is the format\n # matplotlib accepts.\n r, g, b = tableau20[color]\n return (round(r/255.,1), round(g/255.,1), round(b/255.,1))\n\nfrom IPython.core.display import display, HTML\ndisplay(HTML(\"<style>.container { width:100% !important; }</style>\"))\nimport warnings\nwarnings.filterwarnings('ignore')",
"_____no_output_____"
],
[
"dist = list()\nfor _ in range(1000):\n tmp = list()\n for _ in range(1000):\n tmp.append(1 / gauss(1, 0.06))\n dist.append(min(tmp))",
"_____no_output_____"
],
[
"fig, axis = plt.subplots(nrows=1,ncols=1)\nfig.set_size_inches(15,7.5)\n_ = axis.hist(dist,bins=1000,color=tableau20(0))\n_ = axis.set_xticklabels(np.around(axis.get_xticks(), decimals=5).tolist(),fontsize=16)\n_ = axis.set_yticklabels((axis.get_yticks()).tolist(),fontsize=16)\n_ = axis.grid('on')",
"_____no_output_____"
],
[
"print(np.mean(dist), np.std(dist))",
"_____no_output_____"
],
[
"dist2 = list()\nfor _ in range(10000):\n dist2.append(gauss(0 + 1 * np.log10(10000), 0.4))\n\nfig, axis = plt.subplots(nrows=1,ncols=1, figsize=(15,7.5))\n_ = axis.hist(dist2, bins=1000)",
"_____no_output_____"
],
[
"print(np.mean(dist),np.std(dist))",
"_____no_output_____"
],
[
"dist = list()\nfor _ in range(1000):\n tmp = list()\n for _ in range(1000):\n tmp.append(uniform(-10, 10))\n dist.append(max(tmp))",
"_____no_output_____"
],
[
"fig, axis = plt.subplots(nrows=1,ncols=1)\nfig.set_size_inches(15,7.5)\n_ = axis.hist(dist,bins=1000,color=tableau20(0))\n_ = axis.set_xticklabels(np.around(axis.get_xticks(), decimals=5).tolist(),fontsize=16)\n_ = axis.set_yticklabels((axis.get_yticks()).tolist(),fontsize=16)\n_ = axis.grid('on')",
"_____no_output_____"
],
[
"from scipy.stats import invgauss\n\ndist = list()\nfor _ in range(10000):\n r = invgauss.rvs(1, size=10000)\n dist.append(max(r))",
"_____no_output_____"
],
[
"fig, axis = plt.subplots(nrows=1,ncols=1)\nfig.set_size_inches(15,7.5)\n_ = axis.hist(dist,bins=1000,color=tableau20(0))\n_ = axis.set_xticklabels(np.around(axis.get_xticks(), decimals=5).tolist(),fontsize=16)\n_ = axis.set_yticklabels((axis.get_yticks()).tolist(),fontsize=16)\n_ = axis.grid('on')",
"_____no_output_____"
],
[
"for n in [4,8,16,32,64,128,256,1024]:\n dist = list()\n for _ in range(100000):\n tmp = list()\n for _ in range(n):\n tmp.append(uniform(-0.1, 0.1))\n dist.append(max(tmp))\n\n print(n,':',np.mean(dist), np.std(dist))",
"_____no_output_____"
],
[
"for n in [4,8,16,32,64,128,256,1024]:\n dist = list()\n for _ in range(100000):\n tmp = list()\n for _ in range(n):\n tmp.append(uniform(-0.2, 0.2))\n dist.append(max(tmp))\n\n print(n,':',np.mean(dist), np.std(dist))",
"_____no_output_____"
],
[
"fig, axis = plt.subplots(nrows=1,ncols=1)\nfig.set_size_inches(15,7.5)\n_ = axis.hist(dist,bins=1000,color=tableau20(0))\n_ = axis.set_xticklabels(np.around(axis.get_xticks(), decimals=5).tolist(),fontsize=16)\n_ = axis.set_yticklabels((axis.get_yticks()).tolist(),fontsize=16)\n_ = axis.grid('on')",
"_____no_output_____"
],
[
"for n in [4,8,16,32,64,128,256,1024]:\n dist = list()\n for _ in range(100000):\n tmp = list()\n for _ in range(n):\n tmp.append(uniform(-0.3, 0.3))\n dist.append(max(tmp))\n\n print(n,':',np.mean(dist), np.std(dist))",
"_____no_output_____"
],
[
"for n in [4,8,16,32,64,128,256,1024]:\n dist = list()\n for _ in range(100000):\n tmp = list()\n for _ in range(n):\n tmp.append(uniform(-0.4, 0.4))\n dist.append(max(tmp))\n\n print(n,':',np.mean(dist), np.std(dist))",
"_____no_output_____"
],
[
"for n in [4,8,16,32,64,128,256,1024]:\n dist = list()\n for _ in range(100000):\n tmp = list()\n for _ in range(n):\n tmp.append(uniform(-0.5, 0.5))\n dist.append(max(tmp))\n\n print(n,':',np.mean(dist), np.std(dist))",
"_____no_output_____"
],
[
"for n in [4,8,16,32,64,128,256,1024]:\n dist = list()\n for _ in range(100000):\n tmp = list()\n for _ in range(n):\n tmp.append(uniform(-0.1, 0.1))\n dist.append(max(tmp))\n\n print(n,':',np.mean(dist), np.std(dist))",
"4 : 0.06007081792475389 0.03250150576319406\n8 : 0.07784145420032174 0.01977236033226086\n16 : 0.08823836778961511 0.011087132934927106\n32 : 0.09394209518381264 0.0058580700381117585\n64 : 0.09692912168963168 0.003032271759988272\n128 : 0.09845295040228835 0.0015340808160820049\n256 : 0.09921881007170821 0.0007740524254575908\n1024 : 0.09980468207052574 0.0001948260752006027\n"
],
[
"np.sqrt(np.std(tmp)**2 + 0.0036)",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb91dea9be111158f5ed5bbe7bd3e2e8e900a44c
| 732,716 |
ipynb
|
Jupyter Notebook
|
mission_to_mars.ipynb
|
EmmaK0822/web_scraping
|
4cb8b9fb61ee6b909ecdb3d6716a338a852a92df
|
[
"MIT"
] | null | null | null |
mission_to_mars.ipynb
|
EmmaK0822/web_scraping
|
4cb8b9fb61ee6b909ecdb3d6716a338a852a92df
|
[
"MIT"
] | null | null | null |
mission_to_mars.ipynb
|
EmmaK0822/web_scraping
|
4cb8b9fb61ee6b909ecdb3d6716a338a852a92df
|
[
"MIT"
] | null | null | null | 204.840928 | 353,239 | 0.737877 |
[
[
[
"# Use Splinter to navigate the sites when needed and BeautifulSoup to help find and parse out the necessary data.\nfrom splinter import Browser\nfrom bs4 import BeautifulSoup\nimport pandas as pd\nfrom selenium import webdriver",
"_____no_output_____"
]
],
[
[
"# NASA Mars News",
"_____no_output_____"
]
],
[
[
"executable_path = {\"executable_path\": \"chromedriver\"}\nbrowser = Browser(\"chrome\", **executable_path, headless=False)\n\nurl = \"https://mars.nasa.gov/news/?page=0&per_page=40&order=publish_date+desc%2Ccreated_at+desc&search=&category=19%2C165%2C184%2C204&blank_scope=Latest\"\nbrowser.visit(url)\n\nhtml = browser.html\nsoup = BeautifulSoup(html, \"html.parser\")",
"_____no_output_____"
],
[
"print(soup.prettify())",
"<!DOCTYPE html>\n<html class=\"no-flash cookies geolocation svg picture canvas video webgl srcdoc supports no-hiddenscroll fullscreen flexbox cssanimations flexboxlegacy no-flexboxtweener csstransforms csstransforms3d csstransitions preserve3d -webkit- no-touchevents\" lang=\"en\" style=\"\" xml:lang=\"en\" xmlns=\"http://www.w3.org/1999/xhtml\">\n <head>\n <script src=\"//api-public.addthis.com/url/shares.json?url=https%3A%2F%2Fmars.nasa.gov%2Fnews%2F%3Fpage%3D0%26per_page%3D40%26order%3Dpublish_date%2Bdesc%252Ccreated_at%2Bdesc%26search%3D%26category%3D19%252C165%252C184%252C204%26blank_scope%3DLatest&callback=_ate.cbs.rcb_g68x0\" type=\"text/javascript\">\n </script>\n <script src=\"//www.reddit.com/api/info.json?url=https%3A%2F%2Fmars.nasa.gov%2Fnews%2F%3Fpage%3D0%26per_page%3D40%26order%3Dpublish_date%2Bdesc%252Ccreated_at%2Bdesc%26search%3D%26category%3D19%252C165%252C184%252C204%26blank_scope%3DLatest&jsonp=_ate.cbs.rcb_av1x0\" type=\"text/javascript\">\n </script>\n <script src=\"//graph.facebook.com/?id=https%3A%2F%2Fmars.nasa.gov%2Fnews%2F%3Fpage%3D0%26per_page%3D40%26order%3Dpublish_date%2Bdesc%252Ccreated_at%2Bdesc%26search%3D%26category%3D19%252C165%252C184%252C204%26blank_scope%3DLatest&callback=_ate.cbs.rcb_522f0\" type=\"text/javascript\">\n </script>\n <script src=\"https://bam.nr-data.net/1/5e33925808?a=59562082&v=1071.385e752&to=JVcPR0MLWApSRU1eAQVVEhxSC1oSUlkWbBMHXwRAHhdcCUA%3D&rst=3176&ref=https://mars.nasa.gov/news/&ap=236&be=726&fe=2502&dc=1601&af=err,xhr,stn,ins&perf=%7B%22timing%22:%7B%22of%22:1532218140052,%22n%22:0,%22f%22:3,%22dn%22:4,%22dne%22:104,%22c%22:104,%22s%22:120,%22ce%22:204,%22rq%22:205,%22rp%22:621,%22rpe%22:638,%22dl%22:641,%22di%22:1601,%22ds%22:1601,%22de%22:1953,%22dc%22:2501,%22l%22:2501,%22le%22:2570%7D,%22navigation%22:%7B%7D%7D&jsonp=NREUM.setToken\" type=\"text/javascript\">\n </script>\n <script src=\"//m.addthis.com/live/red_lojson/300lo.json?si=5b53cb1e6480f157&bkl=0&bl=1&pdt=1792&sid=5b53cb1e6480f157&pub=ra-5a690e4c1320e328&rev=v8.3.25-wp&ln=en&pc=men&cb=0&ab=-&dp=mars.nasa.gov&fp=news%2F%3Fpage%3D0%26per_page%3D40%26order%3Dpublish_date%2Bdesc%252Ccreated_at%2Bdesc%26search%3D%26category%3D19%252C165%252C184%252C204%26blank_scope%3DLatest&fr=&of=1&pd=0&irt=0&vcl=0&md=0&ct=1&tct=0&abt=0&cdn=0&pi=1&rb=0&gen=100&chr=UTF-8&mk=Mars%2Cmissions%2CNASA%2Crover%2CCuriosity%2COpportunity%2CInSight%2CMars%20Reconnaissance%20Orbiter%2Cfacts&colc=1532218142611&jsl=1&skipb=1&callback=addthis.cbs.oln9_8039031381306510\" type=\"text/javascript\">\n </script>\n <script src=\"//m.addthisedge.com/live/boost/ra-5a690e4c1320e328/_ate.track.config_resp\" type=\"text/javascript\">\n </script>\n <meta content=\"text/html; charset=utf-8\" http-equiv=\"Content-Type\"/>\n <!-- Always force latest IE rendering engine or request Chrome Frame -->\n <meta content=\"IE=edge,chrome=1\" http-equiv=\"X-UA-Compatible\"/>\n <script src=\"https://js-agent.newrelic.com/nr-1071.min.js\">\n </script>\n <script async=\"\" src=\"https://www.google-analytics.com/analytics.js\">\n </script>\n <script type=\"text/javascript\">\n window.NREUM||(NREUM={});NREUM.info={\"beacon\":\"bam.nr-data.net\",\"errorBeacon\":\"bam.nr-data.net\",\"licenseKey\":\"5e33925808\",\"applicationID\":\"59562082\",\"transactionName\":\"JVcPR0MLWApSRU1eAQVVEhxSC1oSUlkWbBMHXwRAHhdcCUA=\",\"queueTime\":0,\"applicationTime\":236,\"agent\":\"\"}\n </script>\n <script type=\"text/javascript\">\n (window.NREUM||(NREUM={})).loader_config={xpid:\"VQcPUlZTDxAFXVRUBQEPVA==\"};window.NREUM||(NREUM={}),__nr_require=function(t,n,e){function r(e){if(!n[e]){var o=n[e]={exports:{}};t[e][0].call(o.exports,function(n){var o=t[e][1][n];return r(o||n)},o,o.exports)}return n[e].exports}if(\"function\"==typeof __nr_require)return __nr_require;for(var o=0;o<e.length;o++)r(e[o]);return r}({1:[function(t,n,e){function r(t){try{s.console&&console.log(t)}catch(n){}}var o,i=t(\"ee\"),a=t(15),s={};try{o=localStorage.getItem(\"__nr_flags\").split(\",\"),console&&\"function\"==typeof console.log&&(s.console=!0,o.indexOf(\"dev\")!==-1&&(s.dev=!0),o.indexOf(\"nr_dev\")!==-1&&(s.nrDev=!0))}catch(c){}s.nrDev&&i.on(\"internal-error\",function(t){r(t.stack)}),s.dev&&i.on(\"fn-err\",function(t,n,e){r(e.stack)}),s.dev&&(r(\"NR AGENT IN DEVELOPMENT MODE\"),r(\"flags: \"+a(s,function(t,n){return t}).join(\", \")))},{}],2:[function(t,n,e){function r(t,n,e,r,s){try{p?p-=1:o(s||new UncaughtException(t,n,e),!0)}catch(f){try{i(\"ierr\",[f,c.now(),!0])}catch(d){}}return\"function\"==typeof u&&u.apply(this,a(arguments))}function UncaughtException(t,n,e){this.message=t||\"Uncaught error with no additional information\",this.sourceURL=n,this.line=e}function o(t,n){var e=n?null:c.now();i(\"err\",[t,e])}var i=t(\"handle\"),a=t(16),s=t(\"ee\"),c=t(\"loader\"),f=t(\"gos\"),u=window.onerror,d=!1,l=\"nr@seenError\",p=0;c.features.err=!0,t(1),window.onerror=r;try{throw new Error}catch(h){\"stack\"in h&&(t(8),t(7),\"addEventListener\"in window&&t(5),c.xhrWrappable&&t(9),d=!0)}s.on(\"fn-start\",function(t,n,e){d&&(p+=1)}),s.on(\"fn-err\",function(t,n,e){d&&!e[l]&&(f(e,l,function(){return!0}),this.thrown=!0,o(e))}),s.on(\"fn-end\",function(){d&&!this.thrown&&p>0&&(p-=1)}),s.on(\"internal-error\",function(t){i(\"ierr\",[t,c.now(),!0])})},{}],3:[function(t,n,e){t(\"loader\").features.ins=!0},{}],4:[function(t,n,e){function r(t){}if(window.performance&&window.performance.timing&&window.performance.getEntriesByType){var o=t(\"ee\"),i=t(\"handle\"),a=t(8),s=t(7),c=\"learResourceTimings\",f=\"addEventListener\",u=\"resourcetimingbufferfull\",d=\"bstResource\",l=\"resource\",p=\"-start\",h=\"-end\",m=\"fn\"+p,w=\"fn\"+h,v=\"bstTimer\",y=\"pushState\",g=t(\"loader\");g.features.stn=!0,t(6);var b=NREUM.o.EV;o.on(m,function(t,n){var e=t[0];e instanceof b&&(this.bstStart=g.now())}),o.on(w,function(t,n){var e=t[0];e instanceof b&&i(\"bst\",[e,n,this.bstStart,g.now()])}),a.on(m,function(t,n,e){this.bstStart=g.now(),this.bstType=e}),a.on(w,function(t,n){i(v,[n,this.bstStart,g.now(),this.bstType])}),s.on(m,function(){this.bstStart=g.now()}),s.on(w,function(t,n){i(v,[n,this.bstStart,g.now(),\"requestAnimationFrame\"])}),o.on(y+p,function(t){this.time=g.now(),this.startPath=location.pathname+location.hash}),o.on(y+h,function(t){i(\"bstHist\",[location.pathname+location.hash,this.startPath,this.time])}),f in window.performance&&(window.performance[\"c\"+c]?window.performance[f](u,function(t){i(d,[window.performance.getEntriesByType(l)]),window.performance[\"c\"+c]()},!1):window.performance[f](\"webkit\"+u,function(t){i(d,[window.performance.getEntriesByType(l)]),window.performance[\"webkitC\"+c]()},!1)),document[f](\"scroll\",r,{passive:!0}),document[f](\"keypress\",r,!1),document[f](\"click\",r,!1)}},{}],5:[function(t,n,e){function r(t){for(var n=t;n&&!n.hasOwnProperty(u);)n=Object.getPrototypeOf(n);n&&o(n)}function o(t){s.inPlace(t,[u,d],\"-\",i)}function i(t,n){return t[1]}var a=t(\"ee\").get(\"events\"),s=t(18)(a,!0),c=t(\"gos\"),f=XMLHttpRequest,u=\"addEventListener\",d=\"removeEventListener\";n.exports=a,\"getPrototypeOf\"in Object?(r(document),r(window),r(f.prototype)):f.prototype.hasOwnProperty(u)&&(o(window),o(f.prototype)),a.on(u+\"-start\",function(t,n){var e=t[1],r=c(e,\"nr@wrapped\",function(){function t(){if(\"function\"==typeof e.handleEvent)return e.handleEvent.apply(e,arguments)}var n={object:t,\"function\":e}[typeof e];return n?s(n,\"fn-\",null,n.name||\"anonymous\"):e});this.wrapped=t[1]=r}),a.on(d+\"-start\",function(t){t[1]=this.wrapped||t[1]})},{}],6:[function(t,n,e){var r=t(\"ee\").get(\"history\"),o=t(18)(r);n.exports=r,o.inPlace(window.history,[\"pushState\",\"replaceState\"],\"-\")},{}],7:[function(t,n,e){var r=t(\"ee\").get(\"raf\"),o=t(18)(r),i=\"equestAnimationFrame\";n.exports=r,o.inPlace(window,[\"r\"+i,\"mozR\"+i,\"webkitR\"+i,\"msR\"+i],\"raf-\"),r.on(\"raf-start\",function(t){t[0]=o(t[0],\"fn-\")})},{}],8:[function(t,n,e){function r(t,n,e){t[0]=a(t[0],\"fn-\",null,e)}function o(t,n,e){this.method=e,this.timerDuration=isNaN(t[1])?0:+t[1],t[0]=a(t[0],\"fn-\",this,e)}var i=t(\"ee\").get(\"timer\"),a=t(18)(i),s=\"setTimeout\",c=\"setInterval\",f=\"clearTimeout\",u=\"-start\",d=\"-\";n.exports=i,a.inPlace(window,[s,\"setImmediate\"],s+d),a.inPlace(window,[c],c+d),a.inPlace(window,[f,\"clearImmediate\"],f+d),i.on(c+u,r),i.on(s+u,o)},{}],9:[function(t,n,e){function r(t,n){d.inPlace(n,[\"onreadystatechange\"],\"fn-\",s)}function o(){var t=this,n=u.context(t);t.readyState>3&&!n.resolved&&(n.resolved=!0,u.emit(\"xhr-resolved\",[],t)),d.inPlace(t,y,\"fn-\",s)}function i(t){g.push(t),h&&(x?x.then(a):w?w(a):(E=-E,O.data=E))}function a(){for(var t=0;t<g.length;t++)r([],g[t]);g.length&&(g=[])}function s(t,n){return n}function c(t,n){for(var e in t)n[e]=t[e];return n}t(5);var f=t(\"ee\"),u=f.get(\"xhr\"),d=t(18)(u),l=NREUM.o,p=l.XHR,h=l.MO,m=l.PR,w=l.SI,v=\"readystatechange\",y=[\"onload\",\"onerror\",\"onabort\",\"onloadstart\",\"onloadend\",\"onprogress\",\"ontimeout\"],g=[];n.exports=u;var b=window.XMLHttpRequest=function(t){var n=new p(t);try{u.emit(\"new-xhr\",[n],n),n.addEventListener(v,o,!1)}catch(e){try{u.emit(\"internal-error\",[e])}catch(r){}}return n};if(c(p,b),b.prototype=p.prototype,d.inPlace(b.prototype,[\"open\",\"send\"],\"-xhr-\",s),u.on(\"send-xhr-start\",function(t,n){r(t,n),i(n)}),u.on(\"open-xhr-start\",r),h){var x=m&&m.resolve();if(!w&&!m){var E=1,O=document.createTextNode(E);new h(a).observe(O,{characterData:!0})}}else f.on(\"fn-end\",function(t){t[0]&&t[0].type===v||a()})},{}],10:[function(t,n,e){function r(t){var n=this.params,e=this.metrics;if(!this.ended){this.ended=!0;for(var r=0;r<d;r++)t.removeEventListener(u[r],this.listener,!1);if(!n.aborted){if(e.duration=a.now()-this.startTime,4===t.readyState){n.status=t.status;var i=o(t,this.lastSize);if(i&&(e.rxSize=i),this.sameOrigin){var c=t.getResponseHeader(\"X-NewRelic-App-Data\");c&&(n.cat=c.split(\", \").pop())}}else n.status=0;e.cbTime=this.cbTime,f.emit(\"xhr-done\",[t],t),s(\"xhr\",[n,e,this.startTime])}}}function o(t,n){var e=t.responseType;if(\"json\"===e&&null!==n)return n;var r=\"arraybuffer\"===e||\"blob\"===e||\"json\"===e?t.response:t.responseText;return h(r)}function i(t,n){var e=c(n),r=t.params;r.host=e.hostname+\":\"+e.port,r.pathname=e.pathname,t.sameOrigin=e.sameOrigin}var a=t(\"loader\");if(a.xhrWrappable){var s=t(\"handle\"),c=t(11),f=t(\"ee\"),u=[\"load\",\"error\",\"abort\",\"timeout\"],d=u.length,l=t(\"id\"),p=t(14),h=t(13),m=window.XMLHttpRequest;a.features.xhr=!0,t(9),f.on(\"new-xhr\",function(t){var n=this;n.totalCbs=0,n.called=0,n.cbTime=0,n.end=r,n.ended=!1,n.xhrGuids={},n.lastSize=null,p&&(p>34||p<10)||window.opera||t.addEventListener(\"progress\",function(t){n.lastSize=t.loaded},!1)}),f.on(\"open-xhr-start\",function(t){this.params={method:t[0]},i(this,t[1]),this.metrics={}}),f.on(\"open-xhr-end\",function(t,n){\"loader_config\"in NREUM&&\"xpid\"in NREUM.loader_config&&this.sameOrigin&&n.setRequestHeader(\"X-NewRelic-ID\",NREUM.loader_config.xpid)}),f.on(\"send-xhr-start\",function(t,n){var e=this.metrics,r=t[0],o=this;if(e&&r){var i=h(r);i&&(e.txSize=i)}this.startTime=a.now(),this.listener=function(t){try{\"abort\"===t.type&&(o.params.aborted=!0),(\"load\"!==t.type||o.called===o.totalCbs&&(o.onloadCalled||\"function\"!=typeof n.onload))&&o.end(n)}catch(e){try{f.emit(\"internal-error\",[e])}catch(r){}}};for(var s=0;s<d;s++)n.addEventListener(u[s],this.listener,!1)}),f.on(\"xhr-cb-time\",function(t,n,e){this.cbTime+=t,n?this.onloadCalled=!0:this.called+=1,this.called!==this.totalCbs||!this.onloadCalled&&\"function\"==typeof e.onload||this.end(e)}),f.on(\"xhr-load-added\",function(t,n){var e=\"\"+l(t)+!!n;this.xhrGuids&&!this.xhrGuids[e]&&(this.xhrGuids[e]=!0,this.totalCbs+=1)}),f.on(\"xhr-load-removed\",function(t,n){var e=\"\"+l(t)+!!n;this.xhrGuids&&this.xhrGuids[e]&&(delete this.xhrGuids[e],this.totalCbs-=1)}),f.on(\"addEventListener-end\",function(t,n){n instanceof m&&\"load\"===t[0]&&f.emit(\"xhr-load-added\",[t[1],t[2]],n)}),f.on(\"removeEventListener-end\",function(t,n){n instanceof m&&\"load\"===t[0]&&f.emit(\"xhr-load-removed\",[t[1],t[2]],n)}),f.on(\"fn-start\",function(t,n,e){n instanceof m&&(\"onload\"===e&&(this.onload=!0),(\"load\"===(t[0]&&t[0].type)||this.onload)&&(this.xhrCbStart=a.now()))}),f.on(\"fn-end\",function(t,n){this.xhrCbStart&&f.emit(\"xhr-cb-time\",[a.now()-this.xhrCbStart,this.onload,n],n)})}},{}],11:[function(t,n,e){n.exports=function(t){var n=document.createElement(\"a\"),e=window.location,r={};n.href=t,r.port=n.port;var o=n.href.split(\"://\");!r.port&&o[1]&&(r.port=o[1].split(\"/\")[0].split(\"@\").pop().split(\":\")[1]),r.port&&\"0\"!==r.port||(r.port=\"https\"===o[0]?\"443\":\"80\"),r.hostname=n.hostname||e.hostname,r.pathname=n.pathname,r.protocol=o[0],\"/\"!==r.pathname.charAt(0)&&(r.pathname=\"/\"+r.pathname);var i=!n.protocol||\":\"===n.protocol||n.protocol===e.protocol,a=n.hostname===document.domain&&n.port===e.port;return r.sameOrigin=i&&(!n.hostname||a),r}},{}],12:[function(t,n,e){function r(){}function o(t,n,e){return function(){return i(t,[f.now()].concat(s(arguments)),n?null:this,e),n?void 0:this}}var i=t(\"handle\"),a=t(15),s=t(16),c=t(\"ee\").get(\"tracer\"),f=t(\"loader\"),u=NREUM;\"undefined\"==typeof window.newrelic&&(newrelic=u);var d=[\"setPageViewName\",\"setCustomAttribute\",\"setErrorHandler\",\"finished\",\"addToTrace\",\"inlineHit\",\"addRelease\"],l=\"api-\",p=l+\"ixn-\";a(d,function(t,n){u[n]=o(l+n,!0,\"api\")}),u.addPageAction=o(l+\"addPageAction\",!0),u.setCurrentRouteName=o(l+\"routeName\",!0),n.exports=newrelic,u.interaction=function(){return(new r).get()};var h=r.prototype={createTracer:function(t,n){var e={},r=this,o=\"function\"==typeof n;return i(p+\"tracer\",[f.now(),t,e],r),function(){if(c.emit((o?\"\":\"no-\")+\"fn-start\",[f.now(),r,o],e),o)try{return n.apply(this,arguments)}catch(t){throw c.emit(\"fn-err\",[arguments,this,t],e),t}finally{c.emit(\"fn-end\",[f.now()],e)}}}};a(\"setName,setAttribute,save,ignore,onEnd,getContext,end,get\".split(\",\"),function(t,n){h[n]=o(p+n)}),newrelic.noticeError=function(t){\"string\"==typeof t&&(t=new Error(t)),i(\"err\",[t,f.now()])}},{}],13:[function(t,n,e){n.exports=function(t){if(\"string\"==typeof t&&t.length)return t.length;if(\"object\"==typeof t){if(\"undefined\"!=typeof ArrayBuffer&&t instanceof ArrayBuffer&&t.byteLength)return t.byteLength;if(\"undefined\"!=typeof Blob&&t instanceof Blob&&t.size)return t.size;if(!(\"undefined\"!=typeof FormData&&t instanceof FormData))try{return JSON.stringify(t).length}catch(n){return}}}},{}],14:[function(t,n,e){var r=0,o=navigator.userAgent.match(/Firefox[\\/\\s](\\d+\\.\\d+)/);o&&(r=+o[1]),n.exports=r},{}],15:[function(t,n,e){function r(t,n){var e=[],r=\"\",i=0;for(r in t)o.call(t,r)&&(e[i]=n(r,t[r]),i+=1);return e}var o=Object.prototype.hasOwnProperty;n.exports=r},{}],16:[function(t,n,e){function r(t,n,e){n||(n=0),\"undefined\"==typeof e&&(e=t?t.length:0);for(var r=-1,o=e-n||0,i=Array(o<0?0:o);++r<o;)i[r]=t[n+r];return i}n.exports=r},{}],17:[function(t,n,e){n.exports={exists:\"undefined\"!=typeof window.performance&&window.performance.timing&&\"undefined\"!=typeof window.performance.timing.navigationStart}},{}],18:[function(t,n,e){function r(t){return!(t&&t instanceof Function&&t.apply&&!t[a])}var o=t(\"ee\"),i=t(16),a=\"nr@original\",s=Object.prototype.hasOwnProperty,c=!1;n.exports=function(t,n){function e(t,n,e,o){function nrWrapper(){var r,a,s,c;try{a=this,r=i(arguments),s=\"function\"==typeof e?e(r,a):e||{}}catch(f){l([f,\"\",[r,a,o],s])}u(n+\"start\",[r,a,o],s);try{return c=t.apply(a,r)}catch(d){throw u(n+\"err\",[r,a,d],s),d}finally{u(n+\"end\",[r,a,c],s)}}return r(t)?t:(n||(n=\"\"),nrWrapper[a]=t,d(t,nrWrapper),nrWrapper)}function f(t,n,o,i){o||(o=\"\");var a,s,c,f=\"-\"===o.charAt(0);for(c=0;c<n.length;c++)s=n[c],a=t[s],r(a)||(t[s]=e(a,f?s+o:o,i,s))}function u(e,r,o){if(!c||n){var i=c;c=!0;try{t.emit(e,r,o,n)}catch(a){l([a,e,r,o])}c=i}}function d(t,n){if(Object.defineProperty&&Object.keys)try{var e=Object.keys(t);return e.forEach(function(e){Object.defineProperty(n,e,{get:function(){return t[e]},set:function(n){return t[e]=n,n}})}),n}catch(r){l([r])}for(var o in t)s.call(t,o)&&(n[o]=t[o]);return n}function l(n){try{t.emit(\"internal-error\",n)}catch(e){}}return t||(t=o),e.inPlace=f,e.flag=a,e}},{}],ee:[function(t,n,e){function r(){}function o(t){function n(t){return t&&t instanceof r?t:t?c(t,s,i):i()}function e(e,r,o,i){if(!l.aborted||i){t&&t(e,r,o);for(var a=n(o),s=h(e),c=s.length,f=0;f<c;f++)s[f].apply(a,r);var d=u[y[e]];return d&&d.push([g,e,r,a]),a}}function p(t,n){v[t]=h(t).concat(n)}function h(t){return v[t]||[]}function m(t){return d[t]=d[t]||o(e)}function w(t,n){f(t,function(t,e){n=n||\"feature\",y[e]=n,n in u||(u[n]=[])})}var v={},y={},g={on:p,emit:e,get:m,listeners:h,context:n,buffer:w,abort:a,aborted:!1};return g}function i(){return new r}function a(){(u.api||u.feature)&&(l.aborted=!0,u=l.backlog={})}var s=\"nr@context\",c=t(\"gos\"),f=t(15),u={},d={},l=n.exports=o();l.backlog=u},{}],gos:[function(t,n,e){function r(t,n,e){if(o.call(t,n))return t[n];var r=e();if(Object.defineProperty&&Object.keys)try{return Object.defineProperty(t,n,{value:r,writable:!0,enumerable:!1}),r}catch(i){}return t[n]=r,r}var o=Object.prototype.hasOwnProperty;n.exports=r},{}],handle:[function(t,n,e){function r(t,n,e,r){o.buffer([t],r),o.emit(t,n,e)}var o=t(\"ee\").get(\"handle\");n.exports=r,r.ee=o},{}],id:[function(t,n,e){function r(t){var n=typeof t;return!t||\"object\"!==n&&\"function\"!==n?-1:t===window?0:a(t,i,function(){return o++})}var o=1,i=\"nr@id\",a=t(\"gos\");n.exports=r},{}],loader:[function(t,n,e){function r(){if(!x++){var t=b.info=NREUM.info,n=l.getElementsByTagName(\"script\")[0];if(setTimeout(u.abort,3e4),!(t&&t.licenseKey&&t.applicationID&&n))return u.abort();f(y,function(n,e){t[n]||(t[n]=e)}),c(\"mark\",[\"onload\",a()+b.offset],null,\"api\");var e=l.createElement(\"script\");e.src=\"https://\"+t.agent,n.parentNode.insertBefore(e,n)}}function o(){\"complete\"===l.readyState&&i()}function i(){c(\"mark\",[\"domContent\",a()+b.offset],null,\"api\")}function a(){return E.exists&&performance.now?Math.round(performance.now()):(s=Math.max((new Date).getTime(),s))-b.offset}var s=(new Date).getTime(),c=t(\"handle\"),f=t(15),u=t(\"ee\"),d=window,l=d.document,p=\"addEventListener\",h=\"attachEvent\",m=d.XMLHttpRequest,w=m&&m.prototype;NREUM.o={ST:setTimeout,SI:d.setImmediate,CT:clearTimeout,XHR:m,REQ:d.Request,EV:d.Event,PR:d.Promise,MO:d.MutationObserver};var v=\"\"+location,y={beacon:\"bam.nr-data.net\",errorBeacon:\"bam.nr-data.net\",agent:\"js-agent.newrelic.com/nr-1071.min.js\"},g=m&&w&&w[p]&&!/CriOS/.test(navigator.userAgent),b=n.exports={offset:s,now:a,origin:v,features:{},xhrWrappable:g};t(12),l[p]?(l[p](\"DOMContentLoaded\",i,!1),d[p](\"load\",r,!1)):(l[h](\"onreadystatechange\",o),d[h](\"onload\",r)),c(\"mark\",[\"firstbyte\",s],null,\"api\");var x=0,E=t(17)},{}]},{},[\"loader\",2,10,4,3]);\n </script>\n <!-- Responsiveness -->\n <meta content=\"width=device-width, initial-scale=1.0\" name=\"viewport\"/>\n <!-- Favicon -->\n <link href=\"/apple-touch-icon.png\" rel=\"apple-touch-icon\" sizes=\"180x180\"/>\n <link href=\"/favicon-32x32.png\" rel=\"icon\" sizes=\"32x32\" type=\"image/png\"/>\n <link href=\"/favicon-16x16.png\" rel=\"icon\" sizes=\"16x16\" type=\"image/png\"/>\n <link href=\"/manifest.json\" rel=\"manifest\"/>\n <link color=\"#e48b55\" href=\"/safari-pinned-tab.svg\" rel=\"mask-icon\"/>\n <meta content=\"#000000\" name=\"theme-color\"/>\n <meta content=\"authenticity_token\" name=\"csrf-param\"/>\n <meta content=\"0MrFxUAngZGm7nCQHmc3XB7LGFHHdo+oGVvlXvrLtMw=\" name=\"csrf-token\"/>\n <title>\n News – NASA’s Mars Exploration Program\n </title>\n <meta content=\"NASA’s Mars Exploration Program \" property=\"og:site_name\"/>\n <meta content=\"mars.nasa.gov\" name=\"author\"/>\n <meta content=\"Mars, missions, NASA, rover, Curiosity, Opportunity, InSight, Mars Reconnaissance Orbiter, facts\" name=\"keywords\"/>\n <meta content=\"NASA’s real-time portal for Mars exploration, featuring the latest news, images, and discoveries from the Red Planet.\" name=\"description\"/>\n <meta content=\"NASA’s real-time portal for Mars exploration, featuring the latest news, images, and discoveries from the Red Planet.\" property=\"og:description\"/>\n <meta content=\"News – NASA’s Mars Exploration Program \" property=\"og:title\"/>\n <meta content=\"https://mars.nasa.gov/news?page=0&per_page=40&order=publish_date+desc%2Ccreated_at+desc&search=&category=19%2C165%2C184%2C204&blank_scope=Latest\" property=\"og:url\"/>\n <meta content=\"article\" property=\"og:type\"/>\n <meta content=\"2017-09-22 19:53:22 UTC\" property=\"og:updated_time\"/>\n <meta content=\"https://mars.nasa.gov/system/site_config_values/meta_share_images/1_142497main_PIA03154-200.jpg\" property=\"og:image\"/>\n <meta content=\"https://mars.nasa.gov/system/site_config_values/meta_share_images/1_142497main_PIA03154-200.jpg\" name=\"twitter:image\"/>\n <link href=\"https://mars.nasa.gov/system/site_config_values/meta_share_images/1_142497main_PIA03154-200.jpg\" rel=\"image_src\"/>\n <meta content=\"195570401081308\" property=\"fb:app_id\"/>\n <style data-href=\"https://fonts.googleapis.com/css?family=Montserrat:200,300,400,500,600,700|Raleway:300,400\" media=\"\">\n /* cyrillic-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 200;\n src: local('Montserrat ExtraLight'), local('Montserrat-ExtraLight'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_aZA3gTD_u50.woff2) format('woff2');\n unicode-range: U+0460-052F, U+1C80-1C88, U+20B4, U+2DE0-2DFF, U+A640-A69F, U+FE2E-FE2F;\n}\n/* cyrillic */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 200;\n src: local('Montserrat ExtraLight'), local('Montserrat-ExtraLight'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_aZA3g3D_u50.woff2) format('woff2');\n unicode-range: U+0400-045F, U+0490-0491, U+04B0-04B1, U+2116;\n}\n/* vietnamese */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 200;\n src: local('Montserrat ExtraLight'), local('Montserrat-ExtraLight'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_aZA3gbD_u50.woff2) format('woff2');\n unicode-range: U+0102-0103, U+0110-0111, U+1EA0-1EF9, U+20AB;\n}\n/* latin-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 200;\n src: local('Montserrat ExtraLight'), local('Montserrat-ExtraLight'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_aZA3gfD_u50.woff2) format('woff2');\n unicode-range: U+0100-024F, U+0259, U+1E00-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;\n}\n/* latin */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 200;\n src: local('Montserrat ExtraLight'), local('Montserrat-ExtraLight'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_aZA3gnD_g.woff2) format('woff2');\n unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;\n}\n/* cyrillic-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 300;\n src: local('Montserrat Light'), local('Montserrat-Light'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_cJD3gTD_u50.woff2) format('woff2');\n unicode-range: U+0460-052F, U+1C80-1C88, U+20B4, U+2DE0-2DFF, U+A640-A69F, U+FE2E-FE2F;\n}\n/* cyrillic */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 300;\n src: local('Montserrat Light'), local('Montserrat-Light'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_cJD3g3D_u50.woff2) format('woff2');\n unicode-range: U+0400-045F, U+0490-0491, U+04B0-04B1, U+2116;\n}\n/* vietnamese */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 300;\n src: local('Montserrat Light'), local('Montserrat-Light'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_cJD3gbD_u50.woff2) format('woff2');\n unicode-range: U+0102-0103, U+0110-0111, U+1EA0-1EF9, U+20AB;\n}\n/* latin-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 300;\n src: local('Montserrat Light'), local('Montserrat-Light'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_cJD3gfD_u50.woff2) format('woff2');\n unicode-range: U+0100-024F, U+0259, U+1E00-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;\n}\n/* latin */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 300;\n src: local('Montserrat Light'), local('Montserrat-Light'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_cJD3gnD_g.woff2) format('woff2');\n unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;\n}\n/* cyrillic-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 400;\n src: local('Montserrat Regular'), local('Montserrat-Regular'), url(https://fonts.gstatic.com/s/montserrat/v12/JTUSjIg1_i6t8kCHKm459WRhyzbi.woff2) format('woff2');\n unicode-range: U+0460-052F, U+1C80-1C88, U+20B4, U+2DE0-2DFF, U+A640-A69F, U+FE2E-FE2F;\n}\n/* cyrillic */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 400;\n src: local('Montserrat Regular'), local('Montserrat-Regular'), url(https://fonts.gstatic.com/s/montserrat/v12/JTUSjIg1_i6t8kCHKm459W1hyzbi.woff2) format('woff2');\n unicode-range: U+0400-045F, U+0490-0491, U+04B0-04B1, U+2116;\n}\n/* vietnamese */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 400;\n src: local('Montserrat Regular'), local('Montserrat-Regular'), url(https://fonts.gstatic.com/s/montserrat/v12/JTUSjIg1_i6t8kCHKm459WZhyzbi.woff2) format('woff2');\n unicode-range: U+0102-0103, U+0110-0111, U+1EA0-1EF9, U+20AB;\n}\n/* latin-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 400;\n src: local('Montserrat Regular'), local('Montserrat-Regular'), url(https://fonts.gstatic.com/s/montserrat/v12/JTUSjIg1_i6t8kCHKm459Wdhyzbi.woff2) format('woff2');\n unicode-range: U+0100-024F, U+0259, U+1E00-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;\n}\n/* latin */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 400;\n src: local('Montserrat Regular'), local('Montserrat-Regular'), url(https://fonts.gstatic.com/s/montserrat/v12/JTUSjIg1_i6t8kCHKm459Wlhyw.woff2) format('woff2');\n unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;\n}\n/* cyrillic-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 500;\n src: local('Montserrat Medium'), local('Montserrat-Medium'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_ZpC3gTD_u50.woff2) format('woff2');\n unicode-range: U+0460-052F, U+1C80-1C88, U+20B4, U+2DE0-2DFF, U+A640-A69F, U+FE2E-FE2F;\n}\n/* cyrillic */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 500;\n src: local('Montserrat Medium'), local('Montserrat-Medium'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_ZpC3g3D_u50.woff2) format('woff2');\n unicode-range: U+0400-045F, U+0490-0491, U+04B0-04B1, U+2116;\n}\n/* vietnamese */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 500;\n src: local('Montserrat Medium'), local('Montserrat-Medium'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_ZpC3gbD_u50.woff2) format('woff2');\n unicode-range: U+0102-0103, U+0110-0111, U+1EA0-1EF9, U+20AB;\n}\n/* latin-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 500;\n src: local('Montserrat Medium'), local('Montserrat-Medium'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_ZpC3gfD_u50.woff2) format('woff2');\n unicode-range: U+0100-024F, U+0259, U+1E00-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;\n}\n/* latin */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 500;\n src: local('Montserrat Medium'), local('Montserrat-Medium'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_ZpC3gnD_g.woff2) format('woff2');\n unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;\n}\n/* cyrillic-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 600;\n src: local('Montserrat SemiBold'), local('Montserrat-SemiBold'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_bZF3gTD_u50.woff2) format('woff2');\n unicode-range: U+0460-052F, U+1C80-1C88, U+20B4, U+2DE0-2DFF, U+A640-A69F, U+FE2E-FE2F;\n}\n/* cyrillic */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 600;\n src: local('Montserrat SemiBold'), local('Montserrat-SemiBold'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_bZF3g3D_u50.woff2) format('woff2');\n unicode-range: U+0400-045F, U+0490-0491, U+04B0-04B1, U+2116;\n}\n/* vietnamese */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 600;\n src: local('Montserrat SemiBold'), local('Montserrat-SemiBold'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_bZF3gbD_u50.woff2) format('woff2');\n unicode-range: U+0102-0103, U+0110-0111, U+1EA0-1EF9, U+20AB;\n}\n/* latin-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 600;\n src: local('Montserrat SemiBold'), local('Montserrat-SemiBold'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_bZF3gfD_u50.woff2) format('woff2');\n unicode-range: U+0100-024F, U+0259, U+1E00-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;\n}\n/* latin */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 600;\n src: local('Montserrat SemiBold'), local('Montserrat-SemiBold'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_bZF3gnD_g.woff2) format('woff2');\n unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;\n}\n/* cyrillic-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 700;\n src: local('Montserrat Bold'), local('Montserrat-Bold'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_dJE3gTD_u50.woff2) format('woff2');\n unicode-range: U+0460-052F, U+1C80-1C88, U+20B4, U+2DE0-2DFF, U+A640-A69F, U+FE2E-FE2F;\n}\n/* cyrillic */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 700;\n src: local('Montserrat Bold'), local('Montserrat-Bold'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_dJE3g3D_u50.woff2) format('woff2');\n unicode-range: U+0400-045F, U+0490-0491, U+04B0-04B1, U+2116;\n}\n/* vietnamese */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 700;\n src: local('Montserrat Bold'), local('Montserrat-Bold'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_dJE3gbD_u50.woff2) format('woff2');\n unicode-range: U+0102-0103, U+0110-0111, U+1EA0-1EF9, U+20AB;\n}\n/* latin-ext */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 700;\n src: local('Montserrat Bold'), local('Montserrat-Bold'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_dJE3gfD_u50.woff2) format('woff2');\n unicode-range: U+0100-024F, U+0259, U+1E00-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;\n}\n/* latin */\n@font-face {\n font-family: 'Montserrat';\n font-style: normal;\n font-weight: 700;\n src: local('Montserrat Bold'), local('Montserrat-Bold'), url(https://fonts.gstatic.com/s/montserrat/v12/JTURjIg1_i6t8kCHKm45_dJE3gnD_g.woff2) format('woff2');\n unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;\n}\n/* latin-ext */\n@font-face {\n font-family: 'Raleway';\n font-style: normal;\n font-weight: 300;\n src: local('Raleway Light'), local('Raleway-Light'), url(https://fonts.gstatic.com/s/raleway/v12/1Ptrg8zYS_SKggPNwIYqWqhPAMif.woff2) format('woff2');\n unicode-range: U+0100-024F, U+0259, U+1E00-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;\n}\n/* latin */\n@font-face {\n font-family: 'Raleway';\n font-style: normal;\n font-weight: 300;\n src: local('Raleway Light'), local('Raleway-Light'), url(https://fonts.gstatic.com/s/raleway/v12/1Ptrg8zYS_SKggPNwIYqWqZPAA.woff2) format('woff2');\n unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;\n}\n/* latin-ext */\n@font-face {\n font-family: 'Raleway';\n font-style: normal;\n font-weight: 400;\n src: local('Raleway'), local('Raleway-Regular'), url(https://fonts.gstatic.com/s/raleway/v12/1Ptug8zYS_SKggPNyCMIT5lu.woff2) format('woff2');\n unicode-range: U+0100-024F, U+0259, U+1E00-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;\n}\n/* latin */\n@font-face {\n font-family: 'Raleway';\n font-style: normal;\n font-weight: 400;\n src: local('Raleway'), local('Raleway-Regular'), url(https://fonts.gstatic.com/s/raleway/v12/1Ptug8zYS_SKggPNyC0ITw.woff2) format('woff2');\n unicode-range: U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;\n}\n </style>\n <style data-href=\"/assets/public_manifest-e05f97cdafb31d988d511960c6c62e08ece83d37c15daa3b9b991e2bc176a1e3.css\" media=\"all\">\n @charset \"UTF-8\";.ui-helper-hidden{display:none}.ui-helper-hidden-accessible{position:absolute !important;clip:rect(1px 1px 1px 1px);clip:rect(1px, 1px, 1px, 1px)}.ui-helper-reset{margin:0;padding:0;border:0;outline:0;line-height:1.3;text-decoration:none;font-size:100%;list-style:none}.ui-helper-clearfix:before,.ui-helper-clearfix:after{content:\"\";display:table}.ui-helper-clearfix:after{clear:both}.ui-helper-clearfix{zoom:1}.ui-helper-zfix{width:100%;height:100%;top:0;left:0;position:absolute;opacity:0;filter:Alpha(Opacity=0)}.ui-state-disabled{cursor:default !important}.ui-icon{display:block;text-indent:-99999px;overflow:hidden;background-repeat:no-repeat}.ui-widget-overlay{position:absolute;top:0;left:0;width:100%;height:100%}.ui-slider{position:relative;text-align:left}.ui-slider .ui-slider-handle{position:absolute;z-index:2;width:6.2em;height:.7em;cursor:default}.ui-slider .ui-slider-range{position:absolute;z-index:1;font-size:.7em;display:block;border:0;background-position:0 0}.ui-slider-horizontal{height:.8em}.ui-slider-horizontal .ui-slider-handle{top:0;margin-left:0}.ui-slider-horizontal .ui-slider-range{top:0;height:100%}.ui-slider-horizontal .ui-slider-range-min{left:0}.ui-slider-horizontal .ui-slider-range-max{right:0}.ui-slider-vertical{width:.8em;height:100px}.ui-slider-vertical .ui-slider-handle{left:-.3em;margin-left:0;margin-bottom:-.6em}.ui-slider-vertical .ui-slider-range{left:0;width:100%}.ui-slider-vertical .ui-slider-range-min{bottom:0}.ui-slider-vertical .ui-slider-range-max{top:0}.ui-widget{font-family:Segoe UI,Arial,sans-serif;font-size:1.1em}.ui-widget .ui-widget{font-size:1em}.ui-widget input,.ui-widget select,.ui-widget textarea,.ui-widget button{font-family:Segoe UI,Arial,sans-serif;font-size:1em}.ui-widget-content{border:1px solid #666666;background:#000000;color:#ffffff}.ui-widget-content a{color:#ffffff}.ui-widget-header{border:1px solid #333333;background:#333333;color:#ffffff;font-weight:700}.ui-widget-header a{color:#ffffff}.ui-state-default,.ui-widget-content .ui-state-default,.ui-widget-header .ui-state-default{border:1px solid #666666;background:#555555;font-weight:700;color:#eeeeee}.ui-state-default a,.ui-state-default a:link,.ui-state-default a:visited{color:#eeeeee;text-decoration:none}.ui-state-hover a,.ui-state-hover a:hover{color:#ffffff;text-decoration:none}.ui-state-active,.ui-widget-content .ui-state-active,.ui-widget-header .ui-state-active{border:1px solid #ffaf0f;background:#f58400;font-weight:700;color:#ffffff}.ui-state-active a,.ui-state-active a:link,.ui-state-active a:visited{color:#ffffff;text-decoration:none}.ui-state-highlight,.ui-widget-content .ui-state-highlight,.ui-widget-header .ui-state-highlight{border:1px solid #cccccc;background:#eeeeee;color:#2e7db2}.ui-state-highlight a,.ui-widget-content .ui-state-highlight a,.ui-widget-header .ui-state-highlight a{color:#2e7db2}.ui-state-error,.ui-widget-content .ui-state-error,.ui-widget-header .ui-state-error{border:1px solid #ffb73d;background:#ffc73d;color:#111111}.ui-state-error a,.ui-widget-content .ui-state-error a,.ui-widget-header .ui-state-error a{color:#111111}.ui-state-error-text,.ui-widget-content .ui-state-error-text,.ui-widget-header .ui-state-error-text{color:#111111}.ui-priority-primary,.ui-widget-content .ui-priority-primary,.ui-widget-header .ui-priority-primary{font-weight:700}.ui-priority-secondary,.ui-widget-content .ui-priority-secondary,.ui-widget-header .ui-priority-secondary{opacity:.7;filter:Alpha(Opacity=70);font-weight:400}.ui-state-disabled,.ui-widget-content .ui-state-disabled,.ui-widget-header .ui-state-disabled{opacity:.35;filter:Alpha(Opacity=35);background-image:none}.ui-icon{width:16px;height:16px}.ui-icon-carat-1-n{background-position:0 0}.ui-icon-carat-1-ne{background-position:-16px 0}.ui-icon-carat-1-e{background-position:-32px 0}.ui-icon-carat-1-se{background-position:-48px 0}.ui-icon-carat-1-s{background-position:-64px 0}.ui-icon-carat-1-sw{background-position:-80px 0}.ui-icon-carat-1-w{background-position:-96px 0}.ui-icon-carat-1-nw{background-position:-112px 0}.ui-icon-carat-2-n-s{background-position:-128px 0}.ui-icon-carat-2-e-w{background-position:-144px 0}.ui-icon-triangle-1-n{background-position:0 -16px}.ui-icon-triangle-1-ne{background-position:-16px -16px}.ui-icon-triangle-1-e{background-position:-32px -16px}.ui-icon-triangle-1-se{background-position:-48px -16px}.ui-icon-triangle-1-s{background-position:-64px -16px}.ui-icon-triangle-1-sw{background-position:-80px -16px}.ui-icon-triangle-1-w{background-position:-96px -16px}.ui-icon-triangle-1-nw{background-position:-112px -16px}.ui-icon-triangle-2-n-s{background-position:-128px -16px}.ui-icon-triangle-2-e-w{background-position:-144px -16px}.ui-icon-arrow-1-n{background-position:0 -32px}.ui-icon-arrow-1-ne{background-position:-16px -32px}.ui-icon-arrow-1-e{background-position:-32px -32px}.ui-icon-arrow-1-se{background-position:-48px -32px}.ui-icon-arrow-1-s{background-position:-64px -32px}.ui-icon-arrow-1-sw{background-position:-80px -32px}.ui-icon-arrow-1-w{background-position:-96px -32px}.ui-icon-arrow-1-nw{background-position:-112px -32px}.ui-icon-arrow-2-n-s{background-position:-128px -32px}.ui-icon-arrow-2-ne-sw{background-position:-144px -32px}.ui-icon-arrow-2-e-w{background-position:-160px -32px}.ui-icon-arrow-2-se-nw{background-position:-176px -32px}.ui-icon-arrowstop-1-n{background-position:-192px -32px}.ui-icon-arrowstop-1-e{background-position:-208px -32px}.ui-icon-arrowstop-1-s{background-position:-224px -32px}.ui-icon-arrowstop-1-w{background-position:-240px -32px}.ui-icon-arrowthick-1-n{background-position:0 -48px}.ui-icon-arrowthick-1-ne{background-position:-16px -48px}.ui-icon-arrowthick-1-e{background-position:-32px -48px}.ui-icon-arrowthick-1-se{background-position:-48px -48px}.ui-icon-arrowthick-1-s{background-position:-64px -48px}.ui-icon-arrowthick-1-sw{background-position:-80px -48px}.ui-icon-arrowthick-1-w{background-position:-96px -48px}.ui-icon-arrowthick-1-nw{background-position:-112px -48px}.ui-icon-arrowthick-2-n-s{background-position:-128px -48px}.ui-icon-arrowthick-2-ne-sw{background-position:-144px -48px}.ui-icon-arrowthick-2-e-w{background-position:-160px -48px}.ui-icon-arrowthick-2-se-nw{background-position:-176px -48px}.ui-icon-arrowthickstop-1-n{background-position:-192px -48px}.ui-icon-arrowthickstop-1-e{background-position:-208px -48px}.ui-icon-arrowthickstop-1-s{background-position:-224px -48px}.ui-icon-arrowthickstop-1-w{background-position:-240px -48px}.ui-icon-arrowreturnthick-1-w{background-position:0 -64px}.ui-icon-arrowreturnthick-1-n{background-position:-16px -64px}.ui-icon-arrowreturnthick-1-e{background-position:-32px -64px}.ui-icon-arrowreturnthick-1-s{background-position:-48px -64px}.ui-icon-arrowreturn-1-w{background-position:-64px -64px}.ui-icon-arrowreturn-1-n{background-position:-80px -64px}.ui-icon-arrowreturn-1-e{background-position:-96px -64px}.ui-icon-arrowreturn-1-s{background-position:-112px -64px}.ui-icon-arrowrefresh-1-w{background-position:-128px -64px}.ui-icon-arrowrefresh-1-n{background-position:-144px -64px}.ui-icon-arrowrefresh-1-e{background-position:-160px -64px}.ui-icon-arrowrefresh-1-s{background-position:-176px -64px}.ui-icon-arrow-4{background-position:0 -80px}.ui-icon-arrow-4-diag{background-position:-16px -80px}.ui-icon-extlink{background-position:-32px -80px}.ui-icon-newwin{background-position:-48px -80px}.ui-icon-refresh{background-position:-64px -80px}.ui-icon-shuffle{background-position:-80px -80px}.ui-icon-transfer-e-w{background-position:-96px -80px}.ui-icon-transferthick-e-w{background-position:-112px -80px}.ui-icon-folder-collapsed{background-position:0 -96px}.ui-icon-folder-open{background-position:-16px -96px}.ui-icon-document{background-position:-32px -96px}.ui-icon-document-b{background-position:-48px -96px}.ui-icon-note{background-position:-64px -96px}.ui-icon-mail-closed{background-position:-80px -96px}.ui-icon-mail-open{background-position:-96px -96px}.ui-icon-suitcase{background-position:-112px -96px}.ui-icon-comment{background-position:-128px -96px}.ui-icon-person{background-position:-144px -96px}.ui-icon-print{background-position:-160px -96px}.ui-icon-trash{background-position:-176px -96px}.ui-icon-locked{background-position:-192px -96px}.ui-icon-unlocked{background-position:-208px -96px}.ui-icon-bookmark{background-position:-224px -96px}.ui-icon-tag{background-position:-240px -96px}.ui-icon-home{background-position:0 -112px}.ui-icon-flag{background-position:-16px -112px}.ui-icon-calendar{background-position:-32px -112px}.ui-icon-cart{background-position:-48px -112px}.ui-icon-pencil{background-position:-64px -112px}.ui-icon-clock{background-position:-80px -112px}.ui-icon-disk{background-position:-96px -112px}.ui-icon-calculator{background-position:-112px -112px}.ui-icon-zoomin{background-position:-128px -112px}.ui-icon-zoomout{background-position:-144px -112px}.ui-icon-search{background-position:-160px -112px}.ui-icon-wrench{background-position:-176px -112px}.ui-icon-gear{background-position:-192px -112px}.ui-icon-heart{background-position:-208px -112px}.ui-icon-star{background-position:-224px -112px}.ui-icon-link{background-position:-240px -112px}.ui-icon-cancel{background-position:0 -128px}.ui-icon-plus{background-position:-16px -128px}.ui-icon-plusthick{background-position:-32px -128px}.ui-icon-minus{background-position:-48px -128px}.ui-icon-minusthick{background-position:-64px -128px}.ui-icon-close{background-position:-80px -128px}.ui-icon-closethick{background-position:-96px -128px}.ui-icon-key{background-position:-112px -128px}.ui-icon-lightbulb{background-position:-128px -128px}.ui-icon-scissors{background-position:-144px -128px}.ui-icon-clipboard{background-position:-160px -128px}.ui-icon-copy{background-position:-176px -128px}.ui-icon-contact{background-position:-192px -128px}.ui-icon-image{background-position:-208px -128px}.ui-icon-video{background-position:-224px -128px}.ui-icon-script{background-position:-240px -128px}.ui-icon-alert{background-position:0 -144px}.ui-icon-info{background-position:-16px -144px}.ui-icon-notice{background-position:-32px -144px}.ui-icon-help{background-position:-48px -144px}.ui-icon-check{background-position:-64px -144px}.ui-icon-bullet{background-position:-80px -144px}.ui-icon-radio-on{background-position:-96px -144px}.ui-icon-radio-off{background-position:-112px -144px}.ui-icon-pin-w{background-position:-128px -144px}.ui-icon-pin-s{background-position:-144px -144px}.ui-icon-play{background-position:0 -160px}.ui-icon-pause{background-position:-16px -160px}.ui-icon-seek-next{background-position:-32px -160px}.ui-icon-seek-prev{background-position:-48px -160px}.ui-icon-seek-end{background-position:-64px -160px}.ui-icon-seek-start{background-position:-80px -160px}.ui-icon-seek-first{background-position:-80px -160px}.ui-icon-stop{background-position:-96px -160px}.ui-icon-eject{background-position:-112px -160px}.ui-icon-volume-off{background-position:-128px -160px}.ui-icon-volume-on{background-position:-144px -160px}.ui-icon-power{background-position:0 -176px}.ui-icon-signal-diag{background-position:-16px -176px}.ui-icon-signal{background-position:-32px -176px}.ui-icon-battery-0{background-position:-48px -176px}.ui-icon-battery-1{background-position:-64px -176px}.ui-icon-battery-2{background-position:-80px -176px}.ui-icon-battery-3{background-position:-96px -176px}.ui-icon-circle-plus{background-position:0 -192px}.ui-icon-circle-minus{background-position:-16px -192px}.ui-icon-circle-close{background-position:-32px -192px}.ui-icon-circle-triangle-e{background-position:-48px -192px}.ui-icon-circle-triangle-s{background-position:-64px -192px}.ui-icon-circle-triangle-w{background-position:-80px -192px}.ui-icon-circle-triangle-n{background-position:-96px -192px}.ui-icon-circle-arrow-e{background-position:-112px -192px}.ui-icon-circle-arrow-s{background-position:-128px -192px}.ui-icon-circle-arrow-w{background-position:-144px -192px}.ui-icon-circle-arrow-n{background-position:-160px -192px}.ui-icon-circle-zoomin{background-position:-176px -192px}.ui-icon-circle-zoomout{background-position:-192px -192px}.ui-icon-circle-check{background-position:-208px -192px}.ui-icon-circlesmall-plus{background-position:0 -208px}.ui-icon-circlesmall-minus{background-position:-16px -208px}.ui-icon-circlesmall-close{background-position:-32px -208px}.ui-icon-squaresmall-plus{background-position:-48px -208px}.ui-icon-squaresmall-minus{background-position:-64px -208px}.ui-icon-squaresmall-close{background-position:-80px -208px}.ui-icon-grip-dotted-vertical{background-position:0 -224px}.ui-icon-grip-dotted-horizontal{background-position:-16px -224px}.ui-icon-grip-solid-vertical{background-position:-32px -224px}.ui-icon-grip-solid-horizontal{background-position:-48px -224px}.ui-icon-gripsmall-diagonal-se{background-position:-64px -224px}.ui-icon-grip-diagonal-se{background-position:-80px -224px}.ui-corner-all,.ui-corner-top,.ui-corner-left,.ui-corner-tl{-moz-border-radius-topleft:6px;-webkit-border-top-left-radius:6px;-khtml-border-top-left-radius:6px;border-top-left-radius:6px}.ui-corner-all,.ui-corner-top,.ui-corner-right,.ui-corner-tr{-moz-border-radius-topright:6px;-webkit-border-top-right-radius:6px;-khtml-border-top-right-radius:6px;border-top-right-radius:6px}.ui-corner-all,.ui-corner-bottom,.ui-corner-left,.ui-corner-bl{-moz-border-radius-bottomleft:6px;-webkit-border-bottom-left-radius:6px;-khtml-border-bottom-left-radius:6px;border-bottom-left-radius:6px}.ui-corner-all,.ui-corner-bottom,.ui-corner-right,.ui-corner-br{-moz-border-radius-bottomright:6px;-webkit-border-bottom-right-radius:6px;-khtml-border-bottom-right-radius:6px;border-bottom-right-radius:6px}#simplemodal-overlay{background-color:#000}#simplemodal-container{height:360px;width:600px;color:#fff;background-color:#000;border:0;padding:0}#simplemodal-container .simplemodal-data{padding:20px 50px}.slick-slider{position:relative;display:block;box-sizing:border-box;-moz-box-sizing:border-box;-webkit-touch-callout:none;-webkit-user-select:none;-khtml-user-select:none;-moz-user-select:none;-ms-user-select:none;user-select:none;-ms-touch-action:pan-y;touch-action:pan-y;-webkit-tap-highlight-color:transparent}.slick-list{position:relative;overflow:hidden;display:block;margin:0;padding:0}.slick-list:focus{outline:none}.slick-loading .slick-list{background:#fff url(\"https://cdnjs.cloudflare.com/ajax/libs/slick-carousel/1.6.0/ajax-loader.gif\") center center no-repeat}.slick-list.dragging{cursor:pointer;cursor:hand}.slick-slider .slick-list,.slick-track,.slick-slide,.slick-slide img{-webkit-transform:translate3d(0, 0, 0);-moz-transform:translate3d(0, 0, 0);-ms-transform:translate3d(0, 0, 0);-o-transform:translate3d(0, 0, 0);transform:translate3d(0, 0, 0)}.slick-track{position:relative;left:0;top:0;display:block;zoom:1}.slick-track:before,.slick-track:after{content:\"\";display:table}.slick-track:after{clear:both}.slick-loading .slick-track{visibility:hidden}.slick-slide{float:left;height:100%;min-height:1px;display:none}.slick-slide img{display:block}.slick-slide.slick-loading img{display:none}.slick-slide.dragging img{pointer-events:none}.slick-initialized .slick-slide{display:block}.slick-loading .slick-slide{visibility:hidden}.slick-vertical .slick-slide{display:block;height:auto;border:1px solid transparent}@font-face{font-family:\"slick\";src:url(\"https://cdnjs.cloudflare.com/ajax/libs/slick-carousel/1.6.0/fonts/slick.eot\");src:url(\"https://cdnjs.cloudflare.com/ajax/libs/slick-carousel/1.6.0/fonts/slick.eot?#iefix\") format(\"embedded-opentype\"),url(\"https://cdnjs.cloudflare.com/ajax/libs/slick-carousel/1.6.0/fonts/slick.woff\") format(\"woff\"),url(\"https://cdnjs.cloudflare.com/ajax/libs/slick-carousel/1.6.0/fonts/slick.ttf\") format(\"truetype\"),url(\"https://cdnjs.cloudflare.com/ajax/libs/slick-carousel/1.6.0/fonts/slick.svg#slick\") format(\"svg\");font-weight:normal;font-style:normal}.slick-prev,.slick-next{position:absolute;display:block;height:20px;width:20px;line-height:0;font-size:0;cursor:pointer;background:transparent;color:transparent;top:50%;margin-top:-10px;padding:0;border:none;outline:none}.slick-prev:hover,.slick-prev:focus,.slick-next:hover,.slick-next:focus{outline:none;background:transparent;color:transparent}.slick-prev:hover:before,.slick-prev:focus:before,.slick-next:hover:before,.slick-next:focus:before{opacity:1}.slick-prev.slick-disabled:before,.slick-next.slick-disabled:before{opacity:0.25}.slick-prev:before,.slick-next:before{font-family:\"slick\";font-size:20px;line-height:1;color:white;opacity:0.75;-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale}.slick-prev{left:-25px}.slick-prev:before{content:\"\\2190\"}.slick-next{right:-25px}.slick-next:before{content:\"\\2192\"}.slick-slider{margin-bottom:30px}.slick-dots{position:absolute;bottom:-45px;list-style:none;display:block;text-align:center;padding:0;width:100%}.slick-dots li{position:relative;display:inline-block;height:20px;width:20px;margin:0 5px;padding:0;cursor:pointer}.slick-dots li button{border:0;background:transparent;display:block;height:20px;width:20px;outline:none;line-height:0;font-size:0;color:transparent;padding:5px;cursor:pointer}.slick-dots li button:hover,.slick-dots li button:focus{outline:none}.slick-dots li button:hover:before,.slick-dots li button:focus:before{opacity:1}.slick-dots li button:before{position:absolute;top:0;left:0;content:\"\\2022\";width:20px;height:20px;font-family:\"slick\";font-size:6px;line-height:20px;text-align:center;color:black;opacity:0.25;-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale}.slick-dots li.slick-active button:before{color:black;opacity:0.75}[dir=\"rtl\"] .slick-next{right:auto;left:-25px}[dir=\"rtl\"] .slick-next:before{content:\"\\2190\"}[dir=\"rtl\"] .slick-prev{right:-25px;left:auto}[dir=\"rtl\"] .slick-prev:before{content:\"\\2192\"}[dir=\"rtl\"] .slick-slide{float:right}@font-face{font-family:\"foundation-icons\";src:url(\"https://mars.nasa.gov/assets/gulp/vendor/fonts/foundation-icons.eot\");src:url(\"https://mars.nasa.gov/assets/gulp/vendor/fonts/foundation-icons.eot?#iefix\") format(\"embedded-opentype\"),url(\"https://mars.nasa.gov/assets/gulp/vendor/fonts/foundation-icons.woff\") format(\"woff\"),url(\"https://mars.nasa.gov/assets/gulp/vendor/fonts/foundation-icons.ttf\") format(\"truetype\"),url(\"https://mars.nasa.gov/assets/gulp/vendor/fonts/foundation-icons.svg#fontcustom\") format(\"svg\");font-weight:normal;font-style:normal}.fi-address-book:before,.fi-alert:before,.fi-align-center:before,.fi-align-justify:before,.fi-align-left:before,.fi-align-right:before,.fi-anchor:before,.fi-annotate:before,.fi-archive:before,.fi-arrow-down:before,.fi-arrow-left:before,.fi-arrow-right:before,.fi-arrow-up:before,.fi-arrows-compress:before,.fi-arrows-expand:before,.fi-arrows-in:before,.fi-arrows-out:before,.fi-asl:before,.fi-asterisk:before,.fi-at-sign:before,.fi-background-color:before,.fi-battery-empty:before,.fi-battery-full:before,.fi-battery-half:before,.fi-bitcoin-circle:before,.fi-bitcoin:before,.fi-blind:before,.fi-bluetooth:before,.fi-bold:before,.fi-book-bookmark:before,.fi-book:before,.fi-bookmark:before,.fi-braille:before,.fi-burst-new:before,.fi-burst-sale:before,.fi-burst:before,.fi-calendar:before,.fi-camera:before,.fi-check:before,.fi-checkbox:before,.fi-clipboard-notes:before,.fi-clipboard-pencil:before,.fi-clipboard:before,.fi-clock:before,.fi-closed-caption:before,.fi-cloud:before,.fi-comment-minus:before,.fi-comment-quotes:before,.fi-comment-video:before,.fi-comment:before,.fi-comments:before,.fi-compass:before,.fi-contrast:before,.fi-credit-card:before,.fi-crop:before,.fi-crown:before,.fi-css3:before,.fi-database:before,.fi-die-five:before,.fi-die-four:before,.fi-die-one:before,.fi-die-six:before,.fi-die-three:before,.fi-die-two:before,.fi-dislike:before,.fi-dollar-bill:before,.fi-dollar:before,.fi-download:before,.fi-eject:before,.fi-elevator:before,.fi-euro:before,.fi-eye:before,.fi-fast-forward:before,.fi-female-symbol:before,.fi-female:before,.fi-filter:before,.fi-first-aid:before,.fi-flag:before,.fi-folder-add:before,.fi-folder-lock:before,.fi-folder:before,.fi-foot:before,.fi-foundation:before,.fi-graph-bar:before,.fi-graph-horizontal:before,.fi-graph-pie:before,.fi-graph-trend:before,.fi-guide-dog:before,.fi-hearing-aid:before,.fi-heart:before,.fi-home:before,.fi-html5:before,.fi-indent-less:before,.fi-indent-more:before,.fi-info:before,.fi-italic:before,.fi-key:before,.fi-laptop:before,.fi-layout:before,.fi-lightbulb:before,.fi-like:before,.fi-link:before,.fi-list-bullet:before,.fi-list-number:before,.fi-list-thumbnails:before,.fi-list:before,.fi-lock:before,.fi-loop:before,.fi-magnifying-glass:before,.fi-mail:before,.fi-male-female:before,.fi-male-symbol:before,.fi-male:before,.fi-map:before,.fi-marker:before,.fi-megaphone:before,.fi-microphone:before,.fi-minus-circle:before,.fi-minus:before,.fi-mobile-signal:before,.fi-mobile:before,.fi-monitor:before,.fi-mountains:before,.fi-music:before,.fi-next:before,.fi-no-dogs:before,.fi-no-smoking:before,.fi-page-add:before,.fi-page-copy:before,.fi-page-csv:before,.fi-page-delete:before,.fi-page-doc:before,.fi-page-edit:before,.fi-page-export-csv:before,.fi-page-export-doc:before,.fi-page-export-pdf:before,.fi-page-export:before,.fi-page-filled:before,.fi-page-multiple:before,.fi-page-pdf:before,.fi-page-remove:before,.fi-page-search:before,.fi-page:before,.fi-paint-bucket:before,.fi-paperclip:before,.fi-pause:before,.fi-paw:before,.fi-paypal:before,.fi-pencil:before,.fi-photo:before,.fi-play-circle:before,.fi-play-video:before,.fi-play:before,.fi-plus:before,.fi-pound:before,.fi-power:before,.fi-previous:before,.fi-price-tag:before,.fi-pricetag-multiple:before,.fi-print:before,.fi-prohibited:before,.fi-projection-screen:before,.fi-puzzle:before,.fi-quote:before,.fi-record:before,.fi-refresh:before,.fi-results-demographics:before,.fi-results:before,.fi-rewind-ten:before,.fi-rewind:before,.fi-rss:before,.fi-safety-cone:before,.fi-save:before,.fi-share:before,.fi-sheriff-badge:before,.fi-shield:before,.fi-shopping-bag:before,.fi-shopping-cart:before,.fi-shuffle:before,.fi-skull:before,.fi-social-500px:before,.fi-social-adobe:before,.fi-social-amazon:before,.fi-social-android:before,.fi-social-apple:before,.fi-social-behance:before,.fi-social-bing:before,.fi-social-blogger:before,.fi-social-delicious:before,.fi-social-designer-news:before,.fi-social-deviant-art:before,.fi-social-digg:before,.fi-social-dribbble:before,.fi-social-drive:before,.fi-social-dropbox:before,.fi-social-evernote:before,.fi-social-facebook:before,.fi-social-flickr:before,.fi-social-forrst:before,.fi-social-foursquare:before,.fi-social-game-center:before,.fi-social-github:before,.fi-social-google-plus:before,.fi-social-hacker-news:before,.fi-social-hi5:before,.fi-social-instagram:before,.fi-social-joomla:before,.fi-social-lastfm:before,.fi-social-linkedin:before,.fi-social-medium:before,.fi-social-myspace:before,.fi-social-orkut:before,.fi-social-path:before,.fi-social-picasa:before,.fi-social-pinterest:before,.fi-social-rdio:before,.fi-social-reddit:before,.fi-social-skillshare:before,.fi-social-skype:before,.fi-social-smashing-mag:before,.fi-social-snapchat:before,.fi-social-spotify:before,.fi-social-squidoo:before,.fi-social-stack-overflow:before,.fi-social-steam:before,.fi-social-stumbleupon:before,.fi-social-treehouse:before,.fi-social-tumblr:before,.fi-social-twitter:before,.fi-social-vimeo:before,.fi-social-windows:before,.fi-social-xbox:before,.fi-social-yahoo:before,.fi-social-yelp:before,.fi-social-youtube:before,.fi-social-zerply:before,.fi-social-zurb:before,.fi-sound:before,.fi-star:before,.fi-stop:before,.fi-strikethrough:before,.fi-subscript:before,.fi-superscript:before,.fi-tablet-landscape:before,.fi-tablet-portrait:before,.fi-target-two:before,.fi-target:before,.fi-telephone-accessible:before,.fi-telephone:before,.fi-text-color:before,.fi-thumbnails:before,.fi-ticket:before,.fi-torso-business:before,.fi-torso-female:before,.fi-torso:before,.fi-torsos-all-female:before,.fi-torsos-all:before,.fi-torsos-female-male:before,.fi-torsos-male-female:before,.fi-torsos:before,.fi-trash:before,.fi-trees:before,.fi-trophy:before,.fi-underline:before,.fi-universal-access:before,.fi-unlink:before,.fi-unlock:before,.fi-upload-cloud:before,.fi-upload:before,.fi-usb:before,.fi-video:before,.fi-volume-none:before,.fi-volume-strike:before,.fi-volume:before,.fi-web:before,.fi-wheelchair:before,.fi-widget:before,.fi-wrench:before,.fi-x-circle:before,.fi-x:before,.fi-yen:before,.fi-zoom-in:before,.fi-zoom-out:before{font-family:\"foundation-icons\";font-style:normal;font-weight:normal;font-variant:normal;text-transform:none;line-height:1;-webkit-font-smoothing:antialiased;display:inline-block;text-decoration:inherit}.fi-address-book:before{content:\"\\f100\"}.fi-alert:before{content:\"\\f101\"}.fi-align-center:before{content:\"\\f102\"}.fi-align-justify:before{content:\"\\f103\"}.fi-align-left:before{content:\"\\f104\"}.fi-align-right:before{content:\"\\f105\"}.fi-anchor:before{content:\"\\f106\"}.fi-annotate:before{content:\"\\f107\"}.fi-archive:before{content:\"\\f108\"}.fi-arrow-down:before{content:\"\\f109\"}.fi-arrow-left:before{content:\"\\f10a\"}.fi-arrow-right:before{content:\"\\f10b\"}.fi-arrow-up:before{content:\"\\f10c\"}.fi-arrows-compress:before{content:\"\\f10d\"}.fi-arrows-expand:before{content:\"\\f10e\"}.fi-arrows-in:before{content:\"\\f10f\"}.fi-arrows-out:before{content:\"\\f110\"}.fi-asl:before{content:\"\\f111\"}.fi-asterisk:before{content:\"\\f112\"}.fi-at-sign:before{content:\"\\f113\"}.fi-background-color:before{content:\"\\f114\"}.fi-battery-empty:before{content:\"\\f115\"}.fi-battery-full:before{content:\"\\f116\"}.fi-battery-half:before{content:\"\\f117\"}.fi-bitcoin-circle:before{content:\"\\f118\"}.fi-bitcoin:before{content:\"\\f119\"}.fi-blind:before{content:\"\\f11a\"}.fi-bluetooth:before{content:\"\\f11b\"}.fi-bold:before{content:\"\\f11c\"}.fi-book-bookmark:before{content:\"\\f11d\"}.fi-book:before{content:\"\\f11e\"}.fi-bookmark:before{content:\"\\f11f\"}.fi-braille:before{content:\"\\f120\"}.fi-burst-new:before{content:\"\\f121\"}.fi-burst-sale:before{content:\"\\f122\"}.fi-burst:before{content:\"\\f123\"}.fi-calendar:before{content:\"\\f124\"}.fi-camera:before{content:\"\\f125\"}.fi-check:before{content:\"\\f126\"}.fi-checkbox:before{content:\"\\f127\"}.fi-clipboard-notes:before{content:\"\\f128\"}.fi-clipboard-pencil:before{content:\"\\f129\"}.fi-clipboard:before{content:\"\\f12a\"}.fi-clock:before{content:\"\\f12b\"}.fi-closed-caption:before{content:\"\\f12c\"}.fi-cloud:before{content:\"\\f12d\"}.fi-comment-minus:before{content:\"\\f12e\"}.fi-comment-quotes:before{content:\"\\f12f\"}.fi-comment-video:before{content:\"\\f130\"}.fi-comment:before{content:\"\\f131\"}.fi-comments:before{content:\"\\f132\"}.fi-compass:before{content:\"\\f133\"}.fi-contrast:before{content:\"\\f134\"}.fi-credit-card:before{content:\"\\f135\"}.fi-crop:before{content:\"\\f136\"}.fi-crown:before{content:\"\\f137\"}.fi-css3:before{content:\"\\f138\"}.fi-database:before{content:\"\\f139\"}.fi-die-five:before{content:\"\\f13a\"}.fi-die-four:before{content:\"\\f13b\"}.fi-die-one:before{content:\"\\f13c\"}.fi-die-six:before{content:\"\\f13d\"}.fi-die-three:before{content:\"\\f13e\"}.fi-die-two:before{content:\"\\f13f\"}.fi-dislike:before{content:\"\\f140\"}.fi-dollar-bill:before{content:\"\\f141\"}.fi-dollar:before{content:\"\\f142\"}.fi-download:before{content:\"\\f143\"}.fi-eject:before{content:\"\\f144\"}.fi-elevator:before{content:\"\\f145\"}.fi-euro:before{content:\"\\f146\"}.fi-eye:before{content:\"\\f147\"}.fi-fast-forward:before{content:\"\\f148\"}.fi-female-symbol:before{content:\"\\f149\"}.fi-female:before{content:\"\\f14a\"}.fi-filter:before{content:\"\\f14b\"}.fi-first-aid:before{content:\"\\f14c\"}.fi-flag:before{content:\"\\f14d\"}.fi-folder-add:before{content:\"\\f14e\"}.fi-folder-lock:before{content:\"\\f14f\"}.fi-folder:before{content:\"\\f150\"}.fi-foot:before{content:\"\\f151\"}.fi-foundation:before{content:\"\\f152\"}.fi-graph-bar:before{content:\"\\f153\"}.fi-graph-horizontal:before{content:\"\\f154\"}.fi-graph-pie:before{content:\"\\f155\"}.fi-graph-trend:before{content:\"\\f156\"}.fi-guide-dog:before{content:\"\\f157\"}.fi-hearing-aid:before{content:\"\\f158\"}.fi-heart:before{content:\"\\f159\"}.fi-home:before{content:\"\\f15a\"}.fi-html5:before{content:\"\\f15b\"}.fi-indent-less:before{content:\"\\f15c\"}.fi-indent-more:before{content:\"\\f15d\"}.fi-info:before{content:\"\\f15e\"}.fi-italic:before{content:\"\\f15f\"}.fi-key:before{content:\"\\f160\"}.fi-laptop:before{content:\"\\f161\"}.fi-layout:before{content:\"\\f162\"}.fi-lightbulb:before{content:\"\\f163\"}.fi-like:before{content:\"\\f164\"}.fi-link:before{content:\"\\f165\"}.fi-list-bullet:before{content:\"\\f166\"}.fi-list-number:before{content:\"\\f167\"}.fi-list-thumbnails:before{content:\"\\f168\"}.fi-list:before{content:\"\\f169\"}.fi-lock:before{content:\"\\f16a\"}.fi-loop:before{content:\"\\f16b\"}.fi-magnifying-glass:before{content:\"\\f16c\"}.fi-mail:before{content:\"\\f16d\"}.fi-male-female:before{content:\"\\f16e\"}.fi-male-symbol:before{content:\"\\f16f\"}.fi-male:before{content:\"\\f170\"}.fi-map:before{content:\"\\f171\"}.fi-marker:before{content:\"\\f172\"}.fi-megaphone:before{content:\"\\f173\"}.fi-microphone:before{content:\"\\f174\"}.fi-minus-circle:before{content:\"\\f175\"}.fi-minus:before{content:\"\\f176\"}.fi-mobile-signal:before{content:\"\\f177\"}.fi-mobile:before{content:\"\\f178\"}.fi-monitor:before{content:\"\\f179\"}.fi-mountains:before{content:\"\\f17a\"}.fi-music:before{content:\"\\f17b\"}.fi-next:before{content:\"\\f17c\"}.fi-no-dogs:before{content:\"\\f17d\"}.fi-no-smoking:before{content:\"\\f17e\"}.fi-page-add:before{content:\"\\f17f\"}.fi-page-copy:before{content:\"\\f180\"}.fi-page-csv:before{content:\"\\f181\"}.fi-page-delete:before{content:\"\\f182\"}.fi-page-doc:before{content:\"\\f183\"}.fi-page-edit:before{content:\"\\f184\"}.fi-page-export-csv:before{content:\"\\f185\"}.fi-page-export-doc:before{content:\"\\f186\"}.fi-page-export-pdf:before{content:\"\\f187\"}.fi-page-export:before{content:\"\\f188\"}.fi-page-filled:before{content:\"\\f189\"}.fi-page-multiple:before{content:\"\\f18a\"}.fi-page-pdf:before{content:\"\\f18b\"}.fi-page-remove:before{content:\"\\f18c\"}.fi-page-search:before{content:\"\\f18d\"}.fi-page:before{content:\"\\f18e\"}.fi-paint-bucket:before{content:\"\\f18f\"}.fi-paperclip:before{content:\"\\f190\"}.fi-pause:before{content:\"\\f191\"}.fi-paw:before{content:\"\\f192\"}.fi-paypal:before{content:\"\\f193\"}.fi-pencil:before{content:\"\\f194\"}.fi-photo:before{content:\"\\f195\"}.fi-play-circle:before{content:\"\\f196\"}.fi-play-video:before{content:\"\\f197\"}.fi-play:before{content:\"\\f198\"}.fi-plus:before{content:\"\\f199\"}.fi-pound:before{content:\"\\f19a\"}.fi-power:before{content:\"\\f19b\"}.fi-previous:before{content:\"\\f19c\"}.fi-price-tag:before{content:\"\\f19d\"}.fi-pricetag-multiple:before{content:\"\\f19e\"}.fi-print:before{content:\"\\f19f\"}.fi-prohibited:before{content:\"\\f1a0\"}.fi-projection-screen:before{content:\"\\f1a1\"}.fi-puzzle:before{content:\"\\f1a2\"}.fi-quote:before{content:\"\\f1a3\"}.fi-record:before{content:\"\\f1a4\"}.fi-refresh:before{content:\"\\f1a5\"}.fi-results-demographics:before{content:\"\\f1a6\"}.fi-results:before{content:\"\\f1a7\"}.fi-rewind-ten:before{content:\"\\f1a8\"}.fi-rewind:before{content:\"\\f1a9\"}.fi-rss:before{content:\"\\f1aa\"}.fi-safety-cone:before{content:\"\\f1ab\"}.fi-save:before{content:\"\\f1ac\"}.fi-share:before{content:\"\\f1ad\"}.fi-sheriff-badge:before{content:\"\\f1ae\"}.fi-shield:before{content:\"\\f1af\"}.fi-shopping-bag:before{content:\"\\f1b0\"}.fi-shopping-cart:before{content:\"\\f1b1\"}.fi-shuffle:before{content:\"\\f1b2\"}.fi-skull:before{content:\"\\f1b3\"}.fi-social-500px:before{content:\"\\f1b4\"}.fi-social-adobe:before{content:\"\\f1b5\"}.fi-social-amazon:before{content:\"\\f1b6\"}.fi-social-android:before{content:\"\\f1b7\"}.fi-social-apple:before{content:\"\\f1b8\"}.fi-social-behance:before{content:\"\\f1b9\"}.fi-social-bing:before{content:\"\\f1ba\"}.fi-social-blogger:before{content:\"\\f1bb\"}.fi-social-delicious:before{content:\"\\f1bc\"}.fi-social-designer-news:before{content:\"\\f1bd\"}.fi-social-deviant-art:before{content:\"\\f1be\"}.fi-social-digg:before{content:\"\\f1bf\"}.fi-social-dribbble:before{content:\"\\f1c0\"}.fi-social-drive:before{content:\"\\f1c1\"}.fi-social-dropbox:before{content:\"\\f1c2\"}.fi-social-evernote:before{content:\"\\f1c3\"}.fi-social-facebook:before{content:\"\\f1c4\"}.fi-social-flickr:before{content:\"\\f1c5\"}.fi-social-forrst:before{content:\"\\f1c6\"}.fi-social-foursquare:before{content:\"\\f1c7\"}.fi-social-game-center:before{content:\"\\f1c8\"}.fi-social-github:before{content:\"\\f1c9\"}.fi-social-google-plus:before{content:\"\\f1ca\"}.fi-social-hacker-news:before{content:\"\\f1cb\"}.fi-social-hi5:before{content:\"\\f1cc\"}.fi-social-instagram:before{content:\"\\f1cd\"}.fi-social-joomla:before{content:\"\\f1ce\"}.fi-social-lastfm:before{content:\"\\f1cf\"}.fi-social-linkedin:before{content:\"\\f1d0\"}.fi-social-medium:before{content:\"\\f1d1\"}.fi-social-myspace:before{content:\"\\f1d2\"}.fi-social-orkut:before{content:\"\\f1d3\"}.fi-social-path:before{content:\"\\f1d4\"}.fi-social-picasa:before{content:\"\\f1d5\"}.fi-social-pinterest:before{content:\"\\f1d6\"}.fi-social-rdio:before{content:\"\\f1d7\"}.fi-social-reddit:before{content:\"\\f1d8\"}.fi-social-skillshare:before{content:\"\\f1d9\"}.fi-social-skype:before{content:\"\\f1da\"}.fi-social-smashing-mag:before{content:\"\\f1db\"}.fi-social-snapchat:before{content:\"\\f1dc\"}.fi-social-spotify:before{content:\"\\f1dd\"}.fi-social-squidoo:before{content:\"\\f1de\"}.fi-social-stack-overflow:before{content:\"\\f1df\"}.fi-social-steam:before{content:\"\\f1e0\"}.fi-social-stumbleupon:before{content:\"\\f1e1\"}.fi-social-treehouse:before{content:\"\\f1e2\"}.fi-social-tumblr:before{content:\"\\f1e3\"}.fi-social-twitter:before{content:\"\\f1e4\"}.fi-social-vimeo:before{content:\"\\f1e5\"}.fi-social-windows:before{content:\"\\f1e6\"}.fi-social-xbox:before{content:\"\\f1e7\"}.fi-social-yahoo:before{content:\"\\f1e8\"}.fi-social-yelp:before{content:\"\\f1e9\"}.fi-social-youtube:before{content:\"\\f1ea\"}.fi-social-zerply:before{content:\"\\f1eb\"}.fi-social-zurb:before{content:\"\\f1ec\"}.fi-sound:before{content:\"\\f1ed\"}.fi-star:before{content:\"\\f1ee\"}.fi-stop:before{content:\"\\f1ef\"}.fi-strikethrough:before{content:\"\\f1f0\"}.fi-subscript:before{content:\"\\f1f1\"}.fi-superscript:before{content:\"\\f1f2\"}.fi-tablet-landscape:before{content:\"\\f1f3\"}.fi-tablet-portrait:before{content:\"\\f1f4\"}.fi-target-two:before{content:\"\\f1f5\"}.fi-target:before{content:\"\\f1f6\"}.fi-telephone-accessible:before{content:\"\\f1f7\"}.fi-telephone:before{content:\"\\f1f8\"}.fi-text-color:before{content:\"\\f1f9\"}.fi-thumbnails:before{content:\"\\f1fa\"}.fi-ticket:before{content:\"\\f1fb\"}.fi-torso-business:before{content:\"\\f1fc\"}.fi-torso-female:before{content:\"\\f1fd\"}.fi-torso:before{content:\"\\f1fe\"}.fi-torsos-all-female:before{content:\"\\f1ff\"}.fi-torsos-all:before{content:\"\\f200\"}.fi-torsos-female-male:before{content:\"\\f201\"}.fi-torsos-male-female:before{content:\"\\f202\"}.fi-torsos:before{content:\"\\f203\"}.fi-trash:before{content:\"\\f204\"}.fi-trees:before{content:\"\\f205\"}.fi-trophy:before{content:\"\\f206\"}.fi-underline:before{content:\"\\f207\"}.fi-universal-access:before{content:\"\\f208\"}.fi-unlink:before{content:\"\\f209\"}.fi-unlock:before{content:\"\\f20a\"}.fi-upload-cloud:before{content:\"\\f20b\"}.fi-upload:before{content:\"\\f20c\"}.fi-usb:before{content:\"\\f20d\"}.fi-video:before{content:\"\\f20e\"}.fi-volume-none:before{content:\"\\f20f\"}.fi-volume-strike:before{content:\"\\f210\"}.fi-volume:before{content:\"\\f211\"}.fi-web:before{content:\"\\f212\"}.fi-wheelchair:before{content:\"\\f213\"}.fi-widget:before{content:\"\\f214\"}.fi-wrench:before{content:\"\\f215\"}.fi-x-circle:before{content:\"\\f216\"}.fi-x:before{content:\"\\f217\"}.fi-yen:before{content:\"\\f218\"}.fi-zoom-in:before{content:\"\\f219\"}.fi-zoom-out:before{content:\"\\f21a\"}/*! normalize.css v1.1.2 | MIT License | git.io/normalize */article,aside,details,figcaption,figure,footer,header,hgroup,main,nav,section,summary{display:block}audio,canvas,video{display:inline-block;*display:inline;*zoom:1}audio:not([controls]){display:none;height:0}[hidden]{display:none}html{font-size:100%;-ms-text-size-adjust:100%;-webkit-text-size-adjust:100%}html,button,input,select,textarea{font-family:sans-serif}body{margin:0}a:focus{outline:thin dotted}a:active,a:hover{outline:0}h1{font-size:2em;margin:0.67em 0}h2{font-size:1.5em;margin:0.83em 0}h3{font-size:1.17em;margin:1em 0}h4{font-size:1em;margin:1.33em 0}h5{font-size:0.83em;margin:1.67em 0}h6{font-size:0.67em;margin:2.33em 0}abbr[title]{border-bottom:1px dotted}b,strong{font-weight:bold}blockquote{margin:1em 40px}dfn{font-style:italic}hr{-moz-box-sizing:content-box;box-sizing:content-box;height:0}mark{background:#ff0;color:#000}p,pre{margin:1em 0}code,kbd,pre,samp{font-family:monospace, serif;_font-family:'courier new', monospace;font-size:1em}pre{white-space:pre;white-space:pre-wrap;word-wrap:break-word}q{quotes:none}q:before,q:after{content:'';content:none}small{font-size:80%}sub,sup{font-size:75%;line-height:0;position:relative;vertical-align:baseline}sup{top:-0.5em}sub{bottom:-0.25em}dl,menu,ol,ul{margin:1em 0}dd{margin:0 0 0 40px}menu,ol,ul{padding:0 0 0 40px}nav ul,nav ol{list-style:none;list-style-image:none}img{border:0;-ms-interpolation-mode:bicubic}svg:not(:root){overflow:hidden}figure{margin:0}form{margin:0}fieldset{border:1px solid #c0c0c0;margin:0 2px;padding:0.35em 0.625em 0.75em}legend{border:0;padding:0;white-space:normal;*margin-left:-7px}button,input,select,textarea{font-size:100%;margin:0;vertical-align:baseline;*vertical-align:middle}button,input{line-height:normal}button,select{text-transform:none}button,html input[type=\"button\"],input[type=\"reset\"],input[type=\"submit\"]{-webkit-appearance:button;cursor:pointer;*overflow:visible}button[disabled],html input[disabled]{cursor:default}input[type=\"checkbox\"],input[type=\"radio\"]{box-sizing:border-box;padding:0;*height:13px;*width:13px}input[type=\"search\"]{-webkit-appearance:textfield;-moz-box-sizing:content-box;-webkit-box-sizing:content-box;box-sizing:content-box}input[type=\"search\"]::-webkit-search-cancel-button,input[type=\"search\"]::-webkit-search-decoration{-webkit-appearance:none}button::-moz-focus-inner,input::-moz-focus-inner{border:0;padding:0}textarea{overflow:auto;vertical-align:top}table{border-collapse:collapse;border-spacing:0}html,button,input,select,textarea{color:#222}body{font-size:1em;line-height:1.4}::-moz-selection{background:#b3d4fc;text-shadow:none}::selection{background:#b3d4fc;text-shadow:none}hr{display:block;height:1px;border:0;border-top:1px solid #ccc;margin:1em 0;padding:0}audio,canvas,img,video{vertical-align:middle}fieldset{border:0;margin:0;padding:0}textarea{resize:vertical}.browsehappy{margin:0.2em 0;background:#ccc;color:#000;padding:0.2em 0}.ir{background-color:transparent;border:0;overflow:hidden;*text-indent:-9999px}.ir:before{content:\"\";display:block;width:0;height:150%}.hidden{display:none !important;visibility:hidden}.visuallyhidden{border:0;clip:rect(0 0 0 0);height:1px;margin:-1px;overflow:hidden;padding:0;position:absolute;width:1px}.visuallyhidden.focusable:active,.visuallyhidden.focusable:focus{clip:auto;height:auto;margin:0;overflow:visible;position:static;width:auto}.invisible{visibility:hidden}.clearfix:before,.body_form fieldset.radio_buttons .option:before,#site_footer .sitemap_directory:before,.main_area_sitemap .sitemap_directory:before,article:before,.grid_gallery.list_view li.slide:before,.main_carousel .slick-nav:before,.main_carousel.module .slick-slider .content_body:before,.advanced_search .filter_bar .search_row:before,.content_page #primary_column:before,#secondary_column aside.list_view_module li:before,.wysiwyg_content .related_content_module ul:before,#secondary_column .related_content_module ul:before,.wysiwyg_content .related_content_module li:before,#secondary_column .related_content_module li:before,blockquote:before,.faq_section ul.q_and_a .text.answer:before,ul.item_list:before,ul.item_list>li:before,ul.item_list .list_content:before,.clearfix:after,.body_form fieldset.radio_buttons .option:after,#site_footer .sitemap_directory:after,.main_area_sitemap .sitemap_directory:after,article:after,.grid_gallery.list_view li.slide:after,.main_carousel .slick-nav:after,.main_carousel.module .slick-slider .content_body:after,.advanced_search .filter_bar .search_row:after,.content_page #primary_column:after,#secondary_column aside.list_view_module li:after,.wysiwyg_content .related_content_module ul:after,#secondary_column .related_content_module ul:after,.wysiwyg_content .related_content_module li:after,#secondary_column .related_content_module li:after,blockquote:after,.faq_section ul.q_and_a .text.answer:after,ul.item_list:after,ul.item_list>li:after,ul.item_list .list_content:after{content:\" \";display:table}.clearfix:after,.body_form fieldset.radio_buttons .option:after,#site_footer .sitemap_directory:after,.main_area_sitemap .sitemap_directory:after,article:after,.grid_gallery.list_view li.slide:after,.main_carousel .slick-nav:after,.main_carousel.module .slick-slider .content_body:after,.advanced_search .filter_bar .search_row:after,.content_page #primary_column:after,#secondary_column aside.list_view_module li:after,.wysiwyg_content .related_content_module ul:after,#secondary_column .related_content_module ul:after,.wysiwyg_content .related_content_module li:after,#secondary_column .related_content_module li:after,blockquote:after,.faq_section ul.q_and_a .text.answer:after,ul.item_list:after,ul.item_list>li:after,ul.item_list .list_content:after{clear:both}.clearfix,.body_form fieldset.radio_buttons .option,#site_footer .sitemap_directory,.main_area_sitemap .sitemap_directory,article,.grid_gallery.list_view li.slide,.main_carousel .slick-nav,.main_carousel.module .slick-slider .content_body,.advanced_search .filter_bar .search_row,.content_page #primary_column,#secondary_column aside.list_view_module li,.wysiwyg_content .related_content_module ul,#secondary_column .related_content_module ul,.wysiwyg_content .related_content_module li,#secondary_column .related_content_module li,blockquote,.faq_section ul.q_and_a .text.answer,ul.item_list,ul.item_list>li,ul.item_list .list_content{*zoom:1}@media print{*{background:transparent !important;color:#000 !important;box-shadow:none !important;text-shadow:none !important}a,a:visited{text-decoration:underline}a[href]:after{content:\" (\" attr(href) \")\"}abbr[title]:after{content:\" (\" attr(title) \")\"}.ir a:after,a[href^=\"javascript:\"]:after,a[href^=\"#\"]:after{content:\"\"}pre,blockquote{border:1px solid #999;page-break-inside:avoid}thead{display:table-header-group}tr,img{page-break-inside:avoid}img{max-width:100% !important}@page{margin:0.5cm}p,h2,h3{orphans:3;widows:3}h2,h3{page-break-after:avoid}}html,button,input,select,textarea{color:#3c3c3c}.browsehappy{background:white;color:#333;padding:1em;position:absolute;top:0;left:0;z-index:9999;width:100%;height:100%}html.touch.-webkit-{-webkit-tap-highlight-color:transparent}.visuallyhidden.focusable:active,.visuallyhidden.focusable:focus{position:absolute}.site_header_area .brand_area{background:url(\"https://mars.nasa.gov/assets/[email protected]\") no-repeat;background-size:100%;display:inline-block;width:54px;height:54px}.site_header_area .brand_area .brand1{height:100%;float:left;text-indent:-9999px}.site_header_area .brand_area .brand2{display:block;float:left;height:100%;text-indent:-9999px}.site_header_area .brand_area a.top_logo,.site_header_area .brand_area a.sub_logo{width:100%;float:left}.site_header_area .brand_area a.top_logo{height:39%;width:30%}.site_header_area .brand_area a.sub_logo{height:45%}.site_header_area .brand_area a.single_logo{width:100%;float:left;height:82%}.site_header_area .brand_area .nasa_logo{width:100%;height:100%;display:block}*,*:before,*:after{-moz-box-sizing:border-box;-webkit-box-sizing:border-box;box-sizing:border-box}body{margin-left:auto;margin-right:auto;margin-top:0;background-color:white}@media (max-width: 1023px){body.nav_overlay_true{overflow:hidden}}img{width:100%}p{line-height:1.4em;margin-bottom:17px;margin-top:0;font-size:16px;color:#222}@media (min-width: 600px), print{p{font-size:18px}}@media (min-width: 769px), print{p{margin-bottom:20px;font-size:16px}}@media (min-width: 1024px), print{p{font-size:17px}}@media (min-width: 1200px){p{font-size:18px}}a{text-decoration:none;color:#257cdf}a:hover{text-decoration:underline}a[name]{position:relative;display:block;visibility:hidden;margin:0;padding:0}@media (max-width: 1023px){a[name]{top:-58px}}@media (max-width: 1023px) and (min-width: 480px){a[name]{top:-58px}}@media (max-width: 1023px) and (min-width: 600px), print and (max-width: 1023px){a[name]{top:-58px}}@media (max-width: 1023px) and (min-width: 769px), print and (max-width: 1023px){a[name]{top:-70px}}@media (min-width: 1024px){a[name]{top:-47px}}dl,menu,ol,ul{margin:0;padding:0}ul{list-style-type:none}ol{list-style-position:inside}hr,.gradient_line,.related.module .gradient_line_module_top{clear:both;margin:1em 0}.print_only{display:none}@font-face{font-family:'Whitney';src:url(\"https://mars.nasa.gov/assets/fonts/Whitney-Book.otf\")}@font-face{font-family:'Whitney-Bold';src:url(\"https://mars.nasa.gov/assets/fonts/Whitney-Bold.otf\")}@font-face{font-family:'WhitneyCondensed-Bold';src:url(\"https://mars.nasa.gov/assets/fonts/WhitneyCondensed-Bold.otf\")}.button,.outline_button,.primary_media_feature .floating_text_area .button,.banner_header_overlay .button{font-weight:700;display:inline-block;margin-bottom:.5em;margin-left:auto;margin-right:auto;background-color:#3b788b;color:white;line-height:1em;border:0;text-decoration:none;border-radius:4px;cursor:pointer;text-shadow:none;font-size:13px;padding:12px 24px;text-transform:uppercase;white-space:nowrap}@media (min-width: 769px), print{.button,.outline_button,.primary_media_feature .floating_text_area .button,.banner_header_overlay .button{font-size:14px}}.button:hover,.outline_button:hover,.primary_media_feature .floating_text_area .button:hover{background-color:#5097ad;text-decoration:none}.outline_button,.primary_media_feature .floating_text_area .button,.primary_media_feature .floating_text_area .outline_button,.banner_header_overlay .button,.banner_header_overlay .outline_button{border-radius:10px;border:2px solid white;background:none;color:#FFF}.outline_button:hover,.primary_media_feature .floating_text_area .button:hover,.primary_media_feature .floating_text_area .outline_button:hover,.banner_header_overlay .button:hover{background-color:#5097ad;border-color:#5097ad}.section_search,.overlay_search{color:white;display:inline-block;position:relative}.section_search .search_field,.overlay_search .search_field{color:white;background-color:#282828;background-color:rgba(255,255,255,0.1);font-weight:500;font-size:16px;border:none;border-radius:4px;height:40px;padding-left:1.1em;padding-right:40px;width:155px}.section_search .search_field.placeholder,.overlay_search .search_field.placeholder{color:rgba(255,255,255,0.8);-webkit-font-smoothing:antialiased;opacity:1 !important;font-family:\"Montserrat\",Helvetica,Arial,sans-serif}.section_search .search_field:-moz-placeholder,.overlay_search .search_field:-moz-placeholder{color:rgba(255,255,255,0.8);-webkit-font-smoothing:antialiased;opacity:1 !important;font-family:\"Montserrat\",Helvetica,Arial,sans-serif}.section_search .search_field::-moz-placeholder,.overlay_search .search_field::-moz-placeholder{color:rgba(255,255,255,0.8);-webkit-font-smoothing:antialiased;opacity:1 !important;font-family:\"Montserrat\",Helvetica,Arial,sans-serif}.section_search .search_field::-webkit-input-placeholder,.overlay_search .search_field::-webkit-input-placeholder{color:rgba(255,255,255,0.8);-webkit-font-smoothing:antialiased;opacity:1 !important;font-family:\"Montserrat\",Helvetica,Arial,sans-serif}.section_search .search_field:-ms-input-placeholder,.overlay_search .search_field:-ms-input-placeholder{color:rgba(255,255,255,0.8);-webkit-font-smoothing:antialiased;opacity:1 !important;font-family:\"Montserrat\",Helvetica,Arial,sans-serif}.section_search .search_submit,.overlay_search .search_submit{padding:0;cursor:pointer;width:42px;height:42px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -127px -5px;background-size:300px;position:absolute;right:-5px;top:-3px;border:none;margin-left:-44px;opacity:.8}.section_search .search_submit:hover,.overlay_search .search_submit:hover,.section_search .search_submit.active,.overlay_search .search_submit.active,.section_search .search_submit.current,.overlay_search .search_submit.current{background-position:-127px -5px}.section_search .search_field{background-color:#F3F4F8;color:#222}.section_search .search_field.placeholder{color:rgba(255,255,255,0.8);opacity:1 !important}.section_search .search_field:-moz-placeholder{color:rgba(255,255,255,0.8);opacity:1 !important}.section_search .search_field::-moz-placeholder{color:rgba(255,255,255,0.8);opacity:1 !important}.section_search .search_field::-webkit-input-placeholder{color:rgba(255,255,255,0.8);opacity:1 !important}.section_search .search_field:-ms-input-placeholder{color:rgba(255,255,255,0.8);opacity:1 !important}.section_search .search_submit{padding:0;cursor:pointer;width:42px;height:42px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -127px -54px;background-size:300px;opacity:.6}.section_search .search_submit:hover,.section_search .search_submit.active,.section_search .search_submit.current{background-position:-127px -54px}form.nav_search .search_field{padding-right:20px;height:34px}form.nav_search input:-webkit-autofill,form.overlay_search input:-webkit-autofill{-webkit-box-shadow:0 0 0px 1000px #989898 inset;-webkit-text-fill-color:white !important}.overlay_search .search_field{color:white;background-color:rgba(255,255,255,0.3)}.overlay_search .search_field.placeholder{color:white}.overlay_search .search_field:-moz-placeholder{color:white}.overlay_search .search_field::-moz-placeholder{color:white}.overlay_search .search_field::-webkit-input-placeholder{color:white}.overlay_search .search_field:-ms-input-placeholder{color:white}.overlay_search label.search_label{display:none}.overlay_search .search_submit{padding:0;cursor:pointer;width:42px;height:42px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -131px -5px;background-size:300px}.overlay_search .search_submit:hover,.overlay_search .search_submit.active,.overlay_search .search_submit.current{background-position:-131px -5px}.body_form label{display:block;margin-bottom:.3em}.body_form input:not([type=\"submit\"]):not([type=\"reset\"]),.body_form textarea{font-size:16px}.body_form input[type=\"text\"]:not(#recaptcha_response_field),.body_form input[type=\"tel\"],.body_form input[type=\"email\"]{height:40px}.body_form input:not(#recaptcha_response_field):not(.inline_button):not([type=\"submit\"]):not([type=\"radio\"]):not([type=\"checkbox\"]),.body_form textarea{width:100%;border:1px solid #a7a8a8;background-color:white;border-radius:4px;padding:10px 12px}.body_form input,.body_form textarea{margin-bottom:1em}.body_form .button,.body_form .outline_button,.body_form .primary_media_feature .floating_text_area .button,.primary_media_feature .floating_text_area .body_form .button{margin-top:1em}.body_form select{position:relative;padding:.5em 2em .5em 1em;font-size:16px;border:0;height:40px;vertical-align:middle;color:white;-webkit-appearance:none;-o-appearance:none;-moz-appearance:none;background:#3b788b url(\"https://mars.nasa.gov/assets/[email protected]\") no-repeat 95% 10px;background-position:right .8em top 10px;background-size:9px;font-weight:700;cursor:pointer;width:100%;border-radius:5px;max-width:304px;margin-bottom:1em}.body_form select::-ms-expand{display:none}.body_form select option{padding:0.5em 1em}.body_form label{font-weight:700}.body_form .radio_title{margin-bottom:.5em;font-weight:700}.body_form fieldset.radio_buttons .option{white-space:nowrap;margin-bottom:1em}.body_form fieldset.radio_buttons label{white-space:normal;vertical-align:middle;display:inline}.body_form fieldset.radio_buttons input[type=\"radio\"],.body_form fieldset.radio_buttons input[type=\"checkbox\"]{display:inline-block;margin:0 .5em 0 0;vertical-align:middle}.body_form fieldset.radio_buttons input[type=\"radio\"]+label,.body_form fieldset.radio_buttons input[type=\"checkbox\"]+label{font-weight:400}.body_form .centered{text-align:center}@media (max-width: 480px){#recaptcha_widget_div{overflow:hidden}#recaptcha_widget_div #recaptcha_area{margin:0 auto}}.event_location,.event_date{margin-bottom:1em}.site_header_area{height:58px}@media (min-width: 600px), print{.site_header_area{height:58px}}@media (min-width: 600px), print{.site_header_area{height:58px}}@media (min-width: 769px), print{.site_header_area{height:70px}}@media (min-width: 1024px), print{.site_header_area{height:74px}}@media (min-width: 1200px){.site_header_area{height:82px}}@media (min-width: 1700px){.site_header_area{height:88px}}.site_header_area .brand_area{top:8px;margin-left:8px;height:49px;width:260px;transition:width .3s, height .3s}.site_header_area .site_logo_container{top:16px;margin-left:0;width:189px}.site_header_area .menu_button,.site_header_area #modal_close{top:8px;right:8px}@media (min-width: 769px), print{.site_header_area .brand_area{top:10px;margin-left:12px;height:60px;width:313px}.site_header_area .site_logo_container{top:23px;margin-left:13px;width:189px}.site_header_area .menu_button,.site_header_area #modal_close{top:12px;right:12px}}@media (min-width: 1024px), print{.site_header_area .brand_area{top:12px;margin-left:10px;height:54px;width:288px}.site_header_area .site_logo_container{width:120px}}@media (min-width: 1200px){.site_header_area .brand_area{top:14px;margin-left:17px;height:60px;width:338px}.site_header_area .site_logo_container{top:30px;width:210px}}@media (min-width: 1700px){.site_header_area .brand_area{top:14px;margin-left:30px}.site_header_area .site_logo_container{top:30px;width:246px}}@media (min-width: 769px), print{#home:not(.nav_is_fixed) .brand_area{height:69px;width:375px}}@media (min-width: 1024px), print{#home:not(.nav_is_fixed) .brand_area{width:300px;height:55px}}@media (min-width: 1200px){#home:not(.nav_is_fixed) .brand_area{width:368px;height:68px}}@media (min-width: 1700px){#home:not(.nav_is_fixed) .brand_area{width:420px;height:78px}}#home .site_header_area{background-color:transparent}#home.nav_overlay_true .site_header_area,#home.nav_is_fixed .site_header_area{background-color:#5a2017}.site_header_area{background-color:#5a2017;width:100%;position:absolute;z-index:21}.site_header_area.opaque{transition:background-color .5s ease-in-out}.main_feature_present .site_header_area{background-color:transparent}.main_feature_present .site_header_area.opaque{background-color:transparent}#home.nav_is_fixed .site_header_area{z-index:42}.nav_is_fixed .site_header_area{box-shadow:0 4px 4px -2px rgba(0,0,0,0.15);transition:background-color .5s ease-in-out;background-color:#5a2017}.site_header_area .site_header{width:100%;height:100%}.site_header_area .brand_area{position:relative;display:inline-block;z-index:100;background-image:url(\"https://mars.nasa.gov/assets/[email protected]\")}.site_header_area .brand_area .brand1{width:22%}.site_header_area .brand_area .brand2{width:78%}@media (min-width: 769px){.site_header_area .brand_area .brand2{display:block}}.site_header_area .site_logo_container{position:relative;display:inline-block;vertical-align:top;z-index:100}@media (min-width: 769px), print{.site_header_area .site_logo_container:before{content:\"\";height:110%;background-color:rgba(255,255,255,0.4);width:1px;position:absolute;left:-9px;top:0px}}@media (min-width: 1024px), print{.site_header_area .site_logo_container:before{height:150%;top:1px}}@media (min-width: 1200px){.site_header_area .site_logo_container:before{height:110%;top:-2px}}@media (min-width: 1700px){.site_header_area .site_logo_container:before{top:0px}}.site_header_area .site_logo_container a{display:block}.site_header_area .site_logo_container a:hover{text-decoration:none}.site_header_area .site_logo_container img.site_logo{display:block;position:relative;width:130px;top:3px}@media (min-width: 769px), print{.site_header_area .site_logo_container img.site_logo{width:170px;top:0px}}@media (min-width: 1024px), print{.site_header_area .site_logo_container img.site_logo{width:120px;top:6px}}@media (min-width: 1200px){.site_header_area .site_logo_container img.site_logo{width:188px;top:0}}@media (min-width: 1700px){.site_header_area .site_logo_container img.site_logo{width:215px}}.site_header_area .site_logo_container img.site_logo_truncated{display:none}@media (min-width: 1024px), print{.site_header_area .site_logo_container img.site_logo_truncated{display:block}}@media (min-width: 1200px){.site_header_area .site_logo_container img.site_logo_truncated{display:none}}.site_header_area img.site_logo_black{display:none}.site_header_area form.nav_search{display:inline-block;vertical-align:middle;margin-right:1em}.header_mask{display:none}@media (min-width: 1024px){.header_mask{height:58px;display:block}}@media (min-width: 1024px) and (min-width: 600px), print and (min-width: 1024px){.header_mask{height:58px}}@media (min-width: 1024px) and (min-width: 600px), print and (min-width: 1024px){.header_mask{height:58px}}@media (min-width: 1024px) and (min-width: 769px), print and (min-width: 1024px){.header_mask{height:70px}}@media (min-width: 1024px) and (min-width: 1024px), print and (min-width: 1024px){.header_mask{height:74px}}@media (min-width: 1024px) and (min-width: 1200px){.header_mask{height:82px}}@media (min-width: 1024px) and (min-width: 1700px){.header_mask{height:88px}}@media (max-width: 1023px){#sticky_nav_spacer{height:58px}}@media (max-width: 1023px) and (min-width: 480px){#sticky_nav_spacer{height:58px}}@media (max-width: 1023px) and (min-width: 600px), print and (max-width: 1023px){#sticky_nav_spacer{height:58px}}@media (max-width: 1023px) and (min-width: 769px), print and (max-width: 1023px){#sticky_nav_spacer{height:70px}}@media (max-width: 1023px){.main_feature_present #sticky_nav_spacer{display:none}}.site_header_area .menu_icon{color:transparent;font-size:0}@media (max-width: 1023px){.site_header_area{position:fixed}.fixfixed .site_header_area{position:absolute;box-shadow:none}.nav_is_fixed .site_header_area{box-shadow:0 4px 4px -2px rgba(0,0,0,0.15)}.nav_overlay_true .site_header_area{background-color:#5a2017;transition:none}.site_header_area img.grace_logo_black{display:none}.site_header_area .right_header_container{width:300px}.site_header_area .right_header_container .menu_button{position:absolute;vertical-align:middle;padding:10px;text-decoration:none;-webkit-touch-callout:none;-webkit-user-select:none;-khtml-user-select:none;-moz-user-select:-moz-none;-ms-user-select:none;user-select:none}.site_header_area .right_header_container .menu_button .menu_icon{display:block;padding:0;cursor:pointer;width:25px;height:25px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") 0 0;background-size:300px}.site_header_area .right_header_container .menu_button .menu_icon:hover,.site_header_area .right_header_container .menu_button .menu_icon.active,.site_header_area .right_header_container .menu_button .menu_icon.current{background-position:0 0}.site_header_area .right_header_container #modal_close{display:none;position:absolute;padding:10px;text-decoration:none;-webkit-touch-callout:none;-webkit-user-select:none;-khtml-user-select:none;-moz-user-select:-moz-none;-ms-user-select:none;user-select:none}.site_header_area .right_header_container #modal_close .modal_close_icon{display:block;padding:0;cursor:pointer;width:25px;height:25px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -25px 0;background-size:300px}.site_header_area .right_header_container #modal_close .modal_close_icon:hover,.site_header_area .right_header_container #modal_close .modal_close_icon.active,.site_header_area .right_header_container #modal_close .modal_close_icon.current{background-position:-25px 0}.site_header_area .right_header_container form.nav_search{display:none}.site_header_area.menu_open #modal_close{display:inline-block}}@media (min-width: 1024px){.site_header_area{display:block !important}.site_header_area form.nav_search{display:inline-block;max-width:216px}.site_header_area form.nav_search .search_field{width:37px;padding-right:0;padding-left:0;height:34px}.site_header_area form.nav_search .search_open{padding-left:.8em;padding-right:38px}.no-touchevents .nav_is_fixed .site_header_area{bottom:auto;top:0;position:fixed;width:100%;box-shadow:0 4px 4px -2px rgba(0,0,0,0.15);margin-top:0px}}#site_footer{padding:0;background:black;background-size:100%;position:relative;line-height:1.4}@media (min-width: 600px), print{#site_footer{background:#000 url(\"https://mars.nasa.gov/assets/footer_bg.png\") center no-repeat;background-size:cover}}@media (min-width: 1200px){#site_footer{background-position:center 70%}}#site_footer .gradient_line,#site_footer .related.module .gradient_line_module_top,.related.module #site_footer .gradient_line_module_top{margin-left:auto;margin-right:auto;content:\" \";width:100%;height:1px;clear:both;background:#a7abd2;background:-moz-linear-gradient(left, rgba(167,171,210,0), #a7abd2, rgba(167,171,210,0));background:-webkit-linear-gradient(left, rgba(167,171,210,0), #a7abd2, rgba(167,171,210,0));background:linear-gradient(left, rgba(167,171,210,0), #a7abd2, rgba(167,171,210,0));width:90%}@media (min-width: 769px), print{#site_footer .gradient_line,#site_footer .related.module .gradient_line_module_top,.related.module #site_footer .gradient_line_module_top{width:50%}}#site_footer .footer_line{display:none;margin-left:auto;margin-right:auto;content:\" \";width:85%;height:1px;clear:both;background-color:rgba(255,255,255,0.25)}@media (min-width: 600px), print{#site_footer .footer_line{display:block;width:65%}}.upper_footer{padding:2em 0 0em;width:100%;margin:0 auto}@media (min-width: 600px), print{.upper_footer{padding:4em 0 4em}}@media (min-width: 769px), print{.upper_footer{width:85%}}@media (min-width: 1024px), print{.upper_footer{width:65%}}.upper_footer .share,.upper_footer .footer_newsletter{text-align:center;margin-bottom:2.7em}@media (min-width: 600px), print{.upper_footer .share,.upper_footer .footer_newsletter{margin-bottom:4em;width:100%;float:left}}.upper_footer .share h2,.upper_footer .footer_newsletter h2{font-size:1.8em;font-weight:300;margin-bottom:0.6em;color:#ccdeef;letter-spacing:-.035em}.lower_footer{padding-bottom:4em}@media (min-width: 769px), print{.lower_footer{padding-bottom:9em}}.lower_footer .nav_container{margin:0 auto 1em;position:relative;left:0;width:100%}@media (min-width: 769px), print{.lower_footer .nav_container{padding-top:0.5em}}.lower_footer nav{font-size:1em;text-transform:uppercase;text-align:center;margin-left:auto;margin-right:auto;color:#98c7fc}.lower_footer nav a{padding:0 .4em;font-weight:600;color:#98c7fc;font-size:.85em;text-decoration:none;line-height:2em}@media (min-width: 769px), print{.lower_footer nav a{padding:0 .6em}}.no-touchevents .lower_footer nav a:hover{color:white}.lower_footer nav li{display:inline}.lower_footer nav li:not(:last-child):after{content:\"|\"}.lower_footer .credits{position:relative;float:none;width:auto;text-align:center}.lower_footer .credits .footer_brands_top,.lower_footer .credits .staff,.lower_footer .credits p{color:#ccdeef;font-weight:700;font-size:1em;text-align:center;line-height:1.3em}@media (min-width: 769px), print{.lower_footer .credits .footer_brands_top,.lower_footer .credits .staff,.lower_footer .credits p{font-size:1em}}.lower_footer .credits .footer_brands_top p{font-weight:400}.lower_footer .credits .footer_brands_top p:last-child{margin-bottom:.4em}.lower_footer .credits .footer_brands{color:#ccdeef;margin-bottom:1em}.lower_footer .credits .footer_brands .caltech{font-weight:300}.lower_footer .credits .staff,.lower_footer .credits .staff p{line-height:1.6em;margin:.3em 0;font-weight:400}.lower_footer .credits a{color:#ccdeef;font-weight:700}.no-touchevents .lower_footer .credits a:hover{color:white}@media (max-width: 1023px){.nav_area{display:none;position:fixed;left:0;width:104%;overflow:hidden;height:100%;min-height:100%;background-color:#5a2017;z-index:10000;top:58px}}@media (max-width: 1023px) and (min-width: 480px){.nav_area{top:58px}}@media (max-width: 1023px) and (min-width: 600px), print and (max-width: 1023px){.nav_area{top:58px}}@media (max-width: 1023px) and (min-width: 769px), print and (max-width: 1023px){.nav_area{top:70px}}#site_nav_container .global_subnav_container{display:none}@media (min-width: 1024px){#site_nav_container .global_subnav_container{display:block !important}}@media (max-width: 1023px){#site_nav_container{width:100%;text-align:center;overflow-y:scroll;padding:0 8.8% 150px 4.8%;height:100%;min-height:100%;-webkit-overflow-scrolling:touch}#site_nav_container .site_nav{display:block}#site_nav_container ul.nav{margin-bottom:2em}#site_nav_container ul.nav>li{display:block;padding:1em 0 0}#site_nav_container ul.nav>li .gradient_line,#site_nav_container ul.nav>li .related.module .gradient_line_module_top,.related.module #site_nav_container ul.nav>li .gradient_line_module_top{margin:1em 0 0 0}#site_nav_container ul.nav>li .arrow_box{padding:20px 20px;width:52px;float:right;cursor:pointer;margin:-0.4em -.8em 0 0;display:block;text-align:center}#site_nav_container ul.nav>li .arrow_box.reverse{transform:rotate(180deg);-ms-filter:\"progid:DXImageTransform.Microsoft.Matrix(M11=-1, M12=1.2246063538223773e-16, M21=-1.2246063538223773e-16, M22=-1, SizingMethod='auto expand')\"}#site_nav_container ul.nav>li .arrow_box .arrow_down{width:0;height:0;border-left:6px solid rgba(255,255,255,0);border-right:6px solid rgba(255,255,255,0);border-top:8px solid #fff;float:right}#site_nav_container .nav_title{margin-bottom:.3em;display:block;line-height:1.4em;font-weight:700;text-align:left;width:80%}#site_nav_container .nav_title a{font-size:1.2em;color:#FFF;display:block;width:100%;height:100%;padding:.4em .4em .4em 0}#site_nav_container .nav_title a:hover{text-decoration:none}#site_nav_container ul.subnav li{text-align:left}#site_nav_container ul.subnav a{color:#84B0DD;font-size:1em;line-height:1.4em;text-decoration:none;display:block;padding:.4em 0;font-weight:600}#site_nav_container ul.nav>li.admin_site_nav_item .arrow_box .arrow_up,#site_nav_container ul.nav>li.admin_site_nav_item .arrow_box .arrow_down{border-top-color:#F45F5F}.no-touchevents #site_nav_container ul.nav>li.admin_site_nav_item .arrow_box:hover .arrow_up,.no-touchevents #site_nav_container ul.nav>li.admin_site_nav_item .arrow_box:hover .arrow_down{border-top-color:white}#site_nav_container ul.nav>li.admin_site_nav_item .nav_title a,#site_nav_container ul.nav>li.admin_site_nav_item ul.subnav a{color:#F45F5F}.no-touchevents #site_nav_container ul.nav>li.admin_site_nav_item .nav_title a:hover,.no-touchevents #site_nav_container ul.nav>li.admin_site_nav_item ul.subnav a:hover{color:white}#site_nav_container .overlay_search{margin-bottom:2em;width:100%;max-width:320px}#site_nav_container .overlay_search .search_field{width:100%}#site_nav_container .social_nav{color:white;display:block;background-color:#394862;font-size:1.3em;width:100%;border-radius:4px;padding:1.3em 0 1.7em;max-width:320px;margin:0 auto}#site_nav_container .social_nav .nav_title{margin-bottom:1em;text-align:center;width:auto}}@media (min-width: 1024px){.nav_area{width:100%;height:100%;float:right;bottom:0;right:0;position:absolute;text-align:right;z-index:50;display:block !important}}.no-touchevents .nav_is_fixed .fancybox-wrap #sticky_nav_spacer{display:none}.no-touchevents .nav_is_fixed .fancybox-wrap .nav_area{position:relative}@media (min-width: 1024px){#site_nav_container{padding:0;position:relative;display:inline-block !important;width:100%;height:100%;position:relative;bottom:0}}@media (min-width: 1024px) and (min-width: 1024px), print and (min-width: 1024px){#site_nav_container{bottom:0}}@media (min-width: 1024px) and (min-width: 1200px){#site_nav_container{bottom:0}}@media (min-width: 1024px){#site_nav_container .site_nav{width:100%;padding:0;position:relative;overflow-y:visible;min-height:0;height:100%;top:auto;left:auto;padding-top:22px}}@media (min-width: 1024px) and (min-width: 1200px){#site_nav_container .site_nav{padding-top:27px}}@media (min-width: 1024px) and (min-width: 1700px){#site_nav_container .site_nav{padding-right:0.8em}}@media (min-width: 1024px){#site_nav_container ul.nav{margin-bottom:0;display:inline-block;margin-right:.6em}#site_nav_container ul.nav>li{display:block}}@media (min-width: 1024px) and (min-width: 1024px){#site_nav_container ul.nav>li{display:inline-block;cursor:pointer;border-radius:2px;position:relative}#site_nav_container ul.nav>li:hover{background-color:#9a4739;border-bottom-left-radius:0;border-bottom-right-radius:0}.main_feature_present #site_nav_container ul.nav>li:hover{background-color:rgba(0,0,0,0.5)}.main_feature_present.nav_is_fixed #site_nav_container ul.nav>li:hover{background-color:#9a4739}}@media (min-width: 1024px) and (min-width: 1024px){#site_nav_container ul.nav>li .global_subnav_container{z-index:20;position:relative}}@media (min-width: 1024px){#site_nav_container ul.nav>li:hover .subnav{display:block}#site_nav_container ul.nav>li .gradient_line,#site_nav_container ul.nav>li .related.module .gradient_line_module_top,.related.module #site_nav_container ul.nav>li .gradient_line_module_top{width:60%;margin-top:1.5em;margin-bottom:1.5em}}@media (min-width: 1024px) and (min-width: 1024px){#site_nav_container ul.nav>li .gradient_line,#site_nav_container ul.nav>li .related.module .gradient_line_module_top,.related.module #site_nav_container ul.nav>li .gradient_line_module_top{display:none}}@media (min-width: 1024px){#site_nav_container ul.nav>li:last-child{margin-left:14px}#site_nav_container ul.nav>li:last-child:before{content:\"\";border:1px solid rgba(124,113,110,0.6);position:absolute;height:20px;left:-8px;top:9px}#site_nav_container ul.nav>li:last-child .global_subnav_container>ul.subnav{border-top-left-radius:2px;right:-52px}#site_nav_container .nav_title{margin-bottom:0;display:block;line-height:1.4em;color:white}#site_nav_container .nav_title a,#site_nav_container .nav_title .main_nav_item{display:block;font-size:.88rem;font-weight:600;padding:.5em 6px;color:white}}@media (min-width: 1024px) and (min-width: 1200px){#site_nav_container .nav_title a,#site_nav_container .nav_title .main_nav_item{padding:.5em 0.8em;font-size:.9rem}}@media (min-width: 1024px) and (min-width: 1700px){#site_nav_container .nav_title a,#site_nav_container .nav_title .main_nav_item{padding:.5em 1em}}@media (min-width: 1024px){#site_nav_container .nav_title a:hover,#site_nav_container .nav_title .main_nav_item:hover{text-decoration:none}}@media (min-width: 1024px) and (min-width: 1024px){#site_nav_container ul.subnav{padding:0.4em 0;margin-bottom:0;min-width:190px;display:none;position:absolute;margin:0;border-bottom-left-radius:2px;border-bottom-right-radius:2px;border-top-right-radius:2px;background-color:#9a4739}.main_feature_present #site_nav_container ul.subnav{background-color:rgba(0,0,0,0.5)}.main_feature_present.nav_is_fixed #site_nav_container ul.subnav{background-color:#9a4739}}@media (min-width: 1024px){#site_nav_container ul.subnav li{text-align:center}}@media (min-width: 1024px) and (min-width: 600px), print and (min-width: 1024px){#site_nav_container ul.subnav li{display:inline-block}}@media (min-width: 1024px) and (min-width: 1024px){#site_nav_container ul.subnav li{text-align:left;clear:both;display:block}#site_nav_container ul.subnav li:hover{background-color:rgba(0,0,0,0.3)}}@media (min-width: 1024px){#site_nav_container ul.subnav a{color:#84B0DD;font-size:1em;line-height:1.4em;text-decoration:none;display:block;padding:.4em 0;font-weight:600;white-space:nowrap}}@media (min-width: 1024px) and (min-width: 600px), print and (min-width: 1024px){#site_nav_container ul.subnav a{padding:.4em 1em}}@media (min-width: 1024px) and (min-width: 1024px){#site_nav_container ul.subnav a{white-space:normal;font-size:.85em;color:white;padding:.4em 1.1em}}@media (min-width: 1024px){.no-touchevents #site_nav_container ul.subnav a:hover{color:white}#site_nav_container .social_nav{display:none}#site_nav_container li.admin_site_nav_item{background-color:#D94F34}#site_nav_container li.admin_site_nav_item .nav_title a{color:white}#site_nav_container li.admin_site_nav_item:hover .nav_title,#site_nav_container li.admin_site_nav_item.current .nav_title{background-color:#FF7054 !important}#site_nav_container li.admin_site_nav_item:hover .subnav{display:block !important}#site_nav_container li.admin_site_nav_item ul.subnav{border:none;background-color:#D94F34}#site_nav_container li.admin_site_nav_item ul.subnav a{color:white}#site_nav_container li.admin_site_nav_item ul.subnav li{background-color:#D94F34;border:none}#site_nav_container li.admin_site_nav_item ul.subnav li:hover{background-color:#FF7054}}#site_nav_container .nav_title,#site_nav_container ul.subnav a{font-weight:600}#site_footer .sitemap{font-weight:400;z-index:10;position:relative;margin-bottom:2em}@media (min-width: large){#site_footer .sitemap .grid_layout{width:97%}}#site_footer .sitemap_directory{margin-bottom:2em}#site_footer .sitemap_directory .footer_sitemap_item{margin-bottom:1.8em}@media (min-width: 600px), print{#site_footer .sitemap_directory .footer_sitemap_item{margin-bottom:2em}}@media (min-width: 1024px), print{#site_footer .sitemap_directory .footer_sitemap_item{margin-left:10%}}#site_footer .sitemap_title{font-weight:400;text-transform:capitalize;font-size:1em;margin-bottom:.4em}#site_footer .sitemap_title a,#site_footer .sitemap_title .no_link_nav_item{color:white;text-decoration:none}@media (min-width: 600px), print{#site_footer .sitemap_title{font-size:1.1em;margin-bottom:.4em}}@media (min-width: 1024px), print{#site_footer .sitemap_title{font-size:1.1em}}#site_footer .sitemap_block{text-align:center;width:100%}@media (min-width: 600px), print{#site_footer .sitemap_block{-webkit-box-sizing:border-box;-moz-box-sizing:border-box;box-sizing:border-box;width:25%;float:left;padding-left:1.66667%;padding-right:1.66667%;text-align:left}}@media (min-width: 1024px), print{#site_footer .sitemap_block{-webkit-box-sizing:border-box;-moz-box-sizing:border-box;box-sizing:border-box;width:16.66667%;float:left;padding-left:1.66667%;padding-right:1.66667%}}#site_footer ul.subnav{margin-bottom:1em}#site_footer ul.subnav li{padding-left:1em;text-indent:-1em;margin:0 0 .25em 0}#site_footer ul.subnav a{color:#98c7fc;text-decoration:none;font-size:1em}@media (min-width: 600px), print{#site_footer ul.subnav a{font-size:.85em}}@media (min-width: 1024px), print{#site_footer ul.subnav a{font-size:.95em}}.no-touchevents #site_footer ul.subnav a:hover{color:white}@media (min-width: 600px), print{.main_area_sitemap .grid_layout{width:100%}}.main_area_sitemap .sitemap_directory{padding:2em 0 0}.main_area_sitemap .sitemap_directory .footer_sitemap_item{margin-bottom:1.8em}@media (min-width: 600px), print{.main_area_sitemap .sitemap_directory .footer_sitemap_item{margin-bottom:2em}}.main_area_sitemap .sitemap_block{text-align:center;width:100%}@media (min-width: 600px), print{.main_area_sitemap .sitemap_block{-webkit-box-sizing:border-box;-moz-box-sizing:border-box;box-sizing:border-box;width:25%;float:left;padding-left:1.66667%;padding-right:1.66667%;text-align:left}}@media (min-width: 1024px), print{.main_area_sitemap .sitemap_block{-webkit-box-sizing:border-box;-moz-box-sizing:border-box;box-sizing:border-box;width:16.66667%;float:left;padding-left:1.66667%;padding-right:1.66667%}}.main_area_sitemap .sitemap_block a{word-wrap:normal}.main_area_sitemap .sitemap_title{margin-top:0}.main_area_sitemap .sitemap_title a,.main_area_sitemap .sitemap_title .no_link_nav_item{color:#222}.main_area_sitemap .subnav a{display:block}@media (min-width: 600px), print{.main_area_sitemap .subnav a{padding-left:1em;text-indent:-1em;margin:.1em 0}}.social_icons{display:block}.social_icons .icon{width:44px !important;height:44px !important;display:inline-block;overflow:hidden}.social_icons .icon+.icon{margin-left:.7em}@media (min-width: 769px), print{.social_icons .icon+.icon{margin-left:.9em}}.social_icons .icon img{opacity:1 !important;height:100%;max-width:none}.triple_teaser .social_icons{max-width:188px;white-space:nowrap}@media (min-width: 769px), print{.triple_teaser .social_icons{max-width:none}}.triple_teaser .social_icons .icon{width:44px;height:44px}.triple_teaser .social_icons .icon+.icon{margin-left:.7em}@media (min-width: 600px), print{.triple_teaser .social_icons .icon{width:38px;height:38px}.triple_teaser .social_icons .icon+.icon{margin-left:.4em;margin-left:calc((100% - 152px)/3)}}@media (min-width: 769px), print{.triple_teaser .social_icons .icon{width:44px;height:44px}.triple_teaser .social_icons .icon+.icon{margin-left:.8em}}.addthis_default_style .at300b,.addthis_default_style .at300bo,.addthis_default_style .at300m{padding:0 !important;float:none !important}#_atssh{display:none}#at4-share,#at4-soc{top:60%;bottom:auto}html,html a,select,input,button{-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale}html.no-touchevents{text-rendering:optimizeLegibility}html.no-touchevents html a,html.no-touchevents select,html.no-touchevents input,html.no-touchevents button{text-rendering:optimizeLegibility}input.placeholder,textarea.placeholder{-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale;text-rendering:optimizeLegibility}input:-moz-placeholder,textarea:-moz-placeholder{-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale;text-rendering:optimizeLegibility}input::-moz-placeholder,textarea::-moz-placeholder{-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale;text-rendering:optimizeLegibility}input::-webkit-input-placeholder,textarea::-webkit-input-placeholder{-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale;text-rendering:optimizeLegibility}input:-ms-input-placeholder,textarea:-ms-input-placeholder{-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale;text-rendering:optimizeLegibility}html,button,input,select,textarea{font-family:\"Montserrat\",Helvetica,Arial,sans-serif;color:#222}html{min-height:100%}body{font-family:\"Montserrat\",Helvetica,Arial,sans-serif;font-weight:300;font-size:96%;line-height:1.4;min-height:100%;position:relative;background-color:transparent}body.noscroll{overflow-y:hidden}@media (min-width: 600px), print{body{font-size:98%}}@media (min-width: 769px), print{body{font-size:100%}}@media (min-width: 1024px), print{body{font-size:102%}}@media (min-width: 1200px){body{font-size:104%}}h1,h2,h3,h4,h5{line-height:1.2em}h1{letter-spacing:-.03em}h2{letter-spacing:-.03em}h3{letter-spacing:-.02em}h4{letter-spacing:-.02em}h1,h2,h3,h4,h5{margin:0}img{width:100%}img,embed,object,video{max-width:100%}.jwplayer video{max-width:none}i{font-style:italic}strong{font-weight:700}p{margin:1em 0;font-size:100%}.gradient_line,.related.module .gradient_line_module_top{margin-left:auto;margin-right:auto;content:\" \";width:100%;height:1px;clear:both;background:#b76b5f;background:-moz-linear-gradient(left, rgba(183,107,95,0), #b76b5f, rgba(183,107,95,0));background:-webkit-linear-gradient(left, rgba(183,107,95,0), #b76b5f, rgba(183,107,95,0));background:linear-gradient(left, rgba(183,107,95,0), #b76b5f, rgba(183,107,95,0))}.gradient_line_extra_margin{margin-left:auto;margin-right:auto;content:\" \";width:100%;height:1px;clear:both;background:#BEBEBE;background:-moz-linear-gradient(left, rgba(190,190,190,0), #bebebe, rgba(190,190,190,0));background:-webkit-linear-gradient(left, rgba(190,190,190,0), #bebebe, rgba(190,190,190,0));background:linear-gradient(left, rgba(190,190,190,0), #bebebe, rgba(190,190,190,0));margin:2em 0}@media (min-width: 769px), print{.gradient_line_extra_margin{margin:3em 0}}.module_title,.main_carousel.module .carousel_header .carousel_title,.media_feature_title,.sitemap_title,.nav_title,.article_title,.sidebar_title,#secondary_column .related_content_module .module_title,#secondary_column .related_content_module .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #secondary_column .related_content_module .carousel_title,.right_col .related_content_module .module_title,.right_col .related_content_module .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .right_col .related_content_module .carousel_title,.rollover_title{letter-spacing:-.02em}.module_title,.main_carousel.module .carousel_header .carousel_title{letter-spacing:-.02em}.rollover_title{font-size:2.34em;margin-bottom:0em}@media (min-width: 600px), print{.rollover_title{font-size:2.7em;margin-bottom:0em}}@media (min-width: 769px), print{.rollover_title{font-size:3.06em;margin-bottom:0em}}@media (min-width: 1024px), print{.rollover_title{font-size:3.24em;margin-bottom:0em}}@media (min-width: 1200px){.rollover_title{font-size:3.42em;margin-bottom:0em}}.content_title{letter-spacing:0;font-weight:600}.module_title,.main_carousel.module .carousel_header .carousel_title{font-size:1.69em;margin-bottom:.35em;text-align:center;font-weight:600}@media (min-width: 600px), print{.module_title,.main_carousel.module .carousel_header .carousel_title{font-size:1.95em;margin-bottom:.63em}}@media (min-width: 769px), print{.module_title,.main_carousel.module .carousel_header .carousel_title{font-size:2.21em;margin-bottom:.91em}}@media (min-width: 1024px), print{.module_title,.main_carousel.module .carousel_header .carousel_title{font-size:2.34em;margin-bottom:1.015em}}@media (min-width: 1200px){.module_title,.main_carousel.module .carousel_header .carousel_title{font-size:2.47em;margin-bottom:1.12em}}@media (min-width: 600px), print{.grid_gallery .module_title,.grid_gallery .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .grid_gallery .carousel_title{text-align:left;width:80%}}.module_title_small,.double_teaser .module_title,.double_teaser .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .double_teaser .carousel_title{font-size:1.4em}@media (min-width: 600px), print{.module_title_small,.double_teaser .module_title,.double_teaser .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .double_teaser .carousel_title{font-size:1.8em;margin-bottom:.85em}}.filter_bar .module_title_small,.filter_bar .double_teaser .module_title,.filter_bar .double_teaser .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .filter_bar .double_teaser .carousel_title{text-align:left;width:90%}@media (min-width: 600px), print{.filter_bar .module_title_small,.filter_bar .double_teaser .module_title,.filter_bar .double_teaser .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .filter_bar .double_teaser .carousel_title{text-align:center}}.category_title{font-size:.9em;font-weight:500;color:#f08d77;text-transform:uppercase;margin-bottom:6px}.multimedia_teaser .category_title{font-size:.8em}.primary_media_feature .media_feature_title{font-size:1.43em;margin-bottom:0em;font-weight:400;color:white}@media (min-width: 600px), print{.primary_media_feature .media_feature_title{font-size:1.65em;margin-bottom:0em}}@media (min-width: 769px), print{.primary_media_feature .media_feature_title{font-size:1.87em;margin-bottom:0em}}@media (min-width: 1024px), print{.primary_media_feature .media_feature_title{font-size:1.98em;margin-bottom:0em}}@media (min-width: 1200px){.primary_media_feature .media_feature_title{font-size:2.09em;margin-bottom:0em}}.image_of_the_day .media_feature_title{font-size:1.43em;margin-bottom:0em;font-weight:600;color:white}@media (min-width: 600px), print{.image_of_the_day .media_feature_title{font-size:1.65em;margin-bottom:0em}}@media (min-width: 769px), print{.image_of_the_day .media_feature_title{font-size:1.87em;margin-bottom:0em}}@media (min-width: 1024px), print{.image_of_the_day .media_feature_title{font-size:1.98em;margin-bottom:0em}}@media (min-width: 1200px){.image_of_the_day .media_feature_title{font-size:2.09em;margin-bottom:0em}}.multimedia_module_gallery .media_feature_title{font-size:1.43em;margin-bottom:0em;color:white;font-weight:600}@media (min-width: 600px), print{.multimedia_module_gallery .media_feature_title{font-size:1.65em;margin-bottom:0em}}@media (min-width: 769px), print{.multimedia_module_gallery .media_feature_title{font-size:1.87em;margin-bottom:0em}}@media (min-width: 1024px), print{.multimedia_module_gallery .media_feature_title{font-size:1.98em;margin-bottom:0em}}@media (min-width: 1200px){.multimedia_module_gallery .media_feature_title{font-size:2.09em;margin-bottom:0em}}.article_title{font-size:1.82em;margin-bottom:0em;font-weight:600}@media (min-width: 600px), print{.article_title{font-size:2.1em;margin-bottom:0em}}@media (min-width: 769px), print{.article_title{font-size:2.38em;margin-bottom:0em}}@media (min-width: 1024px), print{.article_title{font-size:2.52em;margin-bottom:0em}}@media (min-width: 1200px){.article_title{font-size:2.66em;margin-bottom:0em}}.magic_shell_title,#iframe_overlay .magic_shell_title{font-family:WhitneyCondensed-Bold,Helvetica,Arial,sans-serif;background-color:#000;font-size:2.6em;font-weight:normal;padding:.9em .5em;text-align:center;line-height:.8}@media (min-width: 600px), print{.magic_shell_title,#iframe_overlay .magic_shell_title{padding:.9em .5em;text-align:left;font-size:2.6em}}.magic_shell_title .parent_title,#iframe_overlay .magic_shell_title .parent_title{display:block}.magic_shell_title .parent_title a,#iframe_overlay .magic_shell_title .parent_title a{color:#f08d77;transition:color 400ms}.magic_shell_title .parent_title a:hover,#iframe_overlay .magic_shell_title .parent_title a:hover{text-decoration:none;color:white}@media (min-width: 600px), print{.magic_shell_title .parent_title,#iframe_overlay .magic_shell_title .parent_title{display:inline;margin-right:0.1em}}.magic_shell_title .article_title,#iframe_overlay .magic_shell_title .article_title{color:#FFF;font-weight:normal}.magic_shell_title .article_title,#iframe_overlay .magic_shell_title .article_title,.magic_shell_title .parent_title,#iframe_overlay .magic_shell_title .parent_title{font-size:.7em;text-transform:uppercase;letter-spacing:normal}.sidebar_title,#secondary_column .related_content_module .module_title,#secondary_column .related_content_module .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #secondary_column .related_content_module .carousel_title,.right_col .related_content_module .module_title,.right_col .related_content_module .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .right_col .related_content_module .carousel_title{font-size:1.55em;margin-bottom:0.6em;font-weight:700;margin-left:-1px}.links_module a{font-size:1em;cursor:pointer}.module{padding:2.5em 0 2.2em;position:relative}@media (min-width: 769px), print{.module{padding:4.8em 0 5em}}.grid_layout{max-width:100%;margin-left:auto;margin-right:auto;width:95%}.grid_layout:after{content:\" \";display:block;clear:both}@media (min-width: 600px), print{.grid_layout{max-width:100%;margin-left:auto;margin-right:auto;width:95%}.grid_layout:after{content:\" \";display:block;clear:both}}@media (min-width: 769px), print{.grid_layout{max-width:100%;margin-left:auto;margin-right:auto;width:90%}.grid_layout:after{content:\" \";display:block;clear:both}}@media (min-width: 1024px), print{.grid_layout{max-width:1200px;width:97%}.content_page .grid_layout{width:90%}}@media (max-width: 480px){.suggested_features .grid_layout,.news_teaser .grid_layout,.carousel_teaser .grid_layout{width:100%}.suggested_features .grid_layout header,.news_teaser .grid_layout header,.carousel_teaser .grid_layout header{margin-left:auto;margin-right:auto;width:95%}.suggested_features .grid_layout footer,.news_teaser .grid_layout footer,.carousel_teaser .grid_layout footer{margin-left:auto;margin-right:auto;width:95%}}.gradient_container_top,.gradient_container_bottom,.white_gradient_container_bottom{height:200px;width:100%;position:absolute;z-index:1}.homepage_carousel .gradient_container_top,.homepage_carousel .gradient_container_bottom,.homepage_carousel .white_gradient_container_bottom{z-index:7}.gradient_container_left{height:100%;width:70%;position:absolute;z-index:1}.gradient_container_top{background:-owg-linear-gradient(rgba(0,0,0,0.6), transparent);background:-webkit-linear-gradient(rgba(0,0,0,0.6), transparent);background:-moz-linear-gradient(rgba(0,0,0,0.6), transparent);background:-o-linear-gradient(rgba(0,0,0,0.6), transparent);background:linear-gradient(rgba(0,0,0,0.6), transparent);pointer-events:none;top:0}.gradient_container_left{background:-owg-linear-gradient(to right, rgba(0,0,0,0.6), transparent);background:-webkit-linear-gradient(to right, rgba(0,0,0,0.6), transparent);background:-moz-linear-gradient(to right, rgba(0,0,0,0.6), transparent);background:-o-linear-gradient(to right, rgba(0,0,0,0.6), transparent);background:linear-gradient(to right, rgba(0,0,0,0.6), transparent);left:0;top:0}.gradient_container_bottom{background:-owg-linear-gradient(transparent, rgba(0,0,0,0.7));background:-webkit-linear-gradient(transparent, rgba(0,0,0,0.7));background:-moz-linear-gradient(transparent, rgba(0,0,0,0.7));background:-o-linear-gradient(transparent, rgba(0,0,0,0.7));background:linear-gradient(transparent, rgba(0,0,0,0.7));pointer-events:none;bottom:0}.white_gradient_container_bottom{background:url(\"https://mars.nasa.gov/assets/white_gradient.png\") repeat-x bottom left;bottom:0;height:100px;pointer-events:none}.gradient_bottom_grid{background-image:linear-gradient(to bottom, transparent 0%, #000 30%, #000 100%)}.grid_gallery .gallery_header{margin-bottom:2em}@media (min-width: 769px), print{.grid_gallery .gallery_header{margin-bottom:3em}}.grid_gallery .gallery_header .module_title,.grid_gallery .gallery_header .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .grid_gallery .gallery_header .carousel_title{margin-bottom:0.5em;text-align:left}.grid_gallery .list_date{font-size:.9em;margin-bottom:.4em;color:#5A5A5A}.grid_gallery.grid_view{background:white}.grid_gallery.grid_view .content_title{letter-spacing:-.03em;display:none}.grid_gallery.grid_view .image_and_description_container{min-height:0}.grid_gallery.grid_view .article_teaser_body{display:none}.grid_gallery.grid_view .list_date{display:none}.grid_gallery.grid_view .list_image{width:100%;float:none;margin:0}.grid_gallery.grid_view .bottom_gradient{color:#222;display:block;position:relative;margin-top:0.3rem;padding-bottom:0.4rem;text-align:left;min-height:52px}@media (min-width: 769px), print{.grid_gallery.grid_view .bottom_gradient{margin-top:.5rem;min-height:85px}}.grid_gallery.grid_view .bottom_gradient div{text-align:left}.grid_gallery.grid_view .bottom_gradient h3{font-weight:500;font-size:1em}.grid_gallery.grid_view li.slide{margin-bottom:.84034%;width:49.57983%;float:left}.grid_gallery.grid_view li.slide:nth-child(2n+1){margin-left:0;margin-right:-100%;clear:both;margin-left:0}.grid_gallery.grid_view li.slide:nth-child(2n+2){margin-left:50.42017%;margin-right:-100%;clear:none}@media (min-width: 600px), print{.grid_gallery.grid_view li.slide{margin-bottom:.84034%;width:32.77311%;float:left}.grid_gallery.grid_view li.slide:nth-child(3n+1){margin-left:0;margin-right:-100%;clear:both;margin-left:0}.grid_gallery.grid_view li.slide:nth-child(3n+2){margin-left:33.61345%;margin-right:-100%;clear:none}.grid_gallery.grid_view li.slide:nth-child(3n+3){margin-left:67.22689%;margin-right:-100%;clear:none}}@media (min-width: 769px), print{.grid_gallery.grid_view li.slide{width:24.36975%;float:left}.grid_gallery.grid_view li.slide:nth-child(4n+1){margin-left:0;margin-right:-100%;clear:both;margin-left:0}.grid_gallery.grid_view li.slide:nth-child(4n+2){margin-left:25.21008%;margin-right:-100%;clear:none}.grid_gallery.grid_view li.slide:nth-child(4n+3){margin-left:50.42017%;margin-right:-100%;clear:none}.grid_gallery.grid_view li.slide:nth-child(4n+4){margin-left:75.63025%;margin-right:-100%;clear:none}}@media (min-width: 1200px){.grid_gallery.grid_view li.slide{width:19.32773%;float:left}.grid_gallery.grid_view li.slide:nth-child(5n+1){margin-left:0;margin-right:-100%;clear:both;margin-left:0}.grid_gallery.grid_view li.slide:nth-child(5n+2){margin-left:20.16807%;margin-right:-100%;clear:none}.grid_gallery.grid_view li.slide:nth-child(5n+3){margin-left:40.33613%;margin-right:-100%;clear:none}.grid_gallery.grid_view li.slide:nth-child(5n+4){margin-left:60.5042%;margin-right:-100%;clear:none}.grid_gallery.grid_view li.slide:nth-child(5n+5){margin-left:80.67227%;margin-right:-100%;clear:none}}.grid_gallery.grid_view li.slide a{text-decoration:none}.grid_gallery.list_view .list_image{float:right;margin-left:4%;margin-bottom:.5em;width:32%}@media (min-width: 600px), print{.grid_gallery.list_view .list_image{margin-left:0;margin-bottom:0;width:23.07692%;float:left;margin-right:2.5641%}}@media (min-width: 769px), print{.grid_gallery.list_view .list_image{width:23.72881%;float:left;margin-right:1.69492%}}@media (min-width: 1024px), print{.grid_gallery.list_view .list_image{width:23.72881%;float:left;margin-right:1.69492%}}.grid_gallery.list_view .list_text{width:auto}@media (min-width: 600px), print{.grid_gallery.list_view .list_text{width:74.35897%;float:right;margin-right:0}}@media (min-width: 769px), print{.grid_gallery.list_view .list_text{width:74.57627%;float:right;margin-right:0}}@media (min-width: 1024px), print{.grid_gallery.list_view .list_text{width:66.10169%;float:left;margin-right:1.69492%}}.grid_gallery.list_view .content_title a{text-decoration:none;cursor:pointer;color:#222}.grid_gallery.list_view .content_title a:hover{text-decoration:underline}.grid_gallery.list_view .content_title{display:block;font-size:1.17em;margin-bottom:.1em;margin-bottom:.2em;font-weight:700;color:#222;letter-spacing:-.035em}@media (min-width: 600px), print{.grid_gallery.list_view .content_title{font-size:1.35em;margin-bottom:.18em}}@media (min-width: 769px), print{.grid_gallery.list_view .content_title{font-size:1.53em;margin-bottom:.26em}}@media (min-width: 1024px), print{.grid_gallery.list_view .content_title{font-size:1.62em;margin-bottom:.29em}}@media (min-width: 1200px){.grid_gallery.list_view .content_title{font-size:1.71em;margin-bottom:.32em}}.grid_gallery.list_view .bottom_gradient{display:none}@media (min-width: 1024px), print{.grid_gallery.list_view .article_teaser_body{font-size:1.1em}}.grid_gallery.list_view li.slide:first-child{border-top:1px solid #CCC}.grid_gallery.list_view li.slide{border-bottom:1px solid #CCC;padding:1.2em 0}.grid_gallery.list_view li.slide a{text-decoration:none;cursor:pointer}.view_selectors{position:relative;margin:0 auto;text-align:center;width:106px;text-align:right}@media (min-width: 769px), print{.view_selectors{position:absolute;right:0;top:0;height:100%}}.view_selectors .nav_item{display:inline-block;position:relative;background-repeat:no-repeat;width:50px;height:50px;cursor:pointer;background-image:url(\"https://mars.nasa.gov/assets/[email protected]\");background-size:125px;background-color:#eef2f6;border-radius:50%;-webkit-touch-callout:none;-webkit-user-select:none;-khtml-user-select:none;-moz-user-select:-moz-none;-ms-user-select:none;user-select:none}.view_selectors .nav_item.list_icon{background-position:-12px -62px}.no-touchevents .view_selectors .nav_item.list_icon:hover{background-position:-12px -12px}.list_view .view_selectors .nav_item.list_icon{background-position:-12px -12px}.view_selectors .nav_item.grid_icon{background-position:-62px -62px}.no-touchevents .view_selectors .nav_item.grid_icon:hover{background-position:-62px -12px}.grid_view .view_selectors .nav_item.grid_icon{background-position:-62px -12px}.grid_gallery#more_section .module_title,.grid_gallery#more_section .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .grid_gallery#more_section .carousel_title{text-align:center;width:100%}.grid_gallery#more_section li.slide{margin-bottom:1.69492%;width:49.15254%;float:left}.grid_gallery#more_section li.slide:nth-child(2n+1){margin-left:0;margin-right:-100%;clear:both;margin-left:0}.grid_gallery#more_section li.slide:nth-child(2n+2){margin-left:50.84746%;margin-right:-100%;clear:none}@media (min-width: 1200px){.grid_gallery#more_section li.slide{width:32.20339%;float:left}.grid_gallery#more_section li.slide:nth-child(3n+1){margin-left:0;margin-right:-100%;clear:both;margin-left:0}.grid_gallery#more_section li.slide:nth-child(3n+2){margin-left:33.89831%;margin-right:-100%;clear:none}.grid_gallery#more_section li.slide:nth-child(3n+3){margin-left:67.79661%;margin-right:-100%;clear:none}}.grid_gallery#more_section li.slide .image_and_description_container{position:relative}.grid_gallery#more_section li.slide a.slide_title{padding-top:.6em;display:block;color:#222;font-weight:400}.grid_gallery#more_section li.slide:hover a.slide_title{color:#366599}.wysiwyg_content ul,ol{margin-left:1.6em}.feature_pages .wysiwyg_content ol,.feature_pages .wysiwyg_content ul{list-style-position:inside}#secondary_column ul,ol{margin-left:1.2em}.wysiwyg_content ul,#secondary_column ul{margin-left:1.6em;list-style-type:disc;list-style-position:outside}.wysiwyg_content ul ul,#secondary_column ul ul{list-style-type:circle;margin-left:1.6em;margin-top:.5em}.wysiwyg_content ul ul ul,#secondary_column ul ul ul{list-style-type:square;margin-left:1.6em;margin-top:.5em}.wysiwyg_content ul ul ul ul,#secondary_column ul ul ul ul{list-style-type:disc;margin-left:1.6em;margin-top:.5em}.wysiwyg_content ol,#secondary_column ol{margin-left:1.6em;list-style-type:decimal;list-style-position:outside}.wysiwyg_content ol ol,#secondary_column ol ol{list-style-type:decimal;margin-left:1.6em;margin-top:.5em}.wysiwyg_content ol ol ol,#secondary_column ol ol ol{list-style-type:decimal;margin-left:1.6em;margin-top:.5em}.wysiwyg_content ol ol ol ol,#secondary_column ol ol ol ol{list-style-type:decimal;margin-left:1.6em;margin-top:.5em}.wysiwyg_content ol,.wysiwyg_content ul,#secondary_column ol,#secondary_column ul{margin-bottom:2em}.wysiwyg_content ol:last-child,.wysiwyg_content ul:last-child,#secondary_column ol:last-child,#secondary_column ul:last-child{margin-bottom:0}.wysiwyg_content ol li,.wysiwyg_content ul li,#secondary_column ol li,#secondary_column ul li{margin-bottom:.5em}.wysiwyg_content .item_list_module,.wysiwyg_content .item_list,.wysiwyg_content .list_sublist,.wysiwyg_content .footnotes ul,.wysiwyg_content .sidebar_gallery,.wysiwyg_content .related_items,.wysiwyg_content .sitemap_directory ul,.wysiwyg_content .list_view_module ul,.wysiwyg_content .faq_topics ul,.wysiwyg_content .item_grid,.wysiwyg_content .related_content_module ul,.wysiwyg_content .sig_events_module ul,.wysiwyg_content ul.detailed_def_nav,#secondary_column .item_list_module,#secondary_column .item_list,#secondary_column .list_sublist,#secondary_column .footnotes ul,#secondary_column .sidebar_gallery,#secondary_column .related_items,#secondary_column .sitemap_directory ul,#secondary_column .list_view_module ul,#secondary_column .faq_topics ul,#secondary_column .item_grid,#secondary_column .related_content_module ul,#secondary_column .sig_events_module ul,#secondary_column ul.detailed_def_nav{margin-left:0;list-style-type:none;list-style-position:inside}.wysiwyg_content .item_list_module li,.wysiwyg_content .item_list li,.wysiwyg_content .list_sublist li,.wysiwyg_content .footnotes ul li,.wysiwyg_content .sidebar_gallery li,.wysiwyg_content .related_items li,.wysiwyg_content .sitemap_directory ul li,.wysiwyg_content .list_view_module ul li,.wysiwyg_content .faq_topics ul li,.wysiwyg_content .item_grid li,.wysiwyg_content .related_content_module ul li,.wysiwyg_content .sig_events_module ul li,.wysiwyg_content ul.detailed_def_nav li,#secondary_column .item_list_module li,#secondary_column .item_list li,#secondary_column .list_sublist li,#secondary_column .footnotes ul li,#secondary_column .sidebar_gallery li,#secondary_column .related_items li,#secondary_column .sitemap_directory ul li,#secondary_column .list_view_module ul li,#secondary_column .faq_topics ul li,#secondary_column .item_grid li,#secondary_column .related_content_module ul li,#secondary_column .sig_events_module ul li,#secondary_column ul.detailed_def_nav li{margin-bottom:0}.module header{margin-bottom:1em;position:relative}.module footer{text-align:center;position:relative}.module footer a.detail_link{text-transform:uppercase;font-size:.9em;font-weight:400}.module .module_title,.main_carousel.module .carousel_header .carousel_title{font-weight:300;color:#6d3007}.multimedia_teaser{-webkit-touch-callout:none;-webkit-user-select:none;-khtml-user-select:none;-moz-user-select:-moz-none;-ms-user-select:none;user-select:none;overflow:hidden}#secondary_column aside .multimedia_teaser{position:relative}#secondary_column aside .multimedia_teaser .text{position:absolute;width:100%;text-align:center;padding:0 1.4em 2em;bottom:0}#secondary_column aside .multimedia_teaser .text .category_title,#secondary_column aside .multimedia_teaser .text .media_feature_title{color:white}.multimedia_teaser{-webkit-touch-callout:none;-webkit-user-select:none;-khtml-user-select:none;-moz-user-select:-moz-none;-ms-user-select:none;user-select:none;overflow:hidden}.multimedia_teaser .util-carousel{margin-bottom:2em;width:190%}@media (min-width: 480px){.multimedia_teaser .util-carousel{width:90%}}@media (min-width: 769px), print{.multimedia_teaser .util-carousel{margin-bottom:3em}}.suggested_features.module{-webkit-touch-callout:none;-webkit-user-select:none;-khtml-user-select:none;-moz-user-select:-moz-none;-ms-user-select:none;user-select:none;background-color:#eef2f6}.related.module{-webkit-touch-callout:none;-webkit-user-select:none;-khtml-user-select:none;-moz-user-select:-moz-none;-ms-user-select:none;user-select:none;padding-top:1em}.related.module .module_title,.related.module .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .related.module .carousel_title{text-align:left;font-size:2em}.related.module .gradient_line_module_top{margin:0 0 2em}.carousel_teaser.related .module_title,.carousel_teaser.related .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .carousel_teaser.related .carousel_title{text-align:center}@media (min-width: 600px), print{.carousel_teaser.related .module_title,.carousel_teaser.related .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .carousel_teaser.related .carousel_title{text-align:left;width:88%;margin-left:auto;margin-right:auto}}@media (min-width: 769px), print{.carousel_teaser.related .module_title,.carousel_teaser.related .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .carousel_teaser.related .carousel_title{width:88.5%}}@media (min-width: 1024px), print{.carousel_teaser.related .module_title,.carousel_teaser.related .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .carousel_teaser.related .carousel_title{width:89%}}section.site_teaser .img_col{width:100%;margin-bottom:1.5em}@media (min-width: 600px), print{section.site_teaser .img_col{width:40.78947%;float:left;margin-right:5.26316%;margin-bottom:0}}section.site_teaser .text_col{width:100%}@media (min-width: 600px), print{section.site_teaser .text_col{width:53.94737%;float:left;margin-right:5.26316%;float:right;margin-right:0}}section.site_teaser .text_col .category_title{font-size:0.9em}section.site_teaser .text_col p{margin:1em 0 1.7em}section.site_teaser .site_teaser_caption{margin:.5em 1em 0 0;text-align:right;font-size:.8em}section.site_teaser footer{text-align:center}@media (min-width: 600px), print{section.site_teaser footer{text-align:left}}section.site_teaser .button,section.site_teaser .outline_button,section.site_teaser .primary_media_feature .floating_text_area .button,.primary_media_feature .floating_text_area section.site_teaser .button{padding:0.8em 1.2em}section.more_bar{text-align:center;background-color:#4d91a6;color:black;height:36px;cursor:pointer;position:relative}section.more_bar .title,section.more_bar .arrow_down{display:inline-block;vertical-align:middle;margin-top:6px}section.more_bar .arrow_down{padding:0;cursor:pointer;width:25px;height:25px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -50px -125px;background-size:300px}section.more_bar .arrow_down:hover,section.more_bar .arrow_down.active,section.more_bar .arrow_down.current{background-position:-50px -125px}.inline_dashboard_item{display:inline}.inline_dashboard_item div{display:inline-block}.homepage_carousel .floating_text_area{width:100%;padding:1.4em;bottom:120px;text-align:center;margin-left:auto;margin-right:auto;color:white}@media (min-width: 769px), print{.homepage_carousel .floating_text_area{bottom:calc(125px + 3em)}}.homepage_carousel .floating_text_area .description{display:block;max-height:130px;overflow-y:auto;padding:0 1.4em;color:#ffffff;font-weight:300}.homepage_carousel .floating_text_area .description a{color:#69B9FF}@media (min-width: 769px), print{.homepage_carousel .floating_text_area .description{display:block;line-height:1.4em;padding:0;max-height:none;overflow:hidden}}.touchevents .homepage_carousel .floating_text_area .description{display:none}.no-touchevents .homepage_carousel .floating_text_area .description{display:none}@media (min-width: 769px), print{.no-touchevents .homepage_carousel .floating_text_area .description{display:block !important}}.homepage_carousel .floating_text_area .description .detail_link{display:inline-block;color:#69B9FF;text-transform:none}.homepage_carousel .floating_text_area .description .detail_link:hover{text-decoration:none;color:#ffffff}@media (min-width: 769px), print{.homepage_carousel .floating_text_area .description .detail_link{display:none}}@media (orientation: landscape){.homepage_carousel .floating_text_area .description{display:none !important}}.homepage_carousel .floating_text_area footer{margin:0}@media (min-width: 769px), print{.homepage_carousel .floating_text_area footer{margin:1.6em 0 0}}.homepage_carousel .floating_text_area .media_feature_title{color:white;margin-bottom:.4em;font-size:1.6em;width:70%;margin-left:auto;margin-right:auto;position:relative;font-weight:400}.homepage_carousel .floating_text_area .media_feature_title a{color:white;text-decoration:none}@media (min-width: 600px), print{.homepage_carousel .floating_text_area .media_feature_title{font-size:2em;width:80%}}@media (min-width: 769px), print{.homepage_carousel .floating_text_area .media_feature_title{font-size:1.5em;margin-bottom:.4em;width:100%}}@media (min-width: 1024px), print{.homepage_carousel .floating_text_area .media_feature_title{font-size:1.6em}}@media (min-width: 1200px){.homepage_carousel .floating_text_area .media_feature_title{font-size:1.8em}}@media (min-width: 1700px){.homepage_carousel .floating_text_area .media_feature_title{font-size:1.9em}}.homepage_carousel .floating_text_area .media_feature_title span.arrow{background:url(\"https://mars.nasa.gov/assets/[email protected]\") center no-repeat;position:absolute;right:-25%;margin-top:-0.2em;background-size:12px;height:44px;width:44px;bottom:-9px}@media (min-width: 600px), print{.homepage_carousel .floating_text_area .media_feature_title span.arrow{margin-top:0;right:-10%}}@media (min-width: 769px), print{.homepage_carousel .floating_text_area .media_feature_title span.arrow{display:none}}@media (orientation: landscape){.homepage_carousel .floating_text_area .media_feature_title span.arrow{display:none}}.homepage_carousel .floating_text_area .button,.homepage_carousel .floating_text_area .outline_button,.homepage_carousel .floating_text_area .button:hover,.homepage_carousel .floating_text_area .outline_button:hover{display:none;background-color:#3b788b;text-transform:none;font-size:16px;font-weight:400;border-radius:3px;padding:9px 16px}@media (min-width: 769px), print{.homepage_carousel .floating_text_area .button,.homepage_carousel .floating_text_area .outline_button,.homepage_carousel .floating_text_area .button:hover,.homepage_carousel .floating_text_area .outline_button:hover{display:inline-block !important;color:white !important}}.no-touchevents .homepage_carousel .floating_text_area .button,.no-touchevents .homepage_carousel .floating_text_area .outline_button{transition:background 300ms}.no-touchevents .homepage_carousel .floating_text_area .button:hover,.no-touchevents .homepage_carousel .floating_text_area .outline_button:hover{background-color:#569bb1}@media (min-width: 769px), print{.homepage_carousel .floating_text_area.expandable .media_feature_title{font-size:1.5em}}@media (min-width: 1024px), print{.homepage_carousel .floating_text_area.expandable .media_feature_title{font-size:1.7em}}@media (min-width: 1200px){.homepage_carousel .floating_text_area.expandable .media_feature_title{font-size:2.0em}}@media (min-width: 1700px){.homepage_carousel .floating_text_area.expandable .media_feature_title{font-size:2.2em}}@media (min-width: 769px), print{.homepage_carousel .floating_text_area{text-align:left;padding:1.4em;margin:0}.homepage_carousel .floating_text_area.no-box{position:relative;top:40%;transform:translateY(-50%);width:45%;max-width:500px}.homepage_carousel .floating_text_area.no-box.left{left:5%}.homepage_carousel .floating_text_area.no-box.right{left:51%}.homepage_carousel .floating_text_area.box{position:absolute;background-color:rgba(0,0,0,0.6);width:400px}}@media (min-width: 769px) and (min-width: 1024px), print and (min-width: 1024px), print and (min-width: 769px), print{.homepage_carousel .floating_text_area.box{width:480px}}@media (min-width: 769px) and (min-width: 1200px), print and (min-width: 1200px){.homepage_carousel .floating_text_area.box{width:530px}}@media (min-width: 769px), print{.homepage_carousel .floating_text_area.box.left{left:8%}.homepage_carousel .floating_text_area.box.right{right:8%}.homepage_carousel .floating_text_area.expandable,.homepage_carousel .floating_text_area.expandable_light{transition:background-color .5s ease-out;width:40%;top:auto}.homepage_carousel .floating_text_area.expandable footer,.homepage_carousel .floating_text_area.expandable_light footer{margin-bottom:1em}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px), print and (min-width: 769px), print{.homepage_carousel .floating_text_area.expandable,.homepage_carousel .floating_text_area.expandable_light{width:415px}}@media (min-width: 769px) and (min-width: 1200px), print and (min-width: 1200px){.homepage_carousel .floating_text_area.expandable,.homepage_carousel .floating_text_area.expandable_light{width:480px}}@media (min-width: 769px) and (min-width: 1700px), print and (min-width: 1700px){.homepage_carousel .floating_text_area.expandable,.homepage_carousel .floating_text_area.expandable_light{width:600px}}@media (min-width: 769px), print{.homepage_carousel .floating_text_area.expandable.left,.homepage_carousel .floating_text_area.expandable_light.left{left:8%}}@media (min-width: 769px) and (min-width: 1700px), print and (min-width: 1700px){.homepage_carousel .floating_text_area.expandable.left,.homepage_carousel .floating_text_area.expandable_light.left{left:4%}}@media (min-width: 769px), print{.homepage_carousel .floating_text_area.expandable.right,.homepage_carousel .floating_text_area.expandable_light.right{right:8%}}@media (min-width: 769px) and (min-width: 1700px), print and (min-width: 1700px){.homepage_carousel .floating_text_area.expandable.right,.homepage_carousel .floating_text_area.expandable_light.right{right:4%}}@media (min-width: 769px), print{.homepage_carousel .floating_text_area.expandable .description,.homepage_carousel .floating_text_area.expandable_light .description{max-height:0;overflow:hidden;transition:all .7s}.homepage_carousel .floating_text_area.expandable .media_feature_title:after,.homepage_carousel .floating_text_area.expandable_light .media_feature_title:after{content:url(\"https://mars.nasa.gov/assets/arrow_down_prompt.png\");transition:opacity .25s;position:relative;top:-4px;left:10px;opacity:1}.homepage_carousel .floating_text_area.expandable .media_feature_title span.arrow,.homepage_carousel .floating_text_area.expandable_light .media_feature_title span.arrow{display:none}.homepage_carousel .floating_text_area.expandable:hover:before,.homepage_carousel .floating_text_area.expandable_light:hover:before{opacity:0}.homepage_carousel .floating_text_area.expandable:hover .description,.homepage_carousel .floating_text_area.expandable_light:hover .description{max-height:400px}.homepage_carousel .floating_text_area.expandable:hover .media_feature_title:after,.homepage_carousel .floating_text_area.expandable_light:hover .media_feature_title:after{opacity:0}.homepage_carousel .floating_text_area.expandable{background-color:rgba(0,0,0,0.6)}.homepage_carousel .floating_text_area.expandable_light{background-color:rgba(255,255,255,0.9)}.homepage_carousel .floating_text_area.expandable_light .media_feature_title,.homepage_carousel .floating_text_area.expandable_light .description{color:#452520}.homepage_carousel .floating_text_area.expandable_light .category_title{color:#d63e1c}.homepage_carousel .floating_text_area.expandable_light .media_feature_title:after{content:url(\"https://mars.nasa.gov/assets/arrow_down_light.png\")}}.homepage_carousel .floating_text_area.open span.arrow{transform:rotate(180deg)}@media (min-width: 769px), print{.homepage_carousel .floating_text_area.open .description{display:block}}.primary_media_feature .floating_text_area{bottom:2em}.banner_header_overlay{bottom:1em}.primary_media_feature .floating_text_area,.banner_header_overlay{position:absolute;z-index:12;color:white;width:100%;text-align:center;padding:0 1%}.primary_media_feature .floating_text_area .category_title,.banner_header_overlay .category_title{color:white;margin-bottom:0.7em}.primary_media_feature .floating_text_area .description,.banner_header_overlay .description{margin:-.5em auto 1em}@media (min-width: 769px), print{.primary_media_feature .floating_text_area .description,.banner_header_overlay .description{width:500px;margin-bottom:1.5em}}@media (min-width: 1024px), print{.primary_media_feature .floating_text_area .description,.banner_header_overlay .description{width:550px}}.primary_media_feature .floating_text_area .media_feature_title,.banner_header_overlay .media_feature_title{color:white;margin-bottom:.4em;font-size:1.93em}@media (min-width: 600px), print{.primary_media_feature .floating_text_area .media_feature_title,.banner_header_overlay .media_feature_title{font-size:2.8em}}.primary_media_feature .floating_text_area .media_feature_title a,.banner_header_overlay .media_feature_title a{color:white;text-decoration:none}.custom_banner_container{height:190px;width:100%;background-size:cover;background-position:center}@media only screen and (orientation: landscape){.custom_banner_container{height:260px}}@media (min-width: 600px), print{.custom_banner_container{height:420px}}@media only screen and (min-width: 600px) and (orientation: landscape){.custom_banner_container{height:350px}}@media (min-width: 769px), print{.custom_banner_container{height:400px}}@media only screen and (min-width: 769px) and (orientation: landscape){.custom_banner_container{height:400px}}@media (min-width: 1024px), print{.custom_banner_container{height:440px}}@media (min-width: 1200px){.custom_banner_container{height:550px}}@media (min-width: 1700px){.custom_banner_container{height:660px}}.custom_banner_container .banner_header_overlay{position:absolute;width:100%;bottom:0;z-index:2}.custom_banner_container .article_title{margin-bottom:.5em;text-align:center;color:#FFF}.custom_banner_container .secondary_nav_mobile{display:block}.custom_banner_container .secondary_nav_mobile select{position:relative;padding:.5em 2em .5em 1em;font-size:16px;border:0;height:40px;vertical-align:middle;color:white;-webkit-appearance:none;-o-appearance:none;-moz-appearance:none;background:#3b788b url(\"https://mars.nasa.gov/assets/[email protected]\") no-repeat 95% 10px;background-position:right .8em top 10px;background-size:9px;font-weight:700;cursor:pointer;width:100%;border-radius:5px;max-width:304px;margin:.3em 0 2em}.custom_banner_container .secondary_nav_mobile select::-ms-expand{display:none}.custom_banner_container .secondary_nav_mobile select option{padding:0.5em 1em}@media (min-width: 769px), print{.custom_banner_container .secondary_nav_mobile{display:none}}.custom_banner_container .secondary_nav_desktop{display:none;margin:0 0 .8em 0;text-align:center}@media (min-width: 769px), print{.custom_banner_container .secondary_nav_desktop{display:block}}.custom_banner_container .secondary_nav_desktop li{display:inline-block;position:relative}.custom_banner_container .secondary_nav_desktop a{color:#5AA1F5;font-size:1.2em;font-weight:700;display:block;padding:.3em .9em .3em 0}@media (min-width: 1700px){.custom_banner_container .secondary_nav_desktop a{font-size:1.3em}}.custom_banner_container .secondary_nav_desktop li.current a,.custom_banner_container .secondary_nav_desktop li:hover a{text-decoration:none;color:white}.homepage_carousel .master-slider .ms-slide-bgvideocont{transform:none !important}#masterslider{height:480px;width:100%}@media only screen and (orientation: landscape){#masterslider{height:260px}}@media (min-width: 600px), print{#masterslider{height:500px}}@media only screen and (min-width: 600px) and (orientation: landscape){#masterslider{height:350px}}@media (min-width: 769px), print{#masterslider{height:500px}}@media only screen and (min-width: 769px) and (orientation: landscape){#masterslider{height:500px}}@media (min-width: 1024px), print{#masterslider{height:640px}}@media (min-width: 1200px){#masterslider{height:780px}}@media (min-width: 1700px){#masterslider{height:800px}}#masterslider .ms-slide-bgvideocont{background-color:#000}#masterslider .ms-slide-bgvideocont video{max-width:none}#masterslider .ms-nav-next,#masterslider .ms-nav-prev{display:none}@media (min-width: 769px), print{#masterslider .ms-nav-next,#masterslider .ms-nav-prev{display:block}}#masterslider .ms-nav-prev,#masterslider .ms-nav-next{width:40px;height:80px;margin-top:-60px}@media (min-width: 769px), print{#masterslider .ms-nav-prev,#masterslider .ms-nav-next{margin-top:-80px}}#masterslider .ms-nav-prev{background:url(\"https://mars.nasa.gov/assets/arrow_left_slim.png\");background-size:40px 103px;background-position:0;left:15px;border-top-right-radius:6px;border-bottom-right-radius:6px}#masterslider .ms-nav-next{background:url(\"https://mars.nasa.gov/assets/arrow_right_slim.png\");background-size:40px 103px;background-position:0;right:15px;border-top-left-radius:6px;border-bottom-left-radius:6px}#masterslider .ms-bullets{left:0;right:0;margin:0 auto;bottom:120px;z-index:10}@media (min-width: 769px), print{#masterslider .ms-bullets{bottom:125px}}#masterslider .ms-bullet{background-color:white;background-image:none;border-radius:50% 50% 50% 50%;height:8px;width:8px;opacity:0.5;margin:0 10px}#masterslider .ms-bullet:hover{opacity:1.0}#masterslider .ms-bullet-selected{opacity:1.0}.wysiwyg_content .jpl_carousel .master-slider{width:100%;height:400px}.wysiwyg_content .jpl_carousel .slider_caption{margin-top:.8em;color:#5a6470;font-size:0.8em;height:100px;overflow-y:scroll;line-height:1.4em}.wysiwyg_content .jpl_carousel .ms-nav-next,.wysiwyg_content .jpl_carousel .ms-nav-prev{display:none}.wysiwyg_content .jpl_carousel.medium_mid .ms-nav-next,.wysiwyg_content .jpl_carousel.medium_mid .ms-nav-prev,.wysiwyg_content .jpl_carousel.medium_large .ms-nav-next,.wysiwyg_content .jpl_carousel.medium_large .ms-nav-prev,.wysiwyg_content .jpl_carousel.large .ms-nav-next,.wysiwyg_content .jpl_carousel.large .ms-nav-prev,.wysiwyg_content .jpl_carousel.xlarge .ms-nav-next,.wysiwyg_content .jpl_carousel.xlarge .ms-nav-prev,.wysiwyg_content .jpl_carousel.xxlarge .ms-nav-next,.wysiwyg_content .jpl_carousel.xxlarge .ms-nav-prev{display:block}.no-touchevents .wysiwyg_content .jpl_carousel.medium .ms-nav-next,.no-touchevents .wysiwyg_content .jpl_carousel.medium .ms-nav-prev,.no-touchevents .wysiwyg_content .jpl_carousel.small .ms-nav-next,.no-touchevents .wysiwyg_content .jpl_carousel.small .ms-nav-prev{display:block}.wysiwyg_content .jpl_carousel .ms-nav-prev,.wysiwyg_content .jpl_carousel .ms-nav-next{width:40px;height:80px;margin-top:-60px}.wysiwyg_content .jpl_carousel.medium_large .ms-nav-next,.wysiwyg_content .jpl_carousel.medium_large .ms-nav-prev,.wysiwyg_content .jpl_carousel.large .ms-nav-next,.wysiwyg_content .jpl_carousel.large .ms-nav-prev,.wysiwyg_content .jpl_carousel.xlarge .ms-nav-next,.wysiwyg_content .jpl_carousel.xlarge .ms-nav-prev,.wysiwyg_content .jpl_carousel.xxlarge .ms-nav-next,.wysiwyg_content .jpl_carousel.xxlarge .ms-nav-prev{margin-top:-80px}.wysiwyg_content .jpl_carousel .ms-nav-prev{background:url(\"https://mars.nasa.gov/assets/arrow_left_darktheme.png\");background-size:40px 95px;background-color:rgba(32,32,32,0.9);background-position:0;left:0;border-top-right-radius:6px;border-bottom-right-radius:6px}.wysiwyg_content .jpl_carousel .ms-nav-next{background:url(\"https://mars.nasa.gov/assets/arrow_right_darktheme.png\");background-size:40px 95px;background-color:rgba(32,32,32,0.9);background-position:0;right:0;border-top-left-radius:6px;border-bottom-left-radius:6px}.slick-slider .slick-prev,.slick-slider .slick-next{width:30px;height:48px}.slick-slider .slick-prev:before,.slick-slider .slick-next:before{content:'';display:inline-block;padding:0;cursor:pointer;width:17px;height:25px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -30px -100px;background-size:300px;opacity:1}.slick-slider .slick-prev:before:hover,.slick-slider .slick-prev:before.active,.slick-slider .slick-prev:before.current,.slick-slider .slick-next:before:hover,.slick-slider .slick-next:before.active,.slick-slider .slick-next:before.current{background-position:-30px -100px}.slick-slider .slick-prev:not(.slick-disabled):hover:before,.slick-slider .slick-next:not(.slick-disabled):hover:before{opacity:1}.slick-slider .slick-next:before{content:'';display:inline-block;padding:0;cursor:pointer;width:17px;height:25px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -30px -150px;background-size:300px}.slick-slider .slick-next:before:hover,.slick-slider .slick-next:before.active,.slick-slider .slick-next:before.current{background-position:-30px -150px}.slick-slider .slick-prev{left:-33px}.slick-slider .slick-next{right:-33px}.slick-slider .slick-disabled{cursor:default;opacity:.4}.slick-slider .slick-disabled:before{cursor:default}.main_carousel .slick-nav_container{width:100%;text-align:center;padding-top:1em;border-top:1px solid #BEBEBE}@media (min-width: 600px), print{.main_carousel .slick-nav_container{border:none;margin-top:1em}}.main_carousel .slick-dots{position:relative;top:0}.main_carousel .slick-dots li{vertical-align:top}.main_carousel .slick-dots li button:before{content:\"\";border-radius:50%;background-color:black;height:8px;width:8px;top:6px;left:5px;text-align:center}.main_carousel .slick-nav{position:relative;display:inline-block}.main_carousel .slick-prev,.main_carousel .slick-next{top:-4px}.suggested_features .slick-prev,.suggested_features .slick-next{top:38%}.suggested_features .slick-prev{left:-9%}@media (min-width: 600px), print{.suggested_features .slick-prev{left:-7%}}@media (min-width: 769px), print{.suggested_features .slick-prev{left:-5%}}.suggested_features .slick-next{right:-9%}@media (min-width: 600px), print{.suggested_features .slick-next{right:-7%}}@media (min-width: 769px), print{.suggested_features .slick-next{right:-5%}}.related .slick-prev,.related .slick-next{top:31%}@media (min-width: 600px), print{.related .slick-prev,.related .slick-next{top:38%}}@media (min-width: 1024px), print{.related .slick-prev,.related .slick-next{top:31%}}.related .slick-prev:before,.related .slick-next:before{background-image:none}.main_carousel.module{padding-top:1.7em;padding-bottom:2.1em}@media (min-width: 600px), print{.main_carousel.module{padding-top:2em;padding-bottom:2.2em}}@media (min-width: 769px), print{.main_carousel.module{padding-top:3em;padding-bottom:2.9em}}@media (min-width: 769px), print{.main_carousel.module{padding-top:3.5em;padding-bottom:3.2em}}.main_carousel.module .carousel_header .carousel_title{margin-bottom:.8em}@media (min-width: 600px), print{.main_carousel.module .carousel_header .carousel_title{margin-bottom:1.1em}}.main_carousel.module .slick-slider{margin-left:auto;margin-right:auto;width:100%;margin-bottom:0}.main_carousel.module .slick-slider .slick-slide a{text-decoration:none}.main_carousel.module .slick-slider .content_title{margin:.6em 0 0;color:#222}@media (min-width: 600px), print{.main_carousel.module .slick-slider .content_title{margin:0 0 .8em}}.main_carousel.module .slick-slider .content_title h1{font-size:1.209em;margin-bottom:0em}@media (min-width: 600px), print{.main_carousel.module .slick-slider .content_title h1{font-size:1.395em;margin-bottom:0em}}@media (min-width: 769px), print{.main_carousel.module .slick-slider .content_title h1{font-size:1.581em;margin-bottom:0em}}@media (min-width: 1024px), print{.main_carousel.module .slick-slider .content_title h1{font-size:1.674em;margin-bottom:0em}}@media (min-width: 1200px){.main_carousel.module .slick-slider .content_title h1{font-size:1.767em;margin-bottom:0em}}.main_carousel.module .slick-slider .left-col,.main_carousel.module .slick-slider .right-col{display:inline-block;vertical-align:top}@media (min-width: 600px), print{.main_carousel.module .slick-slider .left-col{display:inline-block;width:45.83333%;float:left;margin-right:4.16667%}}.main_carousel.module .slick-slider .right-col{width:100%}@media (min-width: 600px), print{.main_carousel.module .slick-slider .right-col{width:50%;float:right;margin-right:0}}.main_carousel.module .slick-slider .carousel_item_description{display:none}@media (min-width: 600px), print{.main_carousel.module .slick-slider .carousel_item_description{display:block;margin-bottom:1.5em}}.main_carousel.module .slick-slider .content_body{margin-bottom:.9em}@media (min-width: 600px), print{.main_carousel.module .slick-slider .content_body{margin-bottom:0}}.main_carousel.module .slick-slider .content_footer{padding:.6em 1em 1em 0}@media (min-width: 769px), print{.main_carousel.module .slick-slider .content_footer{padding:0}.main_carousel.module .slick-slider .content_footer a+a{margin-top:.6em}}.main_carousel.module .slick-slider .content_footer .full_artical_link,.main_carousel.module .slick-slider .content_footer .full_category_link{padding-right:.8em;display:inline-block;font-size:.9em;white-space:nowrap}@media (min-width: 600px), print{.main_carousel.module .slick-slider .content_footer .full_artical_link,.main_carousel.module .slick-slider .content_footer .full_category_link{font-size:1em}}@media (min-width: 769px), print{.main_carousel.module .slick-slider .content_footer .full_artical_link,.main_carousel.module .slick-slider .content_footer .full_category_link{float:left;clear:both}}.main_carousel.module .slick-slider .content_footer .full_artical_link:before,.main_carousel.module .slick-slider .content_footer .full_category_link:before{content:\"› \"}.suggested_features.module .slick-slider,.related.module .slick-slider{margin-left:auto;margin-right:auto;width:100%;margin-bottom:1em}@media (min-width: 480px){.suggested_features.module .slick-slider,.related.module .slick-slider{width:84%}}@media (min-width: 600px), print{.suggested_features.module .slick-slider,.related.module .slick-slider{margin-bottom:1.5em;width:90%}}.suggested_features.module .slick-slider .category_title,.related.module .slick-slider .category_title{color:white;line-height:1.4em}.suggested_features.module .slick-slider .slick-slide a,.related.module .slick-slider .slick-slide a{text-decoration:none;color:#222}.suggested_features.module .slick-slider .slide,.related.module .slick-slider .slide{margin:0 6px}@media (min-width: 769px), print{.suggested_features.module .slick-slider .slide,.related.module .slick-slider .slide{margin:0 9px}}.no-touchevents .suggested_features.module .slide:hover .content_title a,.no-touchevents .related.module .slide:hover .content_title a{color:#366599}.no-touchevents .suggested_features.module .slide:hover .rollover_description,.no-touchevents .related.module .slide:hover .rollover_description{background-color:rgba(0,0,0,0.7)}.suggested_features.module .content_title,.related.module .content_title{padding:.6em 0 0;color:#222}.suggested_features.module .content_title a,.related.module .content_title a{font-size:.9em}@media (min-width: 769px), print{.suggested_features.module .content_title a,.related.module .content_title a{font-size:1.14em}}.suggested_features.module .image_and_description_container,.related.module .image_and_description_container{position:relative;overflow:hidden;min-height:129px}.carousel_teaser .slick-slider,.news_teaser .slick-slider,.multimedia_teaser .slick-slider{margin-left:auto;margin-right:auto;width:100%;margin-bottom:2.5em}@media (min-width: 480px){.carousel_teaser .slick-slider,.news_teaser .slick-slider,.multimedia_teaser .slick-slider{width:84%}}@media (min-width: 600px), print{.carousel_teaser .slick-slider,.news_teaser .slick-slider,.multimedia_teaser .slick-slider{width:88%;margin-bottom:3em}}@media (min-width: 769px), print{.carousel_teaser .slick-slider,.news_teaser .slick-slider,.multimedia_teaser .slick-slider{margin-bottom:3.5em}}@media (min-width: 1300px){.carousel_teaser .slick-slider,.news_teaser .slick-slider,.multimedia_teaser .slick-slider{width:92%}}.carousel_teaser .slick-slider .category_title,.news_teaser .slick-slider .category_title,.multimedia_teaser .slick-slider .category_title{color:white;line-height:1.4em}.carousel_teaser .slick-slider .slick-slide a,.news_teaser .slick-slider .slick-slide a,.multimedia_teaser .slick-slider .slick-slide a{text-decoration:none;color:#222}.carousel_teaser .slick-slider .slide,.news_teaser .slick-slider .slide,.multimedia_teaser .slick-slider .slide{margin:0 6px}@media (min-width: 769px), print{.carousel_teaser .slick-slider .slide,.news_teaser .slick-slider .slide,.multimedia_teaser .slick-slider .slide{margin:0 6px}}.no-touchevents .carousel_teaser .slick-slider .slide:hover .content_title a,.no-touchevents .news_teaser .slick-slider .slide:hover .content_title a,.no-touchevents .multimedia_teaser .slick-slider .slide:hover .content_title a{color:#366599;cursor:pointer}.carousel_teaser .slick-slider .image_and_description_container,.news_teaser .slick-slider .image_and_description_container,.multimedia_teaser .slick-slider .image_and_description_container{position:relative;overflow:hidden;min-height:129px}.carousel_teaser .slick-slider .content_title,.news_teaser .slick-slider .content_title,.multimedia_teaser .slick-slider .content_title{padding:.6em 0 0;color:#222;font-weight:400}.carousel_teaser .slick-slider .slick-prev,.carousel_teaser .slick-slider .slick-next,.news_teaser .slick-slider .slick-prev,.news_teaser .slick-slider .slick-next,.multimedia_teaser .slick-slider .slick-prev,.multimedia_teaser .slick-slider .slick-next{top:35%;content:'';display:inline-block;padding:0;cursor:pointer;width:35px;height:35px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -32px -208px;background-size:300px}.carousel_teaser .slick-slider .slick-prev:hover,.carousel_teaser .slick-slider .slick-prev.active,.carousel_teaser .slick-slider .slick-prev.current,.carousel_teaser .slick-slider .slick-next:hover,.carousel_teaser .slick-slider .slick-next.active,.carousel_teaser .slick-slider .slick-next.current,.news_teaser .slick-slider .slick-prev:hover,.news_teaser .slick-slider .slick-prev.active,.news_teaser .slick-slider .slick-prev.current,.news_teaser .slick-slider .slick-next:hover,.news_teaser .slick-slider .slick-next.active,.news_teaser .slick-slider .slick-next.current,.multimedia_teaser .slick-slider .slick-prev:hover,.multimedia_teaser .slick-slider .slick-prev.active,.multimedia_teaser .slick-slider .slick-prev.current,.multimedia_teaser .slick-slider .slick-next:hover,.multimedia_teaser .slick-slider .slick-next.active,.multimedia_teaser .slick-slider .slick-next.current{background-position:-32px -258px}.multimedia_module_gallery .carousel_teaser .slick-slider .slick-prev,.multimedia_module_gallery .carousel_teaser .slick-slider .slick-next,.multimedia_module_gallery .news_teaser .slick-slider .slick-prev,.multimedia_module_gallery .news_teaser .slick-slider .slick-next,.multimedia_module_gallery .multimedia_teaser .slick-slider .slick-prev,.multimedia_module_gallery .multimedia_teaser .slick-slider .slick-next{top:45%}.related .carousel_teaser .slick-slider .slick-prev,.related .carousel_teaser .slick-slider .slick-next,.related .news_teaser .slick-slider .slick-prev,.related .news_teaser .slick-slider .slick-next,.related .multimedia_teaser .slick-slider .slick-prev,.related .multimedia_teaser .slick-slider .slick-next{top:25%}@media (min-width: 769px), print{.related .carousel_teaser .slick-slider .slick-prev,.related .carousel_teaser .slick-slider .slick-next,.related .news_teaser .slick-slider .slick-prev,.related .news_teaser .slick-slider .slick-next,.related .multimedia_teaser .slick-slider .slick-prev,.related .multimedia_teaser .slick-slider .slick-next{top:25%}}@media (min-width: 1024px), print{.related .carousel_teaser .slick-slider .slick-prev,.related .carousel_teaser .slick-slider .slick-next,.related .news_teaser .slick-slider .slick-prev,.related .news_teaser .slick-slider .slick-next,.related .multimedia_teaser .slick-slider .slick-prev,.related .multimedia_teaser .slick-slider .slick-next{top:31%}}.news_teaser .carousel_teaser .slick-slider .slick-prev,.suggested_features .carousel_teaser .slick-slider .slick-prev,.news_teaser .carousel_teaser .slick-slider .slick-next,.suggested_features .carousel_teaser .slick-slider .slick-next,.news_teaser .news_teaser .slick-slider .slick-prev,.suggested_features .news_teaser .slick-slider .slick-prev,.news_teaser .news_teaser .slick-slider .slick-next,.suggested_features .news_teaser .slick-slider .slick-next,.news_teaser .multimedia_teaser .slick-slider .slick-prev,.suggested_features .multimedia_teaser .slick-slider .slick-prev,.news_teaser .multimedia_teaser .slick-slider .slick-next,.suggested_features .multimedia_teaser .slick-slider .slick-next{top:38%}.carousel_teaser .slick-slider .slick-prev,.news_teaser .slick-slider .slick-prev,.multimedia_teaser .slick-slider .slick-prev{transform:scaleX(-1)}.carousel_teaser .slick-slider .slick-prev,.news_teaser .slick-slider .slick-prev,.multimedia_teaser .slick-slider .slick-prev{left:-9%}@media (min-width: 600px), print{.carousel_teaser .slick-slider .slick-prev,.news_teaser .slick-slider .slick-prev,.multimedia_teaser .slick-slider .slick-prev{left:-7%}}@media (min-width: 769px), print{.carousel_teaser .slick-slider .slick-prev,.news_teaser .slick-slider .slick-prev,.multimedia_teaser .slick-slider .slick-prev{left:-6.5%}}.carousel_teaser .slick-slider .slick-next,.news_teaser .slick-slider .slick-next,.multimedia_teaser .slick-slider .slick-next{right:-9%}@media (min-width: 600px), print{.carousel_teaser .slick-slider .slick-next,.news_teaser .slick-slider .slick-next,.multimedia_teaser .slick-slider .slick-next{right:-7%}}@media (min-width: 769px), print{.carousel_teaser .slick-slider .slick-next,.news_teaser .slick-slider .slick-next,.multimedia_teaser .slick-slider .slick-next{right:-6.5%}}.carousel_teaser .slick-slider .slick-disabled,.news_teaser .slick-slider .slick-disabled,.multimedia_teaser .slick-slider .slick-disabled{cursor:default;opacity:.4}.carousel_teaser .slick-slider .slick-disabled:before,.news_teaser .slick-slider .slick-disabled:before,.multimedia_teaser .slick-slider .slick-disabled:before{cursor:default}section.vital_signs_menu .readout{color:#222;display:inline-block;text-align:right;position:absolute;right:0;top:0;height:100%}section.vital_signs_menu .readout .title{font-size:1em;font-weight:300;color:#e4e4e4;white-space:nowrap;text-transform:uppercase}@media (min-width: 600px), print{section.vital_signs_menu .readout .title{font-size:1.4em}}.mini_module header h3,.mini_module header h4{color:#B27628;font-weight:400;font-size:0.9em;margin:0 0 0.1em;text-transform:uppercase;letter-spacing:0}.mini_module .description{font-size:1.85rem}.mini_module .transmissions{font-weight:400;color:#B27628}.mini_module.countdown{overflow:hidden}.mini_module.countdown .unit>span{color:#B27628;background-color:#FFF1DE}#iframe_overlay .mini_module header h3,#iframe_overlay .mini_module header h4{color:#f7c585}#iframe_overlay .mini_module.countdown .unit>span{color:#B27628;background-color:#382B19}#iframe_overlay .mini_module .transmissions{color:#f7c585}#secondary_column .boxed hr+.mini_module{border:none;margin:0;padding:0}#secondary_column .mini_module header{margin-bottom:0.1rem}#secondary_column .mini_module .description{font-size:1.4em}@media (min-width: 1024px), print{#secondary_column .mini_module .description{font-size:1.5em}}#secondary_column .countdown .unit span{padding:0 0.8em}#primary_column .mini_module{margin:3rem 0}#primary_column .mini_module header{margin-bottom:.5rem}#primary_column .mini_module header h3{font-size:1.7rem}.homepage_dashboard_modal.vital_signs_container{position:absolute;top:0;left:0;width:100%;height:calc(100vh - 36px);z-index:41;color:#222;overflow-x:hidden;overflow-y:hidden;-webkit-overflow-scrolling:touch;visibility:hidden;opacity:0;transition:opacity .5s;background:rgba(0,0,0,0.9);word-wrap:break-word;transition:opacity 400ms}.homepage_dashboard_modal.vital_signs_container.visible{visibility:visible;opacity:1}.homepage_dashboard_modal.vital_signs_container.hide_background{background:transparent}@media (min-width: 1200px){.homepage_dashboard_modal.vital_signs_container{z-index:30}}.homepage_dashboard_modal.vital_signs_container .loading{position:absolute;left:50%;top:44vh;width:150px;margin-left:-75px;text-align:center}.homepage_dashboard_modal.vital_signs_container .loading p{color:#fcb963;padding-top:67px;font-size:15px;font-weight:300;margin:0;display:none}.homepage_dashboard_modal.vital_signs_container .vital_signs.module{padding:0}.homepage_dashboard_modal.vital_signs_container .countdown_time .unit{color:#efe9e6}.homepage_dashboard_modal.vital_signs_container .countdown_time .unit span{color:#966b35;background-color:#000000}.homepage_dashboard_modal.vital_signs_container .content{transition:opacity .5s;opacity:0}.homepage_dashboard_modal.vital_signs_container .content.visible{opacity:1}.homepage_dashboard_modal.vital_signs_container .vital_signs_header{margin-bottom:.8em}@media (min-width: 600px), print{.homepage_dashboard_modal.vital_signs_container .vital_signs_header{margin-bottom:1.5em}}.homepage_dashboard_modal.vital_signs_container a{color:#6bbed8;font-size:.9em}.homepage_dashboard_modal.vital_signs_container .more_link{margin:1em 0;float:right;font-weight:600;white-space:nowrap}.homepage_dashboard_modal.vital_signs_container .more_link::after{content:\" ›\"}.homepage_dashboard_modal.vital_signs_container .grid_layout{display:flex;flex-direction:column;background:rgba(62,23,0,0.75);position:relative;width:100%;margin:0;padding:2em;height:calc(100vh - 36px);z-index:5;left:0;top:100vh;opacity:0}@media (min-width: 769px), print{.homepage_dashboard_modal.vital_signs_container .grid_layout{padding:2em 15%}}@media (min-width: 1200px){.homepage_dashboard_modal.vital_signs_container .grid_layout{display:block;padding:2em 2em 2.5em;max-height:100%;height:auto;background:rgba(39,20,5,0.95);top:4.5em;max-width:550px;width:48%;left:100%}}@media (min-width: 1200px){.homepage_dashboard_modal.vital_signs_container .grid_layout{max-width:700px;width:48%}}@media (min-width: 1700px){.homepage_dashboard_modal.vital_signs_container .grid_layout{max-width:750px;width:48%}}.homepage_dashboard_modal.vital_signs_container .background_area{width:100%;height:100%;position:absolute;top:0;left:0;filter:none;display:none}.homepage_dashboard_modal.vital_signs_container .background_area.blur{filter:blur(3px)}@media (min-width: 1200px){.homepage_dashboard_modal.vital_signs_container .background_area{display:none}}.homepage_dashboard_modal.vital_signs_container .module_title,.homepage_dashboard_modal.vital_signs_container .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .homepage_dashboard_modal.vital_signs_container .carousel_title{margin-bottom:0.4em;text-align:left;color:#efe9e6;font-weight:400;font-size:1.6em;width:90%}@media (min-width: 1200px){.homepage_dashboard_modal.vital_signs_container .module_title,.homepage_dashboard_modal.vital_signs_container .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .homepage_dashboard_modal.vital_signs_container .carousel_title{width:100%}}.homepage_dashboard_modal.vital_signs_container a.close_button{position:absolute;top:.8em;right:.8em;z-index:99;width:44px;height:44px}.homepage_dashboard_modal.vital_signs_container a.close_button .close_icon{display:block;height:100%;position:relative}.homepage_dashboard_modal.vital_signs_container a.close_button .close_icon:before{transform:rotate(-45deg);content:'';position:absolute;height:2px;width:100%;top:calc(50% - 1px);left:0;background:#fff;opacity:.8}.homepage_dashboard_modal.vital_signs_container a.close_button .close_icon:after{transform:rotate(45deg);content:'';position:absolute;height:2px;width:100%;top:calc(50% - 1px);left:0;background:#fff;opacity:.8}.no-touchevents .homepage_dashboard_modal.vital_signs_container a.close_button:hover{opacity:1}@media (min-width: 600px), print{.homepage_dashboard_modal.vital_signs_container a.close_button{top:1em;right:1em}}@media (min-width: 1200px){.homepage_dashboard_modal.vital_signs_container a.close_button{background:rgba(37,21,8,0.8)}}@media (min-width: 1700px){.homepage_dashboard_modal.vital_signs_container a.close_button{top:1.2em;right:1.2em;width:60px;height:60px}}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser{overflow:visible;padding-bottom:2em}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .loader{font-size:15px;color:#fcb963;display:block;margin:1em 0}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser h3,.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser h4{color:#fcb963;font-weight:400;font-size:0.9em;margin:0 0 0.1em;text-transform:uppercase;letter-spacing:0}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser p{color:#efe9e6;font-weight:300}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .desc_title{font-weight:600;margin-bottom:.3em;display:block}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser article{margin:0 0 0.5em}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser article header{margin:0}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser article header .subtitle{color:#fcb963}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser article .description{color:#efe9e6;font-weight:300}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser article .description:not(p){font-size:1.8em}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list{margin-bottom:0}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_content{padding:0}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_title{margin-bottom:.5em;text-transform:uppercase;font-size:.9em;font-weight:400;color:#fcb963}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_image{width:30%;float:right;margin:0 0 1em 1.8em}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_image .caption{font-size:0.7em;display:block;margin-top:0.4em;color:#beb0a4}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_text{width:100%;float:none}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_text p,.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_text .description{color:#efe9e6;font-weight:300}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_text p,.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_text .description a{font-size:16px}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_text p{margin:1em 0}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_text p:first-of-type{margin-top:0}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_text p:last-of-type{margin-bottom:0}.homepage_dashboard_modal.vital_signs_container .vital_signs_teaser .item_list .list_text a.more_or_less{color:#efe9e6;text-decoration:underline}.homepage_dashboard_modal.vital_signs_container .overlay_nav li{display:inline-block;color:#94623a;font-size:1.2em;cursor:pointer;transition:all 200ms;margin-right:0.8em}.homepage_dashboard_modal.vital_signs_container .overlay_nav li.current,.homepage_dashboard_modal.vital_signs_container .overlay_nav li:hover{color:white}.homepage_dashboard_modal.vital_signs_container .overlay_nav .nav_spacer{font-size:0.4em;vertical-align:middle;margin:0 .5em;color:#94623a}.homepage_dashboard_modal.vital_signs_container .column_container{display:block;overflow:auto;overflow-x:hidden;padding-right:2em;max-width:100%;max-height:calc(90% - 150px)}@media (min-height: 800px){.homepage_dashboard_modal.vital_signs_container .column_container{max-height:calc(97% - 150px)}}@media (min-width: 769px) and (min-height: 700px), print and (min-height: 700px){.homepage_dashboard_modal.vital_signs_container .column_container{max-height:calc(100% - 129px)}}@media (min-width: 1200px){.homepage_dashboard_modal.vital_signs_container .column_container{width:100%;overflow-y:auto;height:auto;max-height:calc(80vh - 320px)}}@media (min-width: 1200px) and (min-height: 700px){.homepage_dashboard_modal.vital_signs_container .column_container{max-height:calc(88vh - 320px)}}.homepage_dashboard_modal.vital_signs_container .column_container::-webkit-scrollbar-track{background-color:rgba(252,185,99,0.2)}.homepage_dashboard_modal.vital_signs_container .column_container::-webkit-scrollbar{width:5px;left:10px}.homepage_dashboard_modal.vital_signs_container .column_container::-webkit-scrollbar-thumb{background-color:#eaeaea}.homepage_dashboard_modal.vital_signs_container .content_col{position:relative;float:left;width:100%}.homepage_dashboard_modal.vital_signs_container .content_col hr{border-top:1px solid rgba(255,255,255,0.2)}.homepage_dashboard_modal.vital_signs_container .content_col .wysiwyg_content>*{color:#efe9e6}.homepage_dashboard_modal.vital_signs_container .content_col .wysiwyg_content>*:last-child{margin-bottom:0}.homepage_dashboard_modal.vital_signs_container .content_col .wysiwyg_content p{color:#efe9e6}.homepage_dashboard_modal.vital_signs_container .content_col .wysiwyg_content h4{font-size:0.9em;font-weight:400;margin-top:0}.homepage_dashboard_modal.vital_signs_container .dsn_status_module+p{margin-top:0}#vital_signs_modal::-webkit-scrollbar-track{background-color:rgba(252,185,99,0.2)}#vital_signs_modal::-webkit-scrollbar{width:4px;left:10px}#vital_signs_modal::-webkit-scrollbar-thumb{background-color:#eaeaea}section.vital_signs_menu{overflow:hidden;position:relative;background-color:transparent;color:white;z-index:20;height:95px;text-align:center;visibility:hidden;margin-top:-95px}section.vital_signs_menu .slick-slider{width:80%;margin-left:auto;margin-right:auto;overflow:hidden}@media (min-width: 480px){section.vital_signs_menu .slick-slider{width:84%}}@media (min-width: 600px), print{section.vital_signs_menu .slick-slider{width:calc(100% - 85px)}}@media (min-width: 769px), print{section.vital_signs_menu .slick-slider{width:90%}}@media (min-width: 1024px), print{section.vital_signs_menu .slick-slider{width:92%}}@media (min-width: 1700px){section.vital_signs_menu .slick-slider{width:1350px}}section.vital_signs_menu .slick-slider .slick-list{z-index:22}section.vital_signs_menu .slick-slider .slick-track{margin-left:auto;margin-right:auto}section.vital_signs_menu .slick-navigation{display:block;position:absolute;top:0;width:100%;height:100%;z-index:21}@media screen and (min-width: 1550px){section.vital_signs_menu .slick-navigation{display:none}}section.vital_signs_menu .slick-navigation .slick-prev,section.vital_signs_menu .slick-navigation .slick-next{position:absolute;background-color:transparent;font-size:1.8em;color:#808080;height:77px;background-image:url(\"https://mars.nasa.gov/assets/[email protected]\");background-repeat:no-repeat;background-size:135px;width:34px;text-indent:-9999px;top:4px;z-index:999999}section.vital_signs_menu .slick-navigation .slick-prev{left:.7%;border-right:1px solid rgba(216,216,216,0.2);background-position:-27px 19px}@media (min-width: 1024px), print{section.vital_signs_menu .slick-navigation .slick-prev{left:.8%}}@media (min-width: 1200px){section.vital_signs_menu .slick-navigation .slick-prev{left:.9%}}section.vital_signs_menu .slick-navigation .slick-prev:hover{background-position:0px 19px}section.vital_signs_menu .slick-navigation .slick-next{right:.7%;border-left:1px solid rgba(216,216,216,0.2);background-position:-75px 19px}@media (min-width: 1024px), print{section.vital_signs_menu .slick-navigation .slick-next{right:.8%}}@media (min-width: 1200px){section.vital_signs_menu .slick-navigation .slick-next{right:.9%}}section.vital_signs_menu .slick-navigation .slick-next:hover{background-position:-102px 19px}section.vital_signs_menu ul .slick-slide{height:95px;margin:0}section.vital_signs_menu li{display:inline-block;vertical-align:top;position:relative;width:100%}section.vital_signs_menu li:last-child .image_and_description_container{border-right:none}section.vital_signs_menu .image_and_description_container{display:block;overflow:visible;cursor:pointer;width:100%;height:75%;padding-right:.5em;padding-left:1em}@media (min-width: 480px){section.vital_signs_menu .image_and_description_container{border-right:1px solid rgba(216,216,216,0.2)}}@media (min-width: 1700px){section.vital_signs_menu .image_and_description_container{border-right:none}}.no-touchevents section.vital_signs_menu .image_and_description_container:hover .readout .overlay_icon{background:url(\"https://mars.nasa.gov/assets/dashboard_expand_hover.png\") no-repeat;background-size:100%}section.vital_signs_menu .image_and_description_container.current .readout .overlay_icon{background:url(\"https://mars.nasa.gov/assets/dashboard_contract.png\") no-repeat;background-size:100%}.no-touchevents section.vital_signs_menu .image_and_description_container.current:hover .readout .overlay_icon{background:url(\"https://mars.nasa.gov/assets/dashboard_contract_hover.png\") no-repeat !important;background-size:100% !important}section.vital_signs_menu .readout .subtitle{color:#4e8fa4;font-size:.78em;font-weight:500;text-transform:uppercase}section.vital_signs_menu .readout{color:white;position:relative;text-align:left;top:11px;margin:0 auto;float:left}section.vital_signs_menu .readout .text_container{float:left;margin-right:10px}section.vital_signs_menu .readout .title{font-size:1.4em;margin-bottom:0;letter-spacing:-.03em}@media (min-width: 480px){section.vital_signs_menu .readout .title{font-size:1.5em}}@media (min-width: 600px), print{section.vital_signs_menu .readout .title{font-size:1.2em}}@media (min-width: 769px), print{section.vital_signs_menu .readout .title{font-size:1.3em}}@media (min-width: 850px){section.vital_signs_menu .readout .title{font-size:1.1em}}@media (min-width: 950px){section.vital_signs_menu .readout .title{font-size:1.2em}}@media (min-width: 1700px){section.vital_signs_menu .readout .title{font-size:1.5em}}section.vital_signs_menu .readout .overlay_icon{position:absolute;width:37px;height:37px;background:url(\"https://mars.nasa.gov/assets/dashboard_expand.png\") no-repeat;background-size:100%;display:inline-block}#vital_signs_modal.visible ~ .vital_signs_menu{z-index:40}div.modal_open{display:none !important}.image_of_the_day{overflow:hidden;z-index:10;padding:0}.image_of_the_day .window{width:100%;position:absolute;overflow:hidden;height:100%;padding:1em}.image_of_the_day .window.mobile{height:100%;min-height:100%}.image_of_the_day #featured_image{z-index:9;top:0;left:0;height:100%;overflow:hidden}.image_of_the_day a.image_day{width:100%;height:100%;position:absolute;top:0;left:0;z-index:11;text-indent:-999px}.image_of_the_day .grid_layout{height:100%;position:relative}@media (min-width: 1024px), print{.image_of_the_day .grid_layout{width:86%}}.image_of_the_day .floating_text_area{position:absolute;bottom:0;width:100%;text-align:left}@media (min-width: 600px), print{.image_of_the_day .floating_text_area{bottom:1.3rem;text-align:left}}.image_of_the_day header{z-index:12;width:100%}.image_of_the_day header .header_link{display:inline-block;width:100%}@media (min-width: 600px), print{.image_of_the_day header .header_link{width:70%}}.image_of_the_day header .header_link:hover{text-decoration:none}.image_of_the_day header .category_title{color:white;font-size:0.9em;margin-bottom:0.3em}@media (min-width: 600px), print{.image_of_the_day header .category_title{margin-bottom:0.7em}}.image_of_the_day header .media_feature_title{font-size:1.75rem;font-weight:200;margin-bottom:.3rem}@media (min-width: 600px), print{.image_of_the_day header .media_feature_title{margin-bottom:0;font-size:3rem}}.image_of_the_day header .multimedia_link{display:inline-block;text-transform:uppercase;font-size:0.9em;color:white;text-align:right;font-weight:600;right:0;position:relative}@media (min-width: 600px), print{.image_of_the_day header .multimedia_link{bottom:10px;width:28%;position:absolute}}.image_of_the_day header .multimedia_link a{color:white}.image_of_the_day .outline_button,.image_of_the_day .primary_media_feature .floating_text_area .button,.primary_media_feature .floating_text_area .image_of_the_day .button,.image_of_the_day .primary_media_feature .floating_text_area .outline_button,.primary_media_feature .floating_text_area .image_of_the_day .outline_button,.image_of_the_day .banner_header_overlay .button,.banner_header_overlay .image_of_the_day .button{opacity:1}.filter_bar{z-index:20}.filter_bar .section_search{padding-bottom:1em;max-width:380px;width:100%;margin:0 auto}@media (min-width: 1024px), print{.filter_bar .section_search{width:auto;max-width:none;display:block !important;padding-bottom:0}}.filter_bar .section_search .search_submit{right:0px;top:-1px}.filter_bar.fixed{position:fixed;top:0;left:0;width:100%;-webkit-box-shadow:0 4px 4px -1px rgba(0,0,0,0.15);-moz-box-shadow:0 4px 4px -2px rgba(0,0,0,0.15);box-shadow:0 4px 4px -2px rgba(0,0,0,0.15)}.filter_bar .search_binder{width:100%;max-width:304px;position:relative;margin-left:auto;margin-right:auto;margin:0 auto .7em 0}@media (min-width: 480px){.filter_bar .search_binder{margin:0 0 .7em 0}}@media (min-width: 1024px), print{.filter_bar .search_binder{position:relative;vertical-align:top;display:inline-block;width:35%;margin-right:1%;max-width:350px}}.filter_bar input.search_field{width:100%}.filter_bar input.search_field::-webkit-input-placeholder{color:#bbb !important}.filter_bar input.search_field::-moz-placeholder{color:#bbb !important}.filter_bar input.search_field:-moz-placeholder{color:#bbb !important}.filter_bar input.search_field:-ms-input-placeholder{color:#bbb !important}.filter_bar select{position:relative;padding:.5em 2em .5em 1em;font-size:16px;border:0;height:40px;vertical-align:middle;color:white;-webkit-appearance:none;-o-appearance:none;-moz-appearance:none;background:#3b788b url(\"https://mars.nasa.gov/assets/[email protected]\") no-repeat 95% 10px;background-position:right .8em top 10px;background-size:9px;font-weight:700;cursor:pointer;width:100%;border-radius:5px;max-width:304px;margin:0 auto .5em;float:none}.filter_bar select::-ms-expand{display:none}.filter_bar select option{padding:0.5em 1em}@media (min-width: 1024px), print{.filter_bar select{margin-bottom:0;width:30%;max-width:284px}.filter_bar select+select{margin-left:1%}}.filter_bar header{display:inline-block;width:100%;text-align:left}@media (min-width: 600px), print{.filter_bar header{text-align:center}}@media (min-width: 1024px), print{.filter_bar header{display:none}}.filter_bar .arrow_box{display:inline-block;position:absolute;padding:4px;cursor:pointer;right:0;bottom:7px;float:none;transition:all .2s}@media (min-width: 600px), print{.filter_bar .arrow_box{text-align:center}}.filter_bar .arrow_box.rotate_up{transform:rotate(180deg)}.filter_bar .arrow_box.rotate_right{transform:rotate(270deg)}.filter_bar .arrow_box.rotate_left{transform:rotate(90deg)}.filter_bar .arrow_box .arrow_down{display:block;border-left:8px solid transparent;border-right:8px solid transparent;border-top:8px solid #8597B1}.advanced_search .filter_bar .section_search{max-width:none}.advanced_search .filter_bar .suggestion_text{margin-top:0}.advanced_search .filter_bar .search_row+.search_row{margin-top:1em}@media (min-width: 600px), print{.advanced_search .filter_bar .search_row+.search_row{margin-top:0}}@media (min-width: 600px), print{.advanced_search .filter_bar .filter1{width:32.20339%;float:left;margin-right:1.69492%}}@media (min-width: 600px), print{.advanced_search .filter_bar .search_binder{width:49.15254%;float:left;margin-right:1.69492%}}@media (min-width: 600px), print{.advanced_search .filter_bar .conjunction{width:15.25424%;float:right;margin-right:0}}.advanced_search .filter_bar .search_field{background-color:transparent;border:1px solid #C1C1C1;padding-right:1.1em}.advanced_search footer{margin-top:1.5em}.rollover_description{opacity:0;height:0;z-index:1;overflow:hidden;transition:opacity .4s}.rollover_description .rollover_description_inner{height:100%;overflow:hidden}.slide{position:relative;min-height:100%}.slide .overlay_arrow{display:none}@media (min-width: 769px), print{.no-touchevents .slide:hover .rollover_description{padding:.9rem;position:absolute;opacity:1;height:auto;top:0;right:0;width:100%;height:100%;color:white;background-color:rgba(0,0,0,0.9);cursor:pointer;font-size:.95rem;line-height:1.3}.no-touchevents .slide:hover .rollover_description p{line-height:inherit;font-size:inherit;color:white}.no-touchevents .slide:hover .rollover_description p:first-child{margin-top:0}.no-touchevents .slide:hover .rollover_title{font-size:1.6em;font-weight:700;margin-bottom:.2em}.no-touchevents .slide:hover .overlay_arrow{height:14px;width:14px;position:absolute;right:14px;bottom:14px;display:block}.no-touchevents .slide:hover .overlay_arrow img{display:block}}.list_view .rollover_description{display:none}.fancybox-overlay,#fancybox-lock{background:#000 !important}.fancybox-mb-video.fancybox-wrap,.fancybox-mb-info.fancybox-wrap{background:#000}.fancybox-mb-video.fancybox-wrap .fancybox-prev span,.fancybox-mb-info.fancybox-wrap .fancybox-prev span{background-image:url(\"https://mars.nasa.gov/assets/arrow_left_darktheme.png\") !important;background-position:0 !important}.fancybox-mb-video.fancybox-wrap .fancybox-next span,.fancybox-mb-info.fancybox-wrap .fancybox-next span{background-image:url(\"https://mars.nasa.gov/assets/arrow_right_darktheme.png\") !important;background-position:0 !important}.fancybox-mb-video.fancybox-wrap .fancybox-inner,.fancybox-mb-info.fancybox-wrap .fancybox-inner{border:0}.fancybox-mb-video.fancybox-wrap .fancybox-title-float-wrap,.fancybox-mb-info.fancybox-wrap .fancybox-title-float-wrap{position:relative;right:auto;left:auto}.fancybox-mb-video.fancybox-wrap .fancybox-title-float-wrap .child,.fancybox-mb-info.fancybox-wrap .fancybox-title-float-wrap .child{display:block;margin:auto;white-space:normal;padding:1em 0;line-height:normal}.fancybox-mb-video.fancybox-wrap .fancybox-skin,.fancybox-mb-info.fancybox-wrap .fancybox-skin{background-color:black}.fancybox-mb-video.fancybox-wrap .fancybox-title-inside,.fancybox-mb-info.fancybox-wrap .fancybox-title-inside{text-align:left}.fancybox-mb-video.fancybox-wrap .fancybox-nav{top:-15%}@media (min-width: 1px) and (print: 769px){.fancybox-mb-video.fancybox-wrap,.fancybox-mb-info.fancybox-wrap{margin:0 !important;width:95% !important;margin:0 auto !important}.fancybox-mb-video.fancybox-wrap .fancybox-nav,.fancybox-mb-info.fancybox-wrap .fancybox-nav{display:none}.fancybox-mb-video.fancybox-wrap .fancybox-inner,.fancybox-mb-info.fancybox-wrap .fancybox-inner{width:100% !important;height:auto !important}.fancybox-mb-video.fancybox-wrap .fancybox-image,.fancybox-mb-info.fancybox-wrap .fancybox-image{width:100% !important;height:auto !important;margin:0 auto !important}.fancybox-mb-video.fancybox-wrap{left:0 !important;right:0 !important;position:relative !important;margin:0 auto !important;padding:0 !important;border:none}.fancybox-mb-video.fancybox-wrap .fancybox-inner{margin:0 !important}.fancybox-mb-video.fancybox-wrap .fancybox-iframe{height:600px !important}}#fancybox_video .player{min-height:200px;margin-bottom:1.5em}@media (min-width: 480px){#fancybox_video .player{min-height:300px}}@media (min-width: 600px), print{#fancybox_video .player{min-height:400px}}#fancybox_info{margin-top:1.5em}#fancybox_info,#fancybox_video{color:white}#fancybox_info p,#fancybox_info .description,#fancybox_video p,#fancybox_video .description{color:white}#fancybox_info .image_caption,#fancybox_info .image_caption p,#fancybox_video .image_caption,#fancybox_video .image_caption p{color:#aaa;font-size:.9em}#fancybox_info .image_details,#fancybox_video .image_details{overflow:hidden;display:inline-block;width:100%;min-width:inherit}#fancybox_info .image_details .text,#fancybox_video .image_details .text{float:left;text-align:left;width:100%;margin-top:10px}@media (min-width: 769px), print{#fancybox_info .image_details .text,#fancybox_video .image_details .text{margin-top:0}}@media (min-width: 1024px), print{#fancybox_info .image_details .text,#fancybox_video .image_details .text{width:60%}}#fancybox_info .image_details .text .title,#fancybox_video .image_details .text .title{font-size:1.235em;margin-bottom:.1em;line-height:1.3em;font-weight:700}@media (min-width: 600px), print{#fancybox_info .image_details .text .title,#fancybox_video .image_details .text .title{font-size:1.425em;margin-bottom:.18em}}@media (min-width: 769px), print{#fancybox_info .image_details .text .title,#fancybox_video .image_details .text .title{font-size:1.615em;margin-bottom:.26em}}@media (min-width: 1024px), print{#fancybox_info .image_details .text .title,#fancybox_video .image_details .text .title{font-size:1.71em;margin-bottom:.29em}}@media (min-width: 1200px){#fancybox_info .image_details .text .title,#fancybox_video .image_details .text .title{font-size:1.805em;margin-bottom:.32em}}#fancybox_info .image_details .buttons,#fancybox_video .image_details .buttons{width:100%;float:right}@media (min-width: 1024px), print{#fancybox_info .image_details .buttons,#fancybox_video .image_details .buttons{width:40%}}#fancybox_info .image_details .buttons .inner_buttons,#fancybox_video .image_details .buttons .inner_buttons{float:left}@media (min-width: 1024px), print{#fancybox_info .image_details .buttons .inner_buttons,#fancybox_video .image_details .buttons .inner_buttons{float:right}}#fancybox_info .image_details .buttons .addthis_toolbox,#fancybox_video .image_details .buttons .addthis_toolbox{border-radius:4px;overflow:hidden}#fancybox_info .image_details .buttons .addthis_toolbox img,#fancybox_video .image_details .buttons .addthis_toolbox img{height:37px !important;width:auto !important}@media (min-width: 1024px), print{#fancybox_info .image_details .buttons .addthis_toolbox img,#fancybox_video .image_details .buttons .addthis_toolbox img{height:38px !important}}#fancybox_info .image_details .buttons .close_button,#fancybox_video .image_details .buttons .close_button{margin-left:12px}#fancybox_info .image_details .buttons a.button,#fancybox_info .image_details .buttons a.outline_button,#fancybox_video .image_details .buttons a.button,#fancybox_video .image_details .buttons a.outline_button{padding-left:16px;padding-right:16px}#fancybox_info .image_details .buttons .addthis_toolbox,#fancybox_info .image_details .buttons a.button,#fancybox_info .image_details .buttons a.outline_button,#fancybox_video .image_details .buttons .addthis_toolbox,#fancybox_video .image_details .buttons a.button,#fancybox_video .image_details .buttons a.outline_button{float:left;margin-left:0;margin-right:12px}@media (min-width: 1024px), print{#fancybox_info .image_details .buttons .addthis_toolbox,#fancybox_info .image_details .buttons a.button,#fancybox_info .image_details .buttons a.outline_button,#fancybox_video .image_details .buttons .addthis_toolbox,#fancybox_video .image_details .buttons a.button,#fancybox_video .image_details .buttons a.outline_button{float:right;margin-left:12px;margin-right:0}}@media (min-width: 1024px), print{#fancybox_info .image_details .buttons .addthis_toolbox,#fancybox_info .image_details .buttons a.button,#fancybox_info .image_details .buttons a.outline_button,#fancybox_info .image_details .buttons .close_button,#fancybox_video .image_details .buttons .addthis_toolbox,#fancybox_video .image_details .buttons a.button,#fancybox_video .image_details .buttons a.outline_button,#fancybox_video .image_details .buttons .close_button{margin-bottom:12px}}#fancybox_info .close_button,#fancybox_video .close_button{padding:0;cursor:pointer;width:25px;height:25px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -25px 0px;background-size:300px;z-index:8060;position:relative;display:block;float:right}#fancybox_info .close_button:hover,#fancybox_info .close_button.active,#fancybox_info .close_button.current,#fancybox_video .close_button:hover,#fancybox_video .close_button.active,#fancybox_video .close_button.current{background-position:-25px 0px}figure{margin-bottom:1em;max-width:100%}@media (min-width: 769px), print{figure{margin-bottom:2em}}figure figcaption,figure figcaption p{margin-top:.8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{figure figcaption,figure figcaption p{font-size:.88em}}.explore_overlay_page figcaption,.explore_overlay_page figcaption p{color:#b0b4b9}@media (max-width: 480px){figure.lede.full_width{width:100%}figure.lede.full_width figcaption{margin-left:auto;margin-right:auto}}#secondary_column aside figure{margin-bottom:1em}#secondary_column aside figure figcaption{margin-bottom:0}.inline_caption{margin-top:.8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.inline_caption{font-size:.88em}}.content_page #page_header{margin-bottom:1.5em}@media (min-width: 769px), print{.content_page #page_header{margin-bottom:2em}}.content_page #page_header .author{margin:.5em 0 1.8em}.content_page .release_date{font-size:1em;color:#222;text-transform:none}.content_page .category_title{color:#222}.content_page .category_title a{color:#257cdf}.content_page .audio_player{margin-bottom:1em}.content_page .main_feature .master-slider,.content_page .jpl_carousel .master-slider{width:100%;height:300px}@media (min-width: 600px), print{.content_page .main_feature .master-slider,.content_page .jpl_carousel .master-slider{height:400px}}.content_page .main_feature .master-slider .gradient_container_bottom,.content_page .jpl_carousel .master-slider .gradient_container_bottom{height:80px}.content_page .main_feature .master-slider .ms-nav-next,.content_page .main_feature .master-slider .ms-nav-prev,.content_page .jpl_carousel .master-slider .ms-nav-next,.content_page .jpl_carousel .master-slider .ms-nav-prev{display:none}@media (min-width: 769px), print{.content_page .main_feature .master-slider .ms-nav-next,.content_page .main_feature .master-slider .ms-nav-prev,.content_page .jpl_carousel .master-slider .ms-nav-next,.content_page .jpl_carousel .master-slider .ms-nav-prev{display:block}}.content_page .main_feature .master-slider .ms-bullets,.content_page .jpl_carousel .master-slider .ms-bullets{bottom:30px}.content_page .main_feature .master-slider .ms-bullets-count,.content_page .jpl_carousel .master-slider .ms-bullets-count{right:-50%;position:absolute}.content_page .main_feature .master-slider .ms-bullet,.content_page .jpl_carousel .master-slider .ms-bullet{background-color:white;background-image:none;border-radius:50% 50% 50% 50%;height:10px;width:10px;opacity:0.5;margin:0 10px}.content_page .main_feature .master-slider .ms-bullet:hover,.content_page .main_feature .master-slider .ms-bullet.ms-bullet-selected,.content_page .jpl_carousel .master-slider .ms-bullet:hover,.content_page .jpl_carousel .master-slider .ms-bullet.ms-bullet-selected{opacity:1.0}.content_page #primary_column{margin-bottom:5.26316%}@media (min-width: 600px), print{.content_page #primary_column{width:61.53846%;float:left;margin-right:2.5641%;margin-bottom:0}}@media (min-width: 769px), print{.content_page #primary_column{width:64.40678%;float:left;margin-right:1.69492%}}@media (min-width: 1024px), print{.content_page #primary_column{width:61.86441%;float:left;margin-right:1.69492%}}@media (min-width: 1200px){.content_page #primary_column{width:59.32203%;float:left;margin-right:1.69492%}}@media (min-width: 600px), print{.content_page #secondary_column{width:35.89744%;float:right;margin-right:0}}@media (min-width: 769px), print{.content_page #secondary_column{width:32.20339%;float:right;margin-right:0}}.content_page.full_width #primary_column,.content_page.full_width #secondary_column{width:100%}.content_page.feature{padding:2em 0 5.3em}.content_page.feature #secondary_column{display:none}.content_page.feature #primary_column{width:64.40678%;float:left;margin-right:1.69492%;margin:auto;padding:1em 0 5.3em;float:none}#secondary_column>:first-child{margin-top:0}#secondary_column{word-wrap:break-word}#secondary_column aside{margin-bottom:7.14286%}#secondary_column aside:last-child{margin-bottom:0}#secondary_column aside.boxed{border:1px solid #C1C1C1;padding:5.26316%}#secondary_column aside.none{border:0;padding:0}#secondary_column aside>:last-child{margin-bottom:0}#secondary_column aside.links_module li{margin-bottom:.5em}#secondary_column aside.downloads_module .download{margin-bottom:1em}#secondary_column aside.downloads_module .download:last-of-type{margin-bottom:0}#secondary_column aside.downloads_module .button,#secondary_column aside.downloads_module .outline_button{margin-top:1em}#secondary_column aside.list_view_module a{text-decoration:none}#secondary_column aside.list_view_module ul{margin-bottom:1.5em}#secondary_column aside.list_view_module li{padding:.6em 0}#secondary_column aside.list_view_module li:last-child{padding-bottom:0}#secondary_column aside.list_view_module .list_image{float:right;margin-left:4%;margin-bottom:.5em;width:32%}@media (min-width: 600px), print{#secondary_column aside.list_view_module .list_image{margin-left:0;margin-bottom:0;float:left;width:31.03448%;float:left;margin-right:3.44828%}}@media (min-width: 600px), print{#secondary_column aside.list_view_module .list_text{width:65.51724%;float:right;margin-right:0}}#secondary_column aside.list_view_module .list_title{letter-spacing:-.01em;font-weight:700}#secondary_column aside.list_view_module .list_title:hover{color:#222}#secondary_column aside.sig_events_module h4{margin-bottom:1em}#secondary_column aside.sig_events_module h4:last-child{margin-bottom:0}#secondary_column aside.sig_events_module ul{margin-bottom:0}#secondary_column aside.sig_events_module ul li{margin-bottom:.5em}#secondary_column .inline_image{margin-bottom:7.14286%}#secondary_column .inline_image .inline_caption{display:block;margin-top:.8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{#secondary_column .inline_image .inline_caption{font-size:.88em}}#secondary_column .related_content_module{margin:0 0 7.14286% 0;padding:5.26316%;width:100%;border:1px solid #bebebe}#secondary_column .related_content_module li{width:100%;border-bottom:1px solid #bebebe}#secondary_column .related_content_module li:last-child{border-bottom:none;padding-bottom:0}#secondary_column .related_content_module li:first-child{border-top:none}#secondary_column .related_content_module>:last-child{margin-bottom:0}a.main_image_enlarge,a.inline_image_enlarge{display:block;position:relative;height:100%}a.main_image_enlarge .enlarge_icon,a.inline_image_enlarge .enlarge_icon{position:absolute;border-radius:6px;border:1px solid rgba(200,200,200,0.8);left:15px;bottom:15px;width:40px;height:40px;background-color:rgba(0,0,0,0.5);background-image:url(\"https://mars.nasa.gov/assets/zoom_icon.png\");background-size:50%;background-repeat:no-repeat;background-position:50%;opacity:0;transition:opacity 0.2s ease-in}a.main_image_enlarge:hover .enlarge_icon,a.inline_image_enlarge:hover .enlarge_icon{opacity:0.8}body #fancybox-lock{z-index:200000}.article_nav{display:none}@media (min-width: 1024px), print{.article_nav{display:block;position:relative;z-index:11}.article_nav .article_nav_block{position:fixed;height:86px;display:inline-block;top:42.5%}.article_nav .article_nav_block .link_box{width:40px;background-color:#e4e9ef;display:inline;height:100%}.article_nav .article_nav_block .article_details{display:inline;width:250px;background-color:#FFF;text-decoration:none;color:#000;padding:10px;background-color:#e4e9ef}.article_nav .article_nav_block .article_details .img{margin-bottom:6px}.article_nav .article_nav_block .article_details .title{font-weight:700;font-size:.9em}.article_nav .article_nav_block.prev{left:0}.article_nav .article_nav_block.prev .link_box{float:left}.article_nav .article_nav_block.prev .article_details{float:left;display:none}.article_nav .article_nav_block.next{right:0}.article_nav .article_nav_block.next .link_box{float:right}.article_nav .article_nav_block.next .article_details{display:none;float:right}.no-touchevents .article_nav .article_nav_block:hover .article_details{display:block}}.feature_pages{padding:1em 0 3.8em}.feature_pages #page_header{width:94%;max-width:600px;margin-left:auto;margin-right:auto}@media (min-width: 769px), print{.feature_pages #page_header{width:80%}}@media (min-width: 1200px){.feature_pages #page_header{width:55%}}.feature_pages #primary_column{width:100%;margin:0}.feature_pages .wysiwyg_content>*{width:94%;max-width:600px;margin-left:auto;margin-right:auto}@media (min-width: 769px), print{.feature_pages .wysiwyg_content>*{width:80%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content>*{width:55%}}.feature_pages .wysiwyg_content p{font-size:18px;line-height:28px}@media (min-width: 769px), print{.feature_pages .wysiwyg_content p{font-size:19px;line-height:30px}}.feature_pages .wysiwyg_content>ul:not(.item_list_module):not(.item_grid_module),.feature_pages .wysiwyg_content>ol{list-style-position:outside;padding-left:1em}.feature_pages .wysiwyg_content>ul:not(.item_list_module):not(.item_grid_module) ul,.feature_pages .wysiwyg_content>ul:not(.item_list_module):not(.item_grid_module) ol,.feature_pages .wysiwyg_content>ol ul,.feature_pages .wysiwyg_content>ol ol{list-style-position:outside}.top_feature_area{text-align:center;position:relative}.top_feature_area .header_overlay{position:absolute;width:100%;padding:0 1%;top:46%;color:white;transform:translateY(-50%)}@media (min-width: 769px), print{.top_feature_area .header_overlay{padding:0 4%}}.top_feature_area .header_overlay .article_title{font-size:1.2em;margin-bottom:0}@media (min-width: 480px){.top_feature_area .header_overlay .article_title{font-size:1.6em}}@media (min-width: 600px), print{.top_feature_area .header_overlay .article_title{font-size:1.9em}}@media (min-width: 769px), print{.top_feature_area .header_overlay .article_title{font-size:2.2em}}@media (min-width: 1024px), print{.top_feature_area .header_overlay .article_title{font-size:2.8em}}@media (min-width: 1200px){.top_feature_area .header_overlay .article_title{font-size:3.2em}}@media (min-width: 1700px){.top_feature_area .header_overlay .article_title{font-size:3.4em}}.top_feature_area .header_overlay .sub_title{font-size:1.2em}@media (min-width: 480px){.top_feature_area .header_overlay .sub_title{font-size:1.5em}}@media (min-width: 769px), print{.top_feature_area .header_overlay .sub_title{font-size:1.9em}}.top_feature_area .header_overlay .author{padding:0.2em 0.5em 0.5em 0.6em;background-color:rgba(0,0,0,0.5);margin:0.2em auto 1em;display:inline-block}@media (min-width: 480px){.top_feature_area .header_overlay .author{margin-top:0.4em}}@media (min-width: 600px), print{.top_feature_area .header_overlay .author{margin-top:0.7em}}@media (min-width: 769px), print{.top_feature_area .header_overlay .author{padding:0.3em 0.5em 0.5em 0.7em;max-width:360px;margin-top:1em}}@media (min-width: 1024px), print{.top_feature_area .header_overlay .author{padding:0.4em 0.6em 0.6em 0.8em;max-width:400px;margin-top:1.5em}}@media (min-width: 1200px){.top_feature_area .header_overlay .author{margin-top:1.8em}}.top_feature_area .header_overlay .author p{color:white;margin:0;font-size:0.8em}@media (min-width: 769px), print{.top_feature_area .header_overlay .author p{font-size:0.95em}}.top_feature_area .article_title{margin-bottom:0.9em}.top_feature_area .category_title{color:#707070;margin-bottom:1em}.top_feature_area a.category_title{color:#257cdf}.top_feature_area .feature_header:first-child{width:94%;max-width:600px;margin-left:auto;margin-right:auto;padding:3em 0 1.7em}@media (min-width: 769px), print{.top_feature_area .feature_header:first-child{width:80%}}@media (min-width: 1200px){.top_feature_area .feature_header:first-child{width:55%}}@media (min-width: 480px){.top_feature_area .feature_header:first-child{padding:4em 0 2em}}.top_feature_area .feature_header:first-child:after{content:\"\";display:block;height:1px;width:56%;border-bottom:5px solid;max-width:200px;margin:1.7em auto 0.3em}.top_feature_area .feature_header.no_main_image{padding-bottom:.9em}.top_feature_area .release_date{text-transform:none;color:#222;margin-left:0.1em}.top_feature_area .header_overlay+.feature_header{width:94%;max-width:600px;margin-left:auto;margin-right:auto;text-align:left}@media (min-width: 769px), print{.top_feature_area .header_overlay+.feature_header{width:80%}}@media (min-width: 1200px){.top_feature_area .header_overlay+.feature_header{width:55%}}.top_feature_area figure.lede.full_width figcaption{margin:.8em;text-align:left;font-size:1em}.feature_pages .wysiwyg_content .mb_expand{width:100%;max-width:none}.feature_pages .wysiwyg_content .mb_expand .expandable_element_link{width:94%;max-width:600px;margin-left:auto;margin-right:auto;display:block}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .mb_expand .expandable_element_link{width:80%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .mb_expand .expandable_element_link{width:55%}}.feature_pages .wysiwyg_content .mb_expand .expandable_element{display:none;max-width:none;width:100%}.feature_pages .wysiwyg_content .mb_expand .expandable_element>*{width:94%;max-width:600px;margin-left:auto;margin-right:auto}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .mb_expand .expandable_element>*{width:80%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .mb_expand .expandable_element>*{width:55%}}.countdown .unit{font-weight:300;display:inline-block;position:relative;padding:0 9px 0 0;vertical-align:middle;text-align:center}.countdown .unit+.unit:before{content:\" : \";position:absolute;left:-4px}.countdown .unit:first-of-type{padding-left:0}.countdown .unit:last-of-type{padding-right:0}.countdown .unit span{font-weight:600;padding:0 1em;clear:both;display:block;font-size:11px;text-transform:uppercase;text-align:center;margin-bottom:.3rem}.countdown .completed{font-size:1.8em;font-weight:300;margin-top:3px}.feature_pages .countdown,.content_page .countdown{border-top:1px solid #E8E8E8;border-bottom:1px solid #E8E8E8;text-align:center;padding:0.7em 0 0.9em;margin-top:2.7em;margin-bottom:2.7em}.feature_pages .countdown>div,.content_page .countdown>div{display:inline-block;vertical-align:middle}.feature_pages .countdown_title,.content_page .countdown_title{width:100%}@media (min-width: 600px), print{.feature_pages .countdown_title,.content_page .countdown_title{width:auto;margin-right:.8em}}#secondary_column .countdown{margin-bottom:7.14286%;text-align:left}#secondary_column .countdown .countdown_title{margin-right:0;width:100%}#explore_overlay .countdown{border-top-color:#212121;border-bottom-color:#212121}.curtain_module{margin:3em 0;clear:both}.curtain_module .curtain_caption_container{background-color:#eee;padding:1em}.curtain_module .curtain_title{margin:0 0 .5em 0}.curtain_module .curtain_subtitle{color:#777777;margin:0 0 0.5em 0}.feature_pages .curtain_module{width:94%;max-width:100%;margin:3em auto;float:none}@media (min-width: 600px), print{.feature_pages .curtain_module{max-width:600px}}.feature_pages .curtain_module.full-bleed,.feature_pages .curtain_module.full_width,.feature_pages .curtain_module.wide,.feature_pages .curtain_module.parallax{clear:both}@media (min-width: 600px), print{.feature_pages .curtain_module.full-bleed,.feature_pages .curtain_module.full_width,.feature_pages .curtain_module.wide,.feature_pages .curtain_module.parallax{margin-top:5em;margin-bottom:5em}}.feature_pages .curtain_module.column-width{max-width:94%;margin-top:3em;margin-bottom:3em;clear:both}@media (min-width: 600px), print{.feature_pages .curtain_module.column-width{max-width:600px}}.feature_pages .curtain_module.full-bleed{width:100%;max-width:none}.feature_pages .curtain_module.full-bleed figcaption{margin:.8em .8em 0 .8em}.feature_pages .curtain_module.full_width{clear:both}@media (min-width: 769px), print{.feature_pages .curtain_module.full_width{width:94%;max-width:600px;margin-left:auto;margin-right:auto}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px), print and (min-width: 769px), print{.feature_pages .curtain_module.full_width{width:80%}}@media (min-width: 769px) and (min-width: 1200px), print and (min-width: 1200px){.feature_pages .curtain_module.full_width{width:55%}}.feature_pages .curtain_module.wide{width:98%;max-width:none}@media (min-width: 769px), print{.feature_pages .curtain_module.wide{width:95%}}@media (min-width: 769px), print{.feature_pages .curtain_module.column-width{max-width:calc(600px + 6%)}}@media (min-width: 1024px), print{.feature_pages .curtain_module.column-width{max-width:calc(600px + 10%)}}@media (min-width: 1200px){.feature_pages .curtain_module.column-width{max-width:calc(600px + 15%)}}.feature_pages .curtain_module.left,.feature_pages .curtain_module.right{max-width:94%}@media (min-width: 600px), print{.feature_pages .curtain_module.left,.feature_pages .curtain_module.right{width:50%;max-width:50%}}@media (min-width: 769px), print{.feature_pages .curtain_module.left,.feature_pages .curtain_module.right{width:27%;max-width:27%}}@media (min-width: 1700px){.feature_pages .curtain_module.left,.feature_pages .curtain_module.right{width:25%;max-width:25%}}@media (min-width: 600px), print{.feature_pages .curtain_module.left{float:left;margin:1em 2.5em 1.5em 0;margin-left:3%}}@media (min-width: 1200px){.feature_pages .curtain_module.left{margin-left:15%}}@media (min-width: 1700px){.feature_pages .curtain_module.left{margin-left:20%}}@media (min-width: 480px){.feature_pages .curtain_module.right{float:right;margin:1em 0 1.5em 2.5em;margin-right:3%}}@media (min-width: 1200px){.feature_pages .curtain_module.right{margin-right:15%}}@media (min-width: 1700px){.feature_pages .curtain_module.right{margin-right:20%}}.feature_pages .curtain_module.parallax_module{position:relative;overflow:hidden;z-index:10;padding-bottom:0;width:100%;max-width:none}.feature_pages .curtain_module.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.feature_pages .curtain_module.parallax_module .caption{font-size:.88em}}.feature_pages .curtain_module.parallax_module img{height:auto !important}.feature_pages .curtain_module.parallax_module .window{width:100%;height:auto;position:absolute;overflow:hidden;padding:2em}.feature_pages .curtain_module.parallax_module .window.mobile{height:auto;min-height:100%}.feature_pages .curtain_module.parallax_module .window .featured_image{z-index:9;top:0;left:0;height:100%;overflow:hidden}@media (min-width: 769px), print{.feature_pages .curtain_module.parallax_module .window .featured_image{position:absolute}}.explore_overlay_page .curtain_module .curtain_caption_container{background-color:#232323}.wysiwyg_content .related_content_module,#secondary_column .related_content_module{font-weight:700}.wysiwyg_content .related_content_module ul,#secondary_column .related_content_module ul{margin:0}.wysiwyg_content .related_content_module li,#secondary_column .related_content_module li{padding:1em 0;border-bottom:1px solid #E5E5E5}.wysiwyg_content .related_content_module li:first-child,#secondary_column .related_content_module li:first-child{border-top:1px solid #E5E5E5}.wysiwyg_content .related_content_module .module_title,.wysiwyg_content .related_content_module .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .wysiwyg_content .related_content_module .carousel_title,#secondary_column .related_content_module .module_title,#secondary_column .related_content_module .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #secondary_column .related_content_module .carousel_title{font-size:1.2em;text-align:left;margin-top:0}.wysiwyg_content .related_content_module .list_image,#secondary_column .related_content_module .list_image{width:25%;display:inline-block}.wysiwyg_content .related_content_module .list_image+.list_text,#secondary_column .related_content_module .list_image+.list_text{width:67%;position:relative;display:inline-block;margin-left:4%;vertical-align:middle}.wysiwyg_content .related_content_module{max-width:100%;margin-top:1.4em;margin-bottom:1.4em}@media (min-width: 769px), print{.wysiwyg_content .related_content_module{margin-top:2em;margin-bottom:2em}}.wysiwyg_content .related_content_module.left,.wysiwyg_content .related_content_module.right{float:none}@media (min-width: 480px){.wysiwyg_content .related_content_module.left,.wysiwyg_content .related_content_module.right{max-width:50%}}@media (min-width: 1200px){.wysiwyg_content .related_content_module.left,.wysiwyg_content .related_content_module.right{max-width:40%}}@media (min-width: 480px){.wysiwyg_content .related_content_module.left{float:left;margin:1em 2.5em 1.5em 0}}@media (min-width: 480px){.wysiwyg_content .related_content_module.right{float:right;margin:1em 0 1.5em 2.5em}}.wysiwyg_content .related_content_module.full-bleed,.wysiwyg_content .related_content_module.full_width,.wysiwyg_content .related_content_module.wide,.wysiwyg_content .related_content_module.parallax,.wysiwyg_content .related_content_module.column-width{clear:both}.wysiwyg_content .related_content_module.parallax_module{width:100%;position:relative}.wysiwyg_content .related_content_module.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.wysiwyg_content .related_content_module.parallax_module .caption{font-size:.88em}}.explore_overlay_page .wysiwyg_content .related_content_module.parallax_module .caption{color:#b0b4b9}.wysiwyg_content .related_content_module .sidebar_title,.wysiwyg_content #secondary_column .related_content_module .module_title,#secondary_column .wysiwyg_content .related_content_module .module_title,.wysiwyg_content #secondary_column .related_content_module .main_carousel.module .carousel_header .carousel_title,#secondary_column .wysiwyg_content .related_content_module .main_carousel.module .carousel_header .carousel_title,.wysiwyg_content .main_carousel.module .carousel_header #secondary_column .related_content_module .carousel_title,.main_carousel.module .carousel_header #secondary_column .wysiwyg_content .related_content_module .carousel_title,.wysiwyg_content .right_col .related_content_module .module_title,.right_col .wysiwyg_content .related_content_module .module_title,.wysiwyg_content .right_col .related_content_module .main_carousel.module .carousel_header .carousel_title,.right_col .wysiwyg_content .related_content_module .main_carousel.module .carousel_header .carousel_title,.wysiwyg_content .main_carousel.module .carousel_header .right_col .related_content_module .carousel_title,.main_carousel.module .carousel_header .right_col .wysiwyg_content .related_content_module .carousel_title{margin-top:0;font-size:1.5em}.wysiwyg_content .related_content_module.full_width{border:1px solid #D2D2D2;padding:5.26316%}@media (min-width: 600px), print{.wysiwyg_content .related_content_module.full_width{padding:2em}}.wysiwyg_content .related_content_module.full_width li{width:100%}.wysiwyg_content .related_content_module.full_width li:first-child{border-top:none}.wysiwyg_content .related_content_module.full_width li:last-child{border-bottom:transparent 0;padding-bottom:0}.wysiwyg_content .related_content_module.full_width .module_title,.wysiwyg_content .related_content_module.full_width .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .wysiwyg_content .related_content_module.full_width .carousel_title{margin-top:0;font-size:1.5em}.feature_pages .wysiwyg_content .related_content_module{width:94%;max-width:100%;margin:3em auto;float:none}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .related_content_module{max-width:600px}}.feature_pages .wysiwyg_content .related_content_module.full-bleed,.feature_pages .wysiwyg_content .related_content_module.full_width,.feature_pages .wysiwyg_content .related_content_module.wide,.feature_pages .wysiwyg_content .related_content_module.parallax{clear:both}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .related_content_module.full-bleed,.feature_pages .wysiwyg_content .related_content_module.full_width,.feature_pages .wysiwyg_content .related_content_module.wide,.feature_pages .wysiwyg_content .related_content_module.parallax{margin-top:5em;margin-bottom:5em}}.feature_pages .wysiwyg_content .related_content_module.column-width{max-width:94%;margin-top:3em;margin-bottom:3em;clear:both}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .related_content_module.column-width{max-width:600px}}.feature_pages .wysiwyg_content .related_content_module.full-bleed{width:100%;max-width:none}.feature_pages .wysiwyg_content .related_content_module.full-bleed figcaption{margin:.8em .8em 0 .8em}.feature_pages .wysiwyg_content .related_content_module.full_width{clear:both}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .related_content_module.full_width{width:94%;max-width:600px;margin-left:auto;margin-right:auto}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px), print and (min-width: 769px), print{.feature_pages .wysiwyg_content .related_content_module.full_width{width:80%}}@media (min-width: 769px) and (min-width: 1200px), print and (min-width: 1200px){.feature_pages .wysiwyg_content .related_content_module.full_width{width:55%}}.feature_pages .wysiwyg_content .related_content_module.wide{width:98%;max-width:none}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .related_content_module.wide{width:95%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .related_content_module.column-width{max-width:calc(600px + 6%)}}@media (min-width: 1024px), print{.feature_pages .wysiwyg_content .related_content_module.column-width{max-width:calc(600px + 10%)}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .related_content_module.column-width{max-width:calc(600px + 15%)}}.feature_pages .wysiwyg_content .related_content_module.left,.feature_pages .wysiwyg_content .related_content_module.right{max-width:94%}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .related_content_module.left,.feature_pages .wysiwyg_content .related_content_module.right{width:50%;max-width:50%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .related_content_module.left,.feature_pages .wysiwyg_content .related_content_module.right{width:27%;max-width:27%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .related_content_module.left,.feature_pages .wysiwyg_content .related_content_module.right{width:25%;max-width:25%}}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .related_content_module.left{float:left;margin:1em 2.5em 1.5em 0;margin-left:3%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .related_content_module.left{margin-left:15%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .related_content_module.left{margin-left:20%}}@media (min-width: 480px){.feature_pages .wysiwyg_content .related_content_module.right{float:right;margin:1em 0 1.5em 2.5em;margin-right:3%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .related_content_module.right{margin-right:15%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .related_content_module.right{margin-right:20%}}.feature_pages .wysiwyg_content .related_content_module.parallax_module{position:relative;overflow:hidden;z-index:10;padding-bottom:0;width:100%;max-width:none}.feature_pages .wysiwyg_content .related_content_module.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .related_content_module.parallax_module .caption{font-size:.88em}}.feature_pages .wysiwyg_content .related_content_module.parallax_module img{height:auto !important}.feature_pages .wysiwyg_content .related_content_module.parallax_module .window{width:100%;height:auto;position:absolute;overflow:hidden;padding:2em}.feature_pages .wysiwyg_content .related_content_module.parallax_module .window.mobile{height:auto;min-height:100%}.feature_pages .wysiwyg_content .related_content_module.parallax_module .window .featured_image{z-index:9;top:0;left:0;height:100%;overflow:hidden}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .related_content_module.parallax_module .window .featured_image{position:absolute}}.feature_pages .wysiwyg_content .related_content_module .module_title,.feature_pages .wysiwyg_content .related_content_module .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .feature_pages .wysiwyg_content .related_content_module .carousel_title{margin-bottom:0.8em}.vital_signs .related_content_module{font-weight:normal;margin-bottom:1em}.vital_signs .related_content_module li{border:none;padding:0.4em 0;font-size:0.9em}.vital_signs .related_content_module .module_title,.vital_signs .related_content_module .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .vital_signs .related_content_module .carousel_title{font-size:1.2em;margin-bottom:.4em}.vital_signs .related_content_module .list_image{display:none}.vital_signs .related_content_module .list_text{width:100%;float:none;margin-left:10px}.vital_signs .related_content_module .list_text:before{content:\"›\";color:#42a0f2;margin-left:-10px}.explore_overlay_page .related_content_module.full_width{border-color:#353535}.explore_overlay_page .related_content_module ul li{border-bottom:1px solid #212121}.explore_overlay_page .related_content_module ul li:first-child{border-top:1px solid #212121}.wysiwyg_content .carousel_module{max-width:100%;margin-top:1.4em;margin-bottom:1.4em;overflow:hidden;clear:both;height:275px}@media (min-width: 769px), print{.wysiwyg_content .carousel_module{margin-top:2em;margin-bottom:2em}}.wysiwyg_content .carousel_module.left,.wysiwyg_content .carousel_module.right{float:none}@media (min-width: 480px){.wysiwyg_content .carousel_module.left,.wysiwyg_content .carousel_module.right{max-width:50%}}@media (min-width: 1200px){.wysiwyg_content .carousel_module.left,.wysiwyg_content .carousel_module.right{max-width:40%}}@media (min-width: 480px){.wysiwyg_content .carousel_module.left{float:left;margin:1em 2.5em 1.5em 0}}@media (min-width: 480px){.wysiwyg_content .carousel_module.right{float:right;margin:1em 0 1.5em 2.5em}}.wysiwyg_content .carousel_module.full-bleed,.wysiwyg_content .carousel_module.full_width,.wysiwyg_content .carousel_module.wide,.wysiwyg_content .carousel_module.parallax,.wysiwyg_content .carousel_module.column-width{clear:both}.wysiwyg_content .carousel_module.parallax_module{width:100%;position:relative}.wysiwyg_content .carousel_module.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.wysiwyg_content .carousel_module.parallax_module .caption{font-size:.88em}}.explore_overlay_page .wysiwyg_content .carousel_module.parallax_module .caption{color:#b0b4b9}.wysiwyg_content .carousel_module .gradient_container_top{display:none}.wysiwyg_content .carousel_module.medium_mid{height:500px}.wysiwyg_content .carousel_module.medium_large,.wysiwyg_content .carousel_module.large{height:600px}.wysiwyg_content .carousel_module.xlarge,.wysiwyg_content .carousel_module.xxlarge{height:700px}.wysiwyg_content .carousel_module .master-slider{width:100%;height:100%}.wysiwyg_content .carousel_module:last-child{margin-bottom:0}.wysiwyg_content .carousel_module header{position:relative;margin-bottom:0;padding:0 22px}.wysiwyg_content .carousel_module header:after{background:url(\"https://mars.nasa.gov/assets/[email protected]\") no-repeat;content:\"\";position:absolute;right:0;top:.9em;background-size:12px;height:7.5px;width:12px}.wysiwyg_content .carousel_module .media_feature_title{color:white;font-size:1.4em;margin:0}.wysiwyg_content .carousel_module .media_feature_title a{color:white;text-decoration:none}.wysiwyg_content .carousel_module .media_feature_title:hover{cursor:pointer}.wysiwyg_content .carousel_module .media_feature_title:after{background:url(\"https://mars.nasa.gov/assets/[email protected]\") no-repeat;content:\"\";margin-left:14px;background-size:12px;height:7.5px;width:12px;display:inline-block}.wysiwyg_content .carousel_module .subtitle{font-size:1.1em;margin:0.4em 0 .4em}.wysiwyg_content .carousel_module .description{display:block;max-height:130px;overflow-y:auto}.wysiwyg_content .carousel_module .description a{color:#69B9FF}.wysiwyg_content .carousel_module.medium_large .description,.wysiwyg_content .carousel_module.large .description,.wysiwyg_content .carousel_module.xlarge .description,.wysiwyg_content .carousel_module.xxlarge .description{max-height:none;overflow:hidden;margin-bottom:0;display:none}.wysiwyg_content .carousel_module.medium_large .floating_text_area.open header,.wysiwyg_content .carousel_module.large .floating_text_area.open header,.wysiwyg_content .carousel_module.xlarge .floating_text_area.open header,.wysiwyg_content .carousel_module.xxlarge .floating_text_area.open header{margin-bottom:.7em}.wysiwyg_content .carousel_module.medium_large .floating_text_area.open header .media_feature_title:after,.wysiwyg_content .carousel_module.large .floating_text_area.open header .media_feature_title:after,.wysiwyg_content .carousel_module.xlarge .floating_text_area.open header .media_feature_title:after,.wysiwyg_content .carousel_module.xxlarge .floating_text_area.open header .media_feature_title:after{transform:rotate(180deg)}.wysiwyg_content .carousel_module.small .floating_text_area,.wysiwyg_content .carousel_module.medium .floating_text_area,.wysiwyg_content .carousel_module.medium_mid .floating_text_area{background:linear-gradient(transparent, rgba(0,0,0,0.6));background-size:100%;background-repeat:no-repeat;background-position:bottom}.wysiwyg_content .carousel_module.small header,.wysiwyg_content .carousel_module.medium header,.wysiwyg_content .carousel_module.medium_mid header{cursor:pointer;margin-bottom:.4em}.wysiwyg_content .carousel_module.small .description,.wysiwyg_content .carousel_module.medium .description,.wysiwyg_content .carousel_module.medium_mid .description{display:none;cursor:pointer}.wysiwyg_content .carousel_module.small .media_feature_title:after,.wysiwyg_content .carousel_module.medium .media_feature_title:after,.wysiwyg_content .carousel_module.medium_mid .media_feature_title:after{display:none}.wysiwyg_content .carousel_module.small .gradient_container_bottom,.wysiwyg_content .carousel_module.medium .gradient_container_bottom,.wysiwyg_content .carousel_module.medium_mid .gradient_container_bottom{display:none}.wysiwyg_content .carousel_module .floating_text_area{width:100%;padding:2em 1.4em 2.8em;bottom:0;text-align:center;margin-left:auto;margin-right:auto;color:white}.wysiwyg_content .carousel_module.medium_large .floating_text_area,.wysiwyg_content .carousel_module.large .floating_text_area,.wysiwyg_content .carousel_module.xlarge .floating_text_area,.wysiwyg_content .carousel_module.xxlarge .floating_text_area{padding:1.4em;text-align:left;bottom:5em;background-color:black;width:auto;max-width:500px}.wysiwyg_content .carousel_module.medium_large .floating_text_area.left,.wysiwyg_content .carousel_module.medium_large .floating_text_area.bottom_left,.wysiwyg_content .carousel_module.large .floating_text_area.left,.wysiwyg_content .carousel_module.large .floating_text_area.bottom_left,.wysiwyg_content .carousel_module.xlarge .floating_text_area.left,.wysiwyg_content .carousel_module.xlarge .floating_text_area.bottom_left,.wysiwyg_content .carousel_module.xxlarge .floating_text_area.left,.wysiwyg_content .carousel_module.xxlarge .floating_text_area.bottom_left{left:5%}.wysiwyg_content .carousel_module.medium_large .floating_text_area.right,.wysiwyg_content .carousel_module.medium_large .floating_text_area.bottom_right,.wysiwyg_content .carousel_module.large .floating_text_area.right,.wysiwyg_content .carousel_module.large .floating_text_area.bottom_right,.wysiwyg_content .carousel_module.xlarge .floating_text_area.right,.wysiwyg_content .carousel_module.xlarge .floating_text_area.bottom_right,.wysiwyg_content .carousel_module.xxlarge .floating_text_area.right,.wysiwyg_content .carousel_module.xxlarge .floating_text_area.bottom_right{right:5%}.wysiwyg_content .carousel_module.medium_large .floating_text_area header,.wysiwyg_content .carousel_module.large .floating_text_area header,.wysiwyg_content .carousel_module.xlarge .floating_text_area header,.wysiwyg_content .carousel_module.xxlarge .floating_text_area header{padding:0}.wysiwyg_content .carousel_module.medium_large .floating_text_area header:after,.wysiwyg_content .carousel_module.large .floating_text_area header:after,.wysiwyg_content .carousel_module.xlarge .floating_text_area header:after,.wysiwyg_content .carousel_module.xxlarge .floating_text_area header:after{content:none}.wysiwyg_content .carousel_module .floating_text_area.open header:after{transform:rotate(180deg)}.wysiwyg_content .carousel_module .ms-nav-prev,.wysiwyg_content .carousel_module .ms-nav-next{margin-top:-40px}.wysiwyg_content .carousel_module .ms-slide-bgvideocont{background-color:#000}.wysiwyg_content .carousel_module .ms-slide-bgvideocont video{max-width:none}.wysiwyg_content .carousel_module .ms-nav-next,.wysiwyg_content .carousel_module .ms-nav-prev{display:none}.wysiwyg_content .carousel_module.medium_mid .ms-nav-next,.wysiwyg_content .carousel_module.medium_mid .ms-nav-prev,.wysiwyg_content .carousel_module.medium_large .ms-nav-next,.wysiwyg_content .carousel_module.medium_large .ms-nav-prev,.wysiwyg_content .carousel_module.large .ms-nav-next,.wysiwyg_content .carousel_module.large .ms-nav-prev,.wysiwyg_content .carousel_module.xlarge .ms-nav-next,.wysiwyg_content .carousel_module.xlarge .ms-nav-prev,.wysiwyg_content .carousel_module.xxlarge .ms-nav-next,.wysiwyg_content .carousel_module.xxlarge .ms-nav-prev{display:block}.no-touchevents .wysiwyg_content .carousel_module.medium .ms-nav-next,.no-touchevents .wysiwyg_content .carousel_module.medium .ms-nav-prev,.no-touchevents .wysiwyg_content .carousel_module.small .ms-nav-next,.no-touchevents .wysiwyg_content .carousel_module.small .ms-nav-prev{display:block}.wysiwyg_content .carousel_module .ms-nav-prev,.wysiwyg_content .carousel_module .ms-nav-next{width:40px;height:80px;margin-top:-60px}.wysiwyg_content .carousel_module.medium_large .ms-nav-next,.wysiwyg_content .carousel_module.medium_large .ms-nav-prev,.wysiwyg_content .carousel_module.large .ms-nav-next,.wysiwyg_content .carousel_module.large .ms-nav-prev,.wysiwyg_content .carousel_module.xlarge .ms-nav-next,.wysiwyg_content .carousel_module.xlarge .ms-nav-prev,.wysiwyg_content .carousel_module.xxlarge .ms-nav-next,.wysiwyg_content .carousel_module.xxlarge .ms-nav-prev{margin-top:-80px}.wysiwyg_content .carousel_module .ms-nav-prev{background:url(\"https://mars.nasa.gov/assets/arrow_left_darktheme.png\");background-size:40px 95px;background-color:rgba(32,32,32,0.9);background-position:0;left:0;border-top-right-radius:6px;border-bottom-right-radius:6px}.wysiwyg_content .carousel_module .ms-nav-next{background:url(\"https://mars.nasa.gov/assets/arrow_right_darktheme.png\");background-size:40px 95px;background-color:rgba(32,32,32,0.9);background-position:0;right:0;border-top-left-radius:6px;border-bottom-left-radius:6px}.wysiwyg_content .carousel_module .ms-bullets{left:0;right:0;margin:0 auto;bottom:1.2em;z-index:10}.wysiwyg_content .carousel_module.medium_mid .ms-bullets{bottom:1.5em}.wysiwyg_content .carousel_module.medium_large .ms-bullets,.wysiwyg_content .carousel_module.large .ms-bullets,.wysiwyg_content .carousel_module.xlarge .ms-bullets,.wysiwyg_content .carousel_module.xxlarge .ms-bullets{bottom:2.2em}.wysiwyg_content .carousel_module .ms-bullet{background-color:white;background-image:none;border-radius:50% 50% 50% 50%;height:8px;width:8px;opacity:0.5;margin:0 10px}.wysiwyg_content .carousel_module .ms-bullet:hover{opacity:1.0}.wysiwyg_content .carousel_module .ms-bullet-selected{opacity:1.0}.wysiwyg_content .carousel_module .ms-slide-layers{left:0 !important}.wysiwyg_content .carousel_module .ms-container,.wysiwyg_content .carousel_module .ms-slide-layers{max-width:none !important}.feature_pages .wysiwyg_content .carousel_module{width:94%;max-width:100%;margin:3em auto;float:none}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .carousel_module{max-width:600px}}.feature_pages .wysiwyg_content .carousel_module.full-bleed,.feature_pages .wysiwyg_content .carousel_module.full_width,.feature_pages .wysiwyg_content .carousel_module.wide,.feature_pages .wysiwyg_content .carousel_module.parallax{clear:both}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .carousel_module.full-bleed,.feature_pages .wysiwyg_content .carousel_module.full_width,.feature_pages .wysiwyg_content .carousel_module.wide,.feature_pages .wysiwyg_content .carousel_module.parallax{margin-top:5em;margin-bottom:5em}}.feature_pages .wysiwyg_content .carousel_module.column-width{max-width:94%;margin-top:3em;margin-bottom:3em;clear:both}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .carousel_module.column-width{max-width:600px}}.feature_pages .wysiwyg_content .carousel_module.full-bleed{width:100%;max-width:none}.feature_pages .wysiwyg_content .carousel_module.full-bleed figcaption{margin:.8em .8em 0 .8em}.feature_pages .wysiwyg_content .carousel_module.full_width{clear:both}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .carousel_module.full_width{width:94%;max-width:600px;margin-left:auto;margin-right:auto}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px), print and (min-width: 769px), print{.feature_pages .wysiwyg_content .carousel_module.full_width{width:80%}}@media (min-width: 769px) and (min-width: 1200px), print and (min-width: 1200px){.feature_pages .wysiwyg_content .carousel_module.full_width{width:55%}}.feature_pages .wysiwyg_content .carousel_module.wide{width:98%;max-width:none}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .carousel_module.wide{width:95%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .carousel_module.column-width{max-width:calc(600px + 6%)}}@media (min-width: 1024px), print{.feature_pages .wysiwyg_content .carousel_module.column-width{max-width:calc(600px + 10%)}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .carousel_module.column-width{max-width:calc(600px + 15%)}}.feature_pages .wysiwyg_content .carousel_module.left,.feature_pages .wysiwyg_content .carousel_module.right{max-width:94%}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .carousel_module.left,.feature_pages .wysiwyg_content .carousel_module.right{width:50%;max-width:50%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .carousel_module.left,.feature_pages .wysiwyg_content .carousel_module.right{width:27%;max-width:27%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .carousel_module.left,.feature_pages .wysiwyg_content .carousel_module.right{width:25%;max-width:25%}}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .carousel_module.left{float:left;margin:1em 2.5em 1.5em 0;margin-left:3%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .carousel_module.left{margin-left:15%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .carousel_module.left{margin-left:20%}}@media (min-width: 480px){.feature_pages .wysiwyg_content .carousel_module.right{float:right;margin:1em 0 1.5em 2.5em;margin-right:3%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .carousel_module.right{margin-right:15%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .carousel_module.right{margin-right:20%}}.feature_pages .wysiwyg_content .carousel_module.parallax_module{position:relative;overflow:hidden;z-index:10;padding-bottom:0;width:100%;max-width:none}.feature_pages .wysiwyg_content .carousel_module.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .carousel_module.parallax_module .caption{font-size:.88em}}.feature_pages .wysiwyg_content .carousel_module.parallax_module img{height:auto !important}.feature_pages .wysiwyg_content .carousel_module.parallax_module .window{width:100%;height:auto;position:absolute;overflow:hidden;padding:2em}.feature_pages .wysiwyg_content .carousel_module.parallax_module .window.mobile{height:auto;min-height:100%}.feature_pages .wysiwyg_content .carousel_module.parallax_module .window .featured_image{z-index:9;top:0;left:0;height:100%;overflow:hidden}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .carousel_module.parallax_module .window .featured_image{position:absolute}}.explore_overlay_page .full_width .carousel_module{width:94%;max-width:100%;margin:3em auto;float:none}@media (min-width: 600px), print{.explore_overlay_page .full_width .carousel_module{max-width:600px}}.explore_overlay_page .full_width .carousel_module.full-bleed,.explore_overlay_page .full_width .carousel_module.full_width,.explore_overlay_page .full_width .carousel_module.wide,.explore_overlay_page .full_width .carousel_module.parallax{clear:both}@media (min-width: 600px), print{.explore_overlay_page .full_width .carousel_module.full-bleed,.explore_overlay_page .full_width .carousel_module.full_width,.explore_overlay_page .full_width .carousel_module.wide,.explore_overlay_page .full_width .carousel_module.parallax{margin-top:5em;margin-bottom:5em}}.explore_overlay_page .full_width .carousel_module.column-width{max-width:94%;margin-top:3em;margin-bottom:3em;clear:both}@media (min-width: 600px), print{.explore_overlay_page .full_width .carousel_module.column-width{max-width:600px}}.explore_overlay_page .full_width .carousel_module.full-bleed{width:100%;max-width:none}.explore_overlay_page .full_width .carousel_module.full-bleed figcaption{margin:.8em .8em 0 .8em}.explore_overlay_page .full_width .carousel_module.full_width{clear:both}@media (min-width: 769px), print{.explore_overlay_page .full_width .carousel_module.full_width{width:94%;max-width:600px;margin-left:auto;margin-right:auto}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px), print and (min-width: 769px), print{.explore_overlay_page .full_width .carousel_module.full_width{width:80%}}@media (min-width: 769px) and (min-width: 1200px), print and (min-width: 1200px){.explore_overlay_page .full_width .carousel_module.full_width{width:55%}}.explore_overlay_page .full_width .carousel_module.wide{width:98%;max-width:none}@media (min-width: 769px), print{.explore_overlay_page .full_width .carousel_module.wide{width:95%}}@media (min-width: 769px), print{.explore_overlay_page .full_width .carousel_module.column-width{max-width:calc(600px + 6%)}}@media (min-width: 1024px), print{.explore_overlay_page .full_width .carousel_module.column-width{max-width:calc(600px + 10%)}}@media (min-width: 1200px){.explore_overlay_page .full_width .carousel_module.column-width{max-width:calc(600px + 15%)}}.explore_overlay_page .full_width .carousel_module.left,.explore_overlay_page .full_width .carousel_module.right{max-width:94%}@media (min-width: 600px), print{.explore_overlay_page .full_width .carousel_module.left,.explore_overlay_page .full_width .carousel_module.right{width:50%;max-width:50%}}@media (min-width: 769px), print{.explore_overlay_page .full_width .carousel_module.left,.explore_overlay_page .full_width .carousel_module.right{width:27%;max-width:27%}}@media (min-width: 1700px){.explore_overlay_page .full_width .carousel_module.left,.explore_overlay_page .full_width .carousel_module.right{width:25%;max-width:25%}}@media (min-width: 600px), print{.explore_overlay_page .full_width .carousel_module.left{float:left;margin:1em 2.5em 1.5em 0;margin-left:3%}}@media (min-width: 1200px){.explore_overlay_page .full_width .carousel_module.left{margin-left:15%}}@media (min-width: 1700px){.explore_overlay_page .full_width .carousel_module.left{margin-left:20%}}@media (min-width: 480px){.explore_overlay_page .full_width .carousel_module.right{float:right;margin:1em 0 1.5em 2.5em;margin-right:3%}}@media (min-width: 1200px){.explore_overlay_page .full_width .carousel_module.right{margin-right:15%}}@media (min-width: 1700px){.explore_overlay_page .full_width .carousel_module.right{margin-right:20%}}.explore_overlay_page .full_width .carousel_module.parallax_module{position:relative;overflow:hidden;z-index:10;padding-bottom:0;width:100%;max-width:none}.explore_overlay_page .full_width .carousel_module.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.explore_overlay_page .full_width .carousel_module.parallax_module .caption{font-size:.88em}}.explore_overlay_page .full_width .carousel_module.parallax_module img{height:auto !important}.explore_overlay_page .full_width .carousel_module.parallax_module .window{width:100%;height:auto;position:absolute;overflow:hidden;padding:2em}.explore_overlay_page .full_width .carousel_module.parallax_module .window.mobile{height:auto;min-height:100%}.explore_overlay_page .full_width .carousel_module.parallax_module .window .featured_image{z-index:9;top:0;left:0;height:100%;overflow:hidden}@media (min-width: 769px), print{.explore_overlay_page .full_width .carousel_module.parallax_module .window .featured_image{position:absolute}}.explore_overlay_page .full_width .carousel_module.wide{width:98%;max-width:none}@media (min-width: 769px), print{.explore_overlay_page .full_width .carousel_module.wide{width:95%;max-width:1200px}}.wysiwyg_content .image_module{max-width:100%;margin-top:1.4em;margin-bottom:1.4em}@media (min-width: 769px), print{.wysiwyg_content .image_module{margin-top:2em;margin-bottom:2em}}.wysiwyg_content .image_module.left,.wysiwyg_content .image_module.right{float:none}@media (min-width: 480px){.wysiwyg_content .image_module.left,.wysiwyg_content .image_module.right{max-width:50%}}@media (min-width: 1200px){.wysiwyg_content .image_module.left,.wysiwyg_content .image_module.right{max-width:40%}}@media (min-width: 480px){.wysiwyg_content .image_module.left{float:left;margin:1em 2.5em 1.5em 0}}@media (min-width: 480px){.wysiwyg_content .image_module.right{float:right;margin:1em 0 1.5em 2.5em}}.wysiwyg_content .image_module.full-bleed,.wysiwyg_content .image_module.full_width,.wysiwyg_content .image_module.wide,.wysiwyg_content .image_module.parallax,.wysiwyg_content .image_module.column-width{clear:both}.wysiwyg_content .image_module.parallax_module{width:100%;position:relative}.wysiwyg_content .image_module.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.wysiwyg_content .image_module.parallax_module .caption{font-size:.88em}}.explore_overlay_page .wysiwyg_content .image_module.parallax_module .caption{color:#b0b4b9}.feature_pages .wysiwyg_content .image_module{width:94%;max-width:100%;margin:3em auto;float:none}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .image_module{max-width:600px}}.feature_pages .wysiwyg_content .image_module.full-bleed,.feature_pages .wysiwyg_content .image_module.full_width,.feature_pages .wysiwyg_content .image_module.wide,.feature_pages .wysiwyg_content .image_module.parallax{clear:both}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .image_module.full-bleed,.feature_pages .wysiwyg_content .image_module.full_width,.feature_pages .wysiwyg_content .image_module.wide,.feature_pages .wysiwyg_content .image_module.parallax{margin-top:5em;margin-bottom:5em}}.feature_pages .wysiwyg_content .image_module.column-width{max-width:94%;margin-top:3em;margin-bottom:3em;clear:both}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .image_module.column-width{max-width:600px}}.feature_pages .wysiwyg_content .image_module.full-bleed{width:100%;max-width:none}.feature_pages .wysiwyg_content .image_module.full-bleed figcaption{margin:.8em .8em 0 .8em}.feature_pages .wysiwyg_content .image_module.full_width{clear:both}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .image_module.full_width{width:94%;max-width:600px;margin-left:auto;margin-right:auto}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px), print and (min-width: 769px), print{.feature_pages .wysiwyg_content .image_module.full_width{width:80%}}@media (min-width: 769px) and (min-width: 1200px), print and (min-width: 1200px){.feature_pages .wysiwyg_content .image_module.full_width{width:55%}}.feature_pages .wysiwyg_content .image_module.wide{width:98%;max-width:none}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .image_module.wide{width:95%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .image_module.column-width{max-width:calc(600px + 6%)}}@media (min-width: 1024px), print{.feature_pages .wysiwyg_content .image_module.column-width{max-width:calc(600px + 10%)}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .image_module.column-width{max-width:calc(600px + 15%)}}.feature_pages .wysiwyg_content .image_module.left,.feature_pages .wysiwyg_content .image_module.right{max-width:94%}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .image_module.left,.feature_pages .wysiwyg_content .image_module.right{width:50%;max-width:50%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .image_module.left,.feature_pages .wysiwyg_content .image_module.right{width:27%;max-width:27%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .image_module.left,.feature_pages .wysiwyg_content .image_module.right{width:25%;max-width:25%}}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .image_module.left{float:left;margin:1em 2.5em 1.5em 0;margin-left:3%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .image_module.left{margin-left:15%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .image_module.left{margin-left:20%}}@media (min-width: 480px){.feature_pages .wysiwyg_content .image_module.right{float:right;margin:1em 0 1.5em 2.5em;margin-right:3%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .image_module.right{margin-right:15%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .image_module.right{margin-right:20%}}.feature_pages .wysiwyg_content .image_module.parallax_module{position:relative;overflow:hidden;z-index:10;padding-bottom:0;width:100%;max-width:none}.feature_pages .wysiwyg_content .image_module.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .image_module.parallax_module .caption{font-size:.88em}}.feature_pages .wysiwyg_content .image_module.parallax_module img{height:auto !important}.feature_pages .wysiwyg_content .image_module.parallax_module .window{width:100%;height:auto;position:absolute;overflow:hidden;padding:2em}.feature_pages .wysiwyg_content .image_module.parallax_module .window.mobile{height:auto;min-height:100%}.feature_pages .wysiwyg_content .image_module.parallax_module .window .featured_image{z-index:9;top:0;left:0;height:100%;overflow:hidden}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .image_module.parallax_module .window .featured_image{position:absolute}}.image_module figure.inline_figure{margin-bottom:0}.item_grid_module{clear:both;margin:3em auto}@media (min-width: 769px), print{.item_grid_module{margin:4em auto}}.item_grid_module li{visibility:hidden}.item_grid_module .grid_content{margin-bottom:20px}.item_grid_module .grid_content .grid_image{text-align:center}@media (min-width: 600px), print{.item_grid_module .grid_content .grid_image{text-align:left}}.item_grid_module .grid_content .grid_image img{width:auto}@media (min-width: 600px), print{.item_grid_module .grid_content .grid_image img{width:100%}}.item_grid_module .grid_content .caption{font-size:.88em;padding:1em;color:#565455}@media (min-width: 600px), print{.item_grid_module .grid_content .caption{background-color:#D8D6D7}}.feature_pages .wysiwyg_content .item_grid_module{max-width:none;width:90% !important;left:5%;margin-left:auto;margin-right:auto}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .item_grid_module{width:98% !important;left:1%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .item_grid_module{width:95% !important;left:2.5%}}@media (min-width: 1024px), print{.feature_pages .wysiwyg_content .item_grid_module{width:85% !important;left:0}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .item_grid_module{width:75% !important}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .item_grid_module{width:65% !important}}.explore_overlay_page .item_grid_module .caption{background-color:transparent;color:#9C9FA4}.grid_gallery.list_view .item_grid{display:none}.grid_gallery.masonry_view .item_list{display:none}.grid_gallery.masonry_view .grid_text .caption{background-color:#e5ecf4}.grid_gallery.masonry_view .detail_link_button{position:absolute;top:50%;left:50%;transform:translate(-50%, -50%);padding:0.7em 1.5em;border:2px solid white;border-radius:12px;font-size:0.9em;font-weight:bold}.view_selectors .nav_item.masonry_icon{background-position:-62px -62px}.no-touchevents .view_selectors .nav_item.masonry_icon:hover{background-position:-62px -12px}.masonry_view .view_selectors .nav_item.masonry_icon{background-position:-62px -12px}blockquote{clear:both;color:#000}blockquote .quote{font-style:italic;margin-bottom:.5em;font-size:1.5em;line-height:1.4em;font-weight:700}blockquote footer{font-size:1em;text-align:left}blockquote cite{font-style:normal}.explore_overlay_page blockquote{color:#FFF !important}.explore_overlay_page blockquote.inspirational{color:#FFF}.feature_pages .wysiwyg_content blockquote{width:80%}@media (min-width: 769px), print{.feature_pages .wysiwyg_content blockquote{width:65%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content blockquote{width:40%}}.wysiwyg_content .video_player_container,.wysiwyg_content figure.embedded_video{max-width:100%;margin-top:1.4em;margin-bottom:1.4em}@media (min-width: 769px), print{.wysiwyg_content .video_player_container,.wysiwyg_content figure.embedded_video{margin-top:2em;margin-bottom:2em}}.wysiwyg_content .video_player_container.left,.wysiwyg_content .video_player_container.right,.wysiwyg_content figure.embedded_video.left,.wysiwyg_content figure.embedded_video.right{float:none}@media (min-width: 480px){.wysiwyg_content .video_player_container.left,.wysiwyg_content .video_player_container.right,.wysiwyg_content figure.embedded_video.left,.wysiwyg_content figure.embedded_video.right{max-width:50%}}@media (min-width: 1200px){.wysiwyg_content .video_player_container.left,.wysiwyg_content .video_player_container.right,.wysiwyg_content figure.embedded_video.left,.wysiwyg_content figure.embedded_video.right{max-width:40%}}@media (min-width: 480px){.wysiwyg_content .video_player_container.left,.wysiwyg_content figure.embedded_video.left{float:left;margin:1em 2.5em 1.5em 0}}@media (min-width: 480px){.wysiwyg_content .video_player_container.right,.wysiwyg_content figure.embedded_video.right{float:right;margin:1em 0 1.5em 2.5em}}.wysiwyg_content .video_player_container.full-bleed,.wysiwyg_content .video_player_container.full_width,.wysiwyg_content .video_player_container.wide,.wysiwyg_content .video_player_container.parallax,.wysiwyg_content .video_player_container.column-width,.wysiwyg_content figure.embedded_video.full-bleed,.wysiwyg_content figure.embedded_video.full_width,.wysiwyg_content figure.embedded_video.wide,.wysiwyg_content figure.embedded_video.parallax,.wysiwyg_content figure.embedded_video.column-width{clear:both}.wysiwyg_content .video_player_container.parallax_module,.wysiwyg_content figure.embedded_video.parallax_module{width:100%;position:relative}.wysiwyg_content .video_player_container.parallax_module .caption,.wysiwyg_content figure.embedded_video.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.wysiwyg_content .video_player_container.parallax_module .caption,.wysiwyg_content figure.embedded_video.parallax_module .caption{font-size:.88em}}.explore_overlay_page .wysiwyg_content .video_player_container.parallax_module .caption,.explore_overlay_page .wysiwyg_content figure.embedded_video.parallax_module .caption{color:#b0b4b9}.feature_pages .wysiwyg_content .video_player_container,.feature_pages .wysiwyg_content figure.embedded_video{width:94%;max-width:100%;margin:3em auto;float:none}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .video_player_container,.feature_pages .wysiwyg_content figure.embedded_video{max-width:600px}}.feature_pages .wysiwyg_content .video_player_container.full-bleed,.feature_pages .wysiwyg_content .video_player_container.full_width,.feature_pages .wysiwyg_content .video_player_container.wide,.feature_pages .wysiwyg_content .video_player_container.parallax,.feature_pages .wysiwyg_content figure.embedded_video.full-bleed,.feature_pages .wysiwyg_content figure.embedded_video.full_width,.feature_pages .wysiwyg_content figure.embedded_video.wide,.feature_pages .wysiwyg_content figure.embedded_video.parallax{clear:both}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .video_player_container.full-bleed,.feature_pages .wysiwyg_content .video_player_container.full_width,.feature_pages .wysiwyg_content .video_player_container.wide,.feature_pages .wysiwyg_content .video_player_container.parallax,.feature_pages .wysiwyg_content figure.embedded_video.full-bleed,.feature_pages .wysiwyg_content figure.embedded_video.full_width,.feature_pages .wysiwyg_content figure.embedded_video.wide,.feature_pages .wysiwyg_content figure.embedded_video.parallax{margin-top:5em;margin-bottom:5em}}.feature_pages .wysiwyg_content .video_player_container.column-width,.feature_pages .wysiwyg_content figure.embedded_video.column-width{max-width:94%;margin-top:3em;margin-bottom:3em;clear:both}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .video_player_container.column-width,.feature_pages .wysiwyg_content figure.embedded_video.column-width{max-width:600px}}.feature_pages .wysiwyg_content .video_player_container.full-bleed,.feature_pages .wysiwyg_content figure.embedded_video.full-bleed{width:100%;max-width:none}.feature_pages .wysiwyg_content .video_player_container.full-bleed figcaption,.feature_pages .wysiwyg_content figure.embedded_video.full-bleed figcaption{margin:.8em .8em 0 .8em}.feature_pages .wysiwyg_content .video_player_container.full_width,.feature_pages .wysiwyg_content figure.embedded_video.full_width{clear:both}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .video_player_container.full_width,.feature_pages .wysiwyg_content figure.embedded_video.full_width{width:94%;max-width:600px;margin-left:auto;margin-right:auto}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px), print and (min-width: 769px), print{.feature_pages .wysiwyg_content .video_player_container.full_width,.feature_pages .wysiwyg_content figure.embedded_video.full_width{width:80%}}@media (min-width: 769px) and (min-width: 1200px), print and (min-width: 1200px){.feature_pages .wysiwyg_content .video_player_container.full_width,.feature_pages .wysiwyg_content figure.embedded_video.full_width{width:55%}}.feature_pages .wysiwyg_content .video_player_container.wide,.feature_pages .wysiwyg_content figure.embedded_video.wide{width:98%;max-width:none}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .video_player_container.wide,.feature_pages .wysiwyg_content figure.embedded_video.wide{width:95%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .video_player_container.column-width,.feature_pages .wysiwyg_content figure.embedded_video.column-width{max-width:calc(600px + 6%)}}@media (min-width: 1024px), print{.feature_pages .wysiwyg_content .video_player_container.column-width,.feature_pages .wysiwyg_content figure.embedded_video.column-width{max-width:calc(600px + 10%)}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .video_player_container.column-width,.feature_pages .wysiwyg_content figure.embedded_video.column-width{max-width:calc(600px + 15%)}}.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.left,.feature_pages .wysiwyg_content figure.embedded_video.right{max-width:94%}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.left,.feature_pages .wysiwyg_content figure.embedded_video.right{width:50%;max-width:50%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.left,.feature_pages .wysiwyg_content figure.embedded_video.right{width:27%;max-width:27%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.left,.feature_pages .wysiwyg_content figure.embedded_video.right{width:25%;max-width:25%}}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content figure.embedded_video.left{float:left;margin:1em 2.5em 1.5em 0;margin-left:3%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content figure.embedded_video.left{margin-left:15%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content figure.embedded_video.left{margin-left:20%}}@media (min-width: 480px){.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.right{float:right;margin:1em 0 1.5em 2.5em;margin-right:3%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.right{margin-right:15%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.right{margin-right:20%}}.feature_pages .wysiwyg_content .video_player_container.parallax_module,.feature_pages .wysiwyg_content figure.embedded_video.parallax_module{position:relative;overflow:hidden;z-index:10;padding-bottom:0;width:100%;max-width:none}.feature_pages .wysiwyg_content .video_player_container.parallax_module .caption,.feature_pages .wysiwyg_content figure.embedded_video.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .video_player_container.parallax_module .caption,.feature_pages .wysiwyg_content figure.embedded_video.parallax_module .caption{font-size:.88em}}.feature_pages .wysiwyg_content .video_player_container.parallax_module img,.feature_pages .wysiwyg_content figure.embedded_video.parallax_module img{height:auto !important}.feature_pages .wysiwyg_content .video_player_container.parallax_module .window,.feature_pages .wysiwyg_content figure.embedded_video.parallax_module .window{width:100%;height:auto;position:absolute;overflow:hidden;padding:2em}.feature_pages .wysiwyg_content .video_player_container.parallax_module .window.mobile,.feature_pages .wysiwyg_content figure.embedded_video.parallax_module .window.mobile{height:auto;min-height:100%}.feature_pages .wysiwyg_content .video_player_container.parallax_module .window .featured_image,.feature_pages .wysiwyg_content figure.embedded_video.parallax_module .window .featured_image{z-index:9;top:0;left:0;height:100%;overflow:hidden}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .video_player_container.parallax_module .window .featured_image,.feature_pages .wysiwyg_content figure.embedded_video.parallax_module .window .featured_image{position:absolute}}@media (min-width: 600px), print{.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.left,.feature_pages .wysiwyg_content figure.embedded_video.right{width:50%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.left,.feature_pages .wysiwyg_content figure.embedded_video.right{width:44%}}@media (min-width: 1024px), print{.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.left,.feature_pages .wysiwyg_content figure.embedded_video.right{width:33%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .video_player_container.left,.feature_pages .wysiwyg_content .video_player_container.right,.feature_pages .wysiwyg_content figure.embedded_video.left,.feature_pages .wysiwyg_content figure.embedded_video.right{width:25%}}.feature_pages .wysiwyg_content .video_player_container.full-width>iframe,.feature_pages .wysiwyg_content .video_player_container.full-bleed>iframe,.feature_pages .wysiwyg_content .video_player_container.wide>iframe,.feature_pages .wysiwyg_content figure.embedded_video.full-width>iframe,.feature_pages .wysiwyg_content figure.embedded_video.full-bleed>iframe,.feature_pages .wysiwyg_content figure.embedded_video.wide>iframe{max-height:98vh}@media (min-height: 400px){.feature_pages .wysiwyg_content .video_player_container.full-width>iframe,.feature_pages .wysiwyg_content .video_player_container.full-bleed>iframe,.feature_pages .wysiwyg_content .video_player_container.wide>iframe,.feature_pages .wysiwyg_content figure.embedded_video.full-width>iframe,.feature_pages .wysiwyg_content figure.embedded_video.full-bleed>iframe,.feature_pages .wysiwyg_content figure.embedded_video.wide>iframe{max-height:300px}}@media (min-height: 600px){.feature_pages .wysiwyg_content .video_player_container.full-width>iframe,.feature_pages .wysiwyg_content .video_player_container.full-bleed>iframe,.feature_pages .wysiwyg_content .video_player_container.wide>iframe,.feature_pages .wysiwyg_content figure.embedded_video.full-width>iframe,.feature_pages .wysiwyg_content figure.embedded_video.full-bleed>iframe,.feature_pages .wysiwyg_content figure.embedded_video.wide>iframe{max-height:400px}}@media (min-height: 800px){.feature_pages .wysiwyg_content .video_player_container.full-width>iframe,.feature_pages .wysiwyg_content .video_player_container.full-bleed>iframe,.feature_pages .wysiwyg_content .video_player_container.wide>iframe,.feature_pages .wysiwyg_content figure.embedded_video.full-width>iframe,.feature_pages .wysiwyg_content figure.embedded_video.full-bleed>iframe,.feature_pages .wysiwyg_content figure.embedded_video.wide>iframe{max-height:600px}}@media (min-height: 1000px){.feature_pages .wysiwyg_content .video_player_container.full-width>iframe,.feature_pages .wysiwyg_content .video_player_container.full-bleed>iframe,.feature_pages .wysiwyg_content .video_player_container.wide>iframe,.feature_pages .wysiwyg_content figure.embedded_video.full-width>iframe,.feature_pages .wysiwyg_content figure.embedded_video.full-bleed>iframe,.feature_pages .wysiwyg_content figure.embedded_video.wide>iframe{max-height:90vh}}.wysiwyg_content figure.embedded_video>iframe{width:100%;min-height:300px}@media (min-width: 769px), print{.wysiwyg_content figure.embedded_video>iframe{min-height:400px}}@media (min-width: 1200px){.wysiwyg_content figure.embedded_video>iframe{min-height:500px}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content .embedded_video iframe.webvr_module{min-height:400px}}@media (min-width: 1024px), print{.feature_pages .wysiwyg_content .embedded_video iframe.webvr_module{min-height:500px}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .embedded_video iframe.webvr_module{min-height:600px}}@media (min-width: 1700px){.feature_pages .wysiwyg_content .embedded_video iframe.webvr_module{min-height:700px}}.content_page:not(.feature_pages) .wysiwyg_content figure.embedded_video.left,.content_page:not(.feature_pages) .wysiwyg_content figure.embedded_video.right{width:100%;max-width:100%}.content_page:not(.feature_pages) .wysiwyg_content figure.embedded_video.left>iframe,.content_page:not(.feature_pages) .wysiwyg_content figure.embedded_video.right>iframe{min-height:300px}@media (min-width: 769px), print{.content_page:not(.feature_pages) .wysiwyg_content figure.embedded_video.left,.content_page:not(.feature_pages) .wysiwyg_content figure.embedded_video.right{width:50%}}.content_page:not(.feature_pages) .wysiwyg_content .video_player_container.left,.content_page:not(.feature_pages) .wysiwyg_content .video_player_container.right{width:100%;max-width:100%}@media (min-width: 769px), print{.content_page:not(.feature_pages) .wysiwyg_content .video_player_container.left,.content_page:not(.feature_pages) .wysiwyg_content .video_player_container.right{width:50%}}.webvr-button{padding:0px !important;margin:12px}img[title=\"Configure viewer\"]{margin-left:-12px !important}.feature_pages .wysiwyg_content #scene_container{width:94%;max-width:100%;margin:3em auto;float:none;position:relative}@media (min-width: 600px), print{.feature_pages .wysiwyg_content #scene_container{max-width:600px}}.feature_pages .wysiwyg_content #scene_container.full-bleed,.feature_pages .wysiwyg_content #scene_container.full_width,.feature_pages .wysiwyg_content #scene_container.wide,.feature_pages .wysiwyg_content #scene_container.parallax{clear:both}@media (min-width: 600px), print{.feature_pages .wysiwyg_content #scene_container.full-bleed,.feature_pages .wysiwyg_content #scene_container.full_width,.feature_pages .wysiwyg_content #scene_container.wide,.feature_pages .wysiwyg_content #scene_container.parallax{margin-top:5em;margin-bottom:5em}}.feature_pages .wysiwyg_content #scene_container.column-width{max-width:94%;margin-top:3em;margin-bottom:3em;clear:both}@media (min-width: 600px), print{.feature_pages .wysiwyg_content #scene_container.column-width{max-width:600px}}.feature_pages .wysiwyg_content #scene_container.full-bleed{width:100%;max-width:none}.feature_pages .wysiwyg_content #scene_container.full-bleed figcaption{margin:.8em .8em 0 .8em}.feature_pages .wysiwyg_content #scene_container.full_width{clear:both}@media (min-width: 769px), print{.feature_pages .wysiwyg_content #scene_container.full_width{width:94%;max-width:600px;margin-left:auto;margin-right:auto}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px), print and (min-width: 769px), print{.feature_pages .wysiwyg_content #scene_container.full_width{width:80%}}@media (min-width: 769px) and (min-width: 1200px), print and (min-width: 1200px){.feature_pages .wysiwyg_content #scene_container.full_width{width:55%}}.feature_pages .wysiwyg_content #scene_container.wide{width:98%;max-width:none}@media (min-width: 769px), print{.feature_pages .wysiwyg_content #scene_container.wide{width:95%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content #scene_container.column-width{max-width:calc(600px + 6%)}}@media (min-width: 1024px), print{.feature_pages .wysiwyg_content #scene_container.column-width{max-width:calc(600px + 10%)}}@media (min-width: 1200px){.feature_pages .wysiwyg_content #scene_container.column-width{max-width:calc(600px + 15%)}}.feature_pages .wysiwyg_content #scene_container.left,.feature_pages .wysiwyg_content #scene_container.right{max-width:94%}@media (min-width: 600px), print{.feature_pages .wysiwyg_content #scene_container.left,.feature_pages .wysiwyg_content #scene_container.right{width:50%;max-width:50%}}@media (min-width: 769px), print{.feature_pages .wysiwyg_content #scene_container.left,.feature_pages .wysiwyg_content #scene_container.right{width:27%;max-width:27%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content #scene_container.left,.feature_pages .wysiwyg_content #scene_container.right{width:25%;max-width:25%}}@media (min-width: 600px), print{.feature_pages .wysiwyg_content #scene_container.left{float:left;margin:1em 2.5em 1.5em 0;margin-left:3%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content #scene_container.left{margin-left:15%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content #scene_container.left{margin-left:20%}}@media (min-width: 480px){.feature_pages .wysiwyg_content #scene_container.right{float:right;margin:1em 0 1.5em 2.5em;margin-right:3%}}@media (min-width: 1200px){.feature_pages .wysiwyg_content #scene_container.right{margin-right:15%}}@media (min-width: 1700px){.feature_pages .wysiwyg_content #scene_container.right{margin-right:20%}}.feature_pages .wysiwyg_content #scene_container.parallax_module{position:relative;overflow:hidden;z-index:10;padding-bottom:0;width:100%;max-width:none}.feature_pages .wysiwyg_content #scene_container.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.feature_pages .wysiwyg_content #scene_container.parallax_module .caption{font-size:.88em}}.feature_pages .wysiwyg_content #scene_container.parallax_module img{height:auto !important}.feature_pages .wysiwyg_content #scene_container.parallax_module .window{width:100%;height:auto;position:absolute;overflow:hidden;padding:2em}.feature_pages .wysiwyg_content #scene_container.parallax_module .window.mobile{height:auto;min-height:100%}.feature_pages .wysiwyg_content #scene_container.parallax_module .window .featured_image{z-index:9;top:0;left:0;height:100%;overflow:hidden}@media (min-width: 769px), print{.feature_pages .wysiwyg_content #scene_container.parallax_module .window .featured_image{position:absolute}}.feature_pages .wysiwyg_content #scene_container:before{display:block;content:\"\";width:100%;padding-top:75%}.feature_pages .wysiwyg_content #scene_container>.content{position:absolute;top:0;left:0;right:0;bottom:0}.content_page:not(.feature_pages) .wysiwyg_content #scene_container{max-width:100%;margin-top:1.4em;margin-bottom:1.4em;position:relative}@media (min-width: 769px), print{.content_page:not(.feature_pages) .wysiwyg_content #scene_container{margin-top:2em;margin-bottom:2em}}.content_page:not(.feature_pages) .wysiwyg_content #scene_container.left,.content_page:not(.feature_pages) .wysiwyg_content #scene_container.right{float:none}@media (min-width: 480px){.content_page:not(.feature_pages) .wysiwyg_content #scene_container.left,.content_page:not(.feature_pages) .wysiwyg_content #scene_container.right{max-width:50%}}@media (min-width: 1200px){.content_page:not(.feature_pages) .wysiwyg_content #scene_container.left,.content_page:not(.feature_pages) .wysiwyg_content #scene_container.right{max-width:40%}}@media (min-width: 480px){.content_page:not(.feature_pages) .wysiwyg_content #scene_container.left{float:left;margin:1em 2.5em 1.5em 0}}@media (min-width: 480px){.content_page:not(.feature_pages) .wysiwyg_content #scene_container.right{float:right;margin:1em 0 1.5em 2.5em}}.content_page:not(.feature_pages) .wysiwyg_content #scene_container.full-bleed,.content_page:not(.feature_pages) .wysiwyg_content #scene_container.full_width,.content_page:not(.feature_pages) .wysiwyg_content #scene_container.wide,.content_page:not(.feature_pages) .wysiwyg_content #scene_container.parallax,.content_page:not(.feature_pages) .wysiwyg_content #scene_container.column-width{clear:both}.content_page:not(.feature_pages) .wysiwyg_content #scene_container.parallax_module{width:100%;position:relative}.content_page:not(.feature_pages) .wysiwyg_content #scene_container.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.content_page:not(.feature_pages) .wysiwyg_content #scene_container.parallax_module .caption{font-size:.88em}}.explore_overlay_page .content_page:not(.feature_pages) .wysiwyg_content #scene_container.parallax_module .caption{color:#b0b4b9}.content_page:not(.feature_pages) .wysiwyg_content #scene_container:before{display:block;content:\"\";width:100%;padding-top:75%}.content_page:not(.feature_pages) .wysiwyg_content #scene_container>.content{position:absolute;top:0;left:0;right:0;bottom:0}.content_page.atlas_detail #scene_container{max-width:100%;margin-top:1.4em;margin-bottom:1.4em;position:relative}@media (min-width: 769px), print{.content_page.atlas_detail #scene_container{margin-top:2em;margin-bottom:2em}}.content_page.atlas_detail #scene_container.left,.content_page.atlas_detail #scene_container.right{float:none}@media (min-width: 480px){.content_page.atlas_detail #scene_container.left,.content_page.atlas_detail #scene_container.right{max-width:50%}}@media (min-width: 1200px){.content_page.atlas_detail #scene_container.left,.content_page.atlas_detail #scene_container.right{max-width:40%}}@media (min-width: 480px){.content_page.atlas_detail #scene_container.left{float:left;margin:1em 2.5em 1.5em 0}}@media (min-width: 480px){.content_page.atlas_detail #scene_container.right{float:right;margin:1em 0 1.5em 2.5em}}.content_page.atlas_detail #scene_container.full-bleed,.content_page.atlas_detail #scene_container.full_width,.content_page.atlas_detail #scene_container.wide,.content_page.atlas_detail #scene_container.parallax,.content_page.atlas_detail #scene_container.column-width{clear:both}.content_page.atlas_detail #scene_container.parallax_module{width:100%;position:relative}.content_page.atlas_detail #scene_container.parallax_module .caption{margin:.8em .8em 0 .8em;font-size:.8em;color:#5a6470}@media (min-width: 769px), print{.content_page.atlas_detail #scene_container.parallax_module .caption{font-size:.88em}}.explore_overlay_page .content_page.atlas_detail #scene_container.parallax_module .caption{color:#b0b4b9}.content_page.atlas_detail #scene_container:before{display:block;content:\"\";width:100%;padding-top:36.36364%}.content_page.atlas_detail #scene_container>.content{position:absolute;top:0;left:0;right:0;bottom:0}.magic_shell_page .wysiwyg_content #scene_container{height:100%}#scene_container{position:relative;overflow:hidden;width:100%}#scene_container .loading_webgl{width:100%;height:100%;top:0;left:0;right:0;bottom:0;position:absolute;z-index:21;opacity:1.0;color:white;background-color:black;text-align:center;-webkit-transition:opacity 2s, visibility 2s;-moz-transition:opacity 2s, visibility 2s;transition:opacity 2s, visibility 2s}#scene_container .loading_webgl .loading_text{position:relative;top:50%;padding-bottom:6px;color:grey}#scene_container .loading_webgl .load_bar{position:relative;top:50%;left:0;width:0%;height:2px;background-color:#FFFFFF;opacity:0.5}#gl_layer{margin:0 auto}#gl_layer,#gl_fallback_layer{-webkit-tap-highlight-color:transparent;-webkit-tap-highlight-color:transparent;position:absolute;overflow:hidden;top:0;left:0;right:0;bottom:0;width:100%;height:100%;display:none}#gl_layer .module_title_wrapper,#gl_fallback_layer .module_title_wrapper{position:absolute;padding-top:8px;top:-120px;left:50%;width:90%;text-align:center;visibility:hidden;-webkit-transform:translateX(-50%);-ms-transform:translateX(-50%);transform:translateX(-50%);-webkit-transition:top 1s, visibility 1s;-moz-transition:top 1s, visibility 1s;transition:top 1s, visibility 1s;z-index:20}@media (min-width: 480px){#gl_layer .module_title_wrapper,#gl_fallback_layer .module_title_wrapper{padding-top:26px}}@media only screen and (min-width: 480px) and (orientation: landscape){#gl_layer .module_title_wrapper,#gl_fallback_layer .module_title_wrapper{padding-top:0px}}@media (min-width: 600px), print{#gl_layer .module_title_wrapper,#gl_fallback_layer .module_title_wrapper{padding-top:34px}}@media only screen and (min-width: 600px) and (orientation: landscape){#gl_layer .module_title_wrapper,#gl_fallback_layer .module_title_wrapper{padding-top:4px}}@media (min-width: 769px), print{#gl_layer .module_title_wrapper,#gl_fallback_layer .module_title_wrapper{padding-top:34px}}#gl_layer .module_title_wrapper.active,#gl_fallback_layer .module_title_wrapper.active{top:0px;visibility:visible}#gl_layer .module_title_wrapper .gradient_overlay_top,#gl_fallback_layer .module_title_wrapper .gradient_overlay_top{position:absolute;width:120%;height:250%;top:0;left:50%;opacity:.8;background:-moz-linear-gradient(top, #000 0%, #000 14%, transparent 99%, transparent 100%);background:-webkit-gradient(linear, left top, left bottom, color-stop(0%, #000), color-stop(14%, #000), color-stop(99%, transparent), color-stop(100%, transparent));background:-webkit-linear-gradient(top, #000 0%, #000 14%, transparent 99%, transparent 100%);background:-o-linear-gradient(top, #000 0%, #000 14%, transparent 99%, transparent 100%);background:-ms-linear-gradient(top, #000 0%, #000 14%, transparent 99%, transparent 100%);background:linear-gradient(to bottom, #000 0%, #000 14%, transparent 99%, transparent 100%);filter:progid:DXImageTransform.Microsoft.gradient( startColorstr='#000000', endColorstr='#00000000',GradientType=0 );-webkit-transform:translateX(-50%);-ms-transform:translateX(-50%);transform:translateX(-50%)}#gl_layer .module_title_wrapper .module_title,#gl_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_layer .module_title_wrapper .carousel_title,#gl_fallback_layer .module_title_wrapper .module_title,#gl_fallback_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_fallback_layer .module_title_wrapper .carousel_title{font-size:20px;margin-right:0.4em;position:relative;color:#FFFFFF;font-family:Whitney,Helvetica,Arial,sans-serif;vertical-align:middle}@media (min-width: 480px){#gl_layer .module_title_wrapper .module_title,#gl_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_layer .module_title_wrapper .carousel_title,#gl_fallback_layer .module_title_wrapper .module_title,#gl_fallback_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_fallback_layer .module_title_wrapper .carousel_title{font-size:22px}}@media (min-width: 600px), print{#gl_layer .module_title_wrapper .module_title,#gl_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_layer .module_title_wrapper .carousel_title,#gl_fallback_layer .module_title_wrapper .module_title,#gl_fallback_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_fallback_layer .module_title_wrapper .carousel_title{font-size:34px}}@media only screen and (min-width: 600px) and (orientation: landscape){#gl_layer .module_title_wrapper .module_title,#gl_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_layer .module_title_wrapper .carousel_title,#gl_fallback_layer .module_title_wrapper .module_title,#gl_fallback_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_fallback_layer .module_title_wrapper .carousel_title{font-size:24px}}@media (min-width: 769px), print{#gl_layer .module_title_wrapper .module_title,#gl_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_layer .module_title_wrapper .carousel_title,#gl_fallback_layer .module_title_wrapper .module_title,#gl_fallback_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_fallback_layer .module_title_wrapper .carousel_title{font-size:28px}}@media (min-width: 1024px), print{#gl_layer .module_title_wrapper .module_title,#gl_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_layer .module_title_wrapper .carousel_title,#gl_fallback_layer .module_title_wrapper .module_title,#gl_fallback_layer .module_title_wrapper .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header #gl_fallback_layer .module_title_wrapper .carousel_title{font-size:34px}}#gl_layer .module_title_wrapper a.close_focus,#gl_fallback_layer .module_title_wrapper a.close_focus{cursor:pointer;position:relative;display:inline-block;padding:.1em;width:26px;height:26px;vertical-align:middle}@media (min-width: 480px){#gl_layer .module_title_wrapper a.close_focus,#gl_fallback_layer .module_title_wrapper a.close_focus{width:28px;height:28px}}@media (min-width: 600px), print{#gl_layer .module_title_wrapper a.close_focus,#gl_fallback_layer .module_title_wrapper a.close_focus{width:36px;height:36px}}@media only screen and (min-width: 600px) and (orientation: landscape){#gl_layer .module_title_wrapper a.close_focus,#gl_fallback_layer .module_title_wrapper a.close_focus{width:30px;height:30px}}@media (min-width: 769px), print{#gl_layer .module_title_wrapper a.close_focus,#gl_fallback_layer .module_title_wrapper a.close_focus{width:36px;height:36px}}@media (min-width: 1024px), print{#gl_layer .module_title_wrapper a.close_focus,#gl_fallback_layer .module_title_wrapper a.close_focus{width:42px;height:42px}}#gl_layer .module_title_wrapper a.close_focus .close_icon,#gl_fallback_layer .module_title_wrapper a.close_focus .close_icon{display:block;height:100%;position:relative;opacity:0.7}#gl_layer .module_title_wrapper a.close_focus .close_icon:before,#gl_fallback_layer .module_title_wrapper a.close_focus .close_icon:before{transform:rotate(-45deg);content:'';position:absolute;height:3px;width:100%;top:calc(50% - 1.5px);left:0;background:#fff;opacity:.8}#gl_layer .module_title_wrapper a.close_focus .close_icon:after,#gl_fallback_layer .module_title_wrapper a.close_focus .close_icon:after{transform:rotate(45deg);content:'';position:absolute;height:3px;width:100%;top:calc(50% - 1.5px);left:0;background:#fff;opacity:.8}#gl_layer .module_title_wrapper a.close_focus:hover .close_icon,#gl_fallback_layer .module_title_wrapper a.close_focus:hover .close_icon{opacity:1.0}#gl_layer .module_description_wrapper,#gl_fallback_layer .module_description_wrapper{position:absolute;padding-bottom:26px;bottom:-110px;left:50%;width:90%;text-align:center;visibility:hidden;-webkit-transform:translateX(-50%);-ms-transform:translateX(-50%);transform:translateX(-50%);-webkit-transition:bottom 1s, visibility 1s;-moz-transition:bottom 1s, visibility 1s;transition:bottom 1s, visibility 1s;z-index:20}@media (min-width: 480px){#gl_layer .module_description_wrapper,#gl_fallback_layer .module_description_wrapper{padding-bottom:66px}}@media only screen and (min-width: 480px) and (orientation: landscape){#gl_layer .module_description_wrapper,#gl_fallback_layer .module_description_wrapper{padding-bottom:12px}}@media (min-width: 769px), print{#gl_layer .module_description_wrapper,#gl_fallback_layer .module_description_wrapper{padding-bottom:66px}}#gl_layer .module_description_wrapper.active,#gl_fallback_layer .module_description_wrapper.active{bottom:0px;visibility:visible}#gl_layer .module_description_wrapper .gradient_overlay,#gl_fallback_layer .module_description_wrapper .gradient_overlay{position:absolute;width:120%;height:250%;bottom:0;left:50%;opacity:.8;background:-moz-linear-gradient(top, transparent 0%, transparent 7%, #000 100%);background:-webkit-gradient(linear, left top, left bottom, color-stop(0%, transparent), color-stop(7%, transparent), color-stop(100%, #000));background:-webkit-linear-gradient(top, transparent 0%, transparent 7%, #000 100%);background:-o-linear-gradient(top, transparent 0%, transparent 7%, #000 100%);background:-ms-linear-gradient(top, transparent 0%, transparent 7%, #000 100%);background:linear-gradient(to bottom, transparent 0%, transparent 7%, #000 100%);filter:progid:DXImageTransform.Microsoft.gradient( startColorstr='#00000000', endColorstr='#000000',GradientType=0 );-webkit-transform:translateX(-50%);-ms-transform:translateX(-50%);transform:translateX(-50%)}#gl_layer .module_description_wrapper .module_description,#gl_fallback_layer .module_description_wrapper .module_description{padding:0px 12px;font-size:14px;font-weight:300;position:relative;z-index:10;color:#FFFFFF;font-family:Whitney,Helvetica,Arial,sans-serif}@media (min-width: 600px), print{#gl_layer .module_description_wrapper .module_description,#gl_fallback_layer .module_description_wrapper .module_description{font-size:20px}}@media only screen and (min-width: 600px) and (orientation: landscape){#gl_layer .module_description_wrapper .module_description,#gl_fallback_layer .module_description_wrapper .module_description{font-size:16px}}@media (min-width: 1200px){#gl_layer .module_description_wrapper .module_description,#gl_fallback_layer .module_description_wrapper .module_description{font-size:22px}}#gl_layer .module_description_wrapper .explore,#gl_fallback_layer .module_description_wrapper .explore{letter-spacing:.1em;font-family:Whitney-Bold,Helvetica,Arial,sans-serif;font-size:0;vertical-align:baseline;margin:0px;color:#78BDFF;background:rgba(0,0,0,0.1);border-radius:2px;cursor:pointer;position:relative;z-index:10}#gl_layer .module_description_wrapper .explore:hover,#gl_fallback_layer .module_description_wrapper .explore:hover{color:#CFE7FF;background:rgba(0,0,0,0.2)}#gl_layer .module_description_wrapper .explore:after,#gl_fallback_layer .module_description_wrapper .explore:after{content:\"››\";font-size:20px}@media (min-width: 1024px), print{#gl_layer .module_description_wrapper .explore,#gl_fallback_layer .module_description_wrapper .explore{font-size:17px;padding:10px;border:1px solid #78BDFF;vertical-align:middle}#gl_layer .module_description_wrapper .explore:after,#gl_fallback_layer .module_description_wrapper .explore:after{content:none}}#gl_layer .label{position:absolute;padding:0;border:0;margin:0;max-width:300px;text-align:center;-webkit-transform:translate(-50%, -110%);-ms-transform:translate(-50%, -110%);transform:translate(-50%, -110%);font-size:16px;color:#FFFFFF;font-family:Whitney,Helvetica,Arial,sans-serif}#gl_fallback_layer .focus_layer{position:absolute;top:0;right:0;width:100%;height:100%;background-size:cover;background-position:center center;visibility:hidden;opacity:0;-webkit-transition:opacity 1s, visibility 1s;-moz-transition:opacity 1s, visibility 1s;transition:opacity 1s, visibility 1s;z-index:15}#gl_fallback_layer .focus_layer.active{visibility:visible;opacity:1}#gl_fallback_layer .overlay_image{position:absolute;top:0;right:0;width:100%;height:100%;height:100%;background-size:cover;background-position:center center}.hotspot_container{position:absolute;height:100%;width:100%}.hotspot_container .hotspot_wrapper{position:absolute;top:0px;left:0px;width:350px;visibility:visible;-webkit-transform:translate(-12px, -12px);-ms-transform:translate(-12px, -12px);transform:translate(-12px, -12px)}.hotspot_container .hotspot_wrapper span{position:relative;visibility:visible;font-size:0.7em;color:#FFFFFF;opacity:0.0;transition:opacity 0.75s;left:6px}.hotspot_container .hotspot_wrapper.hidden{visibility:hidden}.hotspot_container .hotspot_wrapper .hotspot{width:24px;height:24px;opacity:0.8}.fallback_mode #gl_fallback_layer{height:auto}.fallback_mode #gl_fallback_layer img{height:auto}.image{height:300px;width:300px;border-color:#000000;border:4px solid #ffffff;border-radius:8px;transform:translate(-50%, -50%)}.wysiwyg_content p{line-height:1.4em}.wysiwyg_content p,.wysiwyg_content a{word-wrap:break-word}.wysiwyg_content table a{word-break:break-word}#primary_column .wysiwyg_content>:first-child{margin-top:0}#primary_column .wysiwyg_content .inset_box{padding:.5em 2em;margin:2em 0;border:4px solid #DCE0E5}.wysiwyg_content h1,.wysiwyg_content h2,.wysiwyg_content h3,.wysiwyg_content h4{font-weight:700;letter-spacing:-.01em;margin:1.2em 0 .5em}@media (min-width: 769px), print{.wysiwyg_content h1,.wysiwyg_content h2,.wysiwyg_content h3,.wysiwyg_content h4{margin:1.5em 0 .5em}}.wysiwyg_content h1{font-size:2.2em}.wysiwyg_content h2{font-size:1.8em}.wysiwyg_content h3{font-size:1.4em}.wysiwyg_content h4{font-size:1.1em}.wysiwyg_content strong,.wysiwyg_content b,.wysiwyg_content .bold{font-weight:bold}.wysiwyg_content .content_title{font-size:1.04em;margin-bottom:.1em}@media (min-width: 600px), print{.wysiwyg_content .content_title{font-size:1.2em;margin-bottom:.18em}}@media (min-width: 769px), print{.wysiwyg_content .content_title{font-size:1.36em;margin-bottom:.26em}}@media (min-width: 1024px), print{.wysiwyg_content .content_title{font-size:1.44em;margin-bottom:.29em}}@media (min-width: 1200px){.wysiwyg_content .content_title{font-size:1.52em;margin-bottom:.32em}}.wysiwyg_content .article_teaser_body{font-size:1em}.wysiwyg_content .indent1{margin-left:3.5em}.wysiwyg_content .indent2{margin-left:7em}.wysiwyg_content .indent3{margin-left:10.5em}.wysiwyg_content .publish_date{font-weight:700}.wysiwyg_content .item_list_module{clear:both}.wysiwyg_content .expandable_element_link.style_1{font-size:.8em;font-weight:700;text-transform:uppercase}.wysiwyg_content .expandable_element{display:none}.wysiwyg_content table{border-spacing:1px;border-collapse:separate;font-size:15px;line-height:normal}.wysiwyg_content table th,.wysiwyg_content table td{padding:13px}.wysiwyg_content table th{background-color:#ddd;font-weight:600;text-align:left}.wysiwyg_content table td{background-color:#eee}.wysiwyg_content .table_wrapper{width:100%;margin:1em 0;-webkit-overflow-scrolling:touch}.wysiwyg_content .table_wrapper>div::-webkit-scrollbar{height:12px}.wysiwyg_content .table_wrapper>div::-webkit-scrollbar-track{box-shadow:0 0 2px rgba(0,0,0,0.15) inset;background:#f0f0f0}.wysiwyg_content .table_wrapper>div::-webkit-scrollbar-thumb{border-radius:6px;background:#ccc}.wysiwyg_content .table_wrapper.has-scroll{position:relative;overflow:hidden}.wysiwyg_content .table_wrapper.has-scroll:after{position:absolute;top:0;left:100%;width:50px;height:100%;border-radius:10px 0 0 10px / 50% 0 0 50%;box-shadow:-5px 0 10px rgba(0,0,0,0.25);content:''}.wysiwyg_content .table_wrapper.has-scroll>div{overflow-x:auto}.wysiwyg_content table.mb_table{border-collapse:collapse;width:100%}.wysiwyg_content table.mb_table td{background-color:transparent}.wysiwyg_content table.mb_table th{background-color:white;color:#f08d77;font-size:.75em;font-weight:500;text-align:left;padding:13px}@media (min-width: 600px), print{.wysiwyg_content table.mb_table th{font-size:.9em}}.wysiwyg_content table.mb_table tbody td{font-size:.9em}@media (min-width: 600px), print{.wysiwyg_content table.mb_table tbody td{font-size:1.1em}}.wysiwyg_content table.mb_table tr:nth-child(even){background-color:#edf4fb}.wysiwyg_content table.mb_table tr:nth-child(odd){background-color:#ffffff}.wysiwyg_content table.mb_table td{border:1px solid #d2d2d2;padding:.8em}.wysiwyg_content table.mb_table td:first-child{border-left:transparent}.wysiwyg_content table.mb_table td:last-child{border-right:transparent}.wysiwyg_content table.small_table,.wysiwyg_content table.mb_table.small_table{font-size:.75em;padding:.6em}#main_container form.gsc-search-box{padding:0}#main_container form.gsc-search-box td.gsc-input{padding:0}#main_container table[class^=\"gsc-\"] td,#main_container table[class^=\"gcsc-\"] td{background-color:transparent}#main_container.placeholder{-webkit-font-smoothing:antialiased}#main_container:-moz-placeholder{-webkit-font-smoothing:antialiased}#main_container::-moz-placeholder{-webkit-font-smoothing:antialiased}#main_container::-webkit-input-placeholder{-webkit-font-smoothing:antialiased}#main_container:-ms-input-placeholder{-webkit-font-smoothing:antialiased}#main_container .gsc-control-cse table{margin:0}#main_container input.gsc-input{padding:10px 12px;border-radius:6px;font-size:15px}#main_container input.gsc-search-button{border-color:#fff;background-color:#3b788b;padding:10px 14px 10px;height:38px;color:white;font-size:15px;font-weight:500;border-radius:6px;text-transform:uppercase}#main_container input.gsc-search-button:hover{background-color:#5097ad}#main_container .gsc-selected-option-container{width:auto !important;max-width:none}#main_container td.gsc-clear-button{padding-left:4px}#main_container .cse .gsc-control-cse,#main_container .gsc-control-cse{padding:0}#main_container .cse .gsc-control-cse tr,#main_container .gsc-control-cse tr{background:none !important}#main_container td.gsc-result-info-container{padding-left:0}#main_container .gs-no-results-result .gs-snippet,#main_container .gs-error-result .gs-snippet{padding:5px 0;margin:5px 0;border:none;background-color:transparent}#main_container .gsc-webResult.gsc-results{margin-top:0px}#main_container div.gsc-webResult.gsc-result{border-bottom:1px solid #CFD7E1;padding-bottom:16px;padding-top:16px;padding-left:0;margin-bottom:0px;margin-top:0px}#main_container td.gsc-table-cell-snippet-close{padding:0}#main_container div.gs-title{padding:0;height:auto;line-height:1.4em;text-decoration:none}#main_container .gs-result a.gs-title,#main_container .gs-result a.gs-title b{color:#388FDA;text-decoration:none;font-weight:700;letter-spacing:-.035em;height:auto;padding:0}@media (min-width: 600px), print{#main_container .gs-result a.gs-title,#main_container .gs-result a.gs-title b{font-size:18px}}@media (min-width: 769px), print{#main_container .gs-result a.gs-title,#main_container .gs-result a.gs-title b{font-size:20px}}#main_container a.gs-title:hover{color:#115FA3;text-decoration:underline}#main_container a.gs-title:hover b{color:#115FA3}#main_container .gs-webResult .gs-snippet,#main_container .gs-imageResult .gs-snippet,#main_container .gs-fileFormatType{color:#333;line-height:1.4em}@media (min-width: 1024px), print{#main_container .gs-webResult .gs-snippet,#main_container .gs-imageResult .gs-snippet,#main_container .gs-fileFormatType{font-size:15px}}#main_container .gs-webResult div.gs-visibleUrl,#main_container .gs-imageResult div.gs-visibleUrl{color:#888}#main_container .gsc-table-cell-thumbnail{padding:0 6px 0 0}@media (min-width: 600px), print{#main_container .gsc-table-cell-thumbnail{padding:0 12px 0 0}}@media (min-width: 1024px), print{#main_container .gsc-table-cell-thumbnail{padding:0 16px 0 0}}#main_container .gs-web-image-box{width:100px}@media (min-width: 600px), print{#main_container .gs-web-image-box{padding:0;width:125px}}#main_container img.gs-image,#main_container .gs-promotion-image-box img.gs-promotion-image{border:none;width:100%;height:auto;max-width:none;max-height:none}#main_container a.gs-image{display:block}#main_container .gsc-results .gsc-cursor-box{padding-top:2px}#main_container .gsc-results .gsc-cursor-box .gsc-cursor-page{color:#388FDA;font-size:17px}#main_container .gsc-results .gsc-cursor-box .gsc-cursor-current-page{color:#333;background-color:transparent;text-shadow:none;padding:0}#main_container .gsc-adBlock{display:none !important}.aaa{border:0 solid red}.shareline{width:100%;margin:1.7em 0 2.7em;display:block;position:relative;clear:both}.shareline .shareline_heading{margin-bottom:.5em;position:relative;margin-top:0}.shareline.top_attached_sl{margin-bottom:0}.shareline.bottom_attached_sl{margin-top:-1px;margin-top:0}.shareline.bottom_attached_sl article{border-top:none}.shareline.bottom_attached_sl .shareline_heading{display:none}.shareline article{padding:17px 0em 18px;border-top:1px solid #BEBEBE;border-bottom:1px solid #BEBEBE;position:relative;overflow:visible}.shareline .share_container{display:inline-block;vertical-align:top;width:75px}.shareline .share_container .selector{display:inline-block}.shareline .share_container .selected{display:inline-block;position:relative;top:1px;height:25px;width:25px;cursor:pointer}.shareline .share_container .selected:before{font-size:32px;margin-top:-3px}.shareline .share_container .arrow_box{display:inline-block;position:relative;padding:9px 11px;cursor:pointer}.shareline .share_container .arrow_down{width:0;height:0;border-top:6px solid transparent !important;border-bottom:6px solid transparent !important;border-left:8px solid #b2b2b2;display:inline-block;transform:rotate(90deg)}.shareline .share_container .arrow_down:hover{border-color:black}.shareline .share_options{display:none;position:absolute;top:47px;left:0;z-index:2;background-color:#FFF;border:1px solid #BEBEBE;padding:5px 6px 0 6px}.shareline .share_options .share_btn{width:40px;height:40px;font-size:40px;display:inline;margin:0.1em;cursor:pointer}.shareline a.fi-social-twitter,.shareline a.fi-social-facebook{color:#2b2b2b;text-decoration:none}.shareline .share_text{display:inline-block;vertical-align:middle;width:calc(100% - 75px);color:#555;font-size:95%}#explore_overlay .shareline a.fi-social-twitter,#explore_overlay .shareline a.fi-social-facebook,#explore_overlay .shareline .share_text,.explore_overlay_page .shareline a.fi-social-twitter,.explore_overlay_page .shareline a.fi-social-facebook,.explore_overlay_page .shareline .share_text{color:#FFF}#explore_overlay .shareline .share_options a.fi-social-twitter,#explore_overlay .shareline .share_options a.fi-social-facebook,.explore_overlay_page .shareline .share_options a.fi-social-twitter,.explore_overlay_page .shareline .share_options a.fi-social-facebook{color:#2b2b2b}#explore_overlay .shareline .arrow_down:hover,.explore_overlay_page .shareline .arrow_down:hover{border-color:white}#explore_overlay .shareline article,.explore_overlay_page .shareline article{border-color:#6d6b6b}.keypoint .share_container{width:28px;margin-top:-1px;margin-left:1em}.keypoint .keypoint_icon{font-size:1.2em}section.missions_teaser{background:#edecec;z-index:10}section.missions_teaser header{margin-bottom:2em}section.missions_teaser ul.missions_circles{text-align:center;margin-bottom:2em}section.missions_teaser ul.missions_circles li.mission_item{display:block;margin-left:auto;margin-right:auto;width:300px;text-decoration:none;margin-bottom:3%;border-radius:150px;overflow:hidden}@media (max-width: 480px){section.missions_teaser ul.missions_circles li.mission_item{width:290px;border-radius:145px}}@media (min-width: 769px), print{section.missions_teaser ul.missions_circles li.mission_item{width:210px;border-radius:105px;margin-right:2%;display:inline-block;margin-bottom:0}section.missions_teaser ul.missions_circles li.mission_item:last-child{margin-right:0}}@media (min-width: 1024px), print{section.missions_teaser ul.missions_circles li.mission_item{margin-right:3%;width:300px;border-radius:150px}}@media (min-width: 1200px){section.missions_teaser ul.missions_circles li.mission_item{margin-right:5%}}section.missions_teaser ul.missions_circles li.mission_item:first-child{border:5px solid #3b788b}section.missions_teaser ul.missions_circles li.mission_item:nth-child(2){border:5px solid #c25b28}section.missions_teaser ul.missions_circles li.mission_item:nth-child(3){border:5px solid #fda43c}@media (min-width: 769px), print{section.missions_teaser ul.missions_circles li.mission_item .rollover_description{transition:opacity .4s}}@media (min-width: 769px), print{section.missions_teaser ul.missions_circles li.mission_item .rollover_description .rollover_description_inner{font-size:0.9em}}@media (min-width: 1024px), print{section.missions_teaser ul.missions_circles li.mission_item .rollover_description .rollover_description_inner{font-size:1em}}section.missions_teaser ul.missions_circles li.mission_item .rollover_description *{color:white}section.missions_teaser ul.missions_circles li.mission_item:hover .rollover_description{display:none}@media (min-width: 769px), print{section.missions_teaser ul.missions_circles li.mission_item:hover .rollover_description{display:block;opacity:1;height:100%;width:100%;z-index:1;top:0;right:0;overflow:hidden;position:absolute;background-color:rgba(0,0,0,0.6);border-radius:50%;padding:2em;color:white;font-weight:500;font-size:1.1em}}@media (min-width: 1024px), print{section.missions_teaser ul.missions_circles li.mission_item:hover .rollover_description{padding:6em 1.5em}}@media (min-width: 769px), print{section.missions_teaser ul.missions_circles li.mission_item:hover li.mission_item .title{opacity:0}}section.missions_teaser ul.missions_circles a{height:300px;display:block;position:relative;text-decoration:none}@media (max-width: 480px){section.missions_teaser ul.missions_circles a{height:290px}}@media (min-width: 769px), print{section.missions_teaser ul.missions_circles a{height:210px}}@media (min-width: 1024px), print{section.missions_teaser ul.missions_circles a{height:300px}}section.missions_teaser ul.missions_circles a .title{text-align:center;position:relative;display:block;top:80%;color:white;text-transform:uppercase;font-size:1.2em;font-weight:500;transition:opacity .4s}@media (min-width: 769px), print{section.missions_teaser ul.missions_circles a .title{top:72%}}@media (min-width: 1024px), print{section.missions_teaser ul.missions_circles a .title{top:80%}}section.missions_teaser footer .detail_link{display:block;margin:0.5em 1em 0 0;white-space:nowrap}@media (min-width: 769px), print{section.missions_teaser footer{float:right;text-align:right}}section.missions_teaser footer a{color:#4e8fa4}.wysiwyg_content .footnotes li{position:relative;font-size:.85em;margin-bottom:1em}.wysiwyg_content .footnotes li .footnote h2{margin:0;color:#222;font-size:1.2em}.wysiwyg_content .footnotes li p{margin:0.4em 0 0.6em}.wysiwyg_content .footnote{font-size:.8em}#secondary_column .footnote{font-size:.8em}.primary_media_feature{margin-bottom:0}@media (min-width: 769px), print{.primary_media_feature{padding:0}}.primary_media_feature.single{position:relative;margin-bottom:0;overflow:hidden}.primary_media_feature.single .feature_container{height:300px;background-size:cover;position:relative;z-index:3;background-position:center}@media (min-width: 769px), print{.primary_media_feature.single .feature_container{height:700px}}.primary_media_feature.single.video .play{display:none;position:absolute;top:47%;left:47%;top:calc(50%- 30px);left:calc(50%- 30px);top:-webkit-calc(50% - 30px);left:-webkit-calc(50% - 30px);width:60px;height:60px;padding-top:0;cursor:pointer;background:url(\"https://mars.nasa.gov/assets/play-button.png\") 0 0 no-repeat;z-index:10}.primary_media_feature.single.video .player{width:100%;height:100%;position:absolute;top:0;left:0;z-index:2}.primary_media_feature.single .video_header_overlay{position:absolute;bottom:2em;margin:0 auto;left:0;right:0;width:auto;text-align:center;color:white;z-index:5}.primary_media_feature.single .video_header_overlay .media_feature_title{font-size:3em}.custom_banner_container{position:relative}.faq_section h2{margin-top:0}.faq_section ul.q_and_a{margin-bottom:1em}.faq_section ul.q_and_a .question{margin-bottom:1em}.faq_section ul.q_and_a .question:last-child{margin-bottom:0.6em}.faq_section ul.q_and_a .title_container{cursor:pointer}.faq_section ul.q_and_a .title{font-weight:700;font-size:1.1em}.faq_section ul.q_and_a .text.answer{visibility:hidden;position:absolute;left:-9999px}.faq_section ul.q_and_a .text.answer.open{visibility:visible;position:relative;left:0}.faq_section hr:last-child{display:none}.fullscreen_element{position:absolute;top:7px;right:7px;cursor:pointer;background-color:rgba(0,0,0,0.5);width:50px;height:50px;border-radius:5px;z-index:10}@media (min-width: 769px), print{.fullscreen_element{top:20px;right:20px}}.fullscreen_element .fullscreen-icon{height:25px;width:25px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") 1px -25px;background-size:25px;margin:13px 0 0 13px}.fullscreen_element:hover .fullscreen-icon{background:url(\"https://mars.nasa.gov/assets/[email protected]\") 1px 0px;background-size:25px}.fullscreen_element.fullscreen-mode .fullscreen-icon{background:url(\"https://mars.nasa.gov/assets/[email protected]\") 1px -74px;background-size:25px}.fullscreen_element.fullscreen-mode:hover .fullscreen-icon{background:url(\"https://mars.nasa.gov/assets/[email protected]\") 1px -49px;background-size:25px}#timeline-embed:fullscreen{height:100%;width:100%;min-height:none;max-height:none}.triple_teaser{background-color:white;z-index:11}.triple_teaser .column{width:100%}@media (min-width: 769px), print{.triple_teaser .column{width:31.03448%;float:left}.triple_teaser .column:nth-child(3n+1){margin-left:0;margin-right:-100%;clear:both;margin-left:0}.triple_teaser .column:nth-child(3n+2){margin-left:34.48276%;margin-right:-100%;clear:none}.triple_teaser .column:nth-child(3n+3){margin-left:68.96552%;margin-right:-100%;clear:none}}@media (min-width: 1024px), print{.triple_teaser .column{width:28.57143%;float:left}.triple_teaser .column:nth-child(3n+1){margin-left:0;margin-right:-100%;clear:both;margin-left:0}.triple_teaser .column:nth-child(3n+2){margin-left:35.71429%;margin-right:-100%;clear:none}.triple_teaser .column:nth-child(3n+3){margin-left:71.42857%;margin-right:-100%;clear:none}}.triple_teaser .column:last-child{margin-bottom:1em}.triple_teaser .column+.column{margin-top:3em}@media (min-width: 769px), print{.triple_teaser .column+.column{margin-top:0}}.triple_teaser header{margin-bottom:1.3em}.triple_teaser .module_title,.triple_teaser .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .triple_teaser .carousel_title{text-align:left}@media (min-width: 600px), print{.triple_teaser .module_title,.triple_teaser .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .triple_teaser .carousel_title{font-size:2em}}@media (min-width: 600px), print{.triple_teaser footer{text-align:left}}.triple_teaser footer .detail_link{float:left;clear:both;text-align:left;white-space:nowrap}.triple_teaser .img_area{margin-bottom:1em;float:none;width:100%}.triple_teaser .item_list{margin-bottom:1em}.triple_teaser .item_list li{border-bottom:1px solid #BEBEBE;padding:3.44828% 0}.triple_teaser .item_list li:first-child{padding-top:0}.triple_teaser .item_list li:last-child{border-bottom:none}.triple_teaser .item_list .list_image{width:39.65517%;float:left;margin-right:3.44828%;margin-left:0}@media (min-width: 600px), print{.triple_teaser .item_list .list_image{width:31.03448%;float:left;margin-right:3.44828%}}.triple_teaser .item_list .list_text{width:56.89655%;float:right;margin-right:0}@media (min-width: 600px), print{.triple_teaser .item_list .list_text{width:65.51724%;float:right;margin-right:0}}.triple_teaser .item_list .list_text .date{color:#707070;font-size:0.8em;font-weight:500;margin-bottom:0.3em}.triple_teaser .item_list .list_text .date span:before{content:\" \\2022 \"}.triple_teaser .item_list .list_text .title{font-size:1em;font-weight:500}.triple_teaser .upcoming_events .item_list li{padding:3.44828% 0}.triple_teaser .upcoming_events .item_list li:first-child{padding-top:0}.triple_teaser .upcoming_events .item_list .list_text .date{margin-bottom:.7em}.triple_teaser .follow_teaser .text_area{font-weight:300;font-size:1rem}.triple_teaser .follow_teaser .text_area footer{margin-top:2em}ul.item_list{margin-bottom:2em}ul.item_list .list_title{font-size:1.3em;font-weight:700;margin-bottom:.5em}ul.item_list .list_title a{color:#222}ul.item_list .text_only .list_text{width:100%;padding:0}ul.item_list>li hr{margin:0}ul.item_list .list_image{width:37.5%;float:right;margin-right:0;margin-left:4.16667%;margin-bottom:.5em}@media (min-width: 600px), print{ul.item_list .list_image{margin-left:0;margin-bottom:0;width:35.89744%;float:left;margin-right:2.5641%}}@media (min-width: 769px), print{ul.item_list .list_image{width:22.41379%;float:left;margin-right:3.44828%}}@media (min-width: 1024px), print{ul.item_list .list_image{width:31.03448%;float:left;margin-right:3.44828%}}@media (min-width: 600px), print{ul.item_list .list_text{width:61.53846%;float:right;margin-right:0}}@media (min-width: 769px), print{ul.item_list .list_text{width:74.13793%;float:right;margin-right:0}}@media (min-width: 1024px), print{ul.item_list .list_text{width:65.51724%;float:right;margin-right:0}}ul.item_list .list_text h2,ul.item_list .list_text h3,ul.item_list .list_text h4{margin-top:0}ul.item_list .list_content{padding:1em 0}ul.item_list .list_description{margin-top:0}ul.item_list .description .long{display:none}ul.item_list .description .long p:first-of-type{margin-top:0}ul.people.item_list li.person{padding:4.16667% 0}ul.people.item_list li.person:first-child{padding-top:0}ul.people.item_list .person_header{margin-bottom:1.2em}ul.people.item_list .list_title.list_name{padding-top:7%}@media (min-width: 600px), print{ul.people.item_list .list_title.list_name{padding:0}}ul.people.item_list .person_title{font-weight:300}ul.people.item_list .description{clear:both}@media (min-width: 600px), print{ul.people.item_list .description{clear:none}}ul.people.item_list .person+.person{border-top:1px solid #BEBEBE}ul.item_list.text_item_list .list_text{width:100%}ul.item_list.text_item_list .list_text .date{margin-bottom:.3em}ul.item_list.text_item_list a{color:#257cdf}ul.item_list.text_item_list a:hover{text-decoration:underline}ul.item_list.text_item_list .publication_authors{margin-bottom:.4em}ul.item_list.text_item_list .citation{font-size:.85em;margin-bottom:.4em;font-weight:300}ul.item_list.text_item_list .publication_title{font-size:1.1em;font-weight:700;margin-bottom:.4em}ul.item_list.text_item_list .publication_title a{color:#222}ul.item_list.text_item_list .list_title a{color:#222}.explore_overlay_page .feature_pages .wysiwyg_content .item_list_module{margin-left:auto}@media (min-width: 1024px){.secondary_nav_desktop{overflow-x:auto}}@media (min-width: 1024px){.custom_banner_container .secondary_nav_desktop{overflow-x:visible}}@media (min-width: 1024px){.custom_banner_container .fixed_secondary_nav{overflow-x:auto}}nav.secondary_nav{font-weight:400}nav.secondary_nav .grid_layout{width:100%;padding-left:10px;padding-right:10px;max-width:none}@media (min-width: 600px), print{nav.secondary_nav .grid_layout{padding-left:17px;padding-right:17px}}nav.secondary_nav.secondary_nav_mobile{display:block;width:100%}nav.secondary_nav.secondary_nav_mobile select{position:relative;padding:.5em 2em .5em 1em;font-size:16px;border:0;height:40px;vertical-align:middle;color:white;-webkit-appearance:none;-o-appearance:none;-moz-appearance:none;background:#3b788b url(\"https://mars.nasa.gov/assets/[email protected]\") no-repeat 95% 10px;background-position:right .8em top 10px;background-size:9px;font-weight:700;cursor:pointer;width:100%;border-radius:5px;max-width:304px;margin:0;background-color:#AFB3B9;width:100%;max-width:none;border-radius:0}nav.secondary_nav.secondary_nav_mobile select::-ms-expand{display:none}nav.secondary_nav.secondary_nav_mobile select option{padding:0.5em 1em}@media (min-width: 1024px){nav.secondary_nav.secondary_nav_mobile{display:none}}nav.secondary_nav.secondary_nav_desktop{display:none}@media (min-width: 1024px){nav.secondary_nav.secondary_nav_desktop{padding:0.8em 0 0.9em;display:block;margin:0;background-color:#eee;text-align:center}}nav.secondary_nav.secondary_nav_desktop .section_title{display:none}nav.secondary_nav.secondary_nav_desktop .section_title a{padding-left:0;text-decoration:none}nav.secondary_nav.secondary_nav_desktop li{display:inline-block;position:relative}nav.secondary_nav.secondary_nav_desktop a{color:#777;font-size:1em;font-weight:600;display:block;padding:.3em .3em}@media (min-width: 769px), print{nav.secondary_nav.secondary_nav_desktop a{padding:.3em .6em}}@media (min-width: 1200px){nav.secondary_nav.secondary_nav_desktop a{padding:.3em .9em}}@media (min-width: 1700px){nav.secondary_nav.secondary_nav_desktop a{font-size:1.1em}}.custom_banner_container nav.secondary_nav.secondary_nav_desktop a{color:white}nav.secondary_nav.secondary_nav_desktop ul{white-space:nowrap}nav.secondary_nav.secondary_nav_desktop li.current a,nav.secondary_nav.secondary_nav_desktop li:hover a{text-decoration:none;color:#222}nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav .grid_layout{display:flex;flex-wrap:nowrap;justify-content:space-between}@media (min-width: 1024px){nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav{position:fixed;width:100%;top:0;left:0;z-index:100;box-shadow:0 4px 4px -2px rgba(0,0,0,0.15)}nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav.secondary_nav_desktop{padding:1em 0 0.8em;white-space:nowrap}nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav .section_title{display:inline-block;margin-top:3px;margin-right:1.6em;font-size:1.2em;flex-shrink:0}nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav .section_title a{padding:0;color:#2B2B2B}}@media (min-width: 1024px) and (min-width: 1700px){nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav .section_title{margin-top:6px}}@media (min-width: 1024px){nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav ul{display:inline-block;text-align:right;width:100%}}@media (min-width: 1024px) and (min-width: 1024px), print and (min-width: 1024px){nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li a{padding:.3em .5em;font-size:0.9em}}@media (min-width: 1024px) and (min-width: 1200px){nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li a{padding:.3em .6em;font-size:0.95em}}@media (min-width: 1024px) and (min-width: 1700px){nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li a{padding:.3em .8em;font-size:1em}}@media (min-width: 1024px){nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li.current a,nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav a:hover,nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav .section_title a:hover{text-decoration:none;color:#2B2B2B}nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li:last-of-type a{padding-right:0}}.custom_banner_container nav.secondary_nav.secondary_nav_desktop,.homepage_feature_container nav.secondary_nav.secondary_nav_desktop{text-align:center;margin:0;background-color:transparent}.custom_banner_container nav.secondary_nav.secondary_nav_desktop li,.homepage_feature_container nav.secondary_nav.secondary_nav_desktop li{margin-bottom:6px}.custom_banner_container nav.secondary_nav.secondary_nav_desktop li a,.homepage_feature_container nav.secondary_nav.secondary_nav_desktop li a{color:white}.custom_banner_container nav.secondary_nav.secondary_nav_desktop li.current:after,.homepage_feature_container nav.secondary_nav.secondary_nav_desktop li.current:after{bottom:-60px;left:50%;border:solid transparent;content:\" \";height:0;width:0;position:absolute;pointer-events:none;border-top-color:black;border-width:20px;margin-left:-20px}.custom_banner_container nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav,.homepage_feature_container nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav{background-color:#e4e7ec}.custom_banner_container nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li.current:after,.homepage_feature_container nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li.current:after{content:none}.custom_banner_container nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li a,.homepage_feature_container nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li a{color:#fff}.custom_banner_container nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li a:hover,.homepage_feature_container nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li a:hover{color:#2B2B2B}.custom_banner_container nav.secondary_nav.secondary_nav_mobile,.homepage_feature_container nav.secondary_nav.secondary_nav_mobile{text-align:center;padding:0 2.5%}.custom_banner_container nav.secondary_nav.secondary_nav_mobile select,.homepage_feature_container nav.secondary_nav.secondary_nav_mobile select{position:relative;padding:.5em 2em .5em 1em;font-size:16px;border:0;height:40px;vertical-align:middle;color:white;-webkit-appearance:none;-o-appearance:none;-moz-appearance:none;background:#3b788b url(\"https://mars.nasa.gov/assets/[email protected]\") no-repeat 95% 10px;background-position:right .8em top 10px;background-size:9px;font-weight:700;cursor:pointer;width:100%;border-radius:5px;max-width:304px;margin:0.9em 0 1.1em}.custom_banner_container nav.secondary_nav.secondary_nav_mobile select::-ms-expand,.homepage_feature_container nav.secondary_nav.secondary_nav_mobile select::-ms-expand{display:none}.custom_banner_container nav.secondary_nav.secondary_nav_mobile select option,.homepage_feature_container nav.secondary_nav.secondary_nav_mobile select option{padding:0.5em 1em}.homepage_feature_container nav.secondary_nav{position:absolute;bottom:0;z-index:2}.homepage_feature_container nav.secondary_nav.secondary_nav_desktop{width:100%}.homepage_feature_container nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav{bottom:auto}.homepage_feature_container nav.secondary_nav.secondary_nav_desktop.fixed_secondary_nav li.current:after{content:none}#explore_overlay nav.secondary_nav{display:none}nav.tertiary_nav{position:relative;margin-top:1.2em;font-weight:400}nav.tertiary_nav ul li{position:relative;display:inline-block;vertical-align:middle}nav.tertiary_nav ul li:hover a,nav.tertiary_nav ul li.current a{color:black}nav.tertiary_nav ul li+li:before{content:\" | \";padding:0 .5em;vertical-align:middle;font-weight:100;color:#BEBEBE}nav.tertiary_nav ul a{display:inline-block;vertical-align:middle;color:#909090}nav.tertiary_nav ul a:hover{text-decoration:none}@media (min-width: 769px), print{nav.tertiary_nav ul a{font-size:1.2em}}@media (min-width: 1024px), print{nav.tertiary_nav ul a{font-size:1.3em}}@media (min-width: 1200px){nav.tertiary_nav ul a{font-size:1.4em}}.tertiary_nav_mobile{display:block;text-align:center}@media (min-width: 600px), print{.tertiary_nav_mobile{text-align:left}}.tertiary_nav_mobile select{position:relative;padding:.5em 2em .5em 1em;font-size:16px;border:0;height:40px;vertical-align:middle;color:white;-webkit-appearance:none;-o-appearance:none;-moz-appearance:none;background:#3b788b url(\"https://mars.nasa.gov/assets/[email protected]\") no-repeat 95% 10px;background-position:right .8em top 10px;background-size:9px;font-weight:700;cursor:pointer;width:100%;border-radius:5px;max-width:304px;margin:0 auto}.tertiary_nav_mobile select::-ms-expand{display:none}.tertiary_nav_mobile select option{padding:0.5em 1em}@media (min-width: 769px), print{.tertiary_nav_mobile{display:none}}.tertiary_nav_desktop{display:none}@media (min-width: 769px), print{.tertiary_nav_desktop{display:block}}.condense_control{color:#257cdf}.condense_control:hover{text-decoration:underline}.condense_control:before{content:' › ';white-space:nowrap}section.intro{display:block;background-color:#000;position:absolute;top:0;left:0;z-index:99;width:100%;overflow:hidden;height:100vh}section.intro .brand_area{background:url(\"https://mars.nasa.gov/assets/[email protected]\") no-repeat;background-size:100%;z-index:301;position:absolute;top:2%;left:4%;width:60%;height:80%;max-width:500px}@media (min-width: 769px), print{section.intro .brand_area{width:45%;left:2%}}section.intro img{object-fit:cover;height:100%;width:100%}@media (min-width: 480px){section.intro img{margin-top:0}}body.intro_screen_visible .vital_signs_menu,body.intro_screen_visible .more_bar{z-index:100}body.intro_screen_visible .vital_signs_menu .overlay_icon{display:none}body.intro_screen_visible section.more_bar .title,body.intro_screen_visible section.more_bar .arrow_down{display:none}body.intro_screen_visible section.more_bar:after{content:\"loading...\";display:inline-block;padding:0.6em 0;font-size:.9em}html.explore_overlay_open,body.explore_overlay_open{overflow:hidden;width:100%;height:100%;position:fixed;-ms-overflow-style:-ms-autohiding-scrollbar}#explore_overlay{position:fixed;top:0;left:0;height:100vh;width:100%;z-index:1000001;overflow-x:hidden;visibility:hidden;opacity:0;padding:0 0 6px;background-color:rgba(0,0,0,0.9)}.overlay_loaded #explore_overlay{background-color:#000}#explore_overlay.visible{visibility:visible}#explore_overlay .content{position:relative;height:100%;visibility:hidden;opacity:0}#explore_overlay .content.visible{visibility:visible}#explore_overlay .content>iframe{top:0;left:0;position:absolute}#explore_overlay .loading{position:absolute;left:50%;top:42vh;transform:translateX(-50%);width:auto;text-align:center;display:none}#explore_overlay .loading img{width:44px;height:44px}#explore_overlay .loading p{font-family:Whitney, Helvetica, Arial, sans-serif;position:relative;color:#76aee6;font-size:14px;letter-spacing:0.1em}#explore_overlay .loading .spinner div{background:#ccc !important}#explore_overlay .background_area{-webkit-tap-highlight-color:transparent;-webkit-tap-highlight-color:transparent;width:100%;height:100%;position:absolute;top:0;left:0;cursor:pointer;z-index:-1}#explore_overlay.lightbox_overlay{background-color:rgba(0,0,0,0.75);text-align:center;padding:0}#explore_overlay.lightbox_overlay .content{position:fixed;background-color:#1f1f1f;width:95%;height:92% !important;max-width:1400px;margin:2em auto 0;border-radius:4px;left:auto;right:calc(50vw - 47.5%)}@media only screen and (min-width: 1480px){#explore_overlay.lightbox_overlay .content{right:calc(50vw - 697px)}}@media only screen and (max-device-width: 1024px) and (-webkit-min-device-pixel-ratio: 1){#explore_overlay.lightbox_overlay .content{overflow-y:scroll;-webkit-overflow-scrolling:touch}}.overlay_close_button{position:absolute;top:1em;right:1.1em;z-index:1000003;width:40px;height:40px;position:fixed;background:#000;text-decoration:none;text-align:center;line-height:1em;transition:.3s opacity;visibility:hidden;opacity:0}.overlay_close_button.visible{visibility:visible}.no-touchevents .overlay_close_button:hover{opacity:1}.overlay_close_button .close_icon{display:block;height:100%;position:relative}.overlay_close_button .close_icon:before{transform:rotate(-45deg);content:'';position:absolute;height:1px;width:100%;top:calc(50% - .5px);left:0;background:#fff;opacity:.8}.overlay_close_button .close_icon:after{transform:rotate(45deg);content:'';position:absolute;height:1px;width:100%;top:calc(50% - .5px);left:0;background:#fff;opacity:.8}@media (min-width: 769px), print{.overlay_close_button{width:60px;height:60px;top:1.1em;right:1.1em}}@media (min-width: 1700px){.overlay_close_button{width:70px;height:70px;top:1.2em;right:1.2em}}.overlay_close_button.lightbox_overlay{background-color:#1f1f1f;top:2.5em;right:calc(50vw - 45.5%)}@media (min-width: 600px), print{.overlay_close_button.lightbox_overlay{right:calc(50vw - 45%)}}@media (min-width: 769px), print{.overlay_close_button.lightbox_overlay{top:2.6em;width:50px;height:50px}}@media (min-width: 1024px), print{.overlay_close_button.lightbox_overlay{right:calc(50vw - 45.5%)}}@media only screen and (min-width: 1480px){.overlay_close_button.lightbox_overlay{right:calc(50vw - 680px)}}@media (min-width: 1700px){.overlay_close_button.lightbox_overlay{top:2.7em;width:60px;height:60px}}#iframe_overlay,#iframe_overlay body{height:100%;overflow-y:auto;-webkit-overflow-scrolling:touch;font-weight:400}#iframe_overlay{width:1px;min-width:100%;word-wrap:break-word;color:#e4e3e3}#iframe_overlay p,#iframe_overlay .release_date{color:#e4e3e3}#iframe_overlay hr{border-color:#3c3c3c}#iframe_overlay a{color:#42a0f2}#iframe_overlay .header_mask{display:none}#iframe_overlay .explore_overlay_page{padding-bottom:4em}#iframe_overlay .done_btn{text-align:center}#iframe_overlay .done_btn button{color:#6bbed8;margin:2em 0;padding:.3em .7em .4em;background:none;cursor:pointer;letter-spacing:1px;font-weight:300;outline:none;position:relative;font-size:1.8em;border:1px solid #6bbed8;transition:color 200ms, border-color 200ms}#iframe_overlay .done_btn button::after{content:'Close'}#iframe_overlay .done_btn button:hover{color:#82ddf9;border-color:#82ddf9}#iframe_overlay .left_col,#iframe_overlay .right_col{position:relative;float:left}#iframe_overlay .left_col{width:100%}@media (min-width: 769px), print{#iframe_overlay .left_col{width:65%;border-right:1px solid #BEBEBE;padding-right:1em}}@media (min-width: 1200px){#iframe_overlay .left_col{padding-right:3em}}#iframe_overlay .right_col{width:100%}#iframe_overlay .right_col p{color:#868686}#iframe_overlay .right_col p b{color:#222}@media (min-width: 769px), print{#iframe_overlay .right_col{width:35%;padding-left:1em;left:-1px}}@media (min-width: 1200px){#iframe_overlay .right_col{padding-left:3em}}#iframe_overlay .suggested_features{display:none}#iframe_overlay #secondary_column aside.boxed{border-color:#6d6b6b}#iframe_overlay #secondary_column .related_content_module{border-color:#5a5a5a}#iframe_overlay #secondary_column .related_content_module li{border-color:#3c3c3c;padding:.8em 0}#iframe_overlay .article_nav{display:none}.info_tabs_module{position:relative;color:black;background:url(\"https://mars.nasa.gov/assets/mars_landscape.jpg\") center top no-repeat;background-size:cover}@media (min-width: 769px), print{.info_tabs_module{height:620px;background-position:center bottom}}.info_tabs_module div[data-react-class=\"InfoTabs\"]{height:100%}.info_tabs_module .grid_layout{padding-bottom:10em}@media (min-width: 769px), print{.info_tabs_module .grid_layout{padding-bottom:0}}.info_tabs_module .gradient_container_bottom{display:none}.info_tabs_module .info_tabs{padding:2.7em 0 5em;height:100%}@media (min-width: 769px), print{.info_tabs_module .info_tabs{padding:5.3em 0 5em}}.info_tabs_module .col1,.info_tabs_module .col2{width:100%}@media (min-width: 769px), print{.info_tabs_module .col1,.info_tabs_module .col2{width:48%}}.info_tabs_module .col2{display:none;float:right;margin-top:2rem}@media (min-width: 769px), print{.info_tabs_module .col2{display:block;margin-top:0}}.info_tabs_module .col1{float:left}.info_tabs_module .info_tabs_header{width:100%;margin-bottom:1.8em;display:inline-block;text-align:center}@media (min-width: 769px), print{.info_tabs_module .info_tabs_header{text-align:left;margin-bottom:3em}}.info_tabs_module .info_tabs_header h2{font-size:1.69em;margin-bottom:0em;font-weight:300}@media (min-width: 600px), print{.info_tabs_module .info_tabs_header h2{font-size:1.95em;margin-bottom:0em}}@media (min-width: 769px), print{.info_tabs_module .info_tabs_header h2{font-size:2.21em;margin-bottom:0em}}@media (min-width: 1024px), print{.info_tabs_module .info_tabs_header h2{font-size:2.34em;margin-bottom:0em}}@media (min-width: 1200px){.info_tabs_module .info_tabs_header h2{font-size:2.47em;margin-bottom:0em}}.info_tabs_module .info_tabs_links{padding-left:2rem;position:relative;z-index:2;font-size:1.3rem;font-weight:300;width:90%}@media (min-width: 769px), print{.info_tabs_module .info_tabs_links{width:auto}}.info_tabs_module .info_tabs_links li{cursor:pointer;position:relative;margin-bottom:0.3em}.info_tabs_module .info_tabs_links .tab_title{margin-bottom:0.2em;letter-spacing:-0.02em}.info_tabs_module .info_tabs_links .info_tabs_link:before{content:'';width:0;height:0;border-top:7px solid transparent !important;border-bottom:7px solid transparent !important;border-left:11px solid #943b2b;display:inline-block;transform:none;position:absolute;left:-20px;top:7px;opacity:0;transition:all 200ms}.no-touchevents .info_tabs_module .info_tabs_links .info_tabs_link:not(.active):hover:before,.info_tabs_module .info_tabs_links .active:before{opacity:1;left:-28px}.no-touchevents .info_tabs_module .info_tabs_links .info_tabs_link:not(.active):hover:before{opacity:.7}.info_tabs_module .info_tabs_detail,.info_tabs_module .active .mobile_tab_detail{float:right;max-height:320px;overflow-y:auto;padding-right:1em;position:relative;z-index:2;font-weight:300;width:100%;-webkit-overflow-scrolling:touch}.info_tabs_module .info_tabs_detail::-webkit-scrollbar,.info_tabs_module .active .mobile_tab_detail::-webkit-scrollbar{width:5px}.info_tabs_module .info_tabs_detail::-webkit-scrollbar-thumb,.info_tabs_module .active .mobile_tab_detail::-webkit-scrollbar-thumb{background-color:rgba(107,107,107,0.6)}.info_tabs_module .info_tabs_detail::-webkit-scrollbar-track,.info_tabs_module .active .mobile_tab_detail::-webkit-scrollbar-track{background-color:rgba(157,157,157,0.4)}@media (min-width: 769px), print{.info_tabs_module .info_tabs_detail,.info_tabs_module .active .mobile_tab_detail{padding-right:3em;top:0.8em}}.info_tabs_module .info_tabs_detail .info_tabs_title,.info_tabs_module .active .mobile_tab_detail .info_tabs_title{display:none}@media (min-width: 769px), print{.info_tabs_module .info_tabs_detail .info_tabs_title,.info_tabs_module .active .mobile_tab_detail .info_tabs_title{display:block;font-size:1.1em;margin-bottom:1em;margin-top:0;text-transform:uppercase}}.info_tabs_module .mobile_tab_detail{display:none}.info_tabs_module .active .mobile_tab_detail{font-size:0.95rem;float:none;display:block}@media (min-width: 769px), print{.info_tabs_module .active .mobile_tab_detail{display:none}}.info_tabs_module .active .tab_title{margin-bottom:.5em}@media (min-width: 769px), print{.info_tabs_module .active .tab_title{margin-bottom:.2em}}.info_tabs_module .info_tabs_content *:first-child,.info_tabs_module .active .mobile_tab_detail *:first-child{margin-top:0}.info_tabs_module .info_tabs_content>h2,.info_tabs_module .info_tabs_content>h3,.info_tabs_module .info_tabs_content>h4,.info_tabs_module .info_tabs_content>p,.info_tabs_module .active .mobile_tab_detail>h2,.info_tabs_module .active .mobile_tab_detail>h3,.info_tabs_module .active .mobile_tab_detail>h4,.info_tabs_module .active .mobile_tab_detail>p{margin:1em 0}.info_tabs_module .info_tabs_content>h2,.info_tabs_module .info_tabs_content>h3,.info_tabs_module .info_tabs_content>h4,.info_tabs_module .active .mobile_tab_detail>h2,.info_tabs_module .active .mobile_tab_detail>h3,.info_tabs_module .active .mobile_tab_detail>h4{font-size:1.1em;margin-top:2em}@media (min-width: 769px), print{.info_tabs_module .info_tabs_content>h2,.info_tabs_module .info_tabs_content>h3,.info_tabs_module .info_tabs_content>h4,.info_tabs_module .active .mobile_tab_detail>h2,.info_tabs_module .active .mobile_tab_detail>h3,.info_tabs_module .active .mobile_tab_detail>h4{margin-top:0}}.info_tabs_module .info_tabs_content>h3,.info_tabs_module .info_tabs_content>h4,.info_tabs_module .active .mobile_tab_detail>h3,.info_tabs_module .active .mobile_tab_detail>h4{margin-top:1.5em}.info_tabs_module .info_tabs_content>p:last-child{margin-bottom:1em}.info_tabs_module .info_tabs_content ol{margin-bottom:0}.info_tabs_module .active .mobile_tab_detail>p:last-child{margin-bottom:0}.info_tabs_module .less_option,.info_tabs_module .more_option{display:inline-block;font-size:.95rem;margin-bottom:0.5em}@media (min-width: 769px), print{.info_tabs_module .less_option,.info_tabs_module .more_option{display:none}}.info_tabs_module .less_option:after{content:\"- less\";display:block}.info_tabs_module .more_option:after{content:\"+ more\";display:block}.info_tabs_module footer{width:90%;max-width:1330px;position:absolute;bottom:50px;right:0;left:0;margin:auto;text-align:right}.info_tabs_module .more_link{font-size:1rem;font-weight:500;text-transform:uppercase;color:white;cursor:pointer}@media (min-width: 769px){.parallax_categorized_teaser .bubble_container{max-height:788px}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px){.parallax_categorized_teaser .bubble_container .oculus{transform:scale(0.8)}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(1){top:-30px;left:-20px}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(2){top:-30px;left:224px}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(3){top:214px;left:-20px}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(4){top:214px;left:224px}}@media (min-width: 769px) and (min-width: 1024px), print and (min-width: 769px){.parallax_categorized_teaser .bubble_container .oculus{transform:scale(0.9)}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(1){top:0;left:119px}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(2){top:70px;left:408px}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(3){left:0;top:269px}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(4){top:347px;left:290px}}@media (min-width: 769px) and (min-width: 1200px){.parallax_categorized_teaser .bubble_container .oculus{position:absolute;transform:scale(1)}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(1){top:0;left:179px}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(2){top:104px;left:495px}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(3){top:282px;left:0}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(4){top:390px;left:324px}}@media (min-width: 769px) and (min-width: 1700px){.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(1){top:0;left:229px}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(2){top:134px;left:565px}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(3){top:272px;left:0}.parallax_categorized_teaser .bubble_container .oculus:nth-of-type(4){top:420px;left:354px}}@media (min-width: 769px){.parallax_categorized_teaser .bubble_container{height:75vh}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px){.parallax_categorized_teaser .definition_teasers.two .oculus{transform:scale(0.8)}.parallax_categorized_teaser .definition_teasers.two .oculus:nth-of-type(1){top:-30px;left:-20px}.parallax_categorized_teaser .definition_teasers.two .oculus:nth-of-type(2){top:-30px;left:224px}}@media (min-width: 769px) and (min-width: 1024px), print and (min-width: 769px){.parallax_categorized_teaser .definition_teasers.two .oculus{transform:scale(0.9)}.parallax_categorized_teaser .definition_teasers.two .oculus:nth-of-type(1){top:30px;left:70px}.parallax_categorized_teaser .definition_teasers.two .oculus:nth-of-type(2){top:30px;left:378px}}@media (min-width: 769px) and (min-width: 1200px){.parallax_categorized_teaser .definition_teasers.two .oculus{position:absolute;transform:scale(1)}.parallax_categorized_teaser .definition_teasers.two .oculus:nth-of-type(1){top:40px;left:70px}.parallax_categorized_teaser .definition_teasers.two .oculus:nth-of-type(2){top:40px;left:422px}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px){.parallax_categorized_teaser .definition_teasers.three .oculus{transform:scale(0.8)}.parallax_categorized_teaser .definition_teasers.three .oculus:nth-of-type(1){top:-30px;left:100px}.parallax_categorized_teaser .definition_teasers.three .oculus:nth-of-type(2){top:185px;left:-20px}.parallax_categorized_teaser .definition_teasers.three .oculus:nth-of-type(3){top:185px;left:224px}}@media (min-width: 769px) and (min-width: 1024px), print and (min-width: 769px){.parallax_categorized_teaser .definition_teasers.three .oculus{transform:scale(0.9)}.parallax_categorized_teaser .definition_teasers.three .oculus:nth-of-type(1){top:0;left:204px}.parallax_categorized_teaser .definition_teasers.three .oculus:nth-of-type(2){top:280px;left:30px}.parallax_categorized_teaser .definition_teasers.three .oculus:nth-of-type(3){top:280px;left:378px}}@media (min-width: 769px) and (min-width: 1200px){.parallax_categorized_teaser .definition_teasers.three .oculus{position:absolute;transform:scale(1)}.parallax_categorized_teaser .definition_teasers.three .oculus:nth-of-type(1){top:20px;left:226px}.parallax_categorized_teaser .definition_teasers.three .oculus:nth-of-type(2){top:310px;left:30px}.parallax_categorized_teaser .definition_teasers.three .oculus:nth-of-type(3){top:310px;left:432px}}@media (min-width: 769px){.parallax_categorized_teaser .definition_teasers.four{max-height:788px}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px){.parallax_categorized_teaser .definition_teasers.four .oculus{transform:scale(0.8)}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(1){top:-30px;left:-20px}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(2){top:-30px;left:224px}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(3){top:214px;left:-20px}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(4){top:214px;left:224px}}@media (min-width: 769px) and (min-width: 1024px), print and (min-width: 769px){.parallax_categorized_teaser .definition_teasers.four .oculus{transform:scale(0.9)}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(1){top:0;left:119px}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(2){top:70px;left:408px}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(3){left:0;top:269px}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(4){top:347px;left:290px}}@media (min-width: 769px) and (min-width: 1200px){.parallax_categorized_teaser .definition_teasers.four .oculus{position:absolute;transform:scale(1)}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(1){top:0;left:179px}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(2){top:104px;left:495px}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(3){top:282px;left:0}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(4){top:390px;left:324px}}@media (min-width: 769px) and (min-width: 1700px){.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(1){top:0;left:229px}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(2){top:134px;left:565px}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(3){top:272px;left:0}.parallax_categorized_teaser .definition_teasers.four .oculus:nth-of-type(4){top:420px;left:354px}}@media (min-width: 769px) and (min-width: 769px), print and (min-width: 769px){.parallax_categorized_teaser .definition_teasers.five .oculus{transform:scale(0.8)}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(1){top:-30px;left:-20px}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(2){top:-30px;left:224px}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(3){top:214px;left:-20px}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(4){top:214px;left:224px}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(5){top:432px;left:102px}}@media (min-width: 769px) and (min-width: 1024px), print and (min-width: 769px){.parallax_categorized_teaser .definition_teasers.five .oculus{transform:scale(0.9)}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(1){top:0;left:0}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(2){top:0;left:408px}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(3){top:200px;left:204px}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(4){top:400px;left:0}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(5){top:400px;left:408px}}@media (min-width: 769px) and (min-width: 1200px){.parallax_categorized_teaser .definition_teasers.five .oculus{position:absolute;transform:scale(1)}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(1){top:0;left:0}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(2){top:0;left:452px}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(3){top:200px;left:226px}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(4){top:400px;left:0}.parallax_categorized_teaser .definition_teasers.five .oculus:nth-of-type(5){top:400px;left:452px}}.parallax_categorized_teaser{height:auto;background:url(\"https://mars.nasa.gov/assets/red_planet_bg.png\") center no-repeat;background-size:cover;color:#eeaaa1;overflow:hidden;position:relative;padding-top:3em}@media (min-width: 769px){.parallax_categorized_teaser{height:calc(100vh - 74px);padding-top:4em;min-height:860px}}@media (min-width: 1200px){.parallax_categorized_teaser{padding-top:5em}}@media (min-width: 769px){.parallax_categorized_teaser .module_content{height:calc(100% - 36px)}}.parallax_categorized_teaser .module_content.fixed{position:fixed;top:82px;left:0}.parallax_categorized_teaser .module_title,.parallax_categorized_teaser .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .parallax_categorized_teaser .carousel_title{font-size:2em;font-weight:200;text-align:center;margin-bottom:0.85em}@media (min-width: 769px){.parallax_categorized_teaser .module_title,.parallax_categorized_teaser .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .parallax_categorized_teaser .carousel_title{text-align:left;margin-bottom:1em}}@media (min-width: 1200px){.parallax_categorized_teaser .module_title,.parallax_categorized_teaser .main_carousel.module .carousel_header .carousel_title,.main_carousel.module .carousel_header .parallax_categorized_teaser .carousel_title{font-size:2.2em}}.parallax_categorized_teaser .mobile_only{display:block}@media (min-width: 769px){.parallax_categorized_teaser .mobile_only{display:none}}.parallax_categorized_teaser .categorized_content{width:100%;top:0;left:0;right:0;bottom:0;margin-bottom:0}@media (min-width: 769px){.parallax_categorized_teaser .categorized_content{width:55%;max-width:795px;position:absolute;top:80px;left:38%}}@media (min-width: 769px) and (min-width: 600px), print and (min-width: 769px){.parallax_categorized_teaser .categorized_content{width:58%}}@media (min-width: 769px) and (min-width: 1024px), print and (min-width: 769px){.parallax_categorized_teaser .categorized_content{left:33.5%;width:61.2%}}@media (min-width: 769px) and (min-width: 1200px){.parallax_categorized_teaser .categorized_content{left:34.5%;top:100px}}@media (min-width: 769px) and (min-width: 1700px){.parallax_categorized_teaser .categorized_content{position:relative;left:0;top:20px}}.parallax_categorized_teaser .categorized_content .content_for{position:relative;display:none}@media (min-width: 769px){.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents{padding:0 40px 0 0;max-height:71vh;overflow:hidden;overflow-y:auto}.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents::-webkit-scrollbar{width:5px}.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents::-webkit-scrollbar-thumb{background-color:rgba(255,255,255,0.4)}.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents::-webkit-scrollbar-track{background-color:rgba(255,255,255,0.1)}}.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h2,.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h3,.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h4,.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>p{margin:1em 1.5em}.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h2,.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h3,.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h4{color:#eeaaa1;font-size:1.1em;font-weight:400;margin-top:2em}@media (min-width: 769px){.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h2,.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h3,.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h4{margin-top:0}}.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h2{text-transform:uppercase}.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h3,.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>h4{margin-top:1.5em}.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>p{color:#e3e3e3;max-width:640px}.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>p:last-child{margin-bottom:2em}@media (min-width: 769px){.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>p:last-child{margin-bottom:1em}}.parallax_categorized_teaser .categorized_content .content_for .bubble_container_contents>p a{color:#54b3da}.parallax_categorized_teaser .categorized_content .content_for.current{display:block}@media (min-width: 769px){.parallax_categorized_teaser .categorized_content .content_for{position:absolute;display:block;opacity:0;left:1000px;top:0;transition:all 400ms;width:100%}.parallax_categorized_teaser .categorized_content .content_for.current{left:0;top:0;opacity:1;display:block !important}}.parallax_categorized_teaser .carousels{height:100%;width:100%}.parallax_categorized_teaser footer{position:absolute;height:50px;bottom:0;left:0;width:100vw;background-color:rgba(129,33,24,0.6)}.parallax_categorized_teaser .oculus{position:relative;text-align:center;background-size:cover;height:186px;padding:1.5em 0;border-bottom:3px solid #5fb4ce;margin-bottom:0;overflow:hidden;background-color:rgba(32,52,66,0.39);color:#c7d9de}@media (min-width: 769px){.parallax_categorized_teaser .oculus{padding:0;position:absolute;width:275px;height:275px}}.parallax_categorized_teaser .oculus h3.carousel_title{position:absolute;font-size:.9rem;text-align:center;margin:0 auto 12px;width:auto;font-weight:300}@media (max-width: 768px){.parallax_categorized_teaser .oculus h3.carousel_title{left:50%;transform:translateX(-50%)}}@media (min-width: 769px){.parallax_categorized_teaser .oculus h3.carousel_title{position:relative;width:50%;margin:24px auto 12px}}.parallax_categorized_teaser .oculus h3.description_title{font-size:.9rem;font-weight:400;margin:0 0 12px}@media (min-width: 769px){.parallax_categorized_teaser .oculus h3.description_title{margin:17px 0 12px}}.parallax_categorized_teaser .oculus span.mobile_only_title{font-weight:200}@media (min-width: 769px){.parallax_categorized_teaser .oculus span.mobile_only_title{display:none}}.parallax_categorized_teaser .oculus .slide_description{padding:0 2em;font-weight:400}.parallax_categorized_teaser .oculus .cols{display:flex;margin-left:-6px;margin-right:-6px}.parallax_categorized_teaser .oculus .cols .col{margin:0 auto;max-width:55%}.parallax_categorized_teaser .oculus .cols .val{font-size:2rem;font-weight:400;letter-spacing:-.05em}.parallax_categorized_teaser .oculus .cols .val.small_text{font-size:1.6rem}.parallax_categorized_teaser .oculus .cols .val_label{font-size:.9rem;font-weight:300}@media (min-width: 769px){.parallax_categorized_teaser .oculus{border:3px solid #5fb4ce;border-radius:50%}}.parallax_categorized_teaser .oculus .carousel_title{color:#445c64}.parallax_categorized_teaser .oculus .title,.parallax_categorized_teaser .oculus .hover{color:#c7d9de}.parallax_categorized_teaser .oculus .slick-prev:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #5998ac;display:inline-block;transform:rotate(180deg)}.parallax_categorized_teaser .oculus .slick-prev:before,.parallax_categorized_teaser .oculus .slick-next:before{background-image:none}.parallax_categorized_teaser .oculus .slick-next:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #5998ac;display:inline-block;transform:none}.parallax_categorized_teaser .oculus.data{border-bottom:3px solid #5fb4ce;margin-bottom:0;overflow:hidden;background-color:rgba(32,52,66,0.39);color:#c7d9de}@media (min-width: 769px){.parallax_categorized_teaser .oculus.data{border:3px solid #5fb4ce;border-radius:50%}}.parallax_categorized_teaser .oculus.data .carousel_title{color:#445c64}.parallax_categorized_teaser .oculus.data .title,.parallax_categorized_teaser .oculus.data .hover{color:#c7d9de}.parallax_categorized_teaser .oculus.data .slick-prev:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #5998ac;display:inline-block;transform:rotate(180deg)}.parallax_categorized_teaser .oculus.data .slick-prev:before,.parallax_categorized_teaser .oculus.data .slick-next:before{background-image:none}.parallax_categorized_teaser .oculus.data .slick-next:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #5998ac;display:inline-block;transform:none}.parallax_categorized_teaser .oculus.data h3.carousel_title{display:none}@media (min-width: 769px){.parallax_categorized_teaser .oculus.data h3.carousel_title{display:block}}.parallax_categorized_teaser .oculus.data .carousel_title{color:#6b93a0}.parallax_categorized_teaser .oculus.data .val{color:white}.parallax_categorized_teaser .oculus.data .note{margin-top:15px;font-size:.75rem;font-weight:300}@media (min-width: 769px){.parallax_categorized_teaser .oculus.data .slick-slide.slide{background-color:rgba(95,180,206,0.1)}}.parallax_categorized_teaser .oculus.images{border-bottom:3px solid #f56b60;margin-bottom:0;overflow:hidden;background-color:rgba(0,0,0,0.39);color:#f06c60;padding:0}@media (min-width: 769px){.parallax_categorized_teaser .oculus.images{border:3px solid #f56b60;border-radius:50%}}.parallax_categorized_teaser .oculus.images .carousel_title{color:#f06c60}.parallax_categorized_teaser .oculus.images .title,.parallax_categorized_teaser .oculus.images .hover{color:#ffddcf}.parallax_categorized_teaser .oculus.images .slick-prev:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #f06c60;display:inline-block;transform:rotate(180deg)}.parallax_categorized_teaser .oculus.images .slick-prev:before,.parallax_categorized_teaser .oculus.images .slick-next:before{background-image:none}.parallax_categorized_teaser .oculus.images .slick-next:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #f06c60;display:inline-block;transform:none}.parallax_categorized_teaser .oculus.images h3.carousel_title{z-index:1;background-color:rgba(86,42,42,0.65);text-align:center;padding:5px 14px;border-radius:6px;margin-top:1.2em}@media (max-width: 768px){.parallax_categorized_teaser .oculus.images h3.carousel_title{color:#ffbdb7}}@media (min-width: 769px){.parallax_categorized_teaser .oculus.images h3.carousel_title{background-color:transparent;padding:0;margin-top:24px}}.parallax_categorized_teaser .oculus.images a.hover_state{display:block;height:100%;width:100%;position:absolute;top:0}.parallax_categorized_teaser .oculus.images .rollover_description{background-color:rgba(86,42,42,0.85);padding:1em 2em;text-align:left}.parallax_categorized_teaser .oculus.images .rollover_description .title{font-size:.9rem;color:#f5dddb}.parallax_categorized_teaser .oculus.compare{border-bottom:3px solid #fcb963;margin-bottom:0;overflow:hidden;background-color:rgba(0,0,0,0.39);color:#f7ca99;background:linear-gradient(to right, rgba(53,30,1,0.46) 0%, rgba(53,30,1,0.46) 49%, rgba(10,8,9,0.47) 50%, rgba(10,8,9,0.47) 100%)}@media (min-width: 769px){.parallax_categorized_teaser .oculus.compare{border:3px solid #fcb963;border-radius:50%}}.parallax_categorized_teaser .oculus.compare .carousel_title{color:#f7ca99}.parallax_categorized_teaser .oculus.compare .title,.parallax_categorized_teaser .oculus.compare .hover{color:#f7ca99}.parallax_categorized_teaser .oculus.compare .slick-prev:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #f5b460;display:inline-block;transform:rotate(180deg)}.parallax_categorized_teaser .oculus.compare .slick-prev:before,.parallax_categorized_teaser .oculus.compare .slick-next:before{background-image:none}.parallax_categorized_teaser .oculus.compare .slick-next:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #f5b460;display:inline-block;transform:none}.parallax_categorized_teaser .oculus.compare h3.carousel_title{display:none}@media (min-width: 769px){.parallax_categorized_teaser .oculus.compare h3.carousel_title{display:block}}.parallax_categorized_teaser .oculus.compare .thing_one .val{color:white}.parallax_categorized_teaser .oculus.compare .thing_two{color:#66c8eb}.parallax_categorized_teaser .oculus.compare .thing{font-weight:300}.parallax_categorized_teaser .oculus.compare .details{font-size:.9rem;font-weight:300}.parallax_categorized_teaser .oculus.news{border-bottom:3px solid #f56b60;margin-bottom:0;overflow:hidden;background-color:rgba(0,0,0,0.39);color:#f06c60}@media (min-width: 769px){.parallax_categorized_teaser .oculus.news{border:3px solid #f56b60;border-radius:50%}}.parallax_categorized_teaser .oculus.news .carousel_title{color:#f06c60}.parallax_categorized_teaser .oculus.news .title,.parallax_categorized_teaser .oculus.news .hover{color:#ffddcf}.parallax_categorized_teaser .oculus.news .slick-prev:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #f06c60;display:inline-block;transform:rotate(180deg)}.parallax_categorized_teaser .oculus.news .slick-prev:before,.parallax_categorized_teaser .oculus.news .slick-next:before{background-image:none}.parallax_categorized_teaser .oculus.news .slick-next:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #f06c60;display:inline-block;transform:none}.parallax_categorized_teaser .oculus.news .slide_description{text-align:left}@media (max-width: 768px){.parallax_categorized_teaser .oculus.news .slide_description{margin-top:32px;padding:0 18%}}.parallax_categorized_teaser .oculus.news .slide_description a{color:#f5dddb}.parallax_categorized_teaser .oculus.news .slide_description .description{font-size:.9rem;font-weight:400}@media (min-width: 769px){.parallax_categorized_teaser .oculus.news .slick-slide.slide{background-color:rgba(86,42,42,0.58)}}.parallax_categorized_teaser .more_bar{background-color:rgba(129,33,24,0.4);transition:background-color 200ms}.parallax_categorized_teaser .more_bar:hover{background-color:rgba(129,33,24,0.7)}.parallax_categorized_teaser .more_bar .arrow_down{padding:0;cursor:pointer;width:25px;height:25px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -50px -100px;background-size:300px}.parallax_categorized_teaser .more_bar .arrow_down:hover,.parallax_categorized_teaser .more_bar .arrow_down.active,.parallax_categorized_teaser .more_bar .arrow_down.current{background-position:-50px -100px}.parallax_categorized_teaser .slick-prev,.parallax_categorized_teaser .slick-next{top:calc(50% - 20px)}@media (min-width: 769px){.parallax_categorized_teaser .slick-prev,.parallax_categorized_teaser .slick-next{top:83%}}.parallax_categorized_teaser .slick-slider .slick-prev{left:2%}@media (min-width: 769px){.parallax_categorized_teaser .slick-slider .slick-prev{left:calc(50% - 35px)}}.parallax_categorized_teaser .slick-slider .slick-next{right:2%}@media (min-width: 769px){.parallax_categorized_teaser .slick-slider .slick-next{right:calc(50% - 35px)}}.parallax_categorized_teaser .slick-slide.slide{overflow:hidden;height:186px}@media (min-width: 769px){.parallax_categorized_teaser .slick-slide.slide{height:160px}}.parallax_categorized_teaser .definition_teasers .oculus{cursor:pointer}.parallax_categorized_teaser .definition_teasers .oculus.warm_color_theme{border-bottom:3px solid #fcb963;margin-bottom:0;overflow:hidden;background-color:rgba(0,0,0,0.39);color:#f7ca99}@media (min-width: 769px){.parallax_categorized_teaser .definition_teasers .oculus.warm_color_theme{border:3px solid #fcb963;border-radius:50%}}.parallax_categorized_teaser .definition_teasers .oculus.warm_color_theme .carousel_title{color:#f7ca99}.parallax_categorized_teaser .definition_teasers .oculus.warm_color_theme .title,.parallax_categorized_teaser .definition_teasers .oculus.warm_color_theme .hover{color:#f7ca99}.parallax_categorized_teaser .definition_teasers .oculus.warm_color_theme .slick-prev:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #f5b460;display:inline-block;transform:rotate(180deg)}.parallax_categorized_teaser .definition_teasers .oculus.warm_color_theme .slick-prev:before,.parallax_categorized_teaser .definition_teasers .oculus.warm_color_theme .slick-next:before{background-image:none}.parallax_categorized_teaser .definition_teasers .oculus.warm_color_theme .slick-next:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #f5b460;display:inline-block;transform:none}.parallax_categorized_teaser .definition_teasers .oculus.cool_color_theme{border-bottom:3px solid #5fb4ce;margin-bottom:0;overflow:hidden;background-color:rgba(32,52,66,0.39);color:#c7d9de}@media (min-width: 769px){.parallax_categorized_teaser .definition_teasers .oculus.cool_color_theme{border:3px solid #5fb4ce;border-radius:50%}}.parallax_categorized_teaser .definition_teasers .oculus.cool_color_theme .carousel_title{color:#445c64}.parallax_categorized_teaser .definition_teasers .oculus.cool_color_theme .title,.parallax_categorized_teaser .definition_teasers .oculus.cool_color_theme .hover{color:#c7d9de}.parallax_categorized_teaser .definition_teasers .oculus.cool_color_theme .slick-prev:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #5998ac;display:inline-block;transform:rotate(180deg)}.parallax_categorized_teaser .definition_teasers .oculus.cool_color_theme .slick-prev:before,.parallax_categorized_teaser .definition_teasers .oculus.cool_color_theme .slick-next:before{background-image:none}.parallax_categorized_teaser .definition_teasers .oculus.cool_color_theme .slick-next:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #5998ac;display:inline-block;transform:none}.parallax_categorized_teaser .definition_teasers .oculus.sunset_color_theme{border-bottom:3px solid #f56b60;margin-bottom:0;overflow:hidden;background-color:rgba(0,0,0,0.39);color:#f06c60}@media (min-width: 769px){.parallax_categorized_teaser .definition_teasers .oculus.sunset_color_theme{border:3px solid #f56b60;border-radius:50%}}.parallax_categorized_teaser .definition_teasers .oculus.sunset_color_theme .carousel_title{color:#f06c60}.parallax_categorized_teaser .definition_teasers .oculus.sunset_color_theme .title,.parallax_categorized_teaser .definition_teasers .oculus.sunset_color_theme .hover{color:#ffddcf}.parallax_categorized_teaser .definition_teasers .oculus.sunset_color_theme .slick-prev:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #f06c60;display:inline-block;transform:rotate(180deg)}.parallax_categorized_teaser .definition_teasers .oculus.sunset_color_theme .slick-prev:before,.parallax_categorized_teaser .definition_teasers .oculus.sunset_color_theme .slick-next:before{background-image:none}.parallax_categorized_teaser .definition_teasers .oculus.sunset_color_theme .slick-next:before{width:0;height:0;border-top:9px solid transparent !important;border-bottom:9px solid transparent !important;border-left:9px solid #f06c60;display:inline-block;transform:none}.parallax_categorized_teaser .definition_teasers .title{margin-bottom:12px;font-size:1.6em}.no-touchevents .parallax_categorized_teaser .definition_teasers .title{margin-bottom:0}.parallax_categorized_teaser .definition_teasers .hover{font-size:1.2em}.parallax_categorized_teaser .definition_teasers .title,.parallax_categorized_teaser .definition_teasers .hover{margin-top:0;margin-right:10%;margin-left:10%;top:30%;position:relative;font-weight:300}.parallax_categorized_teaser .definition_teasers .title_container{position:relative;top:50%;transform:translateY(-50%)}.no-touchevents .parallax_categorized_teaser .definition_teasers .hover{display:none}.parallax_categorized_teaser .definition_teasers .bg_container{position:absolute;top:0;left:0;width:100%;height:100%;opacity:0.25;background-size:cover;transition:filter .4s;filter:brightness(60%)}@media (min-width: 769px){.parallax_categorized_teaser .definition_teasers .bg_container{filter:brightness(100%)}}.no-touchevents .parallax_categorized_teaser .definition_teasers .oculus:hover .hover{display:block}.no-touchevents .parallax_categorized_teaser .definition_teasers .oculus:hover .title{display:none}.no-touchevents .parallax_categorized_teaser .definition_teasers .oculus:hover .bg_container{filter:brightness(60%)}.parallax_categorized_teaser .definition_teasers .mobile_detailed_definition{display:none;position:relative}.parallax_categorized_teaser .detailed_def_visible .detailed_definition{padding:1.4em .8em;width:98%;position:relative;display:none}@media (min-width: 769px){.parallax_categorized_teaser .detailed_def_visible .detailed_definition{padding:0;max-width:680px;width:90%;left:3%;float:left;display:block}}@media (min-width: 1024px), print{.parallax_categorized_teaser .detailed_def_visible .detailed_definition{max-width:740px;left:6%}}.parallax_categorized_teaser .detailed_def_visible .mobile_detailed_definition{display:none;padding:0 1.5em;text-align:left}@media (min-width: 769px){.parallax_categorized_teaser .detailed_def_visible .mobile_detailed_definition{display:none}}.parallax_categorized_teaser .detailed_def_visible .mobile_detailed_definition .mobile_detailed_definition_contents a{color:#54b3da}.parallax_categorized_teaser .detailed_def_visible .close_button{display:block;height:35px;width:33px;position:absolute;right:0;top:0;padding:3px;text-decoration:none;-webkit-touch-callout:none;-webkit-user-select:none;-khtml-user-select:none;-moz-user-select:-moz-none;-ms-user-select:none;user-select:none}@media (min-width: 769px){.parallax_categorized_teaser .detailed_def_visible .close_button{top:0;right:-5px}}.parallax_categorized_teaser .detailed_def_visible .close_button .close_icon{display:block;padding:0;cursor:pointer;width:25px;height:25px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -25px 0;background-size:300px}.parallax_categorized_teaser .detailed_def_visible .close_button .close_icon:hover,.parallax_categorized_teaser .detailed_def_visible .close_button .close_icon.active,.parallax_categorized_teaser .detailed_def_visible .close_button .close_icon.current{background-position:-25px 0}.parallax_categorized_teaser .detailed_def_visible p{color:#e3e3e3}.parallax_categorized_teaser .detailed_def_visible .detailed_def_title{padding:0 3%;text-decoration:none;cursor:pointer;color:#eeaaa1;font-size:1.1em;font-weight:400;display:block;text-transform:uppercase}@media (min-width: 769px){.parallax_categorized_teaser .detailed_def_visible .detailed_def_title{padding:0 10% 0 0;margin:.1em 0 0.5em;transition:color 200ms}.parallax_categorized_teaser .detailed_def_visible .detailed_def_title:hover{color:white}}.parallax_categorized_teaser .detailed_def_visible .definition_contents{padding:3%}@media (min-width: 769px){.parallax_categorized_teaser .detailed_def_visible .definition_contents{padding:0 20px 0 0}}.parallax_categorized_teaser .detailed_def_visible .definition_contents h1,.parallax_categorized_teaser .detailed_def_visible .definition_contents h2,.parallax_categorized_teaser .detailed_def_visible .definition_contents h3,.parallax_categorized_teaser .detailed_def_visible .definition_contents h4,.parallax_categorized_teaser .detailed_def_visible .definition_contents h5{font-weight:500}.parallax_categorized_teaser .detailed_def_visible .definition_contents h3{font-size:1.1em;margin:0.6em 0 0.6em}.parallax_categorized_teaser .detailed_def_visible .definition_contents h3:first-child{margin-top:0}.parallax_categorized_teaser .detailed_def_visible .definition_contents a{color:#54b3da}.parallax_categorized_teaser .detailed_def_visible .oculus{display:block}@media (min-width: 769px){.parallax_categorized_teaser .detailed_def_visible .oculus{display:none}}.parallax_categorized_teaser .detailed_def_visible .oculus.current{height:auto;pointer-events:none}.parallax_categorized_teaser .detailed_def_visible .oculus.current a{pointer-events:auto}.parallax_categorized_teaser .detailed_def_visible .oculus.current h2.title,.parallax_categorized_teaser .detailed_def_visible .oculus.current .hover{display:none !important}.parallax_categorized_teaser .detailed_def_visible .oculus.current .mobile_detailed_definition{display:block}.parallax_categorized_teaser .detailed_def_visible .oculus.current .mobile_detailed_definition h1,.parallax_categorized_teaser .detailed_def_visible .oculus.current .mobile_detailed_definition h2,.parallax_categorized_teaser .detailed_def_visible .oculus.current .mobile_detailed_definition h3,.parallax_categorized_teaser .detailed_def_visible .oculus.current .mobile_detailed_definition h4,.parallax_categorized_teaser .detailed_def_visible .oculus.current .mobile_detailed_definition h5{font-weight:500}.parallax_categorized_teaser .detailed_def_visible .oculus.current .mobile_detailed_definition .mobile_detailed_definition_title{margin-top:0;width:85%;font-weight:300}.parallax_categorized_teaser .detailed_def_visible .oculus.current .mobile_detailed_definition .close_button{pointer-events:auto;right:1em}.parallax_categorized_teaser .detailed_def_visible .detailed_def_nav{margin-bottom:1.5em;width:93%}.parallax_categorized_teaser .detailed_def_visible .detailed_def_nav li{cursor:pointer;display:inline-block;font-weight:300;font-size:1.1em;line-height:1.7em;margin-right:1em;transition:color 200ms}.parallax_categorized_teaser .detailed_def_visible .detailed_def_nav li:hover{color:white}.parallax_categorized_teaser .detailed_def_visible .detailed_def_nav li:last-child{margin-right:0}.parallax_categorized_teaser .detailed_def_visible .detailed_def_nav .current{color:white}#iframe_overlay .explore_overlay_page{background-color:black}@keyframes pulse{0%{opacity:1}50%{opacity:.3}100%{opacity:1}}.dsn_connection .target_name,.mars_relay_connection .target_name{display:inline-block;width:calc(100% - 61px);vertical-align:middle;white-space:normal}.dsn_connection .info_tip ~ .dsn_title,.dsn_connection .info_tip ~ .mars_relay_title,.mars_relay_connection .info_tip ~ .dsn_title,.mars_relay_connection .info_tip ~ .mars_relay_title{float:left}.dsn_connection .mars_relay_title,.dsn_connection .dsn_title,.mars_relay_connection .mars_relay_title,.mars_relay_connection .dsn_title{max-width:calc(100% - 30px);display:inline-block}.dsn_connection .signal,.mars_relay_connection .signal{font-size:13px;margin:0.7em 0;white-space:nowrap;font-weight:300}@media (min-width: 480px){.dsn_connection .signal,.mars_relay_connection .signal{font-size:15px}}.dsn_connection .signal .dsn_icon,.mars_relay_connection .signal .dsn_icon{display:inline-block;width:54px;height:54px;vertical-align:middle;margin-right:0.8em}.homepage_dashboard_modal .dsn_connection .signal .sending_receiving,.homepage_dashboard_modal .dsn_connection .signal .sending,.homepage_dashboard_modal .dsn_connection .signal .receiving,.homepage_dashboard_modal .mars_relay_connection .signal .sending_receiving,.homepage_dashboard_modal .mars_relay_connection .signal .sending,.homepage_dashboard_modal .mars_relay_connection .signal .receiving{color:white}.homepage_dashboard_modal .dsn_connection .signal .sending_receiving .target_name,.homepage_dashboard_modal .dsn_connection .signal .sending .target_name,.homepage_dashboard_modal .dsn_connection .signal .receiving .target_name,.homepage_dashboard_modal .mars_relay_connection .signal .sending_receiving .target_name,.homepage_dashboard_modal .mars_relay_connection .signal .sending .target_name,.homepage_dashboard_modal .mars_relay_connection .signal .receiving .target_name{color:white}.homepage_dashboard_modal .dsn_connection .signal .transmissions,.homepage_dashboard_modal .dsn_connection .signal .target_name,.homepage_dashboard_modal .mars_relay_connection .signal .transmissions,.homepage_dashboard_modal .mars_relay_connection .signal .target_name{color:#fcb963}.dsn_connection .signal .sending_receiving,.mars_relay_connection .signal .sending_receiving{animation:pulse 1s infinite}.dsn_connection .signal .sending_receiving .dsn_icon,.mars_relay_connection .signal .sending_receiving .dsn_icon{background:url(\"https://mars.nasa.gov/assets/[email protected]\") -146px -130px no-repeat;background-size:300px}.dsn_connection .signal .sending_receiving .target_name:before,.mars_relay_connection .signal .sending_receiving .target_name:before{content:\" SENDING/RECEIVING\"}.dsn_connection .signal .sending,.mars_relay_connection .signal .sending{animation:pulse 1s infinite}.dsn_connection .signal .sending .dsn_icon,.mars_relay_connection .signal .sending .dsn_icon{background:url(\"https://mars.nasa.gov/assets/[email protected]\") 0 -130px no-repeat;background-size:300px}.dsn_connection .signal .sending .target_name:before,.mars_relay_connection .signal .sending .target_name:before{content:\" SENDING\"}.dsn_connection .signal .receiving,.mars_relay_connection .signal .receiving{animation:pulse 1s infinite}.dsn_connection .signal .receiving .dsn_icon,.mars_relay_connection .signal .receiving .dsn_icon{background:url(\"https://mars.nasa.gov/assets/[email protected]\") -73px -130px no-repeat;background-size:300px}.dsn_connection .signal .receiving .target_name:before,.mars_relay_connection .signal .receiving .target_name:before{content:\" RECEIVING\"}.dsn_connection .signal .disconnected .dsn_icon,.mars_relay_connection .signal .disconnected .dsn_icon{background:url(\"https://mars.nasa.gov/assets/[email protected]\") -219px -130px no-repeat;background-size:300px}.dsn_connection .signal .disconnected .target_name:before,.mars_relay_connection .signal .disconnected .target_name:before{content:\" AWAITING NEXT TRANSMISSION\"}.homepage_dashboard_modal .dsn_connection .signal .dsn_icon,.homepage_dashboard_modal .mars_relay_connection .signal .dsn_icon{background-position-y:0px}#iframe_overlay .dsn_connection .signal .dsn_icon,#iframe_overlay .mars_relay_connection .signal .dsn_icon{background-position-y:-62px}.dsn_connection .info_tip,.mars_relay_connection .info_tip{display:inline-block;margin-left:10px}#secondary_column .dsn_connection .info_tip,#secondary_column .mars_relay_connection .info_tip{margin-left:0}@media (min-width: 1024px), print{#secondary_column .dsn_connection .info_tip,#secondary_column .mars_relay_connection .info_tip{margin-left:10px}}.dsn_connection .info_tip .info_icon,.mars_relay_connection .info_tip .info_icon{display:block;width:25px;height:25px;background:url(\"https://mars.nasa.gov/assets/[email protected]\") -100px 0;background-size:300px;position:absolute;top:-1px}#secondary_column .dsn_connection .info_tip .info_icon,#secondary_column .mars_relay_connection .info_tip .info_icon{right:0;top:-3px}#primary_column .dsn_connection .info_tip .info_icon,#primary_column .mars_relay_connection .info_tip .info_icon{top:2px}#primary_column .dsn_connection .info_tip .info_text,#primary_column .mars_relay_connection .info_tip .info_text{top:-13px}.dsn_connection .info_tip .info_text,.mars_relay_connection .info_tip .info_text{display:none}.dsn_connection .info_tip.open .info_icon:before,.mars_relay_connection .info_tip.open .info_icon:before{display:none}@media (min-width: 480px){.dsn_connection .info_tip.open .info_icon:before,.mars_relay_connection .info_tip.open .info_icon:before{display:block;content:'';position:absolute;right:calc(50% - 7.5px);width:15px;height:15px;background:white;border-right:1px solid #c9c9c9;border-bottom:1px solid #c9c9c9;z-index:1;transform:rotate(45deg);top:-23px}}@media (min-width: 769px), print{.dsn_connection .info_tip.open .info_icon:before,.mars_relay_connection .info_tip.open .info_icon:before{top:-23px}}@media (min-width: 1200px){.dsn_connection .info_tip.open .info_icon:before,.mars_relay_connection .info_tip.open .info_icon:before{top:-24px}}.dsn_connection .info_tip.open .info_text,.mars_relay_connection .info_tip.open .info_text{display:block;position:absolute;color:#222;font-size:.9em;background-color:white;padding:1.3em;border:1px solid #c9c9c9;border-radius:4px;top:-17px;transform:translateY(-100%);width:100%;left:0}@media (min-width: 1024px){.dsn_connection .info_tip.open .info_text,.mars_relay_connection .info_tip.open .info_text{font-size:.8em}}.homepage_dashboard_modal .dsn_connection .info_tip.open .info_icon:before,.homepage_dashboard_modal .mars_relay_connection .info_tip.open .info_icon:before{background-color:#221307;border-right:1px solid #5f4326;border-bottom:1px solid #5f4326}.homepage_dashboard_modal .dsn_connection .info_tip.open .info_text,.homepage_dashboard_modal .mars_relay_connection .info_tip.open .info_text{color:#beb0a4;background-color:#221307;border:1px solid #5f4326}.mars_relay_connection .transmissions{display:inline-block;vertical-align:middle;width:calc(100% - 70px);white-space:normal}.parallax_categorized_teaser{font-weight:400}.parallax_categorized_teaser nav.mobile_catcont_nav{display:block}@media (min-width: 769px){.parallax_categorized_teaser nav.mobile_catcont_nav{display:none}}.parallax_categorized_teaser nav.desktop_catcont_nav{display:none}@media (min-width: 769px){.parallax_categorized_teaser nav.desktop_catcont_nav{display:block}}.parallax_categorized_teaser .nav_content_container{display:block}@media (min-width: 769px){.parallax_categorized_teaser .nav_content_container{max-width:1300px;width:94%;margin:auto;display:flex;height:100%}}.parallax_categorized_teaser nav.catcont_nav{width:260px;margin-right:3%}@media (min-width: 1024px), print{.parallax_categorized_teaser nav.catcont_nav{width:300px;margin-right:4%}}@media (min-width: 1200px){.parallax_categorized_teaser nav.catcont_nav{width:340px;margin-right:5.5%}}@media (min-width: 1700px){.parallax_categorized_teaser nav.catcont_nav{margin-right:7.5%}}.parallax_categorized_teaser nav.catcont_nav .section{background-color:rgba(129,33,24,0.4);margin-bottom:1px;cursor:pointer;text-align:center;padding:14px 5%;transition:background-color 200ms}.no-touchevents .parallax_categorized_teaser nav.catcont_nav .section:hover:not(.current){background-color:rgba(129,33,24,0.8)}@media (min-width: 480px){.parallax_categorized_teaser nav.catcont_nav .section{padding:14px 10%}}@media (min-width: 600px), print{.parallax_categorized_teaser nav.catcont_nav .section{padding:14px 15%}}@media (min-width: 769px){.parallax_categorized_teaser nav.catcont_nav .section{text-align:left;padding:10px 15px}}.parallax_categorized_teaser nav.catcont_nav .section .nav_title,.parallax_categorized_teaser nav.catcont_nav .section .title{text-transform:uppercase;font-size:15px;letter-spacing:0}.parallax_categorized_teaser nav.catcont_nav .section.default .category_info .description{display:block}.parallax_categorized_teaser nav.catcont_nav .category_info .description{display:none;font-size:16px;margin:.5em 0}@media (min-width: 480px){.parallax_categorized_teaser nav.catcont_nav .category_info .description{font-size:15px}}@media (min-width: 769px){.parallax_categorized_teaser nav.catcont_nav .current{background-color:#822118}.parallax_categorized_teaser nav.catcont_nav .current .category_info .title{color:white}}.parallax_categorized_teaser nav.mobile_catcont_nav{width:100%}.parallax_categorized_teaser nav.mobile_catcont_nav.current .category_info .description{display:block}.parallax_categorized_teaser nav.mobile_catcont_nav .category_info .description{margin:.5em 0;display:none}.-ms-.no-flexbox .grid_view.grid_gallery .list_image img{height:auto}\n </style>\n <style data-href=\"/assets/gulp/vendor/jquery.fancybox-364352e03618ba5a8da007665b1f0be31795293b22bc4d7c5974891d4976a137.css\" media=\"screen\">\n @charset \"UTF-8\";/*! fancyBox 3.0.0 Beta 1 fancyapps.com | fancyapps.com/fancybox/#license */#fancybox-loading,#fancybox-lock,.fancybox-wrap,.fancybox-skin,.fancybox-inner,.fancybox-error,.fancybox-image{padding:0;margin:0;border:0;outline:none;vertical-align:top;background-color:transparent;background-repeat:no-repeat;background-image:none;text-shadow:none}.fancybox-wrap iframe,.fancybox-wrap object,.fancybox-wrap embed{padding:0;margin:0;border:0;outline:none;vertical-align:top;background-color:transparent;background-repeat:no-repeat;background-image:none;text-shadow:none}a.fancybox-close,a.fancybox-expand{padding:0;margin:0;border:0;outline:none;vertical-align:top;background-color:transparent;background-repeat:no-repeat;background-image:none;text-shadow:none}a.fancybox-nav{padding:0;margin:0;border:0;outline:none;vertical-align:top;background-color:transparent;background-repeat:no-repeat;background-image:none;text-shadow:none}a.fancybox-nav span{padding:0;margin:0;border:0;outline:none;vertical-align:top;background-color:transparent;background-repeat:no-repeat;background-image:none;text-shadow:none}.fancybox-tmp{padding:0;margin:0;border:0;outline:none;vertical-align:top;background-color:transparent;background-repeat:no-repeat;background-image:none;text-shadow:none}#fancybox-lock{position:fixed;top:0;left:0;right:0;bottom:0;z-index:8020;overflow-y:scroll;overflow-y:auto;overflow-x:auto;-webkit-transition:-webkit-transform 0.5s;-webkit-transform:translateX(0px)}.fancybox-lock-test{overflow-y:hidden !important}.fancybox-lock{overflow:hidden !important;width:auto}.fancybox-lock body{overflow:hidden !important}.fancybox-wrap{position:absolute;top:0;left:0;z-index:8020;-webkit-transform:translate3d(0, 0, 0)}.fancybox-opened{z-index:8030}.fancybox-skin{border-style:solid;border-color:#fff;background:#fff;color:#444}.fancybox-inner{position:relative;overflow:hidden;-webkit-overflow-scrolling:touch;width:100%;height:100%;max-width:100%;max-height:100%}.fancybox-spacer{position:absolute;top:100%;left:0;width:1px}.fancybox-image,.fancybox-iframe{display:block;width:100%;height:100%}.fancybox-image{max-width:100%;max-height:100%;zoom:1}a.fancybox-close{position:absolute;top:-23px;right:-23px;width:46px;height:46px;cursor:pointer;background-position:0 0;z-index:8040}a.fancybox-nav{position:absolute;top:0;width:50%;height:100%;cursor:pointer;text-decoration:none;-webkit-tap-highlight-color:transparent;z-index:8040;overflow:hidden}.fancybox-type-iframe a.fancybox-nav,.fancybox-type-inline a.fancybox-nav,.fancybox-type-html a.fancybox-nav{width:70px}a.fancybox-prev{left:-70px}a.fancybox-next{right:-70px}a.fancybox-nav span{position:absolute;top:50%;width:46px;height:46px;margin-top:-23px;cursor:pointer;z-index:8040}a.fancybox-prev span{left:0;background-position:0 -50px}a.fancybox-next span{right:0;background-position:0 -100px}.fancybox-mobile a.fancybox-nav{max-width:80px}.fancybox-desktop a.fancybox-nav{opacity:0.5;filter:alpha(opacity=50)}.fancybox-desktop a.fancybox-nav:hover{opacity:1;filter:alpha(opacity=100)}a.fancybox-expand{position:absolute;bottom:0;right:0;width:46px;height:46px;z-index:8050;opacity:0;filter:alpha(opacity=0);background-position:0 -150px;zoom:1;-webkit-transition:opacity .5s ease;-moz-transition:opacity .5s ease;-o-transition:opacity .5s ease;transition:opacity .5s ease}.fancybox-wrap:hover a.fancybox-expand{opacity:0.5;filter:alpha(opacity=50)}.fancybox-wrap a.fancybox-expand:hover{opacity:1;filter:alpha(opacity=100)}#fancybox-loading{position:fixed;top:50%;left:50%;margin-top:-30px;margin-left:-30px;width:60px;height:60px;background-color:#111;background-image:url(data:image/gif;base64,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);background-position:center center;opacity:0.85;filter:alpha(opacity=85);cursor:pointer;z-index:8060;-webkit-border-radius:8px;-moz-border-radius:8px;border-radius:8px}.fancybox-tmp{position:absolute !important;top:-99999px;left:-99999px;max-width:99999px;max-height:99999px;overflow:visible !important}.fancybox-title{font:normal 14px \"Helvetica Neue\",Helvetica,Arial,sans-serif;line-height:1.5;position:relative;text-shadow:none;z-index:8050;display:block;visibility:hidden}.fancybox-title-float-wrap{position:relative;margin-top:10px;text-align:center;zoom:1;left:-9999px}.fancybox-title-float-wrap>div{display:inline-block;padding:7px 20px;font-weight:bold;color:#FFF;text-shadow:0 1px 2px #222;background:transparent;background:rgba(0,0,0,0.8);-webkit-border-radius:15px;-moz-border-radius:15px;border-radius:15px}.fancybox-title-outside-wrap{position:relative;margin-top:10px;color:#fff;text-shadow:0 1px rgba(0,0,0,0.5)}.fancybox-title-inside-wrap{padding-top:10px}.fancybox-title-over-wrap{position:absolute;bottom:0;left:0;color:#fff;padding:15px;background:#000;background:rgba(0,0,0,0.8);max-height:50%;overflow:auto}.fancybox-overlay{position:absolute;top:0;left:0;overflow:hidden;z-index:8010}.fancybox-overlay-fixed{position:fixed;width:100%;height:100%}.fancybox-default-skin{border-color:#f9f9f9;background:#f9f9f9}.fancybox-default-skin-open{box-shadow:0 10px 25px rgba(0,0,0,0.5)}.fancybox-default-overlay{background:#333;opacity:0.8;filter:alpha(opacity=80)}.fancybox-default a.fancybox-close,.fancybox-default a.fancybox-expand,.fancybox-default a.fancybox-nav span{background-image:url(data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAC4AAADICAYAAACXpNOoAAAAGXRFWHRTb2Z0d2FyZQBBZG9iZSBJbWFnZVJlYWR5ccllPAAAA2ZpVFh0WE1MOmNvbS5hZG9iZS54bXAAAAAAADw/eHBhY2tldCBiZWdpbj0i77u/IiBpZD0iVzVNME1wQ2VoaUh6cmVTek5UY3prYzlkIj8+IDx4OnhtcG1ldGEgeG1sbnM6eD0iYWRvYmU6bnM6bWV0YS8iIHg6eG1wdGs9IkFkb2JlIFhNUCBDb3JlIDUuMy1jMDExIDY2LjE0NTY2MSwgMjAxMi8wMi8wNi0xNDo1NjoyNyAgICAgICAgIj4gPHJkZjpSREYgeG1sbnM6cmRmPSJodHRwOi8vd3d3LnczLm9yZy8xOTk5LzAyLzIyLXJkZi1zeW50YXgtbnMjIj4gPHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9IiIgeG1sbnM6eG1wTU09Imh0dHA6Ly9ucy5hZG9iZS5jb20veGFwLzEuMC9tbS8iIHhtbG5zOnN0UmVmPSJodHRwOi8vbnMuYWRvYmUuY29tL3hhcC8xLjAvc1R5cGUvUmVzb3VyY2VSZWYjIiB4bWxuczp4bXA9Imh0dHA6Ly9ucy5hZG9iZS5jb20veGFwLzEuMC8iIHhtcE1NOk9yaWdpbmFsRG9jdW1lbnRJRD0ieG1wLmRpZDpGNzRGRjc2NzEwNERFMjExQTc0M0U0NzZGQkE0MTM5RSIgeG1wTU06RG9jdW1lbnRJRD0ieG1wLmRpZDo1RkZERjA4NTZBNEMxMUUyOTFGMkY4MEVGREQ0MkRDNCIgeG1wTU06SW5zdGFuY2VJRD0ieG1wLmlpZDo1RkZERjA4NDZBNEMxMUUyOTFGMkY4MEVGREQ0MkRDNCIgeG1wOkNyZWF0b3JUb29sPSJBZG9iZSBQaG90b3Nob3AgQ1M2IChXaW5kb3dzKSI+IDx4bXBNTTpEZXJpdmVkRnJvbSBzdFJlZjppbnN0YW5jZUlEPSJ4bXAuaWlkOkU2OUM1RDBBNEI2QUUyMTE5NTdDREVCQjFFNDc0RjQzIiBzdFJlZjpkb2N1bWVudElEPSJ4bXAuZGlkOkY3NEZGNzY3MTA0REUyMTFBNzQzRTQ3NkZCQTQxMzlFIi8+IDwvcmRmOkRlc2NyaXB0aW9uPiA8L3JkZjpSREY+IDwveDp4bXBtZXRhPiA8P3hwYWNrZXQgZW5kPSJyIj8+qKJVUQAADXpJREFUeNrsXQtMVNkZvsOMPHwAoq2KuiLWiixV8G01qxHwkbVZFTWa6G7bWI22ig/wnWxr4itqdN0mRjemGjXZBGtMs4hPQov4fovUagUVUOsTUN4M0/+7njO9DDN35l5mhpnuOcmfYS7nnvPd//7nf6MGi8Ui+eMIkPx0CODeHiblF4PBoHmBlp4RV/a0t8f/B8e1MusjwwxG+jSytUzsZ86QRiIzUQMjMyOLpYWvyqQTMAcaRBRC1I6oLfs5SLEuwNYSVRNVEVWyn2vpgfmDWDwN3MA42YYomKgDUThRBCg1NXVIUlJSQv/+/ft2odGWBm6qrq6ufPPmTemTJ0/uXLp0KXflypX/oMtlRO+Jaojq2ZuxaD5cnJyANjHOdiWKJRoXHBw8NzMz89zDhw+LLS6OZ8+e3b958+aRjh07/oKt1Y6tbXAFIyeDErCDE85BQwzC2Gaf7NixI2X27Nnju3Xr1gmTioqKpHPnzkl5eXnSo0ePpLKyMvnm8PBwqU+fPtKoUaOkxMREqXfv3vJ1+n3J1atXvxs/fvxf6Gs5E6EGe5y3x1RnwLk847V3JOpB9LPc3Nylo0ePjseEK1euSLt375auX79uXcN2HeUbHTx4sLRkyRJp2LBh8ncSocyoqKjf04/v2DloJvd6gBsZpyHHPYliLl68mDZixIiY2tpaadOmTVJGRsZHvRoQIJPaaGxslAljxowZ0tq1a6WgoCCptLT0XI8ePX5Ll98yzptbAtzANEQ4Ax2bk5OTPmbMmE8hBgsXLpRu3bolgzUajU4NinIfs9ksP0B8fLy0Z88eWZxKSkoye/bsOY8d3Fol17UaICPTHuB2r61bt04DaNIS0oIFC2TQAMxBAzDnOn8YkPIafyj+O6yBtbAmcfxz0jq/YXsa9foq/EBC5XWl19mbDuIY/GLjxo3SnTt3rKA4YFlpNzRINTU18qdSdOrr62Vw+FTegzWwFtbEiI2NXdC1a9dwZ1rGGfBgJiaRhw4dmkGvMQwH8dixY004CIK8v3//XqqqqpJ/rqyslCoqKmSw5eXl8nWAxkN9+PBBFhPlG8KaWLtDhw69SCutZ3vrAs4PJVRd17i4OFmHQXvwV60EDbId4DqA2zuguM7v56LG1yZ5H8H2NuoFDsMQQYdwCFnDzvfv35dVnlJz4NAoQU+fPl3WNLYHdNKkSdLOnTutIAG+rq7ufyBoTayNPSIiIj49derUeGfATSrXobvDR44c2RcXTp8+bd2EH0ZwVQl68+bN1oO3bt06+cEmTpwo7dq1ywp62bJlVs0SGBgoz8Ga+I49YmJiYKR+SVP+qhc4Xld7UlndceHGjRtWTvLXjM34GDRokBUcdDTAwIpu27ZNvo65Z86csc5v06aNdR3ZhNMnN2KdO3ce6syPUgMOHR5MagpmXiosLGwmAiaTySoq4DAAAjRGSkqKTJy7y5cvl7KyspoAtw0o4DZgtGvXrpcz4AHOXNfQ0NBA5ls02whWD+C5vAM8NITtWLVqlcxtLmYhISHWA64cfA96qFC9WsXloeQcwJ8/f77ZHPJrmhxqqEZPBcsWHrmQPq7jXp6tCYcIKFUeDiJk2nZMmTJF2rBhg5XDONQQMVtTzvegB6tw5p87As4jlxryIeByStHR0c02UnIOKo9rDzxQenq6dPz4cevvp02bJoNX6nlbRnCXlwzYE4ZBF3Cw8gP5E6Vca3Dg3E1VAie/2goaB5ECDGn9+vVWmcd1aCaroaC5SncXn9gD4/Xr11edATepAIdvXEZu7MO5c+cOAjB4cjAekFdshM05+LS0NPkThxDag8v06tWrZWMD0EePHm0GnBskjAkTJsifjx8/vugMuCO3FieuM1E/oiEFBQWrYD3nzJkjA4Am4TqY+x5aBrQRiHMcYgNuHz58WHr79u29Tp06JYPxLB7V5Naamai8IXqRn58vK1hELvy1802h2uwFELjOVaUaaG7EFi9ezFXiJXvBhBatUsOc+mckKhnFxcXlCLdg2nkkw811+/btJQqcZdWI4D4sLEwGTjYAxkQGiuvk/TUBzdfBmsOHD8fbezpu3LiNbG+LXuANLIXwglRX4ZEjR3LwizVr1kgDBgywRjEcBEADLNfrHBS4jodSGioOGmtgLayJcffu3T0Ug75zFDS7JXRD5IIgoCWhG0Dv3bvX7aGbhR0OcP0/RP8eO3bszsuXL/8LGx08eFCaOXOm9XDxA2ovB6LUHpiL77j3wIEDMmgKlrMJ9CK2V70rySEt6QnEnt1ZemIZmfGBPD0Bw3Pz5k2X0hMJCQlSamqqLNMsPZEVFRW1iEX4bktP2CaEIDZdeEJo1qxZEyIjIyO49+hKQggWGINCuhJ6aCSEDjDx0JQQanEK7uTJk9kEtMTVFNzz588fkjX+vkuXLh5PwbmU9Fy6dOnQ5OTkhH79+v2cQP1UmfR89+5dKVnDu8Thv69YsUJz0lOvqDhLM7e1oSBFvGhmGqLKhmoV+XKnB9FdwJsk9hlI3Yl9vaWUllQkLAxAI/cpRNXNldctKssCuAAugAvgArgALoAL4O4fmt1aHe1PPOzjUVMIu17FoiBr1kqLw2fyEnN4LwCaGMKYL4/Ez1OiYulj94RZWzTgIA+ilh9x9X4WnyIrMCY2Njbt2rVrBQ0NDea6urr67OzsaxSbIgGENEd7rVg8Bpxx+idEn0VGRqYVFhY+t434CTzy4JOJuvkEcBZ3Ik09KjQ0dMm9e/ee2EtV1H9Mrs8litYK3O1ahXXFQY77BAUFDTx79uwfSEw+UQmCDZKTCpvH1SEDDc3R22g0DsjKylowdOjQvo7mX7hwIZ8dzCrtobqbRIUxAfVJtDz9+vjx4xfVMlolJSWvoqKiUB8f3GqHk4HG5nFEc/bv339WDfTLly/LEhIS/oQ0HtM6Jq8DZ/KJ/F9/otnbt2//mxro8vLyysTExK00dyLT64F63n5LgRtY2g1yPGPNmjXfN9JwBLq6urp26tSp3zAV2Iul6wzeBs67iKDKps6fP38/GZYGR6BhdObNm/cdzZ3C7mnWBeQN4LzMAq79KiUl5Vtw0xFos9ncmJ6efoTmTmdvJ8ReMsobwANZdWIi5LWioqJKTa63bNmC2vgsohgmWgZ7oudp4CamCcbFx8f/8dWrV2VqoPft24fumjlM47RXgvY2cBiYIdHR0cuLi4tfqYHOyMjIg05nuh06PkDtsHsaOByiL/Ly8u6qgSZTfzsgIGABMzCoXBidaSlPA+9D9BX5Rw41CJnyR4GBgegfTGbOltEV9dqqTpYvx5xyO8iVK1f+6WjCyJEjo0+cODGDRCWaqcwOzAFz3/gxHc4m6hAOk7+oQ781QH5t8v3WyfJrt1ZXIFFWVqYMJLq3ViChO3QjjfR1q4Zufh0s2ySC4FANNhqNv8vOzr6tBj4nJwdtRV/4RCaLgUeSc3hQUNAicg0eqGkamvclc9xa18mC2mZJnke1tbW3k5KS/lxQUPBUJWVtkXT8aaRHvEMGHl1AD8iq3kpOTv62qKjohe283NzcWyzdXN1qmSxvp5k1t33oqEi0cTWxrwWLN4B7pJTiDeCaxNZjNSBf6SgSdU4BXAAXwAVwAVwAF8B9eejtEOJ/t9+BJYQk5p7yv3tw+pdTXvcOGegwFhigK6Ij87kRDJSwwAB/0+PZLn4doRvCrIEIuxB+IQxDOIawDOEZwjQWrrXRGgp6o3g1Gd09tukGdAGhG4h+/5n0sTvI5EvAkWmdi+4ee7kSdAOhK4jmjHJWuPJ28cqgFvKhGwhdQegOYomeMLfXf1pQvCpn3T12B7qC0B2ELiH62ttXilc4nIORsETiUi03iMSnTxav0OXjrHiFlLNN8SqgtYtXPXjxCl0/auCR7EfSnyX/2/lK8WoyyiNqxSuUV1BmQbmFlV3a+krxagoKUzBGKinlBhS4UOjyleJVCOPidJQEURpUK16htIgSo68Ur/DqUXydhWKsmryjmOtLxSuDsniFMrgaeHQVoZzuk8UrNCCogUcDAxoZaO4Q5h77RvEK3UHoElIDjy4jXytewblKRpcQuoUcAUeXEc37yieKV34ZcyrasHuhOwhdQugWcjSfdRmVSb7Uhu0Xh1OrOmS1/NZVh/5qgPzS5Pulk+W3bm2TLjh0/fhDIKHsO/zan0I3vw2W5TZsdPeogUZ3ELqEJB9rw/5STYOgKwjdQTQP/8JRhOQjbdhyR4+jZgR0A6ErCN1B9PURkkes8abVnSzkwd+x7p4mA11A6AZCVxB9fQAHyhOg/TrNrKdfxWOJfW802rR6KUV0CIlyoQAugAvgArgALoAL4AK4AC6AC+ACuAAugLfy0NOi+rn0Mddtb2xVywjQvasc3JdPczM1AdGRgltlL0OL687WVrtXKw53ikq+m+Z4RlRsXv1qxdc4WxGyl/VS3oN/JKFVgLdkc5uHFlpFM7fo2mQVbaPUHj+4g+t6gCtVnlKTxBGoYCcPHGcjZluF5RTABXD3HU6H/obt4XNmOZW+i9aDqksdcqNjYwV/cMc6QlQ8bbpb4mv86N1anxeVfAfike/he5uKqPhPXgRwAVwAF8AFcAFcABfABXABXADXOv4rwABAehOixiUV0gAAAABJRU5ErkJggg==)}@media only screen and (-webkit-min-device-pixel-ratio: 2), only screen and (-moz-min-device-pixel-ratio: 2), only screen and (-o-min-device-pixel-ratio: 2 / 1), only screen and (min-device-pixel-ratio: 2), only screen and (min-resolution: 2dppx){.fancybox-default a.fancybox-close,.fancybox-default a.fancybox-expand,.fancybox-default a.fancybox-nav span{background-image:url(data:image/png;base64,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);background-size:46px auto}}.fancybox-dark a.fancybox-close,.fancybox-dark a.fancybox-expand,.fancybox-dark a.fancybox-nav span{background-image:url(data:image/png;base64,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)}.fancybox-dark-skin{background:#2A2A2A;border-color:#2A2A2A;color:#fff;border-radius:4px;box-shadow:0 0 10px rgba(0,0,0,0.3) inset !important}.fancybox-dark-overlay{background:#000;opacity:0.8;filter:alpha(opacity=80)}@media only screen and (-webkit-min-device-pixel-ratio: 2), only screen and (min--moz-device-pixel-ratio: 2), only screen and (-o-min-device-pixel-ratio: 2 / 1), only screen and (min-device-pixel-ratio: 2), only screen and (min-resolution: 192dpi), only screen and (min-resolution: 2dppx){.fancybox-dark a.fancybox-close,.fancybox-dark a.fancybox-expand,.fancybox-dark a.fancybox-nav span{background-image:url(data:image/png;base64,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);background-size:46px auto}}.fancybox-light a.fancybox-close,.fancybox-light a.fancybox-expand,.fancybox-light a.fancybox-nav span{background-image:url(data:image/png;base64,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)}.fancybox-light-skin-open{box-shadow:0 10px 25px rgba(0,0,0,0.5)}@media only screen and (-webkit-min-device-pixel-ratio: 2), only screen and (-moz-min-device-pixel-ratio: 2), only screen and (-o-min-device-pixel-ratio: 2 / 1), only screen and (min-device-pixel-ratio: 2), only screen and (min-resolution: 2dppx){ .fancybox-light a.fancybox-close{background-image:url(data:image/png;base64,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);background-size:46px auto}.fancybox-light a.fancybox-expand,.fancybox-light a.fancybox-nav span{background-image:url(data:image/png;base64,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);background-size:46px auto}}.fancybox-light-overlay{opacity:0.9;filter:alpha(opacity=90);background:#555555;background:-moz-radial-gradient(center, ellipse cover, #999 0%, #555 100%);background:-webkit-gradient(radial, center center, 0px, center center, 100%, color-stop(0%, #999), color-stop(100%, #555));background:-webkit-radial-gradient(center, ellipse cover, #999 0%, #555 100%);background:-o-radial-gradient(center, ellipse cover, #999 0%, #555 100%);background:-ms-radial-gradient(center, ellipse cover, #999 0%, #555 100%)}\n </style>\n <style data-href=\"/assets/gulp/print-240f8bfaa7f6402dfd6c49ee3c1ffea57a89ddd4c8c90e2f2a5c7d63c5753e32.css\" media=\"print\">\n .print_only{display:block}body{overflow:hidden}.print_logo{position:absolute;top:0;left:0}.site_header_area{position:relative}.nav_is_fixed .site_header_area{position:absolute;top:0}.site_header_area .brand1,.site_header_area .brand2{display:none}.site_header_area .brand_area{width:23%}.site_header_area .grace_logo img.grace_logo_white{display:none}.custom_banner_container{height:68px}.custom_banner_container img{display:none}.custom_banner_container .banner_header_overlay{display:none}a[href]:after{content:\"\"}.module{padding:1em 0}#sticky_nav_spacer{display:none}.nav_is_fixed #sticky_nav_spacer{display:block}.main_carousel.module .slick-slider .grid_layout{width:100%}.main_carousel.module .slick-slider .right-col{width:3.75in !important;float:left;margin-right:12px;margin-bottom:1em}.main_carousel.module .slick-slider .left-col{width:3.75in !important;float:left}.definition_teaser{display:none;color:white !important}.double_teaser .column{width:48%;float:left}.double_teaser .column+.column{margin-left:1%;margin-top:0}#home .site_header_area{position:relative}#site_footer{border-top:1px solid gray}#site_footer .upper_footer{padding:1em 0}#site_footer .footer_science_calendar footer{display:none}#site_footer .footer_science_calendar .col1,#site_footer .footer_science_calendar .col2,#site_footer .footer_science_calendar .col3{display:inline-block;width:30%;padding:1%}#site_footer .sitemap,#site_footer .share{display:none}#site_footer .lower_footer{height:auto}#site_footer .lower_footer .nav_container{display:none}#primary_column{width:60%;float:left;overflow:hidden;position:relative;display:block}#secondary_column{width:32%;float:right;position:relative;font-size:80%}.double_teaser .column{width:46%}.double_teaser .column+.column{float:right}.grid_view .module_title{display:block}body #page .grid_gallery.grid_view li.slide{width:19%;margin:1%;float:left;clear:none}body #page .grid_gallery.grid_view .bottom_gradient{margin-top:0}body #page .grid_gallery.grid_view .bottom_gradient div{margin-top:.3em}.gradient_line{display:none}.multi_teaser,.teasers_module,.multimedia_teaser,.filter_bar,.tertiary_nav_container,.secondary_nav_mobile,.carousel_teaser,.image_of_the_day,.view_selectors,.related,.primary_media_feature,.fancybox-overlay,#fancybox-lock,.suggested_features,.homepage_carousel,#site_footer .brand_area{display:none}\n </style>\n <script src=\"/assets/public_manifest-7136bdf7a3a8424b034f333778efdc8dd66c789f18847281907defafc52d017e.js\">\n </script>\n <!--[if gt IE 8]><!-->\n <script src=\"/assets/not_ie8_manifest.js\">\n </script>\n <style>\n </style>\n <!--[if !IE]>-->\n <script src=\"/assets/not_ie8_manifest.js\">\n </script>\n <style>\n </style>\n <!--<![endif]-->\n <script src=\"/assets/vendor/jquery.fancybox-9d361e5f98a5c0f233a25df9252dbadea6897af1c0ef221d7465e1205941ea0d.js\">\n </script>\n <script src=\"/assets/mb_manifest-86cde101f747092d1465039f6da3fd2930c66319387c196503d3f70665195392.js\">\n </script>\n <!-- /twitter cards -->\n <meta content=\"summary_large_image\" name=\"twitter:card\"/>\n <meta content=\"News \" name=\"twitter:title\"/>\n <meta content=\"NASA’s real-time portal for Mars exploration, featuring the latest news, images, and discoveries from the Red Planet.\" name=\"twitter:description\"/>\n <meta content=\"https://mars.nasa.gov/system/site_config_values/meta_share_images/1_142497main_PIA03154-200.jpg\" name=\"twitter:image\"/>\n <style type=\"text/css\">\n .fancybox-margin{margin-right:17px;}\n </style>\n <style type=\"text/css\">\n .at-icon{fill:#fff;border:0}.at-icon-wrapper{display:inline-block;overflow:hidden}a .at-icon-wrapper{cursor:pointer}.at-rounded,.at-rounded-element .at-icon-wrapper{border-radius:12%}.at-circular,.at-circular-element .at-icon-wrapper{border-radius:50%}.addthis_32x32_style .at-icon{width:2pc;height:2pc}.addthis_24x24_style .at-icon{width:24px;height:24px}.addthis_20x20_style .at-icon{width:20px;height:20px}.addthis_16x16_style .at-icon{width:1pc;height:1pc}#at16lb{display:none;position:absolute;top:0;left:0;width:100%;height:100%;z-index:1001;background-color:#000;opacity:.001}#at_complete,#at_error,#at_share,#at_success{position:static!important}.at15dn{display:none}#at15s,#at16p,#at16p form input,#at16p label,#at16p textarea,#at_share .at_item{font-family:arial,helvetica,tahoma,verdana,sans-serif!important;font-size:9pt!important;outline-style:none;outline-width:0;line-height:1em}* html #at15s.mmborder{position:absolute!important}#at15s.mmborder{position:fixed!important;width:250px!important}#at15s{background:url(data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAoAAAAKCAYAAACNMs+9AAAAGXRFWHRTb2Z0d2FyZQBBZG9iZSBJbWFnZVJlYWR5ccllPAAAABtJREFUeNpiZGBgaGAgAjAxEAlGFVJHIUCAAQDcngCUgqGMqwAAAABJRU5ErkJggg==);float:none;line-height:1em;margin:0;overflow:visible;padding:5px;text-align:left;position:absolute}#at15s a,#at15s span{outline:0;direction:ltr;text-transform:none}#at15s .at-label{margin-left:5px}#at15s .at-icon-wrapper{width:1pc;height:1pc;vertical-align:middle}#at15s .at-icon{width:1pc;height:1pc}.at4-icon{display:inline-block;background-repeat:no-repeat;background-position:top left;margin:0;overflow:hidden;cursor:pointer}.addthis_16x16_style .at4-icon,.addthis_default_style .at4-icon,.at4-icon,.at-16x16{width:1pc;height:1pc;line-height:1pc;background-size:1pc!important}.addthis_32x32_style .at4-icon,.at-32x32{width:2pc;height:2pc;line-height:2pc;background-size:2pc!important}.addthis_24x24_style .at4-icon,.at-24x24{width:24px;height:24px;line-height:24px;background-size:24px!important}.addthis_20x20_style .at4-icon,.at-20x20{width:20px;height:20px;line-height:20px;background-size:20px!important}.at4-icon.circular,.circular .at4-icon,.circular.aticon{border-radius:50%}.at4-icon.rounded,.rounded .at4-icon{border-radius:4px}.at4-icon-left{float:left}#at15s .at4-icon{text-indent:20px;padding:0;overflow:visible;white-space:nowrap;background-size:1pc;width:1pc;height:1pc;background-position:top left;display:inline-block;line-height:1pc}.addthis_vertical_style .at4-icon,.at4-follow-container .at4-icon{margin-right:5px}html>body #at15s{width:250px!important}#at15s.atm{background:none!important;padding:0!important;width:10pc!important}#at15s_inner{background:#fff;border:1px solid #fff;margin:0}#at15s_head{position:relative;background:#f2f2f2;padding:4px;cursor:default;border-bottom:1px solid #e5e5e5}.at15s_head_success{background:#cafd99!important;border-bottom:1px solid #a9d582!important}.at15s_head_success a,.at15s_head_success span{color:#000!important;text-decoration:none}#at15s_brand,#at15sptx,#at16_brand{position:absolute}#at15s_brand{top:4px;right:4px}.at15s_brandx{right:20px!important}a#at15sptx{top:4px;right:4px;text-decoration:none;color:#4c4c4c;font-weight:700}#at15sptx:hover{text-decoration:underline}#at16_brand{top:5px;right:30px;cursor:default}#at_hover{padding:4px}#at_hover .at_item,#at_share .at_item{background:#fff!important;float:left!important;color:#4c4c4c!important}#at_share .at_item .at-icon-wrapper{margin-right:5px}#at_hover .at_bold{font-weight:700;color:#000!important}#at_hover .at_item{width:7pc!important;padding:2px 3px!important;margin:1px;text-decoration:none!important}#at_hover .at_item.athov,#at_hover .at_item:focus,#at_hover .at_item:hover{margin:0!important}#at_hover .at_item.athov,#at_hover .at_item:focus,#at_hover .at_item:hover,#at_share .at_item.athov,#at_share .at_item:hover{background:#f2f2f2!important;border:1px solid #e5e5e5;color:#000!important;text-decoration:none}.ipad #at_hover .at_item:focus{background:#fff!important;border:1px solid #fff}.at15t{display:block!important;height:1pc!important;line-height:1pc!important;padding-left:20px!important;background-position:0 0;text-align:left}.addthis_button,.at15t{cursor:pointer}.addthis_toolbox a.at300b,.addthis_toolbox a.at300m{width:auto}.addthis_toolbox a{margin-bottom:5px;line-height:initial}.addthis_toolbox.addthis_vertical_style{width:200px}.addthis_button_facebook_like .fb_iframe_widget{line-height:100%}.addthis_button_facebook_like iframe.fb_iframe_widget_lift{max-width:none}.addthis_toolbox a.addthis_button_counter,.addthis_toolbox a.addthis_button_facebook_like,.addthis_toolbox a.addthis_button_facebook_send,.addthis_toolbox a.addthis_button_facebook_share,.addthis_toolbox a.addthis_button_foursquare,.addthis_toolbox a.addthis_button_google_plusone,.addthis_toolbox a.addthis_button_linkedin_counter,.addthis_toolbox a.addthis_button_pinterest_pinit,.addthis_toolbox a.addthis_button_stumbleupon_badge,.addthis_toolbox a.addthis_button_tweet{display:inline-block}.at-share-tbx-element .google_plusone_iframe_widget>span>div{vertical-align:top!important}.addthis_toolbox span.addthis_follow_label{display:none}.addthis_toolbox.addthis_vertical_style span.addthis_follow_label{display:block;white-space:nowrap}.addthis_toolbox.addthis_vertical_style a{display:block}.addthis_toolbox.addthis_vertical_style.addthis_32x32_style a{line-height:2pc;height:2pc}.addthis_toolbox.addthis_vertical_style .at300bs{margin-right:4px;float:left}.addthis_toolbox.addthis_20x20_style span{line-height:20px}.addthis_toolbox.addthis_32x32_style span{line-height:2pc}.addthis_toolbox.addthis_pill_combo_style .addthis_button_compact .at15t_compact,.addthis_toolbox.addthis_pill_combo_style a{float:left}.addthis_toolbox.addthis_pill_combo_style a.addthis_button_tweet{margin-top:-2px}.addthis_toolbox.addthis_pill_combo_style .addthis_button_compact .at15t_compact{margin-right:4px}.addthis_default_style .addthis_separator{margin:0 5px;display:inline}div.atclear{clear:both}.addthis_default_style .addthis_separator,.addthis_default_style .at4-icon,.addthis_default_style .at300b,.addthis_default_style .at300bo,.addthis_default_style .at300bs,.addthis_default_style .at300m{float:left}.at300b img,.at300bo img{border:0}a.at300b .at4-icon,a.at300m .at4-icon{display:block}.addthis_default_style .at300b,.addthis_default_style .at300bo,.addthis_default_style .at300m{padding:0 2px}.at300b,.at300bo,.at300bs,.at300m{cursor:pointer}.addthis_button_facebook_like.at300b:hover,.addthis_button_facebook_like.at300bs:hover,.addthis_button_facebook_send.at300b:hover,.addthis_button_facebook_send.at300bs:hover{opacity:1}.addthis_20x20_style .at15t,.addthis_20x20_style .at300bs{overflow:hidden;display:block;height:20px!important;width:20px!important;line-height:20px!important}.addthis_32x32_style .at15t,.addthis_32x32_style .at300bs{overflow:hidden;display:block;height:2pc!important;width:2pc!important;line-height:2pc!important}.at300bs{overflow:hidden;display:block;background-position:0 0;height:1pc;width:1pc;line-height:1pc!important}.addthis_default_style .at15t_compact,.addthis_default_style .at15t_expanded{margin-right:4px}#at_share .at_item{width:123px!important;padding:4px;margin-right:2px;border:1px solid #fff}#at16p{background:url(data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAoAAAAKCAYAAACNMs+9AAAAGXRFWHRTb2Z0d2FyZQBBZG9iZSBJbWFnZVJlYWR5ccllPAAAABtJREFUeNpiZGBgaGAgAjAxEAlGFVJHIUCAAQDcngCUgqGMqwAAAABJRU5ErkJggg==);z-index:10000001;position:absolute;top:50%;left:50%;width:300px;padding:10px;margin:0 auto;margin-top:-185px;margin-left:-155px;font-family:arial,helvetica,tahoma,verdana,sans-serif;font-size:9pt;color:#5e5e5e}#at_share{margin:0;padding:0}#at16pt{position:relative;background:#f2f2f2;height:13px;padding:5px 10px}#at16pt a,#at16pt h4{font-weight:700}#at16pt h4{display:inline;margin:0;padding:0;font-size:9pt;color:#4c4c4c;cursor:default}#at16pt a{position:absolute;top:5px;right:10px;color:#4c4c4c;text-decoration:none;padding:2px}#at15sptx:focus,#at16pt a:focus{outline:thin dotted}#at15s #at16pf a{top:1px}#_atssh{width:1px!important;height:1px!important;border:0!important}.atm{width:10pc!important;padding:0;margin:0;line-height:9pt;letter-spacing:normal;font-family:arial,helvetica,tahoma,verdana,sans-serif;font-size:9pt;color:#444;background:url(data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAoAAAAKCAYAAACNMs+9AAAAGXRFWHRTb2Z0d2FyZQBBZG9iZSBJbWFnZVJlYWR5ccllPAAAABtJREFUeNpiZGBgaGAgAjAxEAlGFVJHIUCAAQDcngCUgqGMqwAAAABJRU5ErkJggg==);padding:4px}.atm-f{text-align:right;border-top:1px solid #ddd;padding:5px 8px}.atm-i{background:#fff;border:1px solid #d5d6d6;padding:0;margin:0;box-shadow:1px 1px 5px rgba(0,0,0,.15)}.atm-s{margin:0!important;padding:0!important}.atm-s a:focus{border:transparent;outline:0;transition:none}#at_hover.atm-s a,.atm-s a{display:block;text-decoration:none;padding:4px 10px;color:#235dab!important;font-weight:400;font-style:normal;transition:none}#at_hover.atm-s .at_bold{color:#235dab!important}#at_hover.atm-s a:hover,.atm-s a:hover{background:#2095f0;text-decoration:none;color:#fff!important}#at_hover.atm-s .at_bold{font-weight:700}#at_hover.atm-s a:hover .at_bold{color:#fff!important}.atm-s a .at-label{vertical-align:middle;margin-left:5px;direction:ltr}.at_PinItButton{display:block;width:40px;height:20px;padding:0;margin:0;background-image:url(//s7.addthis.com/static/t00/pinit00.png);background-repeat:no-repeat}.at_PinItButton:hover{background-position:0 -20px}.addthis_toolbox .addthis_button_pinterest_pinit{position:relative}.at-share-tbx-element .fb_iframe_widget span{vertical-align:baseline!important}#at16pf{height:auto;text-align:right;padding:4px 8px}.at-privacy-info{position:absolute;left:7px;bottom:7px;cursor:pointer;text-decoration:none;font-family:helvetica,arial,sans-serif;font-size:10px;line-height:9pt;letter-spacing:.2px;color:#666}.at-privacy-info:hover{color:#000}.body .wsb-social-share .wsb-social-share-button-vert{padding-top:0;padding-bottom:0}.body .wsb-social-share.addthis_counter_style .addthis_button_tweet.wsb-social-share-button{padding-top:40px}.body .wsb-social-share.addthis_counter_style .addthis_button_google_plusone.wsb-social-share-button{padding-top:0}.body .wsb-social-share.addthis_counter_style .addthis_button_facebook_like.wsb-social-share-button{padding-top:21px}@media print{#at4-follow,#at4-share,#at4-thankyou,#at4-whatsnext,#at4m-mobile,#at15s,.at4,.at4-recommended{display:none!important}}@media screen and (max-width:400px){.at4win{width:100%}}@media screen and (max-height:700px) and (max-width:400px){.at4-thankyou-inner .at4-recommended-container{height:122px;overflow:hidden}.at4-thankyou-inner .at4-recommended .at4-recommended-item:first-child{border-bottom:1px solid #c5c5c5}}\n </style>\n <style type=\"text/css\">\n .at-branding-logo{font-family:helvetica,arial,sans-serif;text-decoration:none;font-size:10px;display:inline-block;margin:2px 0;letter-spacing:.2px}.at-branding-logo .at-branding-icon{background-image:url(\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAoAAAAKCAMAAAC67D+PAAAAGXRFWHRTb2Z0d2FyZQBBZG9iZSBJbWFnZVJlYWR5ccllPAAAAAZQTFRF////+GlNUkcc1QAAAB1JREFUeNpiYIQDBjQmAwMmkwEM0JnY1WIxFyDAABGeAFEudiZsAAAAAElFTkSuQmCC\")}.at-branding-logo .at-branding-icon,.at-branding-logo .at-privacy-icon{display:inline-block;height:10px;width:10px;margin-left:4px;margin-right:3px;margin-bottom:-1px;background-repeat:no-repeat}.at-branding-logo .at-privacy-icon{background-image:url(\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAkAAAAKCAMAAABR24SMAAAAGXRFWHRTb2Z0d2FyZQBBZG9iZSBJbWFnZVJlYWR5ccllPAAAABhQTFRF8fr9ot/xXcfn2/P5AKva////////AKTWodjhjAAAAAd0Uk5T////////ABpLA0YAAAA6SURBVHjaJMzBDQAwCAJAQaj7b9xifV0kUKJ9ciWxlzWEWI5gMF65KUTv0VKkjVeTerqE/x7+9BVgAEXbAWI8QDcfAAAAAElFTkSuQmCC\")}.at-branding-logo span{text-decoration:none}.at-branding-logo .at-branding-addthis,.at-branding-logo .at-branding-powered-by{color:#666}.at-branding-logo .at-branding-addthis:hover{color:#333}.at-cv-with-image .at-branding-addthis,.at-cv-with-image .at-branding-addthis:hover{color:#fff}a.at-branding-logo:visited{color:initial}.at-branding-info{display:inline-block;padding:0 5px;color:#666;border:1px solid #666;border-radius:50%;font-size:10px;line-height:9pt;opacity:.7;transition:all .3s ease;text-decoration:none}.at-branding-info span{border:0;clip:rect(0 0 0 0);height:1px;margin:-1px;overflow:hidden;padding:0;position:absolute;width:1px}.at-branding-info:before{content:'i';font-family:Times New Roman}.at-branding-info:hover{color:#0780df;border-color:#0780df}\n </style>\n <script async=\"\" charset=\"utf-8\" src=\"//s7.addthis.com/static/layers.b01bacf303e2cf5c81a0.js\" type=\"text/javascript\">\n </script>\n <style type=\"text/css\">\n .at-share-dock.atss{top:auto;left:0;right:0;bottom:0;width:100%;max-width:100%;z-index:1000200;box-shadow:0 0 1px 1px #e2dfe2}.at-share-dock.at-share-dock-zindex-hide{z-index:-1!important}.at-share-dock.atss-top{bottom:auto;top:0}.at-share-dock a{width:auto;transition:none;color:#fff;text-decoration:none;box-sizing:content-box;-webkit-box-sizing:content-box;-moz-box-sizing:content-box}.at-share-dock a:hover{width:auto}.at-share-dock .at4-count{height:43px;padding:5px 0 0;line-height:20px;background:#fff;font-family:Helvetica neue,arial}.at-share-dock .at4-count span{width:100%}.at-share-dock .at4-count .at4-share-label{color:#848484;font-size:10px;letter-spacing:1px}.at-share-dock .at4-count .at4-counter{top:2px;position:relative;display:block;color:#222;font-size:22px}.at-share-dock.at-shfs-medium .at4-count{height:36px;line-height:1pc;padding-top:4px}.at-share-dock.at-shfs-medium .at4-count .at4-counter{font-size:18px}.at-share-dock.at-shfs-medium .at-share-btn .at-icon-wrapper,.at-share-dock.at-shfs-medium a .at-icon-wrapper{padding:6px 0}.at-share-dock.at-shfs-small .at4-count{height:26px;line-height:1;padding-top:3px}.at-share-dock.at-shfs-small .at4-count .at4-share-label{font-size:8px}.at-share-dock.at-shfs-small .at4-count .at4-counter{font-size:14px}.at-share-dock.at-shfs-small .at-share-btn .at-icon-wrapper,.at-share-dock.at-shfs-small a .at-icon-wrapper{padding:4px 0}\n </style>\n <style type=\"text/css\">\n div.at-share-close-control.ats-dark,div.at-share-open-control-left.ats-dark,div.at-share-open-control-right.ats-dark{background:#262b30}div.at-share-close-control.ats-light,div.at-share-open-control-left.ats-light,div.at-share-open-control-right.ats-light{background:#fff}div.at-share-close-control.ats-gray,div.at-share-open-control-left.ats-gray,div.at-share-open-control-right.ats-gray{background:#f2f2f2}.atss{position:fixed;top:20%;width:3pc;z-index:100020;background:none}.at-share-close-control{position:relative;width:3pc;overflow:auto}.at-share-open-control-left{position:fixed;top:20%;z-index:100020;left:0;width:22px}.at-share-close-control .at4-arrow.at-left{float:right}.atss-left{left:0;float:left;right:auto}.atss-right{left:auto;float:right;right:0}.atss-right.at-share-close-control .at4-arrow.at-right{position:relative;right:0;overflow:auto}.atss-right.at-share-close-control .at4-arrow{float:left}.at-share-open-control-right{position:fixed;top:20%;z-index:100020;right:0;width:22px;float:right}.atss-right .at-share-close-control .at4-arrow{float:left}.atss.atss-right a{float:right}.atss.atss-right .at4-share-title{float:right;overflow:hidden}.atss .at-share-btn,.atss a{position:relative;display:block;width:3pc;margin:0;outline-offset:-1px;text-align:center;float:left;transition:width .15s ease-in-out;overflow:hidden;background:#e8e8e8;z-index:100030;cursor:pointer}.at-share-btn::-moz-focus-inner{border:0;padding:0}.atss-right .at-share-btn{float:right}.atss .at-share-btn{border:0;padding:0}.atss .at-share-btn:focus,.atss .at-share-btn:hover,.atss a:focus,.atss a:hover{width:4pc}.atss .at-share-btn .at-icon-wrapper,.atss a .at-icon-wrapper{display:block;padding:8px 0}.atss .at-share-btn:last-child,.atss a:last-child{border:none}.atss .at-share-btn span .at-icon,.atss a span .at-icon{position:relative;top:0;left:0;display:block;background-repeat:no-repeat;background-position:50% 50%;width:2pc;height:2pc;line-height:2pc;border:none;padding:0;margin:0 auto;overflow:hidden;cursor:pointer;cursor:hand}.at4-share .at-custom-sidebar-counter{font-family:Helvetica neue,arial;vertical-align:top;margin-right:4px;display:inline-block;text-align:center}.at4-share .at-custom-sidebar-count{font-size:17px;line-height:1.25em;color:#222}.at4-share .at-custom-sidebar-text{font-size:9px;line-height:1.25em;color:#888;letter-spacing:1px}.at4-share .at4-share-count-container{position:absolute;left:0;right:auto;top:auto;bottom:0;width:100%;color:#fff;background:inherit}.at4-share .at4-share-count,.at4-share .at4-share-count-container{line-height:1pc;font-size:10px}.at4-share .at4-share-count{text-indent:0;font-family:Arial,Helvetica Neue,Helvetica,sans-serif;font-weight:200;width:100%;height:1pc}.at4-share .at4-share-count-anchor{padding-bottom:8px;text-decoration:none;transition:padding .15s ease-in-out .15s,width .15s ease-in-out}\n </style>\n <style type=\"text/css\">\n #at4-drawer-outer-container{top:0;width:20pc;position:fixed}#at4-drawer-outer-container.at4-drawer-inline{position:relative}#at4-drawer-outer-container.at4-drawer-inline.at4-drawer-right{float:right;right:0;left:auto}#at4-drawer-outer-container.at4-drawer-inline.at4-drawer-left{float:left;left:0;right:auto}#at4-drawer-outer-container.at4-drawer-shown,#at4-drawer-outer-container.at4-drawer-shown *{z-index:999999}#at4-drawer-outer-container,#at4-drawer-outer-container .at4-drawer-outer,#at-drawer{height:100%;overflow-y:auto;overflow-x:hidden}.at4-drawer-push-content-right-back{position:relative;right:0}.at4-drawer-push-content-right{position:relative;left:20pc!important}.at4-drawer-push-content-left-back{position:relative;left:0}.at4-drawer-push-content-left{position:relative;right:20pc!important}#at4-drawer-outer-container.at4-drawer-right{left:auto;right:-20pc}#at4-drawer-outer-container.at4-drawer-left{right:auto;left:-20pc}#at4-drawer-outer-container.at4-drawer-shown.at4-drawer-right{left:auto;right:0}#at4-drawer-outer-container.at4-drawer-shown.at4-drawer-left{right:auto;left:0}#at-drawer{top:0;z-index:9999999;height:100%;animation-duration:.4s}#at-drawer.drawer-push.at-right{right:-20pc}#at-drawer.drawer-push.at-left{left:-20pc}#at-drawer .at-recommended-label{padding:0 0 0 20px;color:#999;line-height:3pc;font-size:18px;font-weight:300;cursor:default}#at-drawer-arrow{width:30px;height:5pc}#at-drawer-arrow.ats-dark{background:#262b30}#at-drawer-arrow.ats-gray{background:#f2f2f2}#at-drawer-open-arrow{background-image:url(data:image/png;base64,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);background-repeat:no-repeat;width:13px;height:23px;margin:28px 0 0 8px}.at-left #at-drawer-open-arrow{background-position:0 -46px}.ats-dark #at-drawer-open-arrow{background-position:0 -23px}.ats-dark.at-left #at-drawer-open-arrow{background-position:0 -69px}#at-drawer-arrow.at4-drawer-modern-browsers{position:fixed;top:40%;background-repeat:no-repeat;background-position:0 0!important;z-index:9999999}.at4-drawer-inline #at-drawer-arrow{position:absolute}#at-drawer-arrow.at4-drawer-modern-browsers.at-right{right:0}#at-drawer-arrow.at4-drawer-modern-browsers.at-left{left:0}.at4-drawer-push-animation-left{transition:left .4s ease-in-out .15s}.at4-drawer-push-animation-right{transition:right .4s ease-in-out .15s}#at-drawer.drawer-push.at4-drawer-push-animation-right{right:0}#at-drawer.drawer-push.at4-drawer-push-animation-right-back{right:-20pc!important}#at-drawer.drawer-push.at4-drawer-push-animation-left{left:0}#at-drawer.drawer-push.at4-drawer-push-animation-left-back{left:-20pc!important}#at-drawer .at4-closebutton.drawer-close{content:'X';color:#999;display:block;position:absolute;margin:0;top:0;right:0;width:3pc;height:45px;line-height:45px;overflow:hidden;opacity:.5}#at-drawer.ats-dark .at4-closebutton.drawer-close{color:#fff}#at-drawer .at4-closebutton.drawer-close:hover{opacity:1}#at-drawer.ats-dark.at4-recommended .at4-logo-container a{color:#666}#at-drawer.at4-recommended .at4-recommended-vertical{padding:0}#at-drawer.at4-recommended .at4-recommended-item .sponsored-label{margin:2px 0 0 21px;color:#ddd}#at-drawer.at4-recommended .at4-recommended-vertical .at4-recommended-item{position:relative;padding:0;width:20pc;height:180px;margin:0}#at-drawer.at4-recommended .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-img a:after{content:'';position:absolute;top:0;left:0;right:0;bottom:0;background:rgba(0,0,0,.65);z-index:1000000;transition:all .2s ease-in-out}#at-drawer.at4-recommended .at4-recommended-vertical .at4-recommended-item.at-hover .at4-recommended-item-img a:after{background:rgba(0,0,0,.8)}#at-drawer .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-img,#at-drawer .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-img a,#at-drawer .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-img img{width:20pc;height:180px;float:none}#at-drawer .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-caption{width:100%;position:absolute;bottom:0;left:0;height:70px}#at-drawer .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-caption .at-h4{color:#fff;position:absolute;height:52px;top:0;left:20px;right:20px;margin:0;padding:0;line-height:25px;font-size:20px;font-weight:600;z-index:1000001;text-decoration:none;text-transform:none}#at-drawer.at4-recommended .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-caption .at-h4 a:hover{text-decoration:none}#at-drawer.at4-recommended .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-caption .at-h4 a:link{color:#fff}#at-drawer.at4-recommended .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-caption small{position:absolute;top:auto;bottom:10px;left:20px;width:auto;color:#ccc}#at-drawer.at4-recommended .at4-logo-container{margin-left:20px}#at-drawer.ats-dark.at4-recommended .at4-logo-container a:hover{color:#fff}#at-drawer.at4-recommended .at-logo{margin:0}\n </style>\n <style type=\"text/css\">\n .at4-follow.at-mobile{display:none!important}.at4-follow{position:fixed;top:0;right:0;font-weight:400;color:#666;cursor:default;z-index:10001}.at4-follow .at4-follow-inner{position:relative;padding:10px 24px 10px 15px}.at4-follow-inner,.at-follow-open-control{border:0 solid #c5c5c5;border-width:1px 0 1px 1px;margin-top:-1px}.at4-follow .at4-follow-container{margin-left:9pt}.at4-follow.at4-follow-24 .at4-follow-container{height:24px;line-height:23px;font-size:13px}.at4-follow.at4-follow-32 .at4-follow-container{width:15pc;height:2pc;line-height:2pc;font-size:14px}.at4-follow .at4-follow-container .at-follow-label{display:inline-block;height:24px;line-height:24px;margin-right:10px;padding:0;cursor:default;float:left}.at4-follow .at4-follow-container .at-icon-wrapper{height:24px;width:24px}.at4-follow.ats-transparent .at4-follow-inner,.at-follow-open-control.ats-transparent{border-color:transparent}.at4-follow.ats-dark .at4-follow-inner,.at-follow-open-control.ats-dark{background:#262b30;border-color:#000;color:#fff}.at4-follow.ats-dark .at-follow-close-control{background-color:#262b30}.at4-follow.ats-light .at4-follow-inner{background:#fff;border-color:#c5c5c5}.at4-follow.ats-gray .at4-follow-inner,.at-follow-open-control.ats-gray{background:#f2f2f2;border-color:#c5c5c5}.at4-follow.ats-light .at4-follow-close-control,.at-follow-open-control.ats-light{background:#e5e5e5}.at4-follow .at4-follow-inner .at4-follow-close-control{position:absolute;top:0;bottom:0;left:0;width:20px;cursor:pointer;display:none}.at4-follow .at4-follow-inner .at4-follow-close-control div{display:block;line-height:20px;text-indent:-9999em;margin-top:calc(50% + 1px);overflow:hidden}.at-follow-open-control div.at4-arrow.at-left{background-position:0 -2px}.at-follow-open-control{position:fixed;height:35px;top:0;right:0;padding-top:10px;z-index:10002}.at-follow-btn{margin:0 5px 5px 0;padding:0;outline-offset:-1px;display:inline-block;box-sizing:content-box;transition:all .2s ease-in-out}.at-follow-btn:focus,.at-follow-btn:hover{transform:translateY(-4px)}.at4-follow-24 .at-follow-btn{height:25px;line-height:0;width:25px}\n </style>\n <style type=\"text/css\">\n .at-follow-tbx-element .at300b,.at-follow-tbx-element .at300m{display:inline-block;width:auto;padding:0;margin:0 2px 5px;outline-offset:-1px;transition:all .2s ease-in-out}.at-follow-tbx-element .at300b:focus,.at-follow-tbx-element .at300b:hover,.at-follow-tbx-element .at300m:focus,.at-follow-tbx-element .at300m:hover{transform:translateY(-4px)}.at-follow-tbx-element .addthis_vertical_style .at300b,.at-follow-tbx-element .addthis_vertical_style .at300m{display:block}.at-follow-tbx-element .addthis_vertical_style .at300b .addthis_follow_label,.at-follow-tbx-element .addthis_vertical_style .at300b .at-icon-wrapper,.at-follow-tbx-element .addthis_vertical_style .at300m .addthis_follow_label,.at-follow-tbx-element .addthis_vertical_style .at300m .at-icon-wrapper{display:inline-block;vertical-align:middle;margin-right:5px}.at-follow-tbx-element .addthis_vertical_style .at300b:focus,.at-follow-tbx-element .addthis_vertical_style .at300b:hover,.at-follow-tbx-element .addthis_vertical_style .at300m:focus,.at-follow-tbx-element .addthis_vertical_style .at300m:hover{transform:none}\n </style>\n <style type=\"text/css\">\n .at4-jumboshare .at-share-btn{display:inline-block;margin-right:13px;margin-top:13px}.at4-jumboshare .at-share-btn .at-icon{float:left}.at4-jumboshare .at-share-btn .at300bs{display:inline-block;float:left;cursor:pointer}.at4-jumboshare .at4-mobile .at-share-btn .at-icon,.at4-jumboshare .at4-mobile .at-share-btn .at-icon-wrapper{margin:0;padding:0}.at4-jumboshare .at4-mobile .at-share-btn{padding:0}.at4-jumboshare .at4-mobile .at-share-btn .at-label{display:none}.at4-jumboshare .at4-count{font-size:60px;line-height:60px;font-family:Helvetica neue,arial;font-weight:700}.at4-jumboshare .at4-count-container{display:table-cell;text-align:center;min-width:200px;vertical-align:middle;border-right:1px solid #ccc;padding-right:20px}.at4-jumboshare .at4-share-container{display:table-cell;vertical-align:middle;padding-left:20px}.at4-jumboshare .at4-share-container.at-share-tbx-element{padding-top:0}.at4-jumboshare .at4-title{position:relative;font-size:18px;line-height:18px;bottom:2px}.at4-jumboshare .at4-spacer{height:1px;display:block;visibility:hidden;opacity:0}.at4-jumboshare .at-share-btn{display:inline-block;margin:0 2px;line-height:0;padding:0;overflow:hidden;text-decoration:none;text-transform:none;color:#fff;cursor:pointer;transition:all .2s ease-in-out;border:0;background-color:transparent}.at4-jumboshare .at-share-btn:focus,.at4-jumboshare .at-share-btn:hover{transform:translateY(-4px);color:#fff;text-decoration:none}.at4-jumboshare .at-label{font-family:helvetica neue,helvetica,arial,sans-serif;font-size:9pt;padding:0 15px 0 0;margin:0;height:2pc;line-height:2pc;background:none}.at4-jumboshare .at-share-btn:hover,.at4-jumboshare .at-share-btn:link{text-decoration:none}.at4-jumboshare .at-share-btn::-moz-focus-inner{border:0;padding:0}.at4-jumboshare.at-mobile .at-label{display:none}\n </style>\n <style type=\"text/css\">\n .at4-recommendedbox-outer-container{display:inline}.at4-recommended-outer{position:static}.at4-recommended{top:20%;margin:0;text-align:center;font-weight:400;font-size:13px;line-height:17px;color:#666}.at4-recommended.at-inline .at4-recommended-horizontal{text-align:left}.at4-recommended-recommendedbox{padding:0;z-index:inherit}.at4-recommended-recommended{padding:40px 0}.at4-recommended-horizontal{max-height:340px}.at4-recommended.at-medium .at4-recommended-horizontal{max-height:15pc}.at4-recommended.at4-minimal.at-medium .at4-recommended-horizontal{padding-top:10px;max-height:230px}.at4-recommended-text-only .at4-recommended-horizontal{max-height:130px}.at4-recommended-horizontal{padding-top:5px;overflow-y:hidden}.at4-minimal{background:none;color:#000;border:none!important;box-shadow:none!important}@media screen and (max-width:900px){.at4-recommended-horizontal .at4-recommended-item,.at4-recommended-horizontal .at4-recommended-item .at4-recommended-item-img{width:15pc}}.at4-recommended.at4-minimal .at4-recommended-horizontal .at4-recommended-item .at4-recommended-item-caption{padding:0 0 10px}.at4-recommended.at4-minimal .at4-recommended-horizontal .at4-recommended-item-caption{padding:20px 0 0!important}.addthis-smartlayers .at4-recommended .at-h3.at-recommended-label{margin:0;padding:0;font-weight:300;font-size:18px;line-height:24px;color:#464646;width:100%;display:inline-block;zoom:1}.addthis-smartlayers .at4-recommended.at-inline .at-h3.at-recommended-label{text-align:left}#at4-thankyou .addthis-smartlayers .at4-recommended.at-inline .at-h3.at-recommended-label{text-align:center}.at4-recommended .at4-recommended-item{display:inline-block;zoom:1;position:relative;background:#fff;border:1px solid #c5c5c5;width:200px;margin:10px}.addthis_recommended_horizontal .at4-recommended-item{border:none}.at4-recommended .at4-recommended-item .sponsored-label{color:#666;font-size:9px;position:absolute;top:-20px}.at4-recommended .at4-recommended-item-img .at-tli,.at4-recommended .at4-recommended-item-img a{position:absolute;left:0}.at4-recommended.at-inline .at4-recommended-horizontal .at4-recommended-item{margin:10px 20px 0 0}.at4-recommended.at-medium .at4-recommended-horizontal .at4-recommended-item{margin:10px 10px 0 0}.at4-recommended.at-medium .at4-recommended-item{width:140px;overflow:hidden}.at4-recommended .at4-recommended-item .at4-recommended-item-img{position:relative;text-align:center;width:100%;height:200px;line-height:0;overflow:hidden}.at4-recommended .at4-recommended-item .at4-recommended-item-img a{display:block;width:100%;height:200px}.at4-recommended.at-medium .at4-recommended-item .at4-recommended-item-img,.at4-recommended.at-medium .at4-recommended-item .at4-recommended-item-img a{height:140px}.at4-recommended .at4-recommended-item .at4-recommended-item-img img{position:absolute;top:0;left:0;min-height:0;min-width:0;max-height:none;max-width:none;margin:0;padding:0}.at4-recommended .at4-recommended-item .at4-recommended-item-caption{height:74px;overflow:hidden;padding:20px;text-align:left;-ms-box-sizing:content-box;-o-box-sizing:content-box;box-sizing:content-box}.at4-recommended.at-medium .at4-recommended-item .at4-recommended-item-caption{height:50px;padding:15px}.at4-recommended .at4-recommended-item .at4-recommended-item-caption .at-h4{height:54px;margin:0 0 5px;padding:0;overflow:hidden;word-wrap:break-word;font-size:14px;font-weight:400;line-height:18px;text-align:left}.at4-recommended.at-medium .at4-recommended-item .at4-recommended-item-caption .at-h4{font-size:9pt;line-height:1pc;height:33px}.at4-recommended .at4-recommended-item:hover .at4-recommended-item-caption .at-h4{text-decoration:underline}.at4-recommended a:link,.at4-recommended a:visited{text-decoration:none;color:#464646}.at4-recommended .at4-recommended-item .at4-recommended-item-caption .at-h4 a:hover{text-decoration:underline;color:#000}.at4-recommended .at4-recommended-item .at4-recommended-item-caption small{display:block;white-space:nowrap;overflow:hidden;text-overflow:ellipsis;font-size:11px;color:#666}.at4-recommended.at-medium .at4-recommended-item .at4-recommended-item-caption small{font-size:9px}.at4-recommended .at4-recommended-vertical{padding:15px 0 0}.at4-recommended .at4-recommended-vertical .at4-recommended-item{display:block;width:auto;max-width:100%;height:60px;border:none;margin:0 0 15px;box-shadow:none;background:none}.at4-recommended-vertical .at4-recommended-item .at4-recommended-item-img,.at4-recommended-vertical .at4-recommended-item .at4-recommended-item-img img{width:60px;height:60px;float:left}.at4-recommended-vertical .at4-recommended-item .at4-recommended-item-caption{border-top:none;margin:0;height:60px;padding:3px 5px}.at4-recommended .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-caption .at-h4{height:38px;margin:0}.at4-recommended .at4-recommended-vertical .at4-recommended-item .at4-recommended-item-caption small{position:absolute;bottom:0}.at4-recommended .at-recommended-label.at-vertical{text-align:left}.at4-no-image-light-recommended,.at4-no-image-minimal-recommended{background-color:#f2f2f2!important}.at4-no-image-gray-recommended{background-color:#e6e6e5!important}.at4-no-image-dark-recommended{background-color:#4e555e!important}.at4-recommended .at4-recommended-item-placeholder-img{background-repeat:no-repeat!important;background-position:center!important;width:100%!important;height:100%!important}.at4-recommended-horizontal .at4-no-image-dark-recommended .at4-recommended-item-placeholder-img{background-image:url(data:image/png;base64,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)!important}.at4-recommended-vertical .at4-no-image-dark-recommended .at4-recommended-item-placeholder-img{background-image:url(data:image/png;base64,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)!important}.at4-recommended-horizontal .at4-no-image-gray-recommended .at4-recommended-item-placeholder-img,.at4-recommended-horizontal .at4-no-image-light-recommended .at4-recommended-item-placeholder-img,.at4-recommended-horizontal .at4-no-image-minimal-recommended .at4-recommended-item-placeholder-img{background-image:url(data:image/png;base64,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)!important}.at4-recommended-vertical .at4-no-image-gray-recommended .at4-recommended-item-placeholder-img,.at4-recommended-vertical .at4-no-image-light-recommended .at4-recommended-item-placeholder-img,.at4-recommended-vertical .at4-no-image-minimal-recommended .at4-recommended-item-placeholder-img{background-image:url(data:image/png;base64,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)!important}#at-drawer.ats-dark,.at4-recommended.ats-dark .at4-recommended-horizontal .at4-recommended-item-caption,.at4-recommended.ats-dark .at4-recommended-vertical .at4-recommended-item-caption{background:#262b30}#at-drawer.ats-gray,.at4-recommended.ats-gray .at4-recommended-horizontal .at4-recommended-item-caption{background:#f2f2f2}#at-drawer.ats-light,.at4-recommended.ats-light .at4-recommended-horizontal .at4-recommended-item-caption{background:#fff}.at4-recommended.ats-dark .at4-recommended-vertical .at4-recommended-item{background:none}.at4-recommended.ats-dark .at4-recommended-item .at4-recommended-item-caption a:hover,.at4-recommended.ats-dark .at4-recommended-item .at4-recommended-item-caption a:link,.at4-recommended.ats-dark .at4-recommended-item .at4-recommended-item-caption a:visited,.at4-recommended.ats-dark .at4-recommended-item .at4-recommended-item-caption small,.at4-recommended.ats-dark .at4-recommended-item-caption,.at4-recommended.ats-dark .at-logo a:hover,.at4-recommended.ats-dark .at-recommended-label.at-vertical{color:#fff}.at4-recommended-vertical-logo{padding-top:0;text-align:left}.at4-recommended-vertical-logo .at4-logo-container{line-height:10px}.at4-recommended-horizontal-logo{text-align:center}.at4-recommended.at-inline .at4-recommended-horizontal-logo{text-align:left}#at4-thankyou .at4-recommended.at-inline .at4-recommended-horizontal{text-align:center}.at4-recommended .at-logo{margin:10px 0 0;padding:0;height:25px;overflow:auto;-ms-box-sizing:content-box;-o-box-sizing:content-box;box-sizing:content-box}.at4-recommended.at-inline .at4-recommended-horizontal .at-logo{text-align:left}.at4-recommended .at4-logo-container a.at-sponsored-link{color:#666}.at4-recommended-class .at4-logo-container a:hover,.at4-recommendedbox-outer-container .at4-recommended-recommendedbox .at4-logo-container a:hover{color:#000}\n </style>\n <style type=\"text/css\">\n .at-recommendedjumbo-outer-container{margin:0;padding:0;border:0;background:none;color:#000}.at-recommendedjumbo-footer{position:relative;width:100%;height:510px;overflow:hidden;transition:all .3s ease-in-out}.at-mobile .at-recommendedjumbo-footer{height:250px}.at-recommendedjumbo-footer #bg-link:after{content:'';position:absolute;top:0;left:0;right:0;bottom:0;background:rgba(0,0,0,.75)}.at-recommendedjumbo-footer:hover #bg-link:after{background:rgba(0,0,0,.85)}.at-recommendedjumbo-footer *,.at-recommendedjumbo-footer :after,.at-recommendedjumbo-footer :before{box-sizing:border-box}.at-recommendedjumbo-footer:hover #at-recommendedjumbo-footer-bg{animation:atRecommendedJumboAnimatedBackground 1s ease-in-out 1;animation-fill-mode:forwards}.at-recommendedjumbo-footer #at-recommendedjumbo-top-holder{position:absolute;top:0;padding:0 40px;width:100%}.at-mobile .at-recommendedjumbo-footer #at-recommendedjumbo-top-holder{padding:0 20px}.at-recommendedjumbo-footer .at-recommendedjumbo-footer-inner{position:relative;text-align:center;font-family:helvetica,arial,sans-serif;z-index:2;width:100%}.at-recommendedjumbo-footer #at-recommendedjumbo-label-holder{margin:40px 0 0;max-height:30px}.at-mobile .at-recommendedjumbo-footer #at-recommendedjumbo-label-holder{margin:20px 0 0;max-height:20px}.at-recommendedjumbo-footer #at-recommendedjumbo-label{font-weight:300;font-size:24px;line-height:24px;color:#fff;margin:0}.at-mobile .at-recommendedjumbo-footer #at-recommendedjumbo-label{font-weight:150;font-size:14px;line-height:14px}.at-recommendedjumbo-footer #at-recommendedjumbo-title-holder{margin:20px 0 0;min-height:3pc;max-height:78pt}.at-mobile .at-recommendedjumbo-footer #at-recommendedjumbo-title-holder{margin:10px 0 0;min-height:24px;max-height:54px}.at-recommendedjumbo-footer #at-recommendedjumbo-content-title{font-size:3pc;line-height:52px;font-weight:700;margin:0}.at-mobile .at-recommendedjumbo-footer #at-recommendedjumbo-content-title{font-size:24px;line-height:27px}.at-recommendedjumbo-footer a{text-decoration:none;color:#fff}.at-recommendedjumbo-footer a:visited{color:#fff}.at-recommendedjumbo-footer small{margin:20px 0 0;display:inline-block;height:2pc;line-height:2pc;font-size:14px;color:#ccc;cursor:default}.at-mobile .at-recommendedjumbo-footer small{margin:10px 0 0;height:14px;line-height:14px;font-size:9pt}.at-recommendedjumbo-footer .at-logo-container{position:absolute;bottom:20px;margin:auto;left:0;right:0}.at-mobile .at-recommendedjumbo-footer .at-logo-container{bottom:10px}.at-recommendedjumbo-footer a.at-sponsored-link{color:#ccc}.at-recommendedjumbo-footer div #at-recommendedjumbo-logo-link{padding:2px 0 0 11px;text-decoration:none;line-height:20px;font-family:helvetica,arial,sans-serif;font-size:9px;color:#ccc}@keyframes atRecommendedJumboAnimatedBackground{0%{transform:scale(1,1)}to{transform:scale(1.1,1.1)}}\n </style>\n <style type=\"text/css\">\n .at-resp-share-element{position:relative;padding:0;margin:0;font-size:0;line-height:0}.at-resp-share-element:after,.at-resp-share-element:before{content:\" \";display:table}.at-resp-share-element.at-mobile .at4-share-count-container,.at-resp-share-element.at-mobile .at-label{display:none}.at-resp-share-element .at-share-btn{display:inline-block;*display:inline;*zoom:1;margin:0 2px 5px;padding:0;overflow:hidden;line-height:0;text-decoration:none;text-transform:none;color:#fff;cursor:pointer;transition:all .2s ease-in-out;border:0;font-family:helvetica neue,helvetica,arial,sans-serif;background-color:transparent}.at-resp-share-element .at-share-btn::-moz-focus-inner{border:0;padding:0}.at-resp-share-element .at-share-btn:focus,.at-resp-share-element .at-share-btn:hover{transform:translateY(-4px);color:#fff;text-decoration:none}.at-resp-share-element .at-share-btn .at-icon-wrapper{float:left}.at-resp-share-element .at-share-btn.at-share-btn.at-svc-compact:hover{transform:none}.at-resp-share-element .at-share-btn .at-label{font-family:helvetica neue,helvetica,arial,sans-serif;font-size:9pt;padding:0 15px 0 0;margin:0 0 0 5px;height:2pc;line-height:2pc;background:none}.at-resp-share-element .at-icon,.at-resp-share-element .at-label{cursor:pointer}.at-resp-share-element .at4-share-count-container{text-decoration:none;float:right;padding-right:15px;font-size:9pt}.at-mobile .at-resp-share-element .at-label{display:none}.at-resp-share-element.at-mobile .at-share-btn{margin-right:5px}.at-mobile .at-resp-share-element .at-share-btn{padding:5px;margin-right:5px}\n </style>\n <style type=\"text/css\">\n .at-share-tbx-element{position:relative;margin:0;color:#fff;font-size:0}.at-share-tbx-element,.at-share-tbx-element .at-share-btn{font-family:helvetica neue,helvetica,arial,sans-serif;padding:0;line-height:0}.at-share-tbx-element .at-share-btn{cursor:pointer;margin:0 5px 5px 0;display:inline-block;overflow:hidden;border:0;text-decoration:none;text-transform:none;background-color:transparent;color:inherit;transition:all .2s ease-in-out}.at-share-tbx-element .at-share-btn:focus,.at-share-tbx-element .at-share-btn:hover{transform:translateY(-4px);outline-offset:-1px;color:inherit}.at-share-tbx-element .at-share-btn::-moz-focus-inner{border:0;padding:0}.at-share-tbx-element .at-share-btn.at-share-btn.at-svc-compact:hover{transform:none}.at-share-tbx-element .at-icon-wrapper{vertical-align:middle}.at-share-tbx-element .at4-share-count,.at-share-tbx-element .at-label{margin:0 7.5px 0 2.5px;text-decoration:none;vertical-align:middle;display:inline-block;background:none;height:0;font-size:inherit;line-height:inherit;color:inherit}.at-share-tbx-element.at-mobile .at4-share-count,.at-share-tbx-element.at-mobile .at-label{display:none}.at-share-tbx-element .at_native_button{vertical-align:middle}.at-share-tbx-element .addthis_counter.addthis_bubble_style{margin:0 2px;vertical-align:middle;display:inline-block}.at-share-tbx-element .fb_iframe_widget{display:block}.at-share-tbx-element.at-share-tbx-native .at300b{vertical-align:middle}.at-style-responsive .at-share-btn{padding:5px}.at-style-jumbo{display:table}.at-style-jumbo .at4-spacer{height:1px;display:block;visibility:hidden;opacity:0}.at-style-jumbo .at4-count-container{display:table-cell;text-align:center;min-width:200px;vertical-align:middle;border-right:1px solid #ccc;padding-right:20px}.at-style-jumbo .at4-count{font-size:60px;line-height:60px;font-weight:700}.at-style-jumbo .at4-count-title{position:relative;font-size:18px;line-height:18px;bottom:2px}.at-style-jumbo .at-share-btn-elements{display:table-cell;vertical-align:middle;padding-left:20px}.at_flat_counter{cursor:pointer;font-family:helvetica,arial,sans-serif;font-weight:700;text-transform:uppercase;display:inline-block;position:relative;vertical-align:top;height:auto;margin:0 5px;padding:0 6px;left:-1px;background:#ebebeb;color:#32363b;transition:all .2s ease}.at_flat_counter:after{top:30%;left:-4px;content:\"\";position:absolute;border-width:5px 8px 5px 0;border-style:solid;border-color:transparent #ebebeb transparent transparent;display:block;width:0;height:0;transform:translateY(360deg)}.at_flat_counter:hover{background:#e1e2e2}\n </style>\n <style type=\"text/css\">\n .at4-thankyou-background{top:0;right:0;left:0;bottom:0;-webkit-overflow-scrolling:touch;z-index:9999999;background-image:url(data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAoAAAAKCAYAAACNMs+9AAAAGXRFWHRTb2Z0d2FyZQBBZG9iZSBJbWFnZVJlYWR5ccllPAAAABtJREFUeNpizCuu/sRABGBiIBKMKqSOQoAAAwC8KgJipENhxwAAAABJRU5ErkJggg==);background:hsla(217,6%,46%,.95)}.at4-thankyou-background.at-thankyou-shown{position:fixed}.at4-thankyou-inner{position:absolute;width:100%;top:10%;left:50%;margin-left:-50%;text-align:center}.at4-thankyou-mobile .at4-thankyou-inner{top:5%}.thankyou-description{font-weight:400}.at4-thankyou-background .at4lb-inner{position:relative;width:100%;height:100%}.at4-thankyou-background .at4lb-inner .at4x{position:absolute;top:15px;right:15px;display:block;width:20px;height:20px;padding:20px;margin:0;cursor:pointer;transition:opacity .25s ease-in;opacity:.4;background:url(\"data:image/gif;base64,iVBORw0KGgoAAAANSUhEUgAAABQAAAAUCAYAAACNiR0NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAABx0RVh0U29mdHdhcmUAQWRvYmUgRmlyZXdvcmtzIENTNui8sowAAAAWdEVYdENyZWF0aW9uIFRpbWUAMTEvMTMvMTKswDp5AAAAd0lEQVQ4jb2VQRLAIAgDE///Z3qqY1FAhalHMCsCIkVEAIAkkVgvp2lDBgYAnAyHkWotLccNrEd4A7X2TqIdqLfnWBAdaF5rJdyJfjtPH5GT37CaGhoVq3nOm/XflUuLUto2pY1d+vRKh0Pp+MrAVtDe2JkvYNQ+jVSEEFmOkggAAAAASUVORK5CYII=\") no-repeat center center;overflow:hidden;text-indent:-99999em;border:1px solid transparent}.at4-thankyou-background .at4lb-inner .at4x:focus,.at4-thankyou-background .at4lb-inner .at4x:hover{border:1px solid #fff;border-radius:50%;outline:0}.at4-thankyou-background .at4lb-inner #at4-palogo{position:absolute;bottom:10px;display:inline-block;text-decoration:none;font-family:helvetica,arial,sans-serif;font-size:11px;cursor:pointer;-webkit-transition:opacity .25s ease-in;moz-transition:opacity .25s ease-in;transition:opacity .25s ease-in;opacity:.5;z-index:100020;color:#fff;padding:2px 0 0 13px}.at4-thankyou-background .at4lb-inner #at4-palogo .at-branding-addthis,.at4-thankyou-background .at4lb-inner #at4-palogo .at-branding-info{color:#fff}.at4-thankyou-background .at4lb-inner #at4-palogo:hover,.at4-thankyou-background.ats-dark .at4lb-inner a#at4-palogo:hover{text-decoration:none;color:#fff;opacity:1}.at4-thankyou-background.ats-dark{background-image:url(\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAoAAAAKCAYAAACNMs+9AAAAGXRFWHRTb2Z0d2FyZQBBZG9iZSBJbWFnZVJlYWR5ccllPAAAABtJREFUeNpiZGBgeMZABGBiIBKMKqSOQoAAAwB+cQD6hqlbCwAAAABJRU5ErkJggg==\");background:rgba(0,0,0,.85)}.at4-thankyou-background .thankyou-title{color:#fff;font-size:38.5px;margin:10px 20px;line-height:38.5px;font-family:helvetica neue,helvetica,arial,sans-serif;font-weight:300}.at4-thankyou-background.ats-dark .thankyou-description,.at4-thankyou-background.ats-dark .thankyou-title{color:#fff}.at4-thankyou-background .thankyou-description{color:#fff;font-size:18px;margin:10px 0;line-height:24px;padding:0;font-family:helvetica neue,helvetica,arial,sans-serif;font-weight:300}.at4-thankyou-background .at4-thanks-icons{padding-top:10px}.at4-thankyou-mobile *{-webkit-overflow-scrolling:touch}#at4-thankyou .at4-recommended-recommendedbox .at-logo{display:none}.at4-thankyou .at-h3{height:49px;line-height:49px;margin:0 50px 0 20px;padding:1px 0 0;font-family:helvetica neue,helvetica,arial,sans-serif;font-size:1pc;font-weight:700;color:#fff;text-shadow:0 1px #000}.at4-thanks{padding-top:50px;text-align:center}.at4-thanks label{display:block;margin:0 0 15px;font-size:1pc;line-height:1pc}.at4-thanks .at4-h2{background:none;border:none;margin:0 0 10px;padding:0;font-family:helvetica neue,helvetica,arial,sans-serif;font-size:28px;font-weight:300;color:#000}.at4-thanks .at4-thanks-icons{position:relative;height:2pc}.at4-thanks .at4-thanks-icons .at-thankyou-label{display:block;padding-bottom:10px;font-size:14px;color:#666}.at4-thankyou-layer .at-follow .at-icon-wrapper{width:2pc;height:2pc}\n </style>\n <style type=\"text/css\">\n .at4-recommended-toaster{position:fixed;top:auto;bottom:0;right:0;z-index:100021}.at4-recommended-toaster.ats-light{border:1px solid #c5c5c5;background:#fff}.at4-recommended-toaster.ats-gray{border:1px solid #c5c5c5;background:#f2f2f2}.at4-recommended-toaster.ats-dark{background:#262b30;color:#fff}.at4-recommended-toaster .at4-recommended-container{padding-top:0;margin:0}.at4-recommended.at4-recommended-toaster div.at-recommended-label{line-height:1pc;font-size:1pc;text-align:left;padding:20px 0 0 20px}.at4-toaster-outer .at4-recommended .at4-recommended-item .at4-recommended-item-caption .at-h4{font-size:11px;line-height:11px;margin:10px 0 6px;height:30px}.at4-recommended.at4-recommended-toaster div.at-recommended-label.ats-gray,.at4-recommended.at4-recommended-toaster div.at-recommended-label.ats-light{color:#464646}.at4-recommended.at4-recommended-toaster div.at-recommended-label.ats-dark{color:#fff}.at4-toaster-close-control{position:absolute;top:0;right:0;display:block;width:20px;height:20px;line-height:20px;margin:5px 5px 0 0;padding:0;text-indent:-9999em}.at4-toaster-open-control{position:fixed;right:0;bottom:0;z-index:100020}.at4-toaster-outer .at4-recommended-item{width:90pt;border:0;margin:9px 10px 0}.at4-toaster-outer .at4-recommended-item:first-child{margin-left:20px}.at4-toaster-outer .at4-recommended-item:last-child{margin-right:20px}.at4-toaster-outer .at4-recommended-item .at4-recommended-item-img{max-height:90pt;max-width:90pt}.at4-toaster-outer .at4-recommended-item .at4-recommended-item-img img{height:90pt;width:90pt}.at4-toaster-outer .at4-recommended-item .at4-recommended-item-caption{height:30px;padding:0;margin:0;height:initial}.at4-toaster-outer .ats-dark .at4-recommended-item .at4-recommended-item-caption{background:#262b30}.at4-toaster-outer .at4-recommended .at4-recommended-item .at4-recommended-item-caption small{width:auto;line-height:14px;margin:0}.at4-toaster-outer .at4-recommended.ats-dark .at4-recommended-item .at4-recommended-item-caption small{color:#fff}.at4-recommended-toaster .at-logo{margin:0 0 3px 20px;text-align:left}.at4-recommended-toaster .at-logo .at4-logo-container.at-sponsored-logo{position:relative}.at4-toaster-outer .at4-recommended-item .sponsored-label{text-align:right;font-size:10px;color:#666;float:right;position:fixed;bottom:6px;right:20px;top:initial;z-index:99999}\n </style>\n <style type=\"text/css\">\n .at4-whatsnext{position:fixed;bottom:0!important;right:0;background:#fff;border:1px solid #c5c5c5;margin:-1px;width:390px;height:90pt;overflow:hidden;font-size:9pt;font-weight:400;color:#000;z-index:1800000000}.at4-whatsnext a{color:#666}.at4-whatsnext .at-whatsnext-content{height:90pt;position:relative}.at4-whatsnext .at-whatsnext-content .at-branding{position:absolute;bottom:15px;right:10px;padding-left:9px;text-decoration:none;line-height:10px;font-family:helvetica,arial,sans-serif;font-size:10px;color:#666}.at4-whatsnext .at-whatsnext-content .at-whatsnext-content-inner{position:absolute;top:15px;right:20px;bottom:15px;left:140px;text-align:left;height:105px}.at4-whatsnext .at-whatsnext-content-inner a{display:inline-block}.at4-whatsnext .at-whatsnext-content-inner div.at-h6{text-align:left;margin:0;padding:0 0 3px;font-size:11px;color:#666;cursor:default}.at4-whatsnext .at-whatsnext-content .at-h3{text-align:left;margin:5px 0;padding:0;line-height:1.2em;font-weight:400;font-size:14px;height:3pc}.at4-whatsnext .at-whatsnext-content-inner a:link,.at4-whatsnext .at-whatsnext-content-inner a:visited{text-decoration:none;font-weight:400;color:#464646}.at4-whatsnext .at-whatsnext-content-inner a:hover{color:#000}.at4-whatsnext .at-whatsnext-content-inner small{position:absolute;bottom:15px;line-height:10px;font-size:11px;color:#666;cursor:default;text-align:left}.at4-whatsnext .at-whatsnext-content .at-whatsnext-content-img{position:absolute;top:0;left:0;width:90pt;height:90pt;overflow:hidden}.at4-whatsnext .at-whatsnext-content .at-whatsnext-content-img img{position:absolute;top:0;left:0;max-height:none;max-width:none}.at4-whatsnext .at-whatsnext-close-control{position:absolute;top:0;right:0;display:block;width:20px;height:20px;line-height:20px;margin:0 5px 0 0;padding:0;text-indent:-9999em}.at-whatsnext-open-control{position:fixed;right:0;bottom:0;z-index:100020}.at4-whatsnext.ats-dark{background:#262b30}.at4-whatsnext.ats-dark .at-whatsnext-content .at-h3,.at4-whatsnext.ats-dark .at-whatsnext-content a.at4-logo:hover,.at4-whatsnext.ats-dark .at-whatsnext-content-inner a:link,.at4-whatsnext.ats-dark .at-whatsnext-content-inner a:visited{color:#fff}.at4-whatsnext.ats-light{background:#fff}.at4-whatsnext.ats-gray{background:#f2f2f2}.at4-whatsnext.at-whatsnext-nophoto{width:270px}.at4-whatsnext.at-whatsnext-nophoto .at-whatsnext-content-img{display:none}.at4-whatsnext.at-whatsnext-nophoto .at-whatsnext-content .at-whatsnext-content-inner{top:15px;right:0;left:20px}.at4-whatsnext.at-whatsnext-nophoto .at-whatsnext-content .at-whatsnext-content-inner.addthis_32x32_style{top:0;right:0;left:0;padding:45px 20px 0;font-size:20px}.at4-whatsnext.at-whatsnext-nophoto .at-whatsnext-content .at-whatsnext-content-inner .at4-icon,.at4-whatsnext.at-whatsnext-nophoto .at-whatsnext-content .at-whatsnext-content-inner .at4-icon-fw,.at4-whatsnext.at-whatsnext-nophoto .at-whatsnext-content .at-whatsnext-content-inner .whatsnext-msg{vertical-align:middle}.at-whatsnext-img,.at-whatsnext-img-lnk{position:absolute;left:0}\n </style>\n <style type=\"text/css\">\n .at4-whatsnextmobile{position:fixed;bottom:0;right:0;left:0;background:#fff;z-index:9999998;height:170px;font-size:28px}.at4-whatsnextmobile .col-2{height:100%;font-size:1em}.at4-whatsnextmobile .col-2:first-child{max-width:200px;display:inline-block;float:left}.at4-whatsnextmobile .col-2:last-child{position:absolute;left:200px;right:50px;top:0;bottom:0;display:inline-block}.at4-whatsnextmobile .at-whatsnext-content-inner{font-size:1em}.at4-whatsnextmobile .at-whatsnext-content-img img{height:100%;width:100%}.at4-whatsnextmobile .at-close-control{font-size:1em;position:absolute;top:0;right:0;width:50px;height:50px}.at4-whatsnextmobile .at-close-control button{width:100%;height:100%;font-size:1em;font-weight:400;text-decoration:none;opacity:.5;padding:0;cursor:pointer;background:0 0;border:0;-webkit-appearance:none}.at4-whatsnextmobile .at-h3,.at4-whatsnextmobile .at-h6{font-size:1em;margin:0;color:#a1a1a1;margin-left:2.5%;margin-top:25px}.at4-whatsnextmobile .at-h3{font-size:1em;line-height:1em;font-weight:500;height:50%}.at4-whatsnextmobile .at-h3 a{font-size:1em;text-decoration:none}.at4-whatsnextmobile .at-h6{font-size:.8em;line-height:.8em;font-weight:500}.at4-whatsnextmobile .footer{position:absolute;bottom:2px;left:200px;right:0;padding-left:2.5%;font-size:1em;line-height:.6em}.at4-whatsnextmobile .footer small{font-size:.6em;color:#a1a1a1}.at4-whatsnextmobile .footer small:first-child{margin-right:5%;float:left}.at4-whatsnextmobile .footer small:last-child{margin-right:2.5%;float:right}.at4-whatsnextmobile .at-whatsnext-content{height:100%}.at4-whatsnextmobile.ats-dark{background:#262b30;color:#fff}.at4-whatsnextmobile .at-close-control button{color:#bfbfbf}.at4-whatsnextmobile.ats-dark a:link,.at4-whatsnextmobile.ats-dark a:visited{color:#fff}.at4-whatsnextmobile.ats-gray{background:#f2f2f2;color:#262b30}.at4-whatsnextmobile.ats-light{background:#fff;color:#262b30}.at4-whatsnextmobile.ats-dark .footer a:link,.at4-whatsnextmobile.ats-dark .footer a:visited,.at4-whatsnextmobile.ats-gray .footer a:link,.at4-whatsnextmobile.ats-gray .footer a:visited,.at4-whatsnextmobile.ats-light .footer a:link,.at4-whatsnextmobile.ats-light .footer a:visited{color:#a1a1a1}.at4-whatsnextmobile.ats-gray a:link,.at4-whatsnextmobile.ats-gray a:visited,.at4-whatsnextmobile.ats-light a:link,.at4-whatsnextmobile.ats-light a:visited{color:#262b30}@media only screen and (min-device-width:320px) and (max-device-width:480px){.at4-whatsnextmobile{height:85px;font-size:14px}.at4-whatsnextmobile .col-2:first-child{width:75pt}.at4-whatsnextmobile .col-2:last-child{right:25px;left:75pt}.at4-whatsnextmobile .footer{left:75pt}.at4-whatsnextmobile .at-close-control{width:25px;height:25px}.at4-whatsnextmobile .at-h3,.at4-whatsnextmobile .at-h6{margin-top:12.5px}}\n </style>\n <style type=\"text/css\">\n .at-custom-mobile-bar{left:0;right:0;width:100%;height:56px;position:fixed;text-align:center;z-index:100020;background:#fff;overflow:hidden;box-shadow:0 0 10px 0 rgba(0,0,0,.2);font:initial;line-height:normal;top:auto;bottom:0}.at-custom-mobile-bar.at-custom-mobile-bar-zindex-hide{z-index:-1!important}.at-custom-mobile-bar.atss-top{top:0;bottom:auto}.at-custom-mobile-bar.atss-bottom{top:auto;bottom:0}.at-custom-mobile-bar .at-custom-mobile-bar-btns{display:inline-block;text-align:center}.at-custom-mobile-bar .at-custom-mobile-bar-counter,.at-custom-mobile-bar .at-share-btn{margin-top:4px}.at-custom-mobile-bar .at-share-btn{display:inline-block;text-decoration:none;transition:none;box-sizing:content-box}.at-custom-mobile-bar .at-custom-mobile-bar-counter{font-family:Helvetica neue,arial;vertical-align:top;margin-left:4px;margin-right:4px;display:inline-block}.at-custom-mobile-bar .at-custom-mobile-bar-count{font-size:26px;line-height:1.25em;color:#222}.at-custom-mobile-bar .at-custom-mobile-bar-text{font-size:9pt;line-height:1.25em;color:#888;letter-spacing:1px}.at-custom-mobile-bar .at-icon-wrapper{text-align:center;height:3pc;width:3pc;margin:0 4px}.at-custom-mobile-bar .at-icon{vertical-align:top;margin:8px;width:2pc;height:2pc}.at-custom-mobile-bar.at-shfs-medium{height:3pc}.at-custom-mobile-bar.at-shfs-medium .at-custom-mobile-bar-counter{margin-top:6px}.at-custom-mobile-bar.at-shfs-medium .at-custom-mobile-bar-count{font-size:18px}.at-custom-mobile-bar.at-shfs-medium .at-custom-mobile-bar-text{font-size:10px}.at-custom-mobile-bar.at-shfs-medium .at-icon-wrapper{height:40px;width:40px}.at-custom-mobile-bar.at-shfs-medium .at-icon{margin:6px;width:28px;height:28px}.at-custom-mobile-bar.at-shfs-small{height:40px}.at-custom-mobile-bar.at-shfs-small .at-custom-mobile-bar-counter{margin-top:3px}.at-custom-mobile-bar.at-shfs-small .at-custom-mobile-bar-count{font-size:1pc}.at-custom-mobile-bar.at-shfs-small .at-custom-mobile-bar-text{font-size:10px}.at-custom-mobile-bar.at-shfs-small .at-icon-wrapper{height:2pc;width:2pc}.at-custom-mobile-bar.at-shfs-small .at-icon{margin:4px;width:24px;height:24px}\n </style>\n <style type=\"text/css\">\n .at-custom-sidebar{top:20%;width:58px;position:fixed;text-align:center;z-index:100020;background:#fff;overflow:hidden;box-shadow:0 0 10px 0 rgba(0,0,0,.2);font:initial;line-height:normal;top:auto;bottom:0}.at-custom-sidebar.at-custom-sidebar-zindex-hide{z-index:-1!important}.at-custom-sidebar.atss-left{left:0;right:auto;float:left;border-radius:0 4px 4px 0}.at-custom-sidebar.atss-right{left:auto;right:0;float:right;border-radius:4px 0 0 4px}.at-custom-sidebar .at-custom-sidebar-btns{display:inline-block;text-align:center;padding-top:4px}.at-custom-sidebar .at-custom-sidebar-counter{margin-bottom:8px}.at-custom-sidebar .at-share-btn{display:inline-block;text-decoration:none;transition:none;box-sizing:content-box}.at-custom-sidebar .at-custom-sidebar-counter{font-family:Helvetica neue,arial;vertical-align:top;margin-left:4px;margin-right:4px;display:inline-block}.at-custom-sidebar .at-custom-sidebar-count{font-size:21px;line-height:1.25em;color:#222}.at-custom-sidebar .at-custom-sidebar-text{font-size:10px;line-height:1.25em;color:#888;letter-spacing:1px}.at-custom-sidebar .at-icon-wrapper{text-align:center;margin:0 4px}.at-custom-sidebar .at-icon{vertical-align:top;margin:9px;width:2pc;height:2pc}.at-custom-sidebar .at-icon-wrapper{position:relative}.at-custom-sidebar .at4-share-count,.at-custom-sidebar .at4-share-count-container{line-height:1pc;font-size:10px}.at-custom-sidebar .at4-share-count{text-indent:0;font-family:Arial,Helvetica Neue,Helvetica,sans-serif;font-weight:200;width:100%;height:1pc}.at-custom-sidebar .at4-share-count-anchor .at-icon{margin-top:3px}.at-custom-sidebar .at4-share-count-container{position:absolute;left:0;right:auto;top:auto;bottom:0;width:100%;color:#fff;background:inherit}\n </style>\n <style type=\"text/css\">\n .at-image-sharing-mobile-icon{position:absolute;background:#000 url(//s7.addthis.com/static/44a36d35bafef33aa9455b7d3039a771.png) no-repeat top center;background-color:rgba(0,0,0,.9);background-image:url(//s7.addthis.com/static/10db525181ee0bbe1a515001be1c7818.svg),none;border-radius:3px;width:50px;height:40px;top:-9999px;left:-9999px}.at-image-sharing-tool{display:block;position:absolute;text-align:center;z-index:9001;background:none;overflow:hidden;top:-9999px;left:-9999px;font:initial;line-height:0}.at-image-sharing-tool.addthis-animated{animation-duration:.15s}.at-image-sharing-tool.at-orientation-vertical .at-share-btn{display:block}.at-image-sharing-tool.at-orientation-horizontal .at-share-btn{display:inline-block}.at-image-sharing-tool.at-image-sharing-tool-size-big .at-icon{width:43px;height:43px}.at-image-sharing-tool.at-image-sharing-tool-size-mobile .at-share-btn{margin:0!important}.at-image-sharing-tool.at-image-sharing-tool-size-mobile .at-icon-wrapper{height:60px;width:100%;border-radius:0!important}.at-image-sharing-tool.at-image-sharing-tool-size-mobile .at-icon{max-width:100%;height:54px!important;width:54px!important}.at-image-sharing-tool .at-custom-shape.at-image-sharing-tool-btns{margin-right:8px;margin-bottom:8px}.at-image-sharing-tool .at-custom-shape .at-share-btn{margin-top:8px;margin-left:8px}.at-image-sharing-tool .at-share-btn{line-height:0;text-decoration:none;transition:none;box-sizing:content-box}.at-image-sharing-tool .at-icon-wrapper{text-align:center;height:100%;width:100%}.at-image-sharing-tool .at-icon{vertical-align:top;width:2pc;height:2pc;margin:3px}\n </style>\n <style type=\"text/css\">\n .at-expanding-share-button{box-sizing:border-box;position:fixed;z-index:9999}.at-expanding-share-button[data-position=bottom-right]{bottom:10px;right:10px}.at-expanding-share-button[data-position=bottom-right] .at-expanding-share-button-toggle-bg,.at-expanding-share-button[data-position=bottom-right] .at-expanding-share-button-toggle-btn[data-name]:after,.at-expanding-share-button[data-position=bottom-right] .at-icon-wrapper,.at-expanding-share-button[data-position=bottom-right] [data-name]:after{float:right}.at-expanding-share-button[data-position=bottom-right] [data-name]:after{margin-right:10px}.at-expanding-share-button[data-position=bottom-right] .at-expanding-share-button-toggle-btn[data-name]:after{margin-right:5px}.at-expanding-share-button[data-position=bottom-right] .at-icon-wrapper{margin-right:-3px}.at-expanding-share-button[data-position=bottom-left]{bottom:10px;left:10px}.at-expanding-share-button[data-position=bottom-left] .at-expanding-share-button-toggle-bg,.at-expanding-share-button[data-position=bottom-left] .at-expanding-share-button-toggle-btn[data-name]:after,.at-expanding-share-button[data-position=bottom-left] .at-icon-wrapper,.at-expanding-share-button[data-position=bottom-left] [data-name]:after{float:left}.at-expanding-share-button[data-position=bottom-left] [data-name]:after{margin-left:10px}.at-expanding-share-button[data-position=bottom-left] .at-expanding-share-button-toggle-btn[data-name]:after{margin-left:5px}.at-expanding-share-button *,.at-expanding-share-button :after,.at-expanding-share-button :before{box-sizing:border-box}.at-expanding-share-button .at-expanding-share-button-services-list{display:none;list-style:none;margin:0 5px;overflow:visible;padding:0}.at-expanding-share-button .at-expanding-share-button-services-list>li{display:block;height:45px;position:relative;overflow:visible}.at-expanding-share-button .at-expanding-share-button-toggle-btn,.at-expanding-share-button .at-share-btn{transition:.1s;text-decoration:none}.at-expanding-share-button .at-share-btn{display:block;height:40px;padding:0 3px 0 0}.at-expanding-share-button .at-expanding-share-button-toggle-btn{position:relative;overflow:auto}.at-expanding-share-button .at-expanding-share-button-toggle-btn.at-expanding-share-button-hidden[data-name]:after{display:none}.at-expanding-share-button .at-expanding-share-button-toggle-bg{box-shadow:0 2px 4px 0 rgba(0,0,0,.3);border-radius:50%;position:relative}.at-expanding-share-button .at-expanding-share-button-toggle-bg>span{background-image:url(\"data:image/svg+xml,%3Csvg%20width%3D%2232px%22%20height%3D%2232px%22%20viewBox%3D%220%200%2032%2032%22%20version%3D%221.1%22%20xmlns%3D%22http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%22%3E%3Ctitle%3Eshare%3C%2Ftitle%3E%3Cg%20stroke%3D%22none%22%20stroke-width%3D%221%22%20fill%3D%22none%22%20fill-rule%3D%22evenodd%22%3E%3Cg%20fill%3D%22%23FFFFFF%22%3E%3Cpath%20d%3D%22M26%2C13.4285714%20C26%2C13.6220248%2025.9293162%2C13.7894338%2025.7879464%2C13.9308036%20L20.0736607%2C19.6450893%20C19.932291%2C19.786459%2019.7648819%2C19.8571429%2019.5714286%2C19.8571429%20C19.3779752%2C19.8571429%2019.2105662%2C19.786459%2019.0691964%2C19.6450893%20C18.9278267%2C19.5037195%2018.8571429%2C19.3363105%2018.8571429%2C19.1428571%20L18.8571429%2C16.2857143%20L16.3571429%2C16.2857143%20C15.6279725%2C16.2857143%2014.9750773%2C16.3080355%2014.3984375%2C16.3526786%20C13.8217977%2C16.3973217%2013.2488868%2C16.477306%2012.6796875%2C16.5926339%20C12.1104882%2C16.7079619%2011.6157015%2C16.8660704%2011.1953125%2C17.0669643%20C10.7749235%2C17.2678581%2010.3824423%2C17.5264121%2010.0178571%2C17.8426339%20C9.65327199%2C18.1588557%209.35565592%2C18.534596%209.125%2C18.9698661%20C8.89434408%2C19.4051361%208.71391434%2C19.9203839%208.58370536%2C20.515625%20C8.45349637%2C21.1108661%208.38839286%2C21.7842224%208.38839286%2C22.5357143%20C8.38839286%2C22.9449425%208.40699386%2C23.4025272%208.44419643%2C23.9084821%20C8.44419643%2C23.9531252%208.45349693%2C24.0405499%208.47209821%2C24.1707589%20C8.4906995%2C24.3009679%208.5%2C24.3995532%208.5%2C24.4665179%20C8.5%2C24.5781256%208.46837829%2C24.6711306%208.40513393%2C24.7455357%20C8.34188956%2C24.8199408%208.25446484%2C24.8571429%208.14285714%2C24.8571429%20C8.02380893%2C24.8571429%207.9196433%2C24.7938994%207.83035714%2C24.6674107%20C7.77827355%2C24.6004461%207.72991094%2C24.5186017%207.68526786%2C24.421875%20C7.64062478%2C24.3251483%207.59040206%2C24.2135423%207.53459821%2C24.0870536%20C7.47879436%2C23.9605648%207.43973225%2C23.87128%207.41741071%2C23.8191964%20C6.47246551%2C21.6986501%206%2C20.0208395%206%2C18.7857143%20C6%2C17.3050521%206.19717065%2C16.0662252%206.59151786%2C15.0691964%20C7.79688103%2C12.0706695%2011.0520568%2C10.5714286%2016.3571429%2C10.5714286%20L18.8571429%2C10.5714286%20L18.8571429%2C7.71428571%20C18.8571429%2C7.52083237%2018.9278267%2C7.35342333%2019.0691964%2C7.21205357%20C19.2105662%2C7.07068382%2019.3779752%2C7%2019.5714286%2C7%20C19.7648819%2C7%2019.932291%2C7.07068382%2020.0736607%2C7.21205357%20L25.7879464%2C12.9263393%20C25.9293162%2C13.067709%2026%2C13.2351181%2026%2C13.4285714%20L26%2C13.4285714%20Z%22%3E%3C%2Fpath%3E%3C%2Fg%3E%3C%2Fg%3E%3C%2Fsvg%3E\");background-position:center center;background-repeat:no-repeat;transition:transform .4s ease;border-radius:50%;display:block}.at-expanding-share-button .at-icon-wrapper{box-shadow:0 2px 4px 0 rgba(0,0,0,.3);border-radius:50%;display:inline-block;height:40px;line-height:40px;text-align:center;width:40px}.at-expanding-share-button .at-icon{display:inline-block;height:34px;margin:3px 0;vertical-align:top;width:34px}.at-expanding-share-button [data-name]:after{box-shadow:0 2px 4px 0 rgba(0,0,0,.3);transform:translate(0, -50%);transition:.4s;background-color:#fff;border-radius:3px;color:#666;content:attr(data-name);font-family:Helvetica Neue,Helvetica,Arial,sans-serif;font-size:9pt;line-height:9pt;font-weight:500;opacity:0;padding:3px 5px;position:relative;top:20px;white-space:nowrap}.at-expanding-share-button.at-expanding-share-button-show-icons .at-expanding-share-button-services-list{display:block}.at-expanding-share-button.at-expanding-share-button-animate-in .at-expanding-share-button-toggle-bg>span{transform:rotate(270deg);background-image:url(\"data:image/svg+xml,%3Csvg%20xmlns%3D%22http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%22%20xmlns%3Axlink%3D%22http%3A%2F%2Fwww.w3.org%2F1999%2Fxlink%22%20viewBox%3D%220%200%2032%2032%22%3E%3Cg%3E%3Cpath%20d%3D%22M18%2014V8h-4v6H8v4h6v6h4v-6h6v-4h-6z%22%20fill-rule%3D%22evenodd%22%20fill%3D%22white%22%3E%3C%2Fpath%3E%3C%2Fg%3E%3C%2Fsvg%3E\");background-position:center center;background-repeat:no-repeat}.at-expanding-share-button.at-expanding-share-button-animate-in [data-name]:after{opacity:1}.at-expanding-share-button.at-hide-label [data-name]:after{display:none}.at-expanding-share-button.at-expanding-share-button-desktop .at-expanding-share-button-toggle{height:50px}.at-expanding-share-button.at-expanding-share-button-desktop .at-icon-wrapper:hover{box-shadow:0 2px 5px 0 rgba(0,0,0,.5)}.at-expanding-share-button.at-expanding-share-button-desktop .at-expanding-share-button-toggle-bg{height:50px;line-height:50px;width:50px}.at-expanding-share-button.at-expanding-share-button-desktop .at-expanding-share-button-toggle-bg>span{height:50px;width:50px}.at-expanding-share-button.at-expanding-share-button-desktop .at-expanding-share-button-toggle-bg:after{box-shadow:0 2px 5px 0 rgba(0,0,0,.2);transition:opacity .2s ease;border-radius:50%;content:'';height:100%;opacity:0;position:absolute;top:0;left:0;width:100%}.at-expanding-share-button.at-expanding-share-button-desktop .at-expanding-share-button-toggle-bg:hover:after{opacity:1}.at-expanding-share-button.at-expanding-share-button-desktop .at-expanding-share-button-toggle-btn[data-name]:after{top:25px}.at-expanding-share-button.at-expanding-share-button-mobile .at-expanding-share-button-services-list{margin:0}.at-expanding-share-button.at-expanding-share-button-mobile .at-expanding-share-button-toggle-btn,.at-expanding-share-button.at-expanding-share-button-mobile .at-share-btn{outline:0}.at-expanding-share-button.at-expanding-share-button-mobile .at-expanding-share-button-toggle{height:40px;-webkit-tap-highlight-color:transparent}.at-expanding-share-button.at-expanding-share-button-mobile .at-expanding-share-button-toggle-bg,.at-expanding-share-button.at-expanding-share-button-mobile .at-expanding-share-button-toggle-bg span{height:40px;line-height:40px;width:40px}.at-expanding-share-button.at-expanding-share-button-mobile .at-expanding-share-button-click-flash{transform:scale(0);transition:transform ease,opacity ease-in;background-color:hsla(0,0%,100%,.3);border-radius:50%;height:40px;opacity:1;position:absolute;width:40px;z-index:10000}.at-expanding-share-button.at-expanding-share-button-mobile .at-expanding-share-button-click-flash.at-expanding-share-button-click-flash-animate{transform:scale(1);opacity:0}.at-expanding-share-button.at-expanding-share-button-mobile+.at-expanding-share-button-mobile-overlay{transition:opacity ease;bottom:0;background-color:hsla(0,0%,87%,.7);display:block;height:auto;left:0;opacity:0;position:fixed;right:0;top:0;width:auto;z-index:9998}.at-expanding-share-button.at-expanding-share-button-mobile+.at-expanding-share-button-mobile-overlay.at-expanding-share-button-hidden{height:0;width:0;z-index:-10000}.at-expanding-share-button.at-expanding-share-button-mobile.at-expanding-share-button-animate-in+.at-expanding-share-button-mobile-overlay{transition:opacity ease;opacity:1}\n </style>\n <style type=\"text/css\">\n .at-tjin-element .at300b,.at-tjin-element .at300m{display:inline-block;width:auto;padding:0;margin:0 2px 5px;outline-offset:-1px;transition:all .2s ease-in-out}.at-tjin-element .at300b:focus,.at-tjin-element .at300b:hover,.at-tjin-element .at300m:focus,.at-tjin-element .at300m:hover{transform:translateY(-4px)}.at-tjin-element .addthis_tjin_label{display:none}.at-tjin-element .addthis_vertical_style .at300b,.at-tjin-element .addthis_vertical_style .at300m{display:block}.at-tjin-element .addthis_vertical_style .at300b .addthis_tjin_label,.at-tjin-element .addthis_vertical_style .at300b .at-icon-wrapper,.at-tjin-element .addthis_vertical_style .at300m .addthis_tjin_label,.at-tjin-element .addthis_vertical_style .at300m .at-icon-wrapper{display:inline-block;vertical-align:middle;margin-right:5px}.at-tjin-element .addthis_vertical_style .at300b:focus,.at-tjin-element .addthis_vertical_style .at300b:hover,.at-tjin-element .addthis_vertical_style .at300m:focus,.at-tjin-element .addthis_vertical_style .at300m:hover{transform:none}.at-tjin-element .at-tjin-btn{margin:0 5px 5px 0;padding:0;outline-offset:-1px;display:inline-block;box-sizing:content-box;transition:all .2s ease-in-out}.at-tjin-element .at-tjin-btn:focus,.at-tjin-element .at-tjin-btn:hover{transform:translateY(-4px)}.at-tjin-element .at-tjin-title{margin:0 0 15px}\n </style>\n <style type=\"text/css\">\n #addthissmartlayerscssready{color:#bada55!important}.addthis-smartlayers,div#at4-follow,div#at4-share,div#at4-thankyou,div#at4-whatsnext{padding:0;margin:0}#at4-follow-label,#at4-share-label,#at4-whatsnext-label,.at4-recommended-label.hidden{padding:0;border:none;background:none;position:absolute;top:0;left:0;height:0;width:0;overflow:hidden;text-indent:-9999em}.addthis-smartlayers .at4-arrow:hover{cursor:pointer}.addthis-smartlayers .at4-arrow:after,.addthis-smartlayers .at4-arrow:before{content:none}a.at4-logo{background:url(data:image/gif;base64,R0lGODlhBwAHAJEAAP9uQf///wAAAAAAACH5BAkKAAIALAAAAAAHAAcAAAILFH6Ge8EBH2MKiQIAOw==) no-repeat left center}.at4-minimal a.at4-logo{background:url(data:image/gif;base64,R0lGODlhBwAHAJEAAP9uQf///wAAAAAAACH5BAkKAAIALAAAAAAHAAcAAAILFH6Ge8EBH2MKiQIAOw==) no-repeat left center!important}button.at4-closebutton{position:absolute;top:0;right:0;padding:0;margin-right:10px;cursor:pointer;background:transparent;border:0;-webkit-appearance:none;font-size:19px;line-height:1;color:#000;text-shadow:0 1px 0 #fff;opacity:.2}button.at4-closebutton:hover{color:#000;text-decoration:none;cursor:pointer;opacity:.5}div.at4-arrow{background-image:url(data:image/png;base64,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);background-repeat:no-repeat;width:20px;height:20px;margin:0;padding:0;overflow:hidden;text-indent:-9999em;text-align:left;cursor:pointer}#at4-recommendedpanel-outer-container .at4-arrow.at-right,div.at4-arrow.at-right{background-position:-20px 0}#at4-recommendedpanel-outer-container .at4-arrow.at-left,div.at4-arrow.at-left{background-position:0 0}div.at4-arrow.at-down{background-position:-60px 0}div.at4-arrow.at-up{background-position:-40px 0}.ats-dark div.at4-arrow.at-right{background-position:-20px -20px}.ats-dark div.at4-arrow.at-left{background-position:0 -20px}.ats-dark div.at4-arrow.at-down{background-position:-60px -20px}.ats-dark div.at4-arrow.at-up{background-position:-40px -20}.at4-opacity-hidden{opacity:0!important}.at4-opacity-visible{opacity:1!important}.at4-visually-hidden{position:absolute;clip:rect(1px,1px,1px,1px);padding:0;border:0;overflow:hidden}.at4-hidden-off-screen,.at4-hidden-off-screen *{position:absolute!important;top:-9999px!important;left:-9999px!important}.at4-show{display:block!important;opacity:1!important}.at4-show-content{opacity:1!important;visibility:visible}.at4-hide{display:none!important;opacity:0!important}.at4-hide-content{opacity:0!important;visibility:hidden}.at4-visible{display:block!important;opacity:0!important}.at-wordpress-hide{display:none!important;opacity:0!important}.addthis-animated{animation-fill-mode:both;animation-timing-function:ease-out;animation-duration:.3s}.slideInDown.addthis-animated,.slideInLeft.addthis-animated,.slideInRight.addthis-animated,.slideInUp.addthis-animated,.slideOutDown.addthis-animated,.slideOutLeft.addthis-animated,.slideOutRight.addthis-animated,.slideOutUp.addthis-animated{animation-duration:.4s}@keyframes fadeIn{0%{opacity:0}to{opacity:1}}.fadeIn{animation-name:fadeIn}@keyframes fadeInUp{0%{opacity:0;transform:translateY(20px)}to{opacity:1;transform:translateY(0)}}.fadeInUp{animation-name:fadeInUp}@keyframes fadeInDown{0%{opacity:0;transform:translateY(-20px)}to{opacity:1;transform:translateY(0)}}.fadeInDown{animation-name:fadeInDown}@keyframes fadeInLeft{0%{opacity:0;transform:translateX(-20px)}to{opacity:1;transform:translateX(0)}}.fadeInLeft{animation-name:fadeInLeft}@keyframes fadeInRight{0%{opacity:0;transform:translateX(20px)}to{opacity:1;transform:translateX(0)}}.fadeInRight{animation-name:fadeInRight}@keyframes fadeOut{0%{opacity:1}to{opacity:0}}.fadeOut{animation-name:fadeOut}@keyframes fadeOutUp{0%{opacity:1;transform:translateY(0)}to{opacity:0;transform:translateY(-20px)}}.fadeOutUp{animation-name:fadeOutUp}@keyframes fadeOutDown{0%{opacity:1;transform:translateY(0)}to{opacity:0;transform:translateY(20px)}}.fadeOutDown{animation-name:fadeOutDown}@keyframes fadeOutLeft{0%{opacity:1;transform:translateX(0)}to{opacity:0;transform:translateX(-20px)}}.fadeOutLeft{animation-name:fadeOutLeft}@keyframes fadeOutRight{0%{opacity:1;transform:translateX(0)}to{opacity:0;transform:translateX(20px)}}.fadeOutRight{animation-name:fadeOutRight}@keyframes slideInUp{0%{transform:translateY(1500px)}0%,to{opacity:1}to{transform:translateY(0)}}.slideInUp{animation-name:slideInUp}.slideInUp.addthis-animated{animation-duration:.4s}@keyframes slideInDown{0%{transform:translateY(-850px)}0%,to{opacity:1}to{transform:translateY(0)}}.slideInDown{animation-name:slideInDown}@keyframes slideOutUp{0%{transform:translateY(0)}0%,to{opacity:1}to{transform:translateY(-250px)}}.slideOutUp{animation-name:slideOutUp}@keyframes slideOutUpFast{0%{transform:translateY(0)}0%,to{opacity:1}to{transform:translateY(-1250px)}}#at4m-menu.slideOutUp{animation-name:slideOutUpFast}@keyframes slideOutDown{0%{transform:translateY(0)}0%,to{opacity:1}to{transform:translateY(350px)}}.slideOutDown{animation-name:slideOutDown}@keyframes slideOutDownFast{0%{transform:translateY(0)}0%,to{opacity:1}to{transform:translateY(1250px)}}#at4m-menu.slideOutDown{animation-name:slideOutDownFast}@keyframes slideInLeft{0%{opacity:0;transform:translateX(-850px)}to{transform:translateX(0)}}.slideInLeft{animation-name:slideInLeft}@keyframes slideInRight{0%{opacity:0;transform:translateX(1250px)}to{transform:translateX(0)}}.slideInRight{animation-name:slideInRight}@keyframes slideOutLeft{0%{transform:translateX(0)}to{opacity:0;transform:translateX(-350px)}}.slideOutLeft{animation-name:slideOutLeft}@keyframes slideOutRight{0%{transform:translateX(0)}to{opacity:0;transform:translateX(350px)}}.slideOutRight{animation-name:slideOutRight}.at4win{margin:0 auto;background:#fff;border:1px solid #ebeced;width:25pc;box-shadow:0 0 10px rgba(0,0,0,.3);border-radius:8px;font-family:helvetica neue,helvetica,arial,sans-serif;text-align:left;z-index:9999}.at4win .at4win-header{position:relative;border-bottom:1px solid #f2f2f2;background:#fff;height:49px;-webkit-border-top-left-radius:8px;-webkit-border-top-right-radius:8px;-moz-border-radius-topleft:8px;-moz-border-radius-topright:8px;border-top-left-radius:8px;border-top-right-radius:8px;cursor:default}.at4win .at4win-header .at-h3,.at4win .at4win-header h3{height:49px;line-height:49px;margin:0 50px 0 0;padding:1px 0 0;margin-left:20px;font-family:helvetica neue,helvetica,arial,sans-serif;font-size:1pc;font-weight:700;text-shadow:0 1px #fff;color:#333}.at4win .at4win-header .at-h3 img,.at4win .at4win-header h3 img{display:inline-block;margin-right:4px}.at4win .at4win-header .at4-close{display:block;position:absolute;top:0;right:0;background:url(\"data:image/gif;base64,R0lGODlhFAAUAIABAAAAAP///yH5BAEAAAEALAAAAAAUABQAAAIzBIKpG+YMm5Enpodw1HlCfnkKOIqU1VXk55goVb2hi7Y0q95lfG70uurNaqLgTviyyUoFADs=\") no-repeat center center;background-repeat:no-repeat;background-position:center center;border-left:1px solid #d2d2d1;width:49px;height:49px;line-height:49px;overflow:hidden;text-indent:-9999px;text-shadow:none;cursor:pointer;opacity:.5;border:0;transition:opacity .15s ease-in}.at4win .at4win-header .at4-close::-moz-focus-inner{border:0;padding:0}.at4win .at4win-header .at4-close:hover{opacity:1;background-color:#ebeced;border-top-right-radius:7px}.at4win .at4win-content{position:relative;background:#fff;min-height:220px}#at4win-footer{position:relative;background:#fff;border-top:1px solid #d2d2d1;-webkit-border-bottom-right-radius:8px;-webkit-border-bottom-left-radius:8px;-moz-border-radius-bottomright:8px;-moz-border-radius-bottomleft:8px;border-bottom-right-radius:8px;border-bottom-left-radius:8px;height:11px;line-height:11px;padding:5px 20px;font-size:11px;color:#666;-ms-box-sizing:content-box;-o-box-sizing:content-box;box-sizing:content-box}#at4win-footer a{margin-right:10px;text-decoration:none;color:#666}#at4win-footer a:hover{text-decoration:none;color:#000}#at4win-footer a.at4-logo{top:5px;padding-left:10px}#at4win-footer a.at4-privacy{position:absolute;top:5px;right:10px;padding-right:14px}.at4win.ats-dark{border-color:#555;box-shadow:none}.at4win.ats-dark .at4win-header{background:#1b1b1b;-webkit-border-top-left-radius:6px;-webkit-border-top-right-radius:6px;-moz-border-radius-topleft:6px;-moz-border-radius-topright:6px;border-top-left-radius:6px;border-top-right-radius:6px}.at4win.ats-dark .at4win-header .at4-close{background:url(\"data:image/gif;base64,iVBORw0KGgoAAAANSUhEUgAAABQAAAAUCAYAAACNiR0NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAABx0RVh0U29mdHdhcmUAQWRvYmUgRmlyZXdvcmtzIENTNui8sowAAAAWdEVYdENyZWF0aW9uIFRpbWUAMTEvMTMvMTKswDp5AAAAd0lEQVQ4jb2VQRLAIAgDE///Z3qqY1FAhalHMCsCIkVEAIAkkVgvp2lDBgYAnAyHkWotLccNrEd4A7X2TqIdqLfnWBAdaF5rJdyJfjtPH5GT37CaGhoVq3nOm/XflUuLUto2pY1d+vRKh0Pp+MrAVtDe2JkvYNQ+jVSEEFmOkggAAAAASUVORK5CYII=\") no-repeat center center;background-image:url(//s7.addthis.com/static/fb08f6d50887bd0caacc86a62bcdcf68.svg),none;border-color:#333}.at4win.ats-dark .at4win-header .at4-close:hover{background-color:#000}.at4win.ats-dark .at4win-header .at-h3,.at4win.ats-dark .at4win-header h3{color:#fff;text-shadow:0 1px #000}.at4win.ats-gray .at4win-header{background:#fff;border-color:#d2d2d1;-webkit-border-top-left-radius:6px;-webkit-border-top-right-radius:6px;-moz-border-radius-topleft:6px;-moz-border-radius-topright:6px;border-top-left-radius:6px;border-top-right-radius:6px}.at4win.ats-gray .at4win-header a.at4-close{border-color:#d2d2d1}.at4win.ats-gray .at4win-header a.at4-close:hover{background-color:#ebeced}.at4win.ats-gray #at4win-footer{border-color:#ebeced}.at4win .clear{clear:both}.at4win ::selection{background:#fe6d4c;color:#fff}.at4win ::-moz-selection{background:#fe6d4c;color:#fff}.at4-icon-fw{display:inline-block;background-repeat:no-repeat;background-position:0 0;margin:0 5px 0 0;overflow:hidden;text-indent:-9999em;cursor:pointer;padding:0;border-radius:50%;-moz-border-radius:50%;-webkit-border-radius:50%}.at44-follow-container a.aticon{height:2pc;margin:0 5px 5px 0}.at44-follow-container .at4-icon-fw{margin:0}\n </style>\n <style id=\"at4-share-offset\" media=\"screen\" type=\"text/css\">\n #at4-share,#at4-soc {bottom:25px !important;top:auto;}\n </style>\n </head>\n <body id=\"news\" style=\"\">\n <div data-react-class=\"BrowseHappier\" data-react-props='{\"gt\":1,\"lt\":11}'>\n <!-- react-empty: 1 -->\n </div>\n <div id=\"main_container\">\n <div id=\"site_body\">\n <div class=\"site_header_area\">\n <header class=\"site_header\">\n <div class=\"brand_area\">\n <div class=\"brand1\">\n <a class=\"nasa_logo\" href=\"http://www.nasa.gov\" target=\"_blank\" title=\"visit nasa.gov\">\n NASA\n </a>\n </div>\n <div class=\"brand2\">\n <a class=\"top_logo\" href=\"https://science.nasa.gov/\" target=\"_blank\" title=\"Explore NASA Science\">\n NASA Science\n </a>\n <a class=\"sub_logo\" href=\"/mars-exploration/#\" title=\"Mars\">\n Mars Exploration Program\n </a>\n </div>\n <img alt=\"\" class=\"print_only print_logo\" src=\"/assets/[email protected]\"/>\n </div>\n <a class=\"visuallyhidden focusable\" href=\"#page\">\n Skip Navigation\n </a>\n <div class=\"right_header_container\">\n <a class=\"menu_button\" href=\"javascript:void(0);\" id=\"menu_button\">\n <span class=\"menu_icon\">\n menu\n </span>\n </a>\n <a class=\"modal_close\" id=\"modal_close\">\n <span class=\"modal_close_icon\">\n </span>\n </a>\n </div>\n <div class=\"nav_area\">\n <div id=\"site_nav_container\">\n <nav class=\"site_nav\">\n <ul class=\"nav\">\n <li>\n <div class=\"arrow_box\">\n <span class=\"arrow_down\">\n </span>\n </div>\n <div class=\"nav_title\">\n <a class=\"main_nav_item\" href=\"/#red_planet\" target=\"_self\">\n The Red Planet\n </a>\n </div>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a href=\"/#red_planet/0\" target=\"_self\">\n Dashboard\n </a>\n </li>\n <li>\n <a href=\"/#red_planet/1\" target=\"_self\">\n Science Goals\n </a>\n </li>\n <li>\n <a href=\"/#red_planet/2\" target=\"_self\">\n The Planet\n </a>\n </li>\n <li>\n <a href=\"/#red_planet/3\" target=\"_self\">\n Atmosphere\n </a>\n </li>\n <li>\n <a href=\"/#red_planet/4\" target=\"_self\">\n Astrobiology\n </a>\n </li>\n <li>\n <a href=\"/#red_planet/5\" target=\"_self\">\n Past, Present, Future, Timeline\n </a>\n </li>\n </ul>\n </div>\n <div class=\"gradient_line\">\n </div>\n </li>\n <li>\n <div class=\"arrow_box\">\n <span class=\"arrow_down\">\n </span>\n </div>\n <div class=\"nav_title\">\n <a class=\"main_nav_item\" href=\"/#mars_exploration_program\" target=\"_self\">\n The Program\n </a>\n </div>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a href=\"/#mars_exploration_program/0\" target=\"_self\">\n Mission Statement\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/1\" target=\"_self\">\n About the Program\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/2\" target=\"_self\">\n Organization\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/3\" target=\"_self\">\n Why Mars?\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/4\" target=\"_self\">\n Research Programs\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/5\" target=\"_self\">\n Planetary Resources\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/6\" target=\"_self\">\n Technologies\n </a>\n </li>\n </ul>\n </div>\n <div class=\"gradient_line\">\n </div>\n </li>\n <li>\n <div class=\"arrow_box\">\n <span class=\"arrow_down\">\n </span>\n </div>\n <div class=\"nav_title\">\n <a class=\"main_nav_item\" href=\"/#news_and_events\" target=\"_self\">\n News & Events\n </a>\n </div>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a href=\"/news\" target=\"_self\">\n News\n </a>\n </li>\n <li>\n <a href=\"/events\" target=\"_self\">\n Events\n </a>\n </li>\n </ul>\n </div>\n <div class=\"gradient_line\">\n </div>\n </li>\n <li>\n <div class=\"arrow_box\">\n <span class=\"arrow_down\">\n </span>\n </div>\n <div class=\"nav_title\">\n <a class=\"main_nav_item\" href=\"/#multimedia\" target=\"_self\">\n Multimedia\n </a>\n </div>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a href=\"/multimedia/images/\" target=\"_self\">\n Images\n </a>\n </li>\n <li>\n <a href=\"/multimedia/videos/\" target=\"_self\">\n Videos\n </a>\n </li>\n </ul>\n </div>\n <div class=\"gradient_line\">\n </div>\n </li>\n <li>\n <div class=\"arrow_box\">\n <span class=\"arrow_down\">\n </span>\n </div>\n <div class=\"nav_title\">\n <a class=\"main_nav_item\" href=\"/#missions\" target=\"_self\">\n Missions\n </a>\n </div>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a href=\"/mars-exploration/missions/?category=167\" target=\"_self\">\n Past\n </a>\n </li>\n <li>\n <a href=\"/mars-exploration/missions/?category=170\" target=\"_self\">\n Present\n </a>\n </li>\n <li>\n <a href=\"/mars-exploration/missions/?category=171\" target=\"_self\">\n Future\n </a>\n </li>\n <li>\n <a href=\"/mars-exploration/partners\" target=\"_self\">\n International Partners\n </a>\n </li>\n </ul>\n </div>\n <div class=\"gradient_line\">\n </div>\n </li>\n <li>\n <div class=\"nav_title\">\n <a class=\"main_nav_item\" href=\"/#more\" target=\"_self\">\n More\n </a>\n </div>\n <div class=\"gradient_line\">\n </div>\n </li>\n <li>\n <div class=\"nav_title\">\n <a class=\"main_nav_item\" href=\"/legacy\" target=\"_self\">\n Legacy Site\n </a>\n </div>\n <div class=\"gradient_line\">\n </div>\n </li>\n </ul>\n <form action=\"https://mars.nasa.gov/search/\" class=\"overlay_search nav_search\">\n <label class=\"search_label\">\n Search\n </label>\n <input class=\"search_field\" name=\"q\" type=\"text\" value=\"\"/>\n <div class=\"search_submit\">\n </div>\n </form>\n </nav>\n </div>\n </div>\n </header>\n </div>\n <div id=\"sticky_nav_spacer\">\n </div>\n <div id=\"page\">\n <!-- title to go in the page_header -->\n <div class=\"header_mask\">\n </div>\n <div class=\"react_grid_list\" data-react-class=\"GridListPage\" data-react-props='{\"left_column\":false,\"class_name\":\"\",\"default_view\":\"list_view\",\"model\":\"news_items\",\"view_toggle\":false,\"search\":\"true\",\"list_item\":\"News\",\"title\":\"News\",\"categories\":[\"19,165,184,204\"],\"order\":\"publish_date desc,created_at desc\",\"no_items_text\":\"There are no items matching these criteria.\",\"per_page\":null,\"filters\":\"[ [ \\\"date\\\", [ [ \\\"2018\\\", \\\"2018\\\" ], [ \\\"2017\\\", \\\"2017\\\" ], [ \\\"2016\\\", \\\"2016\\\" ], [ \\\"2015\\\", \\\"2015\\\" ], [ \\\"2014\\\", \\\"2014\\\" ], [ \\\"2013\\\", \\\"2013\\\" ], [ \\\"2012\\\", \\\"2012\\\" ], [ \\\"2011\\\", \\\"2011\\\" ], [ \\\"2010\\\", \\\"2010\\\" ], [ \\\"2009\\\", \\\"2009\\\" ], [ \\\"2008\\\", \\\"2008\\\" ], [ \\\"2007\\\", \\\"2007\\\" ], [ \\\"2006\\\", \\\"2006\\\" ], [ \\\"2005\\\", \\\"2005\\\" ], [ \\\"2004\\\", \\\"2004\\\" ], [ \\\"2003\\\", \\\"2003\\\" ], [ \\\"2002\\\", \\\"2002\\\" ], [ \\\"2001\\\", \\\"2001\\\" ], [ \\\"2000\\\", \\\"2000\\\" ] ], [ \\\"Latest\\\", \\\"\\\" ], false ], [ \\\"categories\\\", [ [ \\\"Feature Stories\\\", 165 ], [ \\\"Press Releases\\\", 19 ], [ \\\"Spotlights\\\", 184 ], [ \\\"Status Reports\\\", 204 ] ], [ \\\"All Categories\\\", \\\"\\\" ], false ] ]\",\"conditions\":null,\"scope_in_title\":true,\"options\":{\"blank_scope\":\"Latest\"},\"results_in_title\":false}'>\n <section class=\"grid_gallery module list_view\" data-reactroot=\"\">\n <div class=\"grid_layout\">\n <header class=\"gallery_header\">\n <h2 class=\"module_title\">\n News\n </h2>\n <section class=\"filter_bar\">\n <div class=\"section_search\">\n <div class=\"search_binder\">\n <input class=\"search_field\" name=\"search\" placeholder=\"search\" type=\"text\" value=\"\"/>\n <input class=\"search_submit\" type=\"submit\" value=\"\"/>\n </div>\n <select class=\"filter\" id=\"date\" name=\"date\">\n <option value=\"\">\n Latest\n </option>\n <option value=\"2018\">\n 2018\n </option>\n <option value=\"2017\">\n 2017\n </option>\n <option value=\"2016\">\n 2016\n </option>\n <option value=\"2015\">\n 2015\n </option>\n <option value=\"2014\">\n 2014\n </option>\n <option value=\"2013\">\n 2013\n </option>\n <option value=\"2012\">\n 2012\n </option>\n <option value=\"2011\">\n 2011\n </option>\n <option value=\"2010\">\n 2010\n </option>\n <option value=\"2009\">\n 2009\n </option>\n <option value=\"2008\">\n 2008\n </option>\n <option value=\"2007\">\n 2007\n </option>\n <option value=\"2006\">\n 2006\n </option>\n <option value=\"2005\">\n 2005\n </option>\n <option value=\"2004\">\n 2004\n </option>\n <option value=\"2003\">\n 2003\n </option>\n <option value=\"2002\">\n 2002\n </option>\n <option value=\"2001\">\n 2001\n </option>\n <option value=\"2000\">\n 2000\n </option>\n </select>\n <select class=\"filter\" id=\"categories\" name=\"categories\">\n <option value=\"\">\n All Categories\n </option>\n <option value=\"165\">\n Feature Stories\n </option>\n <option value=\"19\">\n Press Releases\n </option>\n <option value=\"184\">\n Spotlights\n </option>\n <option value=\"204\">\n Status Reports\n </option>\n </select>\n </div>\n </section>\n <!-- react-text: 37 -->\n <!-- /react-text -->\n </header>\n <ul class=\"item_list \">\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8354/storm-chasers-on-mars-searching-for-dusty-secrets/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n Scientists with NASA's Mars orbiters have been waiting years for an event like the current Mars global dust storm.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"\" src=\"/system/news_items/list_view_images/8354_Mars_Dust_Storm_PIA22487_thumb.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n 'Storm Chasers' on Mars Searching for Dusty Secrets\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n July 19, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8354/storm-chasers-on-mars-searching-for-dusty-secrets/\" target=\"_self\">\n 'Storm Chasers' on Mars Searching for Dusty Secrets\n </a>\n </div>\n <div class=\"article_teaser_body\">\n Scientists with NASA's Mars orbiters have been waiting years for an event like the current Mars global dust storm.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8353/nasa-mars-mission-adds-southern-california-dates/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n Looking for summer fun? Southern California families have their choice of the beach, movies, museums -- and even NASA's next mission to Mars.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"The Mars InSight Roadshow van at San Francisco's Exploratorium in April 2018. The Roadshow van will stop at different California venues to share public exhibits and lectures about NASA's InSight mission. \" src=\"/system/news_items/list_view_images/8353_list_image.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA Mars Mission Adds Southern California Dates\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n June 26, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8353/nasa-mars-mission-adds-southern-california-dates/\" target=\"_self\">\n NASA Mars Mission Adds Southern California Dates\n </a>\n </div>\n <div class=\"article_teaser_body\">\n Looking for summer fun? Southern California families have their choice of the beach, movies, museums -- and even NASA's next mission to Mars.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8351/curiosity-captures-photos-of-thickening-dust/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n A storm of tiny dust particles has engulfed much of Mars over the last two weeks.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"A self-portrait by NASA's Curiosity rover taken on Sol 2082 (June 15, 2018). A Martian dust storm has reduced sunlight and visibility at the rover's location in Gale Crater. \" src=\"/system/news_items/list_view_images/8351_PIA22486-320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Curiosity Captures Photos of Thickening Dust\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n June 20, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8351/curiosity-captures-photos-of-thickening-dust/\" target=\"_self\">\n Curiosity Captures Photos of Thickening Dust\n </a>\n </div>\n <div class=\"article_teaser_body\">\n A storm of tiny dust particles has engulfed much of Mars over the last two weeks.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8348/opportunity-hunkers-down-during-dust-storm/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n As of Tuesday morning, June 19, the Martian dust storm had grown in size and was officially a \"planet-encircling\" (or \"global\") dust event.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"This series of images shows simulated views of a darkening Martian sky blotting out the Sun from NASA’s Opportunity rover’s point of view, with the right side simulating Opportunity’s current view in the global dust storm (June 2018). \" src=\"/system/news_items/list_view_images/8348_PIA22521-320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Opportunity Hunkers Down During Dust Storm\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n June 20, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8348/opportunity-hunkers-down-during-dust-storm/\" target=\"_self\">\n Opportunity Hunkers Down During Dust Storm\n </a>\n </div>\n <div class=\"article_teaser_body\">\n As of Tuesday morning, June 19, the Martian dust storm had grown in size and was officially a \"planet-encircling\" (or \"global\") dust event.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8350/nasa-encounters-the-perfect-storm-for-science/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n One of the most intense Martian dust storms ever observed is being studied by a record number of NASA spacecraft.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"This set of images from NASA’s Mars Reconnaissance Orbiter shows a fierce dust storm is kicking up on Mars, with rovers on the surface indicated as icons.\" src=\"/system/news_items/list_view_images/8350_marci-dgm-v04-for-home-page-5-br.gif\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA Encounters the Perfect Storm for Science\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n June 13, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8350/nasa-encounters-the-perfect-storm-for-science/\" target=\"_self\">\n NASA Encounters the Perfect Storm for Science\n </a>\n </div>\n <div class=\"article_teaser_body\">\n One of the most intense Martian dust storms ever observed is being studied by a record number of NASA spacecraft.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8349/media-telecon-about-mars-dust-storm-opportunity/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA will host a media telecon on Wednesday, June 13, about a massive Martian dust storm affecting the Opportunity rover, and how various missions can obtain unique science.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"Mars, as seen by Mars Global Surveyor in 2003.\" src=\"/system/news_items/list_view_images/8349_PIA04591-br.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Media Telecon About Mars Dust Storm, Opportunity\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n June 12, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8349/media-telecon-about-mars-dust-storm-opportunity/\" target=\"_self\">\n Media Telecon About Mars Dust Storm, Opportunity\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA will host a media telecon on Wednesday, June 13, about a massive Martian dust storm affecting the Opportunity rover, and how various missions can obtain unique science.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8347/nasa-finds-ancient-organic-material-mysterious-methane-on-mars/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA’s Curiosity rover has found evidence on Mars with implications for NASA’s search for life.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"NASA's Curiosity rover has discovered ancient organic molecules on Mars, embedded within sedimentary rocks that are billions of years old.\" src=\"/system/news_items/list_view_images/8347_curiosity_methane-320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA Finds Ancient Organic Material, Mysterious Methane on Mars\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n June 7, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8347/nasa-finds-ancient-organic-material-mysterious-methane-on-mars/\" target=\"_self\">\n NASA Finds Ancient Organic Material, Mysterious Methane on Mars\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA’s Curiosity rover has found evidence on Mars with implications for NASA’s search for life.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8346/nasa-to-host-live-discussion-on-new-mars-science-results/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n Questions are welcome during a live discussion at 11 a.m. PDT (2 p.m. EDT) Thursday, June 7, on new science results from NASA's Mars Curiosity rover.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"Selfie of the Curiosity rover\" src=\"/system/news_items/list_view_images/8346_PIA22207-br.png\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA to Host Live Discussion on New Mars Science Results\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n June 6, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8346/nasa-to-host-live-discussion-on-new-mars-science-results/\" target=\"_self\">\n NASA to Host Live Discussion on New Mars Science Results\n </a>\n </div>\n <div class=\"article_teaser_body\">\n Questions are welcome during a live discussion at 11 a.m. PDT (2 p.m. EDT) Thursday, June 7, on new science results from NASA's Mars Curiosity rover.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8345/mars-curiositys-labs-are-back-in-action/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA's Curiosity rover is analyzing drilled samples on Mars in one of its onboard labs for the first time in more than a year.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"The drill bit of NASA's Curiosity Mars rover over one of the sample inlets on the rover's deck. The inlets lead to Curiosity's onboard laboratories. This image was taken on Sol 2068 by the rover's Mast Camera (Mastcam). It has been white balanced and contrast-enhanced. \" src=\"/system/news_items/list_view_images/8345_PIA22327-br.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Mars Curiosity's Labs Are Back in Action\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n June 4, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8345/mars-curiositys-labs-are-back-in-action/\" target=\"_self\">\n Mars Curiosity's Labs Are Back in Action\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA's Curiosity rover is analyzing drilled samples on Mars in one of its onboard labs for the first time in more than a year.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8344/nasa-cubesats-steer-toward-mars/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA has achieved a first for the class of tiny spacecraft known as CubeSats, which are opening new access to space.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"An artist's concept of one of NASA's MarCO CubeSats. \" src=\"/system/news_items/list_view_images/8344_marco_tcm_20180601-320x240.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA CubeSats Steer Toward Mars\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n June 1, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8344/nasa-cubesats-steer-toward-mars/\" target=\"_self\">\n NASA CubeSats Steer Toward Mars\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA has achieved a first for the class of tiny spacecraft known as CubeSats, which are opening new access to space.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8343/scientists-shrink-chemistry-lab-to-seek-evidence-of-life-on-mars/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n An international team of scientists has created a tiny chemistry lab for a rover that will drill beneath the Martian surface looking for signs of past or present life.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"\" src=\"/system/news_items/list_view_images/8343_wilkinson-moma_320x240.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Scientists Shrink Chemistry Lab to Seek Evidence of Life on Mars\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n June 1, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8343/scientists-shrink-chemistry-lab-to-seek-evidence-of-life-on-mars/\" target=\"_self\">\n Scientists Shrink Chemistry Lab to Seek Evidence of Life on Mars\n </a>\n </div>\n <div class=\"article_teaser_body\">\n An international team of scientists has created a tiny chemistry lab for a rover that will drill beneath the Martian surface looking for signs of past or present life.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8342/insight-steers-toward-mars/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n The spacecraft has completed its first trajectory correction maneuver.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"NASA's InSight spacecraft is currently cruising to Mars. Yesterday, it performed its first course correction guiding it to the Red Planet.\" src=\"/system/news_items/list_view_images/8342_insight20180523-320x240o.gif\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n InSight Steers Toward Mars\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n May 23, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8342/insight-steers-toward-mars/\" target=\"_self\">\n InSight Steers Toward Mars\n </a>\n </div>\n <div class=\"article_teaser_body\">\n The spacecraft has completed its first trajectory correction maneuver.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8341/drilling-success-curiosity-is-collecting-mars-rocks/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n Engineers will now test delivering samples to instruments inside NASA's Curiosity Mars rover.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"NASA's Curiosity rover successfully drilled a 2-inch-deep hole in a target called "Duluth" on May 20. It was the first rock sample captured by the drill since October 2016. This image was taken by Curiosity's Mast Camera (Mastcam) on Sol 2057. \" src=\"/system/news_items/list_view_images/8341_PIA22325-320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Drilling Success: Curiosity is Collecting Mars Rocks\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n May 23, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8341/drilling-success-curiosity-is-collecting-mars-rocks/\" target=\"_self\">\n Drilling Success: Curiosity is Collecting Mars Rocks\n </a>\n </div>\n <div class=\"article_teaser_body\">\n Engineers will now test delivering samples to instruments inside NASA's Curiosity Mars rover.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8339/nasas-curiosity-rover-aims-to-get-its-rhythm-back/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n Rover engineers at JPL will try to restore percussive drilling on Mars this week, part of a larger series of tests that will last through summer.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"A test of a new percussive drilling technique at NASA's JPL. Later this week, NASA's Curiosity rover will test percussive drilling on Mars for the first time since December 2016.\" src=\"/system/news_items/list_view_images/8339_Curiosity_drill_PIA22324-th.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA's Curiosity Rover Aims to Get Its Rhythm Back\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n May 17, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8339/nasas-curiosity-rover-aims-to-get-its-rhythm-back/\" target=\"_self\">\n NASA's Curiosity Rover Aims to Get Its Rhythm Back\n </a>\n </div>\n <div class=\"article_teaser_body\">\n Rover engineers at JPL will try to restore percussive drilling on Mars this week, part of a larger series of tests that will last through summer.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8338/a-pale-blue-dot-as-seen-by-a-cubesat/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n One of NASA's MarCO CubeSats has taken its first image.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"This image taken by NASA's Mars Cube One (MarCO) CubeSats contains a photograph of Earth and Mars at a distance. \" src=\"/system/news_items/list_view_images/8338_PIA22323_main-320x240.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n A Pale Blue Dot, As Seen by a CubeSat\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n May 15, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8338/a-pale-blue-dot-as-seen-by-a-cubesat/\" target=\"_self\">\n A Pale Blue Dot, As Seen by a CubeSat\n </a>\n </div>\n <div class=\"article_teaser_body\">\n One of NASA's MarCO CubeSats has taken its first image.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8335/mars-helicopter-to-fly-on-nasas-next-red-planet-rover-mission/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA is adding a Mars helicopter to the agency’s next mission to the Red Planet, Mars 2020.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"\" src=\"/system/news_items/list_view_images/8335_helicopter20180511-16-th.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Mars Helicopter to Fly on NASA’s Next Red Planet Rover Mission\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n May 11, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8335/mars-helicopter-to-fly-on-nasas-next-red-planet-rover-mission/\" target=\"_self\">\n Mars Helicopter to Fly on NASA’s Next Red Planet Rover Mission\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA is adding a Mars helicopter to the agency’s next mission to the Red Planet, Mars 2020.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8334/nasas-first-deep-space-cubesats-say-polo/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n MarCO is a pair of tiny spacecraft that launched with NASA's InSight lander today.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"An artist's rendering of the twin Mars Cube One (MarCO) spacecraft on their cruise to Mars. \" src=\"/system/news_items/list_view_images/8334_PIA22314-th.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA's First Deep-Space CubeSats Say: 'Polo!'\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n May 5, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8334/nasas-first-deep-space-cubesats-say-polo/\" target=\"_self\">\n NASA's First Deep-Space CubeSats Say: 'Polo!'\n </a>\n </div>\n <div class=\"article_teaser_body\">\n MarCO is a pair of tiny spacecraft that launched with NASA's InSight lander today.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8333/nasa-ula-launch-mission-to-study-how-mars-was-made/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA’s Mars InSight mission launched this morning on a 300-million-mile trip to Mars to study for the first time what lies deep beneath the surface of the Red Planet.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt='\tThe NASA InSight spacecraft launches onboard a United Launch Alliance Atlas-V rocket, Saturday, May 5, 2018, from Vandenberg Air Force Base in California. InSight, short for Interior Exploration using Seismic Investigations, Geodesy and Heat Transport, is a Mars lander designed to study the \"inner space\" of Mars: its crust, mantle, and core.' src=\"/system/news_items/list_view_images/8333_41864015862_4eb1b8de31_o-th.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA, ULA Launch Mission to Study How Mars Was Made\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n May 5, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8333/nasa-ula-launch-mission-to-study-how-mars-was-made/\" target=\"_self\">\n NASA, ULA Launch Mission to Study How Mars Was Made\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA’s Mars InSight mission launched this morning on a 300-million-mile trip to Mars to study for the first time what lies deep beneath the surface of the Red Planet.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8332/nasas-first-mission-to-study-the-interior-of-mars-awaits-may-5-launch/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n All systems are go for NASA’s next launch to the Red Planet.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"An artist's impression of the InSight lander on Mars. Credit: NASA/JPL-Caltech\" src=\"/system/news_items/list_view_images/8332_21438_PIA22226_320x240.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA’s First Mission to Study the Interior of Mars Awaits May 5 Launch\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n May 3, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8332/nasas-first-mission-to-study-the-interior-of-mars-awaits-may-5-launch/\" target=\"_self\">\n NASA’s First Mission to Study the Interior of Mars Awaits May 5 Launch\n </a>\n </div>\n <div class=\"article_teaser_body\">\n All systems are go for NASA’s next launch to the Red Planet.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8331/vice-president-pence-visits-jpl-previews-nasas-next-mars-mission-launch/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n A week before NASA's next Mars launch, Vice President Mike Pence toured the birthplace of the InSight Mars Lander and numerous other past, present and future space missions.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"U.S. Vice President Mike Pence, right, is presented a plaque by JPL Director Michael Watkins during a tour of NASA's Jet Propulsion Laboratory, Saturday, April 28, 2018 in Pasadena, California. The plaque presents a view of the Mars Science Laboratory rover Curiosity on the surface of Mars. Photo Credit: (NASA/Bill Ingalls)\" src=\"/system/news_items/list_view_images/8331_vicepresidentpencejplvisit-th.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Vice President Pence Visits JPL, Previews NASA’s Next Mars Mission Launch\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n April 30, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8331/vice-president-pence-visits-jpl-previews-nasas-next-mars-mission-launch/\" target=\"_self\">\n Vice President Pence Visits JPL, Previews NASA’s Next Mars Mission Launch\n </a>\n </div>\n <div class=\"article_teaser_body\">\n A week before NASA's next Mars launch, Vice President Mike Pence toured the birthplace of the InSight Mars Lander and numerous other past, present and future space missions.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8330/nasa-sets-sights-on-may-5-launch-of-insight-to-mars/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA’s next mission to Mars, InSight, is scheduled to launch Saturday, May 5, on a first-ever mission to study the heart of the Red Planet.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"An artist's rendering of a rocket launching with the InSight spacecraft in May.\" src=\"/system/news_items/list_view_images/8330_insight20180329-th.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA Sets Sights on May 5 Launch of InSight to Mars\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n April 27, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8330/nasa-sets-sights-on-may-5-launch-of-insight-to-mars/\" target=\"_self\">\n NASA Sets Sights on May 5 Launch of InSight to Mars\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA’s next mission to Mars, InSight, is scheduled to launch Saturday, May 5, on a first-ever mission to study the heart of the Red Planet.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8329/results-of-heat-shield-testing/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n A post-test inspection of the composite structure for a heat shield to be used on the Mars 2020 mission revealed that a fracture occurred during structural testing.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"Artist's concept of the Mars Science Laboratory entry into the Martian atmosphere.\" src=\"/system/news_items/list_view_images/8329_PIA14835_320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Results of Heat Shield Testing\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n April 26, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8329/results-of-heat-shield-testing/\" target=\"_self\">\n Results of Heat Shield Testing\n </a>\n </div>\n <div class=\"article_teaser_body\">\n A post-test inspection of the composite structure for a heat shield to be used on the Mars 2020 mission revealed that a fracture occurred during structural testing.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8328/nasa-engineers-dream-big-with-small-spacecraft/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n The first CubeSat mission to deep space will launch in May.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"\" src=\"/system/news_items/list_view_images/8328_PIA22314_320.png\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA Engineers Dream Big with Small Spacecraft\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n April 19, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8328/nasa-engineers-dream-big-with-small-spacecraft/\" target=\"_self\">\n NASA Engineers Dream Big with Small Spacecraft\n </a>\n </div>\n <div class=\"article_teaser_body\">\n The first CubeSat mission to deep space will launch in May.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8327/bound-for-mars-countdown-to-first-interplanetary-launch-from-california/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n On May 5, millions of Californians may witness the historic first interplanetary launch from America’s West Coast.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"\" src=\"/system/news_items/list_view_images/8327_InterplanetaryLaunch-320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Bound for Mars: Countdown to First Interplanetary Launch from California\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n April 6, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8327/bound-for-mars-countdown-to-first-interplanetary-launch-from-california/\" target=\"_self\">\n Bound for Mars: Countdown to First Interplanetary Launch from California\n </a>\n </div>\n <div class=\"article_teaser_body\">\n On May 5, millions of Californians may witness the historic first interplanetary launch from America’s West Coast.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8326/nasa-invests-in-visionary-technology/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA is investing in technology concepts, including several from JPL, that may one day be used for future space exploration missions.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"NASA is investing in technology concepts, including several from JPL, that may one day be used for future space exploration missions.\" src=\"/system/news_items/list_view_images/8326_niac320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA Invests in Visionary Technology\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n March 30, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8326/nasa-invests-in-visionary-technology/\" target=\"_self\">\n NASA Invests in Visionary Technology\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA is investing in technology concepts, including several from JPL, that may one day be used for future space exploration missions.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8325/nasa-is-ready-to-study-the-heart-of-mars/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA is about to go on a journey to study the center of Mars.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"An artist's rendering of the InSight spacecraft's cruise stage entering the Martian atmosphere.\" src=\"/system/news_items/list_view_images/8325_insight20180329b_320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA is Ready to Study the Heart of Mars\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n March 29, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8325/nasa-is-ready-to-study-the-heart-of-mars/\" target=\"_self\">\n NASA is Ready to Study the Heart of Mars\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA is about to go on a journey to study the center of Mars.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8324/marsquakes-could-shake-up-planetary-science/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n InSight, the next mission to the Red Planet, will use seismology to see into the depths of Mars.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"Inner Structure of Mars\" src=\"/system/news_items/list_view_images/8324_PIA16078-320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n ‘Marsquakes’ Could Shake Up Planetary Science\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n March 28, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8324/marsquakes-could-shake-up-planetary-science/\" target=\"_self\">\n ‘Marsquakes’ Could Shake Up Planetary Science\n </a>\n </div>\n <div class=\"article_teaser_body\">\n InSight, the next mission to the Red Planet, will use seismology to see into the depths of Mars.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8323/mars-curiosity-celebrates-sol-2000/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA's Mars Curiosity rover just hit a new milestone: its two-thousandth Martian day on the Red Planet. An image mosaic taken recently offers a preview of what comes next.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"rugged landscape of hills \" src=\"/system/news_items/list_view_images/8323_PIA22313_320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Mars Curiosity Celebrates Sol 2,000\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n March 22, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8323/mars-curiosity-celebrates-sol-2000/\" target=\"_self\">\n Mars Curiosity Celebrates Sol 2,000\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA's Mars Curiosity rover just hit a new milestone: its two-thousandth Martian day on the Red Planet. An image mosaic taken recently offers a preview of what comes next.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8322/nasa-briefing-on-first-mission-to-study-mars-interior/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA’s next mission to Mars will be the topic of a media briefing Thursday, March 29, at JPL. The briefing will air live on NASA Television and the agency’s website.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"An artist's rendition of the InSight lander operating on the surface of Mars. \" src=\"/system/news_items/list_view_images/8322_PIA22228_320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA Briefing on First Mission to Study Mars Interior\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n March 22, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8322/nasa-briefing-on-first-mission-to-study-mars-interior/\" target=\"_self\">\n NASA Briefing on First Mission to Study Mars Interior\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA’s next mission to Mars will be the topic of a media briefing Thursday, March 29, at JPL. The briefing will air live on NASA Television and the agency’s website.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8321/new-ar-mobile-app-features-3-d-nasa-spacecraft/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA spacecraft travel to far-off destinations in space, but a new mobile app produced by NASA's Jet Propulsion Laboratory, Pasadena, California, brings spacecraft to users.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"Free Spacecraft AR app uses Google ARCore technology\" src=\"/system/news_items/list_view_images/8321_list_image.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n New 'AR' Mobile App Features 3-D NASA Spacecraft\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n March 20, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8321/new-ar-mobile-app-features-3-d-nasa-spacecraft/\" target=\"_self\">\n New 'AR' Mobile App Features 3-D NASA Spacecraft\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA spacecraft travel to far-off destinations in space, but a new mobile app produced by NASA's Jet Propulsion Laboratory, Pasadena, California, brings spacecraft to users.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8319/nasa-mars-mission-tours-california/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n Scientists and engineers with NASA's next mission to Mars will be touring California cities starting this month.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"This artist's concept shows the InSight lander, its sensors, cameras and instruments\" src=\"/system/news_items/list_view_images/8319_PIA22227_320x240.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA Mars Mission Tours California\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n March 14, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8319/nasa-mars-mission-tours-california/\" target=\"_self\">\n NASA Mars Mission Tours California\n </a>\n </div>\n <div class=\"article_teaser_body\">\n Scientists and engineers with NASA's next mission to Mars will be touring California cities starting this month.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8318/next-nasa-mars-rover-reaches-key-manufacturing-milestone/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA's Mars 2020 mission has begun the assembly, test and launch operations (ATLO) phase of its development, on track for a July 2020 launch to Mars.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"A technician works on the descent stage for NASA’s Mars 2020 mission inside JPL’s Spacecraft Assembly Facility. \" src=\"/system/news_items/list_view_images/8318_PIA22342_320x240.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Next NASA Mars Rover Reaches Key Manufacturing Milestone\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n March 13, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8318/next-nasa-mars-rover-reaches-key-manufacturing-milestone/\" target=\"_self\">\n Next NASA Mars Rover Reaches Key Manufacturing Milestone\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA's Mars 2020 mission has begun the assembly, test and launch operations (ATLO) phase of its development, on track for a July 2020 launch to Mars.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8317/witness-first-mars-launch-from-west-coast/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA invites digital creators to apply for social media credentials to cover the launch of the InSight mission to Mars, May 3-5, at California's Vandenberg Air Force Base.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"Launch of LDCM from Vandenberg AFB, California\" src=\"/system/news_items/list_view_images/8317_list_image.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Witness First Mars Launch from West Coast\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n March 12, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8317/witness-first-mars-launch-from-west-coast/\" target=\"_self\">\n Witness First Mars Launch from West Coast\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA invites digital creators to apply for social media credentials to cover the launch of the InSight mission to Mars, May 3-5, at California's Vandenberg Air Force Base.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8316/360-video-tour-a-mars-robot-test-lab/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n Engineers are practicing operations for NASA's Mars InSight lander, which is launching this spring.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"\" src=\"/system/news_items/list_view_images/8316_InSight_320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n 360 Video: Tour a Mars Robot Test Lab\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n March 8, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8316/360-video-tour-a-mars-robot-test-lab/\" target=\"_self\">\n 360 Video: Tour a Mars Robot Test Lab\n </a>\n </div>\n <div class=\"article_teaser_body\">\n Engineers are practicing operations for NASA's Mars InSight lander, which is launching this spring.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8315/nasa-insight-mission-to-mars-arrives-at-launch-site/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA's InSight spacecraft has arrived at Vandenberg Air Force Base in central California to begin final preparations for a launch this May.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"crate being loaded into military transport plane\" src=\"/system/news_items/list_view_images/8315_list_image.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n NASA InSight Mission to Mars Arrives at Launch Site\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n February 28, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8315/nasa-insight-mission-to-mars-arrives-at-launch-site/\" target=\"_self\">\n NASA InSight Mission to Mars Arrives at Launch Site\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA's InSight spacecraft has arrived at Vandenberg Air Force Base in central California to begin final preparations for a launch this May.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8314/curiosity-tests-a-new-way-to-drill-on-mars/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA's Mars Curiosity rover has conducted the first test of a new drilling technique on the Red Planet since its drill stopped working reliably.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"\" src=\"/system/news_items/list_view_images/8314_PIA22224_320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Curiosity Tests a New Way to Drill on Mars\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n February 28, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8314/curiosity-tests-a-new-way-to-drill-on-mars/\" target=\"_self\">\n Curiosity Tests a New Way to Drill on Mars\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA's Mars Curiosity rover has conducted the first test of a new drilling technique on the Red Planet since its drill stopped working reliably.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8313/seven-ways-mars-insight-is-different/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA has a long and successful track record at Mars. Since 1965, it has flown by, orbited, landed and roved across the surface of the Red Planet. What can InSight -- planned for launch in May -- do that hasn’t been done before?\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"\" src=\"/system/news_items/list_view_images/8313_PIA22228_320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Seven Ways Mars InSight is Different\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n February 22, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8313/seven-ways-mars-insight-is-different/\" target=\"_self\">\n Seven Ways Mars InSight is Different\n </a>\n </div>\n <div class=\"article_teaser_body\">\n NASA has a long and successful track record at Mars. Since 1965, it has flown by, orbited, landed and roved across the surface of the Red Planet. What can InSight -- planned for launch in May -- do that hasn’t been done before?\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8312/nearly-a-decade-after-mars-phoenix-landed-another-look/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n A recent view from Mars orbit of the site where NASA's Phoenix Mars mission landed on far-northern Mars nearly a decade ago captures changes.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"animation of two alternating views of barren martian landscape\" src=\"/system/news_items/list_view_images/8312_PIA22223_320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Nearly a Decade After Mars Phoenix Landed, Another Look\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n February 20, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8312/nearly-a-decade-after-mars-phoenix-landed-another-look/\" target=\"_self\">\n Nearly a Decade After Mars Phoenix Landed, Another Look\n </a>\n </div>\n <div class=\"article_teaser_body\">\n A recent view from Mars orbit of the site where NASA's Phoenix Mars mission landed on far-northern Mars nearly a decade ago captures changes.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8311/spacecraft-exits-safe-mode/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n Diagnostic work is the focus for resuming service and exiting safe standby status.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"spacecraft high over martian surface\" src=\"/system/news_items/list_view_images/8311_PIA05490_320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n Spacecraft Exits Safe Mode\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n February 16, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8311/spacecraft-exits-safe-mode/\" target=\"_self\">\n Spacecraft Exits Safe Mode\n </a>\n </div>\n <div class=\"article_teaser_body\">\n Diagnostic work is the focus for resuming service and exiting safe standby status.\n </div>\n </div>\n </div>\n </li>\n <li class=\"slide\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8310/5000-days-on-mars-solar-powered-rover-approaching-5000th-martian-dawn/\" target=\"_self\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n The Sun will rise on NASA's solar-powered Mars rover Opportunity for the 5,000th time on Saturday, sending rays of energy to a robot that continues to provide revelations.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <div class=\"list_image\">\n <img alt=\"map of rugged slope\" src=\"/system/news_items/list_view_images/8310_pia22221_320.jpg\"/>\n </div>\n <div class=\"bottom_gradient\">\n <div>\n <h3>\n 5,000 Days on Mars; Solar-Powered Rover Approaching 5,000th Martian Dawn\n </h3>\n </div>\n </div>\n </a>\n <div class=\"list_text\">\n <div class=\"list_date\">\n February 15, 2018\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8310/5000-days-on-mars-solar-powered-rover-approaching-5000th-martian-dawn/\" target=\"_self\">\n 5,000 Days on Mars; Solar-Powered Rover Approaching 5,000th Martian Dawn\n </a>\n </div>\n <div class=\"article_teaser_body\">\n The Sun will rise on NASA's solar-powered Mars rover Opportunity for the 5,000th time on Saturday, sending rays of energy to a robot that continues to provide revelations.\n </div>\n </div>\n </div>\n </li>\n </ul>\n <footer class=\"list_footer more_button\">\n <div class=\"loading\">\n </div>\n <a class=\"button\" href=\"#\" type=\"button\">\n More\n </a>\n </footer>\n </div>\n </section>\n </div>\n <section class=\"module suggested_features\">\n <div class=\"grid_layout\">\n <header>\n <h2 class=\"module_title\">\n You Might Also Like\n </h2>\n </header>\n <section>\n <script>\n $(document).ready(function(){\n $(\".features\").slick({\n dots: false,\n infinite: true,\n speed: 300,\n slide: '.features .slide',\n slidesToShow: 3,\n slidesToScroll: 3,\n lazyLoad: 'ondemand',\n centerMode: false,\n arrows: true,\n appendArrows: '.features .slick-nav',\n appendDots: \".features .slick-nav\",\n responsive: [{\"breakpoint\":953,\"settings\":{\"slidesToShow\":2,\"slidesToScroll\":2,\"centerMode\":false}},{\"breakpoint\":480,\"settings\":{\"slidesToShow\":1,\"slidesToScroll\":1,\"centerMode\":true,\"arrows\":false,\"centerPadding\":\"25px\"}}]\n });\n });\n </script>\n <div class=\"features slick-initialized slick-slider\">\n <div class=\"slick-list draggable\" tabindex=\"0\">\n <div class=\"slick-track\" style=\"opacity: 1; width: 3552px; transform: translate3d(-888px, 0px, 0px);\">\n <div class=\"slide slick-slide slick-cloned\" index=\"-3\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8325/nasa-is-ready-to-study-the-heart-of-mars/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA is about to go on a journey to study the center of Mars.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"NASA is Ready to Study the Heart of Mars\" class=\"img-lazy\" data-lazy=\"/system/news_items/list_view_images/8325_insight20180329b_320.jpg\" src=\"/assets/loading_320x240.png\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8325/nasa-is-ready-to-study-the-heart-of-mars/\">\n NASA is Ready to Study the Heart of Mars\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide slick-cloned\" index=\"-2\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8322/nasa-briefing-on-first-mission-to-study-mars-interior/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA’s next mission to Mars will be the topic of a media briefing Thursday, March 29, at JPL. The briefing will air live on NASA Television and the agency’s website.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"NASA Briefing on First Mission to Study Mars Interior\" class=\"img-lazy\" data-lazy=\"/system/news_items/list_view_images/8322_PIA22228_320.jpg\" src=\"/assets/loading_320x240.png\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8322/nasa-briefing-on-first-mission-to-study-mars-interior/\">\n NASA Briefing on First Mission to Study Mars Interior\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide slick-cloned\" index=\"-1\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8321/new-ar-mobile-app-features-3-d-nasa-spacecraft/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA spacecraft travel to far-off destinations in space, but a new mobile app produced by NASA's Jet Propulsion Laboratory, Pasadena, California, brings spacecraft to users.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"New 'AR' Mobile App Features 3-D NASA Spacecraft\" class=\"img-lazy\" data-lazy=\"/system/news_items/list_view_images/8321_list_image.jpg\" src=\"/assets/loading_320x240.png\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8321/new-ar-mobile-app-features-3-d-nasa-spacecraft/\">\n New 'AR' Mobile App Features 3-D NASA Spacecraft\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide slick-active\" index=\"0\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8348/opportunity-hunkers-down-during-dust-storm/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n As of Tuesday morning, June 19, the Martian dust storm had grown in size and was officially a \"planet-encircling\" (or \"global\") dust event.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"Opportunity Hunkers Down During Dust Storm\" class=\"img-lazy\" src=\"/system/news_items/list_view_images/8348_PIA22521-320.jpg?1532218141872\" style=\"opacity: 1;\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8348/opportunity-hunkers-down-during-dust-storm/\">\n Opportunity Hunkers Down During Dust Storm\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide slick-active\" index=\"1\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8347/nasa-finds-ancient-organic-material-mysterious-methane-on-mars/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA’s Curiosity rover has found evidence on Mars with implications for NASA’s search for life.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"NASA Finds Ancient Organic Material, Mysterious Methane on Mars\" class=\"img-lazy\" src=\"/system/news_items/list_view_images/8347_curiosity_methane-320.jpg?1532218141873\" style=\"opacity: 1;\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8347/nasa-finds-ancient-organic-material-mysterious-methane-on-mars/\">\n NASA Finds Ancient Organic Material, Mysterious Methane on Mars\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide slick-active\" index=\"2\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8326/nasa-invests-in-visionary-technology/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA is investing in technology concepts, including several from JPL, that may one day be used for future space exploration missions.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"NASA Invests in Visionary Technology \" class=\"img-lazy\" src=\"/system/news_items/list_view_images/8326_niac320.jpg?1532218141873\" style=\"opacity: 1;\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8326/nasa-invests-in-visionary-technology/\">\n NASA Invests in Visionary Technology\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide\" index=\"3\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8325/nasa-is-ready-to-study-the-heart-of-mars/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA is about to go on a journey to study the center of Mars.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"NASA is Ready to Study the Heart of Mars\" class=\"img-lazy\" data-lazy=\"/system/news_items/list_view_images/8325_insight20180329b_320.jpg\" src=\"/assets/loading_320x240.png\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8325/nasa-is-ready-to-study-the-heart-of-mars/\">\n NASA is Ready to Study the Heart of Mars\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide\" index=\"4\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8322/nasa-briefing-on-first-mission-to-study-mars-interior/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA’s next mission to Mars will be the topic of a media briefing Thursday, March 29, at JPL. The briefing will air live on NASA Television and the agency’s website.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"NASA Briefing on First Mission to Study Mars Interior\" class=\"img-lazy\" data-lazy=\"/system/news_items/list_view_images/8322_PIA22228_320.jpg\" src=\"/assets/loading_320x240.png\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8322/nasa-briefing-on-first-mission-to-study-mars-interior/\">\n NASA Briefing on First Mission to Study Mars Interior\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide\" index=\"5\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8321/new-ar-mobile-app-features-3-d-nasa-spacecraft/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA spacecraft travel to far-off destinations in space, but a new mobile app produced by NASA's Jet Propulsion Laboratory, Pasadena, California, brings spacecraft to users.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"New 'AR' Mobile App Features 3-D NASA Spacecraft\" class=\"img-lazy\" data-lazy=\"/system/news_items/list_view_images/8321_list_image.jpg\" src=\"/assets/loading_320x240.png\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8321/new-ar-mobile-app-features-3-d-nasa-spacecraft/\">\n New 'AR' Mobile App Features 3-D NASA Spacecraft\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide slick-cloned\" index=\"6\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8348/opportunity-hunkers-down-during-dust-storm/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n As of Tuesday morning, June 19, the Martian dust storm had grown in size and was officially a \"planet-encircling\" (or \"global\") dust event.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"Opportunity Hunkers Down During Dust Storm\" class=\"img-lazy\" src=\"/system/news_items/list_view_images/8348_PIA22521-320.jpg?1532218141874\" style=\"opacity: 1;\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8348/opportunity-hunkers-down-during-dust-storm/\">\n Opportunity Hunkers Down During Dust Storm\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide slick-cloned\" index=\"7\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8347/nasa-finds-ancient-organic-material-mysterious-methane-on-mars/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA’s Curiosity rover has found evidence on Mars with implications for NASA’s search for life.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"NASA Finds Ancient Organic Material, Mysterious Methane on Mars\" class=\"img-lazy\" src=\"/system/news_items/list_view_images/8347_curiosity_methane-320.jpg?1532218141874\" style=\"opacity: 1;\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8347/nasa-finds-ancient-organic-material-mysterious-methane-on-mars/\">\n NASA Finds Ancient Organic Material, Mysterious Methane on Mars\n </a>\n </div>\n </div>\n <div class=\"slide slick-slide slick-cloned\" index=\"8\" style=\"width: 278px;\">\n <div class=\"image_and_description_container\">\n <a href=\"/news/8326/nasa-invests-in-visionary-technology/\">\n <div class=\"rollover_description\">\n <div class=\"rollover_description_inner\">\n NASA is investing in technology concepts, including several from JPL, that may one day be used for future space exploration missions.\n </div>\n <div class=\"overlay_arrow\">\n <img alt=\"More\" src=\"/assets/overlay-arrow.png\"/>\n </div>\n </div>\n <img alt=\"NASA Invests in Visionary Technology \" class=\"img-lazy\" src=\"/system/news_items/list_view_images/8326_niac320.jpg?1532218141874\" style=\"opacity: 1;\"/>\n </a>\n </div>\n <div class=\"content_title\">\n <a href=\"/news/8326/nasa-invests-in-visionary-technology/\">\n NASA Invests in Visionary Technology\n </a>\n </div>\n </div>\n </div>\n </div>\n <div class=\"grid_layout\">\n <div class=\"slick-nav_container\">\n <div class=\"slick-nav\">\n <button class=\"slick-prev\" data-role=\"none\" style=\"display: block;\" type=\"button\">\n Previous\n </button>\n <button class=\"slick-next\" data-role=\"none\" style=\"display: block;\" type=\"button\">\n Next\n </button>\n </div>\n </div>\n </div>\n </div>\n </section>\n </div>\n </section>\n </div>\n <footer id=\"site_footer\">\n <div class=\"grid_layout\">\n <section class=\"upper_footer\">\n <div class=\"share\">\n <h2>\n Follow the Journey\n </h2>\n <div class=\"social_icons\">\n <!-- AddThis Button BEGIN -->\n <div class=\"addthis_toolbox addthis_default_style addthis_32x32_style\">\n <a addthis:userid=\"NASABeAMartian\" class=\"addthis_button_twitter_follow icon at300b\" href=\"//twitter.com/NASABeAMartian\" target=\"_blank\" title=\"Follow on Twitter\">\n <img alt=\"twitter\" src=\"/assets/[email protected]\"/>\n <span class=\"addthis_follow_label\">\n Twitter\n </span>\n </a>\n <a addthis:userid=\"NASABEAM\" class=\"addthis_button_facebook_follow icon at300b\" href=\"http://www.facebook.com/NASABEAM\" target=\"_blank\" title=\"Follow on Facebook\">\n <img alt=\"facebook\" src=\"/assets/[email protected]\"/>\n <span class=\"addthis_follow_label\">\n Facebook\n </span>\n </a>\n <a addthis:userid=\"nasa\" class=\"addthis_button_instagram_follow icon at300b\" href=\"http://instagram.com/nasa\" target=\"_blank\" title=\"Follow on Instagram\">\n <img alt=\"instagram\" src=\"/assets/[email protected]\"/>\n <span class=\"addthis_follow_label\">\n Instagram\n </span>\n </a>\n <a addthis:url=\"https://mars.nasa.gov/rss/api/?feed=news&category=all&feedtype=rss\" class=\"addthis_button_rss_follow icon at300b\" href=\"https://mars.nasa.gov/rss/api/?feed=news&category=all&feedtype=rss\" target=\"_blank\" title=\"Follow on RSS\">\n <img alt=\"rss\" src=\"/assets/[email protected]\"/>\n <span class=\"addthis_follow_label\">\n RSS\n </span>\n </a>\n <div class=\"atclear\">\n </div>\n </div>\n <script>\n addthis_loader.init(\"//s7.addthis.com/js/300/addthis_widget.js#pubid=ra-5a690e4c1320e328\", {follow: true})\n </script>\n <!-- AddThis Button END -->\n </div>\n </div>\n <div class=\"gradient_line\">\n </div>\n </section>\n <section class=\"sitemap\">\n <div class=\"sitemap_directory\" id=\"sitemap_directory\" style=\"position: relative; height: 327.266px;\">\n <div class=\"sitemap_block\" style=\"position: absolute; left: 0px; top: 0px;\">\n <div class=\"footer_sitemap_item\">\n <h3 class=\"sitemap_title\">\n <a href=\"/#red_planet\">\n The Red Planet\n </a>\n </h3>\n <ul>\n <li>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a href=\"/#red_planet/0\" target=\"_self\">\n Dashboard\n </a>\n </li>\n <li>\n <a href=\"/#red_planet/1\" target=\"_self\">\n Science Goals\n </a>\n </li>\n <li>\n <a href=\"/#red_planet/2\" target=\"_self\">\n The Planet\n </a>\n </li>\n <li>\n <a href=\"/#red_planet/3\" target=\"_self\">\n Atmosphere\n </a>\n </li>\n <li>\n <a href=\"/#red_planet/4\" target=\"_self\">\n Astrobiology\n </a>\n </li>\n <li>\n <a href=\"/#red_planet/5\" target=\"_self\">\n Past, Present, Future, Timeline\n </a>\n </li>\n </ul>\n </div>\n </li>\n </ul>\n </div>\n </div>\n <div class=\"sitemap_block\" style=\"position: absolute; left: 164px; top: 0px;\">\n <div class=\"footer_sitemap_item\">\n <h3 class=\"sitemap_title\">\n <a href=\"/#mars_exploration_program\">\n The Program\n </a>\n </h3>\n <ul>\n <li>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a href=\"/#mars_exploration_program/0\" target=\"_self\">\n Mission Statement\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/1\" target=\"_self\">\n About the Program\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/2\" target=\"_self\">\n Organization\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/3\" target=\"_self\">\n Why Mars?\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/4\" target=\"_self\">\n Research Programs\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/5\" target=\"_self\">\n Planetary Resources\n </a>\n </li>\n <li>\n <a href=\"/#mars_exploration_program/6\" target=\"_self\">\n Technologies\n </a>\n </li>\n </ul>\n </div>\n </li>\n </ul>\n </div>\n </div>\n <div class=\"sitemap_block\" style=\"position: absolute; left: 328px; top: 0px;\">\n <div class=\"footer_sitemap_item\">\n <h3 class=\"sitemap_title\">\n <a href=\"/#news_and_events\">\n News & Events\n </a>\n </h3>\n <ul>\n <li>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a href=\"/news\" target=\"_self\">\n News\n </a>\n </li>\n <li>\n <a href=\"/events\" target=\"_self\">\n Events\n </a>\n </li>\n </ul>\n </div>\n </li>\n </ul>\n </div>\n </div>\n <div class=\"sitemap_block\" style=\"position: absolute; left: 493px; top: 0px;\">\n <div class=\"footer_sitemap_item\">\n <h3 class=\"sitemap_title\">\n <a href=\"/#multimedia\">\n Multimedia\n </a>\n </h3>\n <ul>\n <li>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a href=\"/multimedia/images/\" target=\"_self\">\n Images\n </a>\n </li>\n <li>\n <a href=\"/multimedia/videos/\" target=\"_self\">\n Videos\n </a>\n </li>\n </ul>\n </div>\n </li>\n </ul>\n </div>\n </div>\n <div class=\"sitemap_block\" style=\"position: absolute; left: 657px; top: 0px;\">\n <div class=\"footer_sitemap_item\">\n <h3 class=\"sitemap_title\">\n <a href=\"/#missions\">\n Missions\n </a>\n </h3>\n <ul>\n <li>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a href=\"/mars-exploration/missions/?category=167\" target=\"_self\">\n Past\n </a>\n </li>\n <li>\n <a href=\"/mars-exploration/missions/?category=170\" target=\"_self\">\n Present\n </a>\n </li>\n <li>\n <a href=\"/mars-exploration/missions/?category=171\" target=\"_self\">\n Future\n </a>\n </li>\n <li>\n <a href=\"/mars-exploration/partners\" target=\"_self\">\n International Partners\n </a>\n </li>\n </ul>\n </div>\n </li>\n </ul>\n </div>\n </div>\n <div class=\"sitemap_block\" style=\"position: absolute; left: 822px; top: 0px;\">\n <div class=\"footer_sitemap_item\">\n <h3 class=\"sitemap_title\">\n <a href=\"/#more\">\n More\n </a>\n </h3>\n <ul>\n <li>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n </ul>\n </div>\n </li>\n </ul>\n </div>\n </div>\n <div class=\"sitemap_block\" style=\"position: absolute; left: 822px; top: 53px;\">\n <div class=\"footer_sitemap_item\">\n <h3 class=\"sitemap_title\">\n <a href=\"/legacy\">\n Legacy Site\n </a>\n </h3>\n <ul>\n <li>\n <div class=\"global_subnav_container\">\n <ul class=\"subnav\">\n <li>\n <a class=\"\" href=\"/legacy\" target=\"_self\">\n Legacy Site\n </a>\n </li>\n </ul>\n </div>\n </li>\n </ul>\n </div>\n </div>\n </div>\n <div class=\"gradient_line\">\n </div>\n </section>\n <section class=\"lower_footer\">\n <div class=\"nav_container\">\n <nav>\n <ul>\n <li>\n <a href=\"http://science.nasa.gov/\" target=\"_blank\">\n NASA Science Mission Directorate\n </a>\n </li>\n <li>\n <a href=\"https://www.jpl.nasa.gov/copyrights.php\" target=\"_blank\">\n Privacy\n </a>\n </li>\n <li>\n <a href=\"http://www.jpl.nasa.gov/imagepolicy/\" target=\"_blank\">\n Image Policy\n </a>\n </li>\n <li>\n <a href=\"https://mars.nasa.gov/feedback/\" target=\"_self\">\n Feedback\n </a>\n </li>\n <li>\n <a href=\"http://mars.nasa.gov/legacy\" target=\"_blank\">\n Legacy Mars Site\n </a>\n </li>\n </ul>\n </nav>\n </div>\n <div class=\"credits\">\n <div class=\"footer_brands_top\">\n <p>\n Managed by the Mars Exploration Program and the Jet Propulsion Laboratory for NASA’s Science Mission Directorate\n </p>\n </div>\n <!-- .footer_brands -->\n <!-- %a.jpl{href: \"\", target: \"_blank\"}Institution -->\n <!-- -->\n <!-- %a.caltech{href: \"\", target: \"_blank\"}Institution -->\n <!-- .staff -->\n <!-- %p -->\n <!-- - get_staff_for_category(get_field_from_admin_config(:web_staff_category_id)) -->\n <!-- - @staff.each_with_index do |staff, idx| -->\n <!-- - unless staff.is_in_footer == 0 -->\n <!-- = staff.title + \": \" -->\n <!-- - if staff.contact_link =~ /@/ -->\n <!-- = mail_to staff.contact_link, staff.name, :subject => \"[#{@site_title}]\" -->\n <!-- - elsif staff.contact_link.present? -->\n <!-- = link_to staff.name, staff.contact_link -->\n <!-- - else -->\n <!-- = staff.name -->\n <!-- - unless (idx + 1 == @staff.size) -->\n <!-- %br -->\n </div>\n </section>\n </div>\n </footer>\n </div>\n </div>\n <script id=\"_fed_an_ua_tag\" src=\"https://dap.digitalgov.gov/Universal-Federated-Analytics-Min.js?agency=NASA&subagency=JPL-Mars-MEPJPL&pua=UA-9453474-9,UA-118212757-11&dclink=true&sp=searchbox&exts=tif,tiff,wav\" type=\"text/javascript\">\n </script>\n <div id=\"_atssh\" style=\"visibility: hidden; height: 1px; width: 1px; position: absolute; top: -9999px; z-index: 100000;\">\n <iframe id=\"_atssh106\" src=\"https://s7.addthis.com/static/sh.e4e8af4de595fdb10ec1459d.html#rand=0.2123387918842583&iit=1532218142590&tmr=load%3D1532218142440%26core%3D1532218142526%26main%3D1532218142575%26ifr%3D1532218142598&cb=0&cdn=0&md=0&kw=Mars%2Cmissions%2CNASA%2Crover%2CCuriosity%2COpportunity%2CInSight%2CMars%20Reconnaissance%20Orbiter%2Cfacts&ab=-&dh=mars.nasa.gov&dr=&du=https%3A%2F%2Fmars.nasa.gov%2Fnews%2F%3Fpage%3D0%26per_page%3D40%26order%3Dpublish_date%2Bdesc%252Ccreated_at%2Bdesc%26search%3D%26category%3D19%252C165%252C184%252C204%26blank_scope%3DLatest&href=https%3A%2F%2Fmars.nasa.gov%2Fnews%2F&dt=News%20%20%E2%80%93%20NASA%E2%80%99s%20Mars%20Exploration%20Program&dbg=0&cap=tc%3D0%26ab%3D0&inst=1&jsl=1&prod=undefined&lng=en&ogt=image%2Cupdated_time%2Ctype%3Darticle%2Curl%2Ctitle%2Cdescription%2Csite_name&pc=men&pub=ra-5a690e4c1320e328&ssl=1&sid=5b53cb1e6480f157&srf=0.01&ver=300&xck=1&xtr=0&og=site_name%3DNASA%25E2%2580%2599s%2520Mars%2520Exploration%2520Program%26description%3DNASA%25E2%2580%2599s%2520real-time%2520portal%2520for%2520Mars%2520exploration%252C%2520featuring%2520the%2520latest%2520news%252C%2520images%252C%2520and%2520discoveries%2520from%2520the%2520Red%2520Planet.%26title%3DNews%2520%2520%25E2%2580%2593%2520NASA%25E2%2580%2599s%2520Mars%2520Exploration%2520Program%26url%3Dhttps%253A%252F%252Fmars.nasa.gov%252Fnews%253Fpage%253D0%2526per_page%253D40%2526order%253Dpublish_date%252Bdesc%25252Ccreated_at%252Bdesc%2526search%253D%2526category%253D19%25252C165%25252C184%25252C204%2526blank_scope%253DLatest%26type%3Darticle%26updated_time%3D2017-09-22%252019%253A53%253A22%2520UTC%26image%3Dhttps%253A%252F%252Fmars.nasa.gov%252Fsystem%252Fsite_config_values%252Fmeta_share_images%252F1_142497main_PIA03154-200.jpg&csi=undefined&rev=v8.3.25-wp&ct=1&xld=1&xd=1\" style=\"height: 1px; width: 1px; position: absolute; top: 0px; z-index: 100000; border: 0px; left: 0px;\" title=\"AddThis utility frame\">\n </iframe>\n </div>\n <style id=\"service-icons-0\">\n </style>\n <div aria-labelledby=\"at4-share-label\" class=\"addthis-smartlayers addthis-smartlayers-desktop\" role=\"region\">\n <div id=\"at4-share-label\">\n AddThis Sharing Sidebar\n </div>\n <div class=\"at4-share addthis_32x32_style atss atss-left addthis-animated slideInLeft\" id=\"at4-share\">\n <a class=\"at-share-btn at-svc-facebook\" role=\"button\" tabindex=\"1\">\n <span class=\"at4-visually-hidden\">\n Share to Facebook\n </span>\n <span class=\"at-icon-wrapper\" style=\"background-color: rgb(59, 89, 152);\">\n <svg aria-labelledby=\"at-svg-facebook-1\" class=\"at-icon at-icon-facebook\" role=\"img\" style=\"fill: rgb(255, 255, 255);\" version=\"1.1\" viewbox=\"0 0 32 32\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n <title id=\"at-svg-facebook-1\" xmlns=\"http://www.w3.org/1999/xhtml\">\n Facebook\n </title>\n <g>\n <path d=\"M22 5.16c-.406-.054-1.806-.16-3.43-.16-3.4 0-5.733 1.825-5.733 5.17v2.882H9v3.913h3.837V27h4.604V16.965h3.823l.587-3.913h-4.41v-2.5c0-1.123.347-1.903 2.198-1.903H22V5.16z\" fill-rule=\"evenodd\">\n </path>\n </g>\n </svg>\n </span>\n </a>\n <a class=\"at-share-btn at-svc-twitter\" role=\"button\" tabindex=\"1\">\n <span class=\"at4-visually-hidden\">\n Share to Twitter\n </span>\n <span class=\"at-icon-wrapper\" style=\"background-color: rgb(29, 161, 242);\">\n <svg aria-labelledby=\"at-svg-twitter-2\" class=\"at-icon at-icon-twitter\" role=\"img\" style=\"fill: rgb(255, 255, 255);\" version=\"1.1\" viewbox=\"0 0 32 32\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n <title id=\"at-svg-twitter-2\" xmlns=\"http://www.w3.org/1999/xhtml\">\n Twitter\n </title>\n <g>\n <path d=\"M27.996 10.116c-.81.36-1.68.602-2.592.71a4.526 4.526 0 0 0 1.984-2.496 9.037 9.037 0 0 1-2.866 1.095 4.513 4.513 0 0 0-7.69 4.116 12.81 12.81 0 0 1-9.3-4.715 4.49 4.49 0 0 0-.612 2.27 4.51 4.51 0 0 0 2.008 3.755 4.495 4.495 0 0 1-2.044-.564v.057a4.515 4.515 0 0 0 3.62 4.425 4.52 4.52 0 0 1-2.04.077 4.517 4.517 0 0 0 4.217 3.134 9.055 9.055 0 0 1-5.604 1.93A9.18 9.18 0 0 1 6 23.85a12.773 12.773 0 0 0 6.918 2.027c8.3 0 12.84-6.876 12.84-12.84 0-.195-.005-.39-.014-.583a9.172 9.172 0 0 0 2.252-2.336\" fill-rule=\"evenodd\">\n </path>\n </g>\n </svg>\n </span>\n </a>\n <a class=\"at-share-btn at-svc-reddit\" role=\"button\" tabindex=\"1\">\n <span class=\"at4-visually-hidden\">\n Share to Reddit\n </span>\n <span class=\"at-icon-wrapper\" style=\"background-color: rgb(255, 87, 0);\">\n <svg aria-labelledby=\"at-svg-reddit-3\" class=\"at-icon at-icon-reddit\" role=\"img\" style=\"fill: rgb(255, 255, 255);\" version=\"1.1\" viewbox=\"0 0 32 32\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n <title id=\"at-svg-reddit-3\" xmlns=\"http://www.w3.org/1999/xhtml\">\n Reddit\n </title>\n <g>\n <path d=\"M27 15.5a2.452 2.452 0 0 1-1.338 2.21c.098.38.147.777.147 1.19 0 1.283-.437 2.47-1.308 3.563-.872 1.092-2.06 1.955-3.567 2.588-1.506.634-3.143.95-4.91.95-1.768 0-3.403-.316-4.905-.95-1.502-.632-2.69-1.495-3.56-2.587-.872-1.092-1.308-2.28-1.308-3.562 0-.388.045-.777.135-1.166a2.47 2.47 0 0 1-1.006-.912c-.253-.4-.38-.842-.38-1.322 0-.678.237-1.26.712-1.744a2.334 2.334 0 0 1 1.73-.726c.697 0 1.29.26 1.78.782 1.785-1.258 3.893-1.928 6.324-2.01l1.424-6.467a.42.42 0 0 1 .184-.26.4.4 0 0 1 .32-.063l4.53 1.006c.147-.306.368-.553.662-.74a1.78 1.78 0 0 1 .97-.278c.508 0 .94.18 1.302.54.36.36.54.796.54 1.31 0 .512-.18.95-.54 1.315-.36.364-.794.546-1.302.546-.507 0-.94-.18-1.295-.54a1.793 1.793 0 0 1-.533-1.308l-4.1-.92-1.277 5.86c2.455.074 4.58.736 6.37 1.985a2.315 2.315 0 0 1 1.757-.757c.68 0 1.256.242 1.73.726.476.484.713 1.066.713 1.744zm-16.868 2.47c0 .513.178.95.534 1.315.356.365.787.547 1.295.547.508 0 .942-.182 1.302-.547.36-.364.54-.802.54-1.315 0-.513-.18-.95-.54-1.31-.36-.36-.794-.54-1.3-.54-.5 0-.93.183-1.29.547a1.79 1.79 0 0 0-.54 1.303zm9.944 4.406c.09-.09.135-.2.135-.323a.444.444 0 0 0-.44-.447c-.124 0-.23.042-.32.124-.336.348-.83.605-1.486.77a7.99 7.99 0 0 1-1.964.248 7.99 7.99 0 0 1-1.964-.248c-.655-.165-1.15-.422-1.486-.77a.456.456 0 0 0-.32-.124.414.414 0 0 0-.306.124.41.41 0 0 0-.135.317.45.45 0 0 0 .134.33c.352.355.837.636 1.455.843.617.207 1.118.33 1.503.366a11.6 11.6 0 0 0 1.117.056c.36 0 .733-.02 1.117-.056.385-.037.886-.16 1.504-.366.62-.207 1.104-.488 1.456-.844zm-.037-2.544c.507 0 .938-.182 1.294-.547.356-.364.534-.802.534-1.315 0-.505-.18-.94-.54-1.303a1.75 1.75 0 0 0-1.29-.546c-.506 0-.94.18-1.3.54-.36.36-.54.797-.54 1.31s.18.95.54 1.315c.36.365.794.547 1.3.547z\" fill-rule=\"evenodd\">\n </path>\n </g>\n </svg>\n </span>\n </a>\n <a class=\"at-share-btn at-svc-email\" role=\"button\" tabindex=\"1\">\n <span class=\"at4-visually-hidden\">\n Share to Email\n </span>\n <span class=\"at-icon-wrapper\" style=\"background-color: rgb(132, 132, 132);\">\n <svg aria-labelledby=\"at-svg-email-4\" class=\"at-icon at-icon-email\" role=\"img\" style=\"fill: rgb(255, 255, 255);\" version=\"1.1\" viewbox=\"0 0 32 32\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n <title id=\"at-svg-email-4\" xmlns=\"http://www.w3.org/1999/xhtml\">\n Email\n </title>\n <g>\n <g fill-rule=\"evenodd\">\n </g>\n <path d=\"M27 22.757c0 1.24-.988 2.243-2.19 2.243H7.19C5.98 25 5 23.994 5 22.757V13.67c0-.556.39-.773.855-.496l8.78 5.238c.782.467 1.95.467 2.73 0l8.78-5.238c.472-.28.855-.063.855.495v9.087z\">\n </path>\n <path d=\"M27 9.243C27 8.006 26.02 7 24.81 7H7.19C5.988 7 5 8.004 5 9.243v.465c0 .554.385 1.232.857 1.514l9.61 5.733c.267.16.8.16 1.067 0l9.61-5.733c.473-.283.856-.96.856-1.514v-.465z\">\n </path>\n </g>\n </svg>\n </span>\n </a>\n <a class=\"at-share-btn at-svc-compact\" role=\"button\" tabindex=\"1\">\n <span class=\"at4-visually-hidden\">\n More AddThis Share options\n </span>\n <span class=\"at-icon-wrapper\" style=\"background-color: rgb(255, 101, 80);\">\n <svg aria-labelledby=\"at-svg-addthis-5\" class=\"at-icon at-icon-addthis\" role=\"img\" style=\"fill: rgb(255, 255, 255);\" version=\"1.1\" viewbox=\"0 0 32 32\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n <title id=\"at-svg-addthis-5\" xmlns=\"http://www.w3.org/1999/xhtml\">\n Addthis\n </title>\n <g>\n <path d=\"M18 14V8h-4v6H8v4h6v6h4v-6h6v-4h-6z\" fill-rule=\"evenodd\">\n </path>\n </g>\n </svg>\n </span>\n </a>\n <div class=\"at-custom-sidebar-counter\" style=\"width: 48px; word-wrap: break-word;\">\n <div class=\"at-custom-sidebar-count\" style=\"color: rgb(34, 34, 34);\">\n 36\n </div>\n <div class=\"at-custom-sidebar-text\" style=\"color: rgb(34, 34, 34);\">\n SHARES\n </div>\n </div>\n <div class=\"at-share-close-control ats-transparent at4-hide-content at4-show\" id=\"at4-scc\" title=\"Hide\">\n <div class=\"at4-arrow at-left\">\n Hide\n </div>\n </div>\n </div>\n <div class=\"at-share-open-control at-share-open-control-left ats-transparent at4-hide\" id=\"at4-soc\" title=\"Show\">\n <div class=\"at4-arrow at-right\">\n Show\n </div>\n </div>\n </div>\n <div aria-labelledby=\"at-thankyou-label\" class=\"at4-thankyou at4-thankyou-background at4-hide ats-transparent at4-thankyou-desktop addthis-smartlayers addthis-animated fadeIn at4-show\" id=\"at4-thankyou\" role=\"dialog\">\n <div class=\"at4lb-inner\">\n <button class=\"at4x\" title=\"Close\">\n Close\n </button>\n <a id=\"at4-palogo\">\n <div>\n <a class=\"at-branding-logo\" href=\"//www.addthis.com/website-tools/overview?utm_source=AddThis%20Tools&utm_medium=image\" target=\"_blank\" title=\"Powered by AddThis\">\n <div class=\"at-branding-icon\">\n </div>\n <span class=\"at-branding-addthis\">\n AddThis\n </span>\n </a>\n </div>\n </a>\n <div class=\"at4-thankyou-inner\">\n <div class=\"thankyou-title\" id=\"at-thankyou-label\">\n </div>\n <div class=\"thankyou-description\">\n </div>\n <div class=\"at4-thankyou-layer\">\n </div>\n </div>\n </div>\n </div>\n <div aria-labelledby=\"at-share-dock-label\" class=\"at-share-dock-outer at4-hide addthis-smartlayers at4-visually-hidden addthis-smartlayers-mobile\" role=\"region\">\n <div class=\"at4-hide\" id=\"at-share-dock-label\">\n AddThis Sharing\n </div>\n <div class=\"at-share-dock atss atss-bottom at-shfs-small addthis-animated slideInUp at4-show at4-hide\" id=\"at-share-dock\">\n <a class=\"at4-count\" href=\"#\" style=\"width: 16.6667%;\">\n <span class=\"at4-counter\">\n </span>\n <span class=\"at4-share-label\">\n SHARES\n </span>\n </a>\n <a class=\"at-share-btn at-svc-facebook\" role=\"button\" style=\"width: 16.6667%;\" tabindex=\"1\" title=\"Facebook\">\n <span class=\"at-icon-wrapper\" style=\"background-color: rgb(59, 89, 152);\">\n <svg alt=\"Facebook\" aria-labelledby=\"at-svg-facebook-6\" class=\"at-icon at-icon-facebook\" role=\"img\" style=\"fill: rgb(255, 255, 255); width: 24px; height: 24px;\" title=\"Facebook\" version=\"1.1\" viewbox=\"0 0 32 32\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n <title id=\"at-svg-facebook-6\" xmlns=\"http://www.w3.org/1999/xhtml\">\n Facebook\n </title>\n <g>\n <path d=\"M22 5.16c-.406-.054-1.806-.16-3.43-.16-3.4 0-5.733 1.825-5.733 5.17v2.882H9v3.913h3.837V27h4.604V16.965h3.823l.587-3.913h-4.41v-2.5c0-1.123.347-1.903 2.198-1.903H22V5.16z\" fill-rule=\"evenodd\">\n </path>\n </g>\n </svg>\n </span>\n </a>\n <a class=\"at-share-btn at-svc-twitter\" role=\"button\" style=\"width: 16.6667%;\" tabindex=\"1\" title=\"Twitter\">\n <span class=\"at-icon-wrapper\" style=\"background-color: rgb(29, 161, 242);\">\n <svg alt=\"Twitter\" aria-labelledby=\"at-svg-twitter-7\" class=\"at-icon at-icon-twitter\" role=\"img\" style=\"fill: rgb(255, 255, 255); width: 24px; height: 24px;\" title=\"Twitter\" version=\"1.1\" viewbox=\"0 0 32 32\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n <title id=\"at-svg-twitter-7\" xmlns=\"http://www.w3.org/1999/xhtml\">\n Twitter\n </title>\n <g>\n <path d=\"M27.996 10.116c-.81.36-1.68.602-2.592.71a4.526 4.526 0 0 0 1.984-2.496 9.037 9.037 0 0 1-2.866 1.095 4.513 4.513 0 0 0-7.69 4.116 12.81 12.81 0 0 1-9.3-4.715 4.49 4.49 0 0 0-.612 2.27 4.51 4.51 0 0 0 2.008 3.755 4.495 4.495 0 0 1-2.044-.564v.057a4.515 4.515 0 0 0 3.62 4.425 4.52 4.52 0 0 1-2.04.077 4.517 4.517 0 0 0 4.217 3.134 9.055 9.055 0 0 1-5.604 1.93A9.18 9.18 0 0 1 6 23.85a12.773 12.773 0 0 0 6.918 2.027c8.3 0 12.84-6.876 12.84-12.84 0-.195-.005-.39-.014-.583a9.172 9.172 0 0 0 2.252-2.336\" fill-rule=\"evenodd\">\n </path>\n </g>\n </svg>\n </span>\n </a>\n <a class=\"at-share-btn at-svc-reddit\" role=\"button\" style=\"width: 16.6667%;\" tabindex=\"1\" title=\"Reddit\">\n <span class=\"at-icon-wrapper\" style=\"background-color: rgb(255, 87, 0);\">\n <svg alt=\"Reddit\" aria-labelledby=\"at-svg-reddit-8\" class=\"at-icon at-icon-reddit\" role=\"img\" style=\"fill: rgb(255, 255, 255); width: 24px; height: 24px;\" title=\"Reddit\" version=\"1.1\" viewbox=\"0 0 32 32\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n <title id=\"at-svg-reddit-8\" xmlns=\"http://www.w3.org/1999/xhtml\">\n Reddit\n </title>\n <g>\n <path d=\"M27 15.5a2.452 2.452 0 0 1-1.338 2.21c.098.38.147.777.147 1.19 0 1.283-.437 2.47-1.308 3.563-.872 1.092-2.06 1.955-3.567 2.588-1.506.634-3.143.95-4.91.95-1.768 0-3.403-.316-4.905-.95-1.502-.632-2.69-1.495-3.56-2.587-.872-1.092-1.308-2.28-1.308-3.562 0-.388.045-.777.135-1.166a2.47 2.47 0 0 1-1.006-.912c-.253-.4-.38-.842-.38-1.322 0-.678.237-1.26.712-1.744a2.334 2.334 0 0 1 1.73-.726c.697 0 1.29.26 1.78.782 1.785-1.258 3.893-1.928 6.324-2.01l1.424-6.467a.42.42 0 0 1 .184-.26.4.4 0 0 1 .32-.063l4.53 1.006c.147-.306.368-.553.662-.74a1.78 1.78 0 0 1 .97-.278c.508 0 .94.18 1.302.54.36.36.54.796.54 1.31 0 .512-.18.95-.54 1.315-.36.364-.794.546-1.302.546-.507 0-.94-.18-1.295-.54a1.793 1.793 0 0 1-.533-1.308l-4.1-.92-1.277 5.86c2.455.074 4.58.736 6.37 1.985a2.315 2.315 0 0 1 1.757-.757c.68 0 1.256.242 1.73.726.476.484.713 1.066.713 1.744zm-16.868 2.47c0 .513.178.95.534 1.315.356.365.787.547 1.295.547.508 0 .942-.182 1.302-.547.36-.364.54-.802.54-1.315 0-.513-.18-.95-.54-1.31-.36-.36-.794-.54-1.3-.54-.5 0-.93.183-1.29.547a1.79 1.79 0 0 0-.54 1.303zm9.944 4.406c.09-.09.135-.2.135-.323a.444.444 0 0 0-.44-.447c-.124 0-.23.042-.32.124-.336.348-.83.605-1.486.77a7.99 7.99 0 0 1-1.964.248 7.99 7.99 0 0 1-1.964-.248c-.655-.165-1.15-.422-1.486-.77a.456.456 0 0 0-.32-.124.414.414 0 0 0-.306.124.41.41 0 0 0-.135.317.45.45 0 0 0 .134.33c.352.355.837.636 1.455.843.617.207 1.118.33 1.503.366a11.6 11.6 0 0 0 1.117.056c.36 0 .733-.02 1.117-.056.385-.037.886-.16 1.504-.366.62-.207 1.104-.488 1.456-.844zm-.037-2.544c.507 0 .938-.182 1.294-.547.356-.364.534-.802.534-1.315 0-.505-.18-.94-.54-1.303a1.75 1.75 0 0 0-1.29-.546c-.506 0-.94.18-1.3.54-.36.36-.54.797-.54 1.31s.18.95.54 1.315c.36.365.794.547 1.3.547z\" fill-rule=\"evenodd\">\n </path>\n </g>\n </svg>\n </span>\n </a>\n <a class=\"at-share-btn at-svc-email\" role=\"button\" style=\"width: 16.6667%;\" tabindex=\"1\" title=\"Email\">\n <span class=\"at-icon-wrapper\" style=\"background-color: rgb(132, 132, 132);\">\n <svg alt=\"Email\" aria-labelledby=\"at-svg-email-9\" class=\"at-icon at-icon-email\" role=\"img\" style=\"fill: rgb(255, 255, 255); width: 24px; height: 24px;\" title=\"Email\" version=\"1.1\" viewbox=\"0 0 32 32\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n <title id=\"at-svg-email-9\" xmlns=\"http://www.w3.org/1999/xhtml\">\n Email\n </title>\n <g>\n <g fill-rule=\"evenodd\">\n </g>\n <path d=\"M27 22.757c0 1.24-.988 2.243-2.19 2.243H7.19C5.98 25 5 23.994 5 22.757V13.67c0-.556.39-.773.855-.496l8.78 5.238c.782.467 1.95.467 2.73 0l8.78-5.238c.472-.28.855-.063.855.495v9.087z\">\n </path>\n <path d=\"M27 9.243C27 8.006 26.02 7 24.81 7H7.19C5.988 7 5 8.004 5 9.243v.465c0 .554.385 1.232.857 1.514l9.61 5.733c.267.16.8.16 1.067 0l9.61-5.733c.473-.283.856-.96.856-1.514v-.465z\">\n </path>\n </g>\n </svg>\n </span>\n </a>\n <a class=\"at-share-btn at-svc-compact\" role=\"button\" style=\"width: 16.6667%;\" tabindex=\"1\" title=\"More\">\n <span class=\"at-icon-wrapper\" style=\"background-color: rgb(255, 101, 80);\">\n <svg alt=\"More\" aria-labelledby=\"at-svg-addthis-10\" class=\"at-icon at-icon-addthis\" role=\"img\" style=\"fill: rgb(255, 255, 255); width: 24px; height: 24px;\" title=\"More\" version=\"1.1\" viewbox=\"0 0 32 32\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n <title id=\"at-svg-addthis-10\" xmlns=\"http://www.w3.org/1999/xhtml\">\n Addthis\n </title>\n <g>\n <path d=\"M18 14V8h-4v6H8v4h6v6h4v-6h6v-4h-6z\" fill-rule=\"evenodd\">\n </path>\n </g>\n </svg>\n </span>\n </a>\n </div>\n </div>\n </body>\n</html>\n"
],
[
"news_title = soup.find(\"div\", class_=\"content_title\").get_text()\nnews_p = soup.find(\"div\", class_=\"article_teaser_body\").get_text()",
"_____no_output_____"
],
[
"print(f\"{news_title}:{news_p}\")",
"'Storm Chasers' on Mars Searching for Dusty Secrets:Scientists with NASA's Mars orbiters have been waiting years for an event like the current Mars global dust storm.\n"
]
],
[
[
"# JPL Mars Space Images",
"_____no_output_____"
]
],
[
[
"executable_path = {\"executable_path\": \"chromedriver\"}\nbrowser = Browser(\"chrome\", **executable_path, headless=False)\n\nurl = \"https://www.jpl.nasa.gov/spaceimages/?search=&category=Mars\"\nbrowser.visit(url)\n\nhtml = browser.html\nsoup = BeautifulSoup(html, \"html.parser\")",
"_____no_output_____"
],
[
"image_url = soup.footer.find(\"a\", class_=\"button fancybox\")[\"data-fancybox-href\"]\nfeatured_image_url = \"https://www.jpl.nasa.gov\" + image_url\nprint(featured_image_url)",
"https://www.jpl.nasa.gov/spaceimages/images/mediumsize/PIA18284_ip.jpg\n"
]
],
[
[
"# Mars Weather",
"_____no_output_____"
]
],
[
[
"executable_path = {\"executable_path\": \"chromedriver\"}\nbrowser = Browser(\"chrome\", **executable_path, headless=False)\n\nurl = \"https://twitter.com/marswxreport?lang=en\"\nbrowser.visit(url)\n\nhtml = browser.html\nsoup = BeautifulSoup(html, \"html.parser\")",
"_____no_output_____"
],
[
"tweets = soup.find_all(\"p\", class_=\"TweetTextSize TweetTextSize--normal js-tweet-text tweet-text\")",
"_____no_output_____"
],
[
"for tweet in tweets:\n tweet_parent = tweet.find_parent(\"div\", class_=\"content\")\n tweet_id = tweet_parent.find(\"a\", class_=\"account-group js-account-group js-action-profile js-user-profile-link js-nav\")[\"href\"]\n \n if tweet_id == '/MarsWxReport':\n mars_weather = tweet_parent.find(\"p\", class_=\"TweetTextSize TweetTextSize--normal js-tweet-text tweet-text\").get_text()\n break",
"_____no_output_____"
],
[
"mars_weather",
"_____no_output_____"
]
],
[
[
"# Mars Facts",
"_____no_output_____"
]
],
[
[
"url = 'https://space-facts.com/mars/'",
"_____no_output_____"
],
[
"tables = pd.read_html(url)\ntables",
"_____no_output_____"
],
[
"df = tables[0]\ndf.columns = [\"Description\", \"Value\"]\ndf.set_index(df[\"Description\"], inplace=True)",
"_____no_output_____"
],
[
"df = df[[\"Value\"]]",
"_____no_output_____"
],
[
"html_table = df.to_html()\nhtml_table = html_table.replace('\\n', '')\nhtml_table",
"_____no_output_____"
]
],
[
[
"# Mars Hemispheres",
"_____no_output_____"
]
],
[
[
"executable_path = {\"executable_path\": \"chromedriver\"}\nbrowser = Browser(\"chrome\", **executable_path, headless=False)\n\nurl = \"https://astrogeology.usgs.gov/search/results?q=hemisphere+enhanced&k1=target&v1=Mars\"\nbrowser.visit(url)\n\nhtml = browser.html\nsoup = BeautifulSoup(html, \"html.parser\")\n\nh3s = soup.find_all(\"h3\")",
"_____no_output_____"
],
[
"titles = []\nfor h3 in h3s:\n h3 = str(h3)\n h3 = h3[4:-14]\n titles.append(h3)\ntitles",
"_____no_output_____"
],
[
"img_urls = []\nfor title in titles:\n browser.click_link_by_partial_text(title)\n\n html = browser.html\n soup = BeautifulSoup(html, \"html.parser\")\n\n img_urls.append(soup.find(\"div\", class_=\"downloads\").find(\"a\")[\"href\"])\nimg_urls",
"_____no_output_____"
],
[
"hemisphere_image_urls = []\nfor title, img_url in zip(titles, img_urls):\n hemisphere_image_urls.append({\"title\": title, \"img_url\":img_url})\n\nhemisphere_image_urls",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
]
] |
cb91f72ad4b4d9ac717e4817ae78c12e5738fad8
| 20,159 |
ipynb
|
Jupyter Notebook
|
study/fft_scratch/UnderstandingTheFFT.ipynb
|
ooshyun/filterdesign
|
59dbea191b8cd44aa9f2d02d3787b5805d486ae2
|
[
"MIT"
] | 1 |
2021-12-27T00:38:32.000Z
|
2021-12-27T00:38:32.000Z
|
study/fft_scratch/UnderstandingTheFFT.ipynb
|
ooshyun/filterdesign
|
59dbea191b8cd44aa9f2d02d3787b5805d486ae2
|
[
"MIT"
] | null | null | null |
study/fft_scratch/UnderstandingTheFFT.ipynb
|
ooshyun/filterdesign
|
59dbea191b8cd44aa9f2d02d3787b5805d486ae2
|
[
"MIT"
] | null | null | null | 37.47026 | 851 | 0.605536 |
[
[
[
"# Understanding the FFT Algorithm\nCopy from http://jakevdp.github.io/blog/2013/08/28/understanding-the-fft/",
"_____no_output_____"
],
[
"*This notebook first appeared as a post by Jake Vanderplas on [Pythonic Perambulations](http://jakevdp.github.io/blog/2013/08/28/understanding-the-fft/). The notebook content is BSD-licensed.*",
"_____no_output_____"
],
[
"<!-- PELICAN_BEGIN_SUMMARY -->\nThe Fast Fourier Transform (FFT) is one of the most important algorithms in signal processing and data analysis. I've used it for years, but having no formal computer science background, It occurred to me this week that I've never thought to ask *how* the FFT computes the discrete Fourier transform so quickly. I dusted off an old algorithms book and looked into it, and enjoyed reading about the deceptively simple computational trick that JW Cooley and John Tukey outlined in their classic [1965 paper](http://www.ams.org/journals/mcom/1965-19-090/S0025-5718-1965-0178586-1/) introducing the subject.",
"_____no_output_____"
],
[
"The goal of this post is to dive into the Cooley-Tukey FFT algorithm, explaining the symmetries that lead to it, and to show some straightforward Python implementations putting the theory into practice. My hope is that this exploration will give data scientists like myself a more complete picture of what's going on in the background of the algorithms we use.\n<!-- PELICAN_END_SUMMARY -->",
"_____no_output_____"
],
[
"## The Discrete Fourier Transform",
"_____no_output_____"
],
[
"The FFT is a fast, $\\mathcal{O}[N\\log N]$ algorithm to compute the Discrete Fourier Transform (DFT), which\nnaively is an $\\mathcal{O}[N^2]$ computation. The DFT, like the more familiar continuous version of the Fourier transform, has a forward and inverse form which are defined as follows:",
"_____no_output_____"
],
[
"**Forward Discrete Fourier Transform (DFT):**\n$$X_k = \\sum_{n=0}^{N-1} x_n \\cdot e^{-i~2\\pi~k~n~/~N}$$",
"_____no_output_____"
],
[
"**Inverse Discrete Fourier Transform (IDFT):**\n$$x_n = \\frac{1}{N}\\sum_{k=0}^{N-1} X_k e^{i~2\\pi~k~n~/~N}$$",
"_____no_output_____"
],
[
"The transformation from $x_n \\to X_k$ is a translation from configuration space to frequency space, and can be very useful in both exploring the power spectrum of a signal, and also for transforming certain problems for more efficient computation. For some examples of this in action, you can check out Chapter 10 of our upcoming Astronomy/Statistics book, with figures and Python source code available [here](http://www.astroml.org/book_figures/chapter10/). For an example of the FFT being used to simplify an otherwise difficult differential equation integration, see my post on [Solving the Schrodinger Equation in Python](http://jakevdp.github.io/blog/2012/09/05/quantum-python/).\n\nBecause of the importance of the FFT in so many fields, Python contains many standard tools and wrappers to compute this. Both NumPy and SciPy have wrappers of the extremely well-tested FFTPACK library, found in the submodules ``numpy.fft`` and ``scipy.fftpack`` respectively. The fastest FFT I am aware of is in the [FFTW](http://www.fftw.org/) package, which is also available in Python via the [PyFFTW](https://pypi.python.org/pypi/pyFFTW) package.\n\nFor the moment, though, let's leave these implementations aside and ask how we might compute the FFT in Python from scratch.",
"_____no_output_____"
],
[
"## Computing the Discrete Fourier Transform",
"_____no_output_____"
],
[
"For simplicity, we'll concern ourself only with the forward transform, as the inverse transform can be implemented in a very similar manner. Taking a look at the DFT expression above, we see that it is nothing more than a straightforward linear operation: a matrix-vector multiplication of $\\vec{x}$,\n\n$$\\vec{X} = M \\cdot \\vec{x}$$\n\nwith the matrix $M$ given by\n\n$$M_{kn} = e^{-i~2\\pi~k~n~/~N}.$$\n\nWith this in mind, we can compute the DFT using simple matrix multiplication as follows:",
"_____no_output_____"
]
],
[
[
"import numpy as np\ndef DFT_slow(x):\n \"\"\"Compute the discrete Fourier Transform of the 1D array x\"\"\"\n x = np.asarray(x, dtype=float)\n N = x.shape[0]\n n = np.arange(N)\n k = n.reshape((N, 1))\n M = np.exp(-2j * np.pi * k * n / N)\n return np.dot(M, x)",
"_____no_output_____"
]
],
[
[
"We can double-check the result by comparing to numpy's built-in FFT function:",
"_____no_output_____"
]
],
[
[
"x = np.random.random(1024)\nnp.allclose(DFT_slow(x), np.fft.fft(x))",
"_____no_output_____"
]
],
[
[
"Just to confirm the sluggishness of our algorithm, we can compare the execution times\nof these two approaches:",
"_____no_output_____"
]
],
[
[
"%timeit DFT_slow(x)\n%timeit np.fft.fft(x)",
"10 loops, best of 3: 75.4 ms per loop\n10000 loops, best of 3: 25.5 µs per loop\n"
]
],
[
[
"We are over 1000 times slower, which is to be expected for such a simplistic implementation. But that's not the worst of it. For an input vector of length $N$, the FFT algorithm scales as $\\mathcal{O}[N\\log N]$, while our slow algorithm scales as $\\mathcal{O}[N^2]$. That means that for $N=10^6$ elements, we'd expect the FFT to complete in somewhere around 50 ms, while our slow algorithm would take nearly 20 hours!\n\nSo how does the FFT accomplish this speedup? The answer lies in exploiting symmetry.",
"_____no_output_____"
],
[
"## Symmetries in the Discrete Fourier Transform",
"_____no_output_____"
],
[
"One of the most important tools in the belt of an algorithm-builder is to exploit symmetries of a problem. If you can show analytically that one piece of a problem is simply related to another, you can compute the subresult\nonly once and save that computational cost. Cooley and Tukey used exactly this approach in deriving the FFT.\n\nWe'll start by asking what the value of $X_{N+k}$ is. From our above expression:",
"_____no_output_____"
],
[
"$$\n\\begin{align*}\nX_{N + k} &= \\sum_{n=0}^{N-1} x_n \\cdot e^{-i~2\\pi~(N + k)~n~/~N}\\\\\n &= \\sum_{n=0}^{N-1} x_n \\cdot e^{- i~2\\pi~n} \\cdot e^{-i~2\\pi~k~n~/~N}\\\\\n &= \\sum_{n=0}^{N-1} x_n \\cdot e^{-i~2\\pi~k~n~/~N}\n\\end{align*}\n$$",
"_____no_output_____"
],
[
"where we've used the identity $\\exp[2\\pi~i~n] = 1$ which holds for any integer $n$.\n\nThe last line shows a nice symmetry property of the DFT:\n\n$$X_{N+k} = X_k.$$\n\nBy a simple extension,\n\n$$X_{k + i \\cdot N} = X_k$$\n\nfor any integer $i$. As we'll see below, this symmetry can be exploited to compute the DFT much more quickly.",
"_____no_output_____"
],
[
"## DFT to FFT: Exploiting Symmetry",
"_____no_output_____"
],
[
"Cooley and Tukey showed that it's possible to divide the DFT computation into two smaller parts. From\nthe definition of the DFT we have:",
"_____no_output_____"
],
[
"$$\n\\begin{align}\nX_k &= \\sum_{n=0}^{N-1} x_n \\cdot e^{-i~2\\pi~k~n~/~N} \\\\\n &= \\sum_{m=0}^{N/2 - 1} x_{2m} \\cdot e^{-i~2\\pi~k~(2m)~/~N} + \\sum_{m=0}^{N/2 - 1} x_{2m + 1} \\cdot e^{-i~2\\pi~k~(2m + 1)~/~N} \\\\\n &= \\sum_{m=0}^{N/2 - 1} x_{2m} \\cdot e^{-i~2\\pi~k~m~/~(N/2)} + e^{-i~2\\pi~k~/~N} \\sum_{m=0}^{N/2 - 1} x_{2m + 1} \\cdot e^{-i~2\\pi~k~m~/~(N/2)}\n\\end{align}\n$$",
"_____no_output_____"
],
[
"We've split the single Discrete Fourier transform into two terms which themselves look very similar to smaller Discrete Fourier Transforms, one on the odd-numbered values, and one on the even-numbered values. So far, however, we haven't saved any computational cycles. Each term consists of $(N/2)*N$ computations, for a total of $N^2$.\n\nThe trick comes in making use of symmetries in each of these terms. Because the range of $k$ is $0 \\le k < N$, while the range of $n$ is $0 \\le n < M \\equiv N/2$, we see from the symmetry properties above that we need only perform half the computations for each sub-problem. Our $\\mathcal{O}[N^2]$ computation has become $\\mathcal{O}[M^2]$, with $M$ half the size of $N$.\n\nBut there's no reason to stop there: as long as our smaller Fourier transforms have an even-valued $M$, we can reapply this divide-and-conquer approach, halving the computational cost each time, until our arrays are small enough that the strategy is no longer beneficial. In the asymptotic limit, this recursive approach scales as $\\mathcal{O}[N\\log N]$.\n\nThis recursive algorithm can be implemented very quickly in Python, falling-back on our slow DFT code when the size of the sub-problem becomes suitably small:",
"_____no_output_____"
]
],
[
[
"def FFT(x):\n \"\"\"A recursive implementation of the 1D Cooley-Tukey FFT\"\"\"\n x = np.asarray(x, dtype=float)\n N = x.shape[0]\n \n if N % 2 > 0:\n raise ValueError(\"size of x must be a power of 2\")\n elif N <= 32: # this cutoff should be optimized\n return DFT_slow(x)\n else:\n X_even = FFT(x[::2])\n X_odd = FFT(x[1::2])\n factor = np.exp(-2j * np.pi * np.arange(N) / N)\n return np.concatenate([X_even + factor[:N / 2] * X_odd,\n X_even + factor[N / 2:] * X_odd])",
"_____no_output_____"
]
],
[
[
"Here we'll do a quick check that our algorithm produces the correct result:",
"_____no_output_____"
]
],
[
[
"x = np.random.random(1024)\nnp.allclose(FFT(x), np.fft.fft(x))",
"_____no_output_____"
]
],
[
[
"And we'll time this algorithm against our slow version:",
"_____no_output_____"
]
],
[
[
"%timeit DFT_slow(x)\n%timeit FFT(x)\n%timeit np.fft.fft(x)",
"10 loops, best of 3: 77.6 ms per loop\n100 loops, best of 3: 4.07 ms per loop\n10000 loops, best of 3: 24.7 µs per loop\n"
]
],
[
[
"Our calculation is faster than the naive version by over an order of magnitude! What's more, our recursive algorithm is asymptotically $\\mathcal{O}[N\\log N]$: we've implemented the Fast Fourier Transform.\n\nNote that we still haven't come close to the speed of the built-in FFT algorithm in numpy, and this is to be expected. The FFTPACK algorithm behind numpy's ``fft`` is a Fortran implementation which has received years of tweaks and optimizations. Furthermore, our NumPy solution involves both Python-stack recursions and the allocation of many temporary arrays, which adds significant computation time.\n\nA good strategy to speed up code when working with Python/NumPy is to vectorize repeated computations where possible. We can do this, and in the process remove our recursive function calls, and make our Python FFT even more efficient.",
"_____no_output_____"
],
[
"## Vectorized Numpy Version",
"_____no_output_____"
],
[
"Notice that in the above recursive FFT implementation, at the lowest recursion level we perform $N~/~32$ identical matrix-vector products. The efficiency of our algorithm would benefit by computing these matrix-vector products all at once as a single matrix-matrix product. At each subsequent level of recursion, we also perform duplicate operations which can be vectorized. NumPy excels at this sort of operation, and we can make use of that fact to create this vectorized version of the Fast Fourier Transform:",
"_____no_output_____"
]
],
[
[
"def FFT_vectorized(x):\n \"\"\"A vectorized, non-recursive version of the Cooley-Tukey FFT\"\"\"\n x = np.asarray(x, dtype=float)\n N = x.shape[0]\n\n if np.log2(N) % 1 > 0:\n raise ValueError(\"size of x must be a power of 2\")\n\n # N_min here is equivalent to the stopping condition above,\n # and should be a power of 2\n N_min = min(N, 32)\n \n # Perform an O[N^2] DFT on all length-N_min sub-problems at once\n n = np.arange(N_min)\n k = n[:, None]\n M = np.exp(-2j * np.pi * n * k / N_min)\n X = np.dot(M, x.reshape((N_min, -1)))\n\n # build-up each level of the recursive calculation all at once\n while X.shape[0] < N:\n X_even = X[:, :X.shape[1] / 2]\n X_odd = X[:, X.shape[1] / 2:]\n factor = np.exp(-1j * np.pi * np.arange(X.shape[0])\n / X.shape[0])[:, None]\n X = np.vstack([X_even + factor * X_odd,\n X_even - factor * X_odd])\n\n return X.ravel()",
"_____no_output_____"
]
],
[
[
"Though the algorithm is a bit more opaque, it is simply a rearrangement of the operations used in the recursive version with one exception: we exploit a symmetry in the ``factor`` computation and construct only half of the array. Again, we'll confirm that our function yields the correct result:",
"_____no_output_____"
]
],
[
[
"x = np.random.random(1024)\nnp.allclose(FFT_vectorized(x), np.fft.fft(x))",
"_____no_output_____"
]
],
[
[
"Because our algorithms are becoming much more efficient, we can use a larger array to compare the timings,\nleaving out ``DFT_slow``:",
"_____no_output_____"
]
],
[
[
"x = np.random.random(1024 * 16)\n%timeit FFT(x)\n%timeit FFT_vectorized(x)\n%timeit np.fft.fft(x)",
"10 loops, best of 3: 72.8 ms per loop\n100 loops, best of 3: 4.11 ms per loop\n1000 loops, best of 3: 505 µs per loop\n"
]
],
[
[
"We've improved our implementation by another order of magnitude! We're now within about a factor of 10 of the FFTPACK benchmark, using only a couple dozen lines of pure Python + NumPy. Though it's still no match computationally speaking, readibility-wise the Python version is far superior to the FFTPACK source, which you can browse [here](http://www.netlib.org/fftpack/fft.c).",
"_____no_output_____"
],
[
"So how does FFTPACK attain this last bit of speedup? Well, mainly it's just a matter of detailed bookkeeping. FFTPACK spends a lot of time making sure to reuse any sub-computation that can be reused. Our numpy version still involves an excess of memory allocation and copying; in a low-level language like Fortran it's easier to control and minimize memory use. In addition, the Cooley-Tukey algorithm can be extended to use splits of size other than 2 (what we've implemented here is known as the *radix-2* Cooley-Tukey FFT). Also, other more sophisticated FFT algorithms may be used, including fundamentally distinct approaches based on convolutions (see, e.g. Bluestein's algorithm and Rader's algorithm). The combination of the above extensions and techniques can lead to very fast FFTs even on arrays whose size is not a power of two.",
"_____no_output_____"
],
[
"Though the pure-Python functions are probably not useful in practice, I hope they've provided a bit of an intuition into what's going on in the background of FFT-based data analysis. As data scientists, we can make-do with black-box implementations of fundamental tools constructed by our more algorithmically-minded colleagues, but I am a firm believer that the more understanding we have about the low-level algorithms we're applying to our data, the better practitioners we'll be.",
"_____no_output_____"
],
[
"*This blog post was written entirely in the IPython Notebook. The full notebook can be downloaded\n[here](http://jakevdp.github.io/downloads/notebooks/UnderstandingTheFFT.ipynb),\nor viewed statically\n[here](http://nbviewer.ipython.org/url/jakevdp.github.io/downloads/notebooks/UnderstandingTheFFT.ipynb).*",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
]
] |
cb91fed9760dfe48c3f0d80129311ab15be7cde7
| 11,422 |
ipynb
|
Jupyter Notebook
|
nbs/15_label.ipynb
|
MannLabs/alphapept
|
66fdc5a96e89bcfbbe886611159e21347c92dbe0
|
[
"Apache-2.0"
] | 97 |
2021-07-24T20:30:09.000Z
|
2022-03-30T08:33:47.000Z
|
nbs/15_label.ipynb
|
MannLabs/alphapept
|
66fdc5a96e89bcfbbe886611159e21347c92dbe0
|
[
"Apache-2.0"
] | 93 |
2021-07-26T15:03:50.000Z
|
2022-03-16T14:37:26.000Z
|
nbs/15_label.ipynb
|
MannLabs/alphapept
|
66fdc5a96e89bcfbbe886611159e21347c92dbe0
|
[
"Apache-2.0"
] | 17 |
2021-07-26T07:40:49.000Z
|
2022-03-15T20:04:34.000Z
| 33.495601 | 159 | 0.55682 |
[
[
[
"# default_exp label",
"_____no_output_____"
]
],
[
[
"# Label\n\n> A collection of functions to do label-based quantification",
"_____no_output_____"
]
],
[
[
"#hide\nfrom nbdev.showdoc import *",
"_____no_output_____"
]
],
[
[
"## Label search\n\nThe label search is implemented based on the compare_frags from the search. \nWe have a fixed number of reporter channels and check if we find a respective peak within the search tolerance.\n\nUseful resources:\n\n- [IsobaricAnalyzer](https://abibuilder.informatik.uni-tuebingen.de/archive/openms/Documentation/nightly/html/TOPP_IsobaricAnalyzer.html)\n\n- [TMT Talk from Hupo 2015](https://assets.thermofisher.com/TFS-Assets/CMD/Reference-Materials/PP-TMT-Multiplexed-Protein-Quantification-HUPO2015-EN.pdf)",
"_____no_output_____"
]
],
[
[
"#export\nfrom numba import njit\nfrom alphapept.search import compare_frags\nimport numpy as np\n\n@njit\ndef label_search(query_frag: np.ndarray, query_int: np.ndarray, label: np.ndarray, reporter_frag_tol:float, ppm:bool)-> (np.ndarray, np.ndarray):\n \"\"\"Function to search for a label for a given spectrum.\n\n Args:\n query_frag (np.ndarray): Array with query fragments.\n query_int (np.ndarray): Array with query intensities.\n label (np.ndarray): Array with label masses.\n reporter_frag_tol (float): Fragment tolerance for search.\n ppm (bool): Flag to use ppm instead of Dalton.\n\n Returns:\n np.ndarray: Array with intensities for the respective label channel. \n np.ndarray: Array with offset masses.\n\n \"\"\"\n \n report = np.zeros(len(label))\n off_mass = np.zeros_like(label)\n \n hits = compare_frags(query_frag, label, reporter_frag_tol, ppm)\n for idx, _ in enumerate(hits):\n if _ > 0:\n report[idx] = query_int[_-1]\n off_mass[idx] = query_frag[_-1] - label[idx]\n\n if ppm:\n off_mass[idx] = off_mass[idx] / (query_frag[_-1] + label[idx]) *2 * 1e6\n \n return report, off_mass",
"_____no_output_____"
],
[
"def test_label_search():\n query_frag = np.array([1,2,3,4,5])\n query_int = np.array([1,2,3,4,5])\n label = np.array([1.0, 2.0, 3.0, 4.0, 5.0])\n frag_tolerance = 0.1\n ppm= False\n\n assert np.allclose(label_search(query_frag, query_int, label, frag_tolerance, ppm)[0], query_int)\n\n query_frag = np.array([1,2,3,4,6])\n query_int = np.array([1,2,3,4,5])\n\n assert np.allclose(label_search(query_frag, query_int, label, frag_tolerance, ppm)[0], np.array([1,2,3,4,0]))\n \n query_frag = np.array([1,2,3,4,6])\n query_int = np.array([5,4,3,2,1])\n\n assert np.allclose(label_search(query_frag, query_int, label, frag_tolerance, ppm)[0], np.array([5,4,3,2,0]))\n \n query_frag = np.array([1.1, 2.2, 3.3, 4.4, 6.6])\n query_int = np.array([1,2,3,4,5])\n \n frag_tolerance = 0.5\n ppm= False\n \n assert np.allclose(label_search(query_frag, query_int, label, frag_tolerance, ppm)[1], np.array([0.1, 0.2, 0.3, 0.4, 0.0]))\n \ntest_label_search()",
"_____no_output_____"
],
[
"#Example usage\n\nquery_frag = np.array([127, 128, 129.1, 132])\nquery_int = np.array([100, 200, 300, 400, 500])\n\nlabel = np.array([127.0, 128.0, 129.0, 130.0])\n\nfrag_tolerance = 0.1\nppm = False\n\nreport, offset = label_search(query_frag, query_int, label, frag_tolerance, ppm)\n\nprint(f'Reported intensities {report}, Offset {offset}')",
"Reported intensities [100. 200. 300. 0.], Offset [0. 0. 0.1 0. ]\n"
]
],
[
[
"## MS2 Search",
"_____no_output_____"
]
],
[
[
"#export \nfrom typing import NamedTuple\nimport alphapept.io\n\ndef search_label_on_ms_file(file_name:str, label:NamedTuple, reporter_frag_tol:float, ppm:bool):\n \"\"\"Wrapper function to search labels on an ms_file and write results to the peptide_fdr of the file.\n\n Args:\n file_name (str): Path to ms_file:\n label (NamedTuple): Label with channels, mod_name and masses.\n reporter_frag_tol (float): Fragment tolerance for search.\n ppm (bool): Flag to use ppm instead of Dalton.\n\n \"\"\"\n\n ms_file = alphapept.io.MS_Data_File(file_name, is_read_only = False)\n \n df = ms_file.read(dataset_name='peptide_fdr')\n label_intensities = np.zeros((len(df), len(label.channels)))\n off_masses = np.zeros((len(df), len(label.channels)))\n labeled = df['sequence'].str.startswith(label.mod_name).values\n query_data = ms_file.read_DDA_query_data()\n\n query_indices = query_data[\"indices_ms2\"]\n query_frags = query_data['mass_list_ms2']\n query_ints = query_data['int_list_ms2']\n\n for idx, query_idx in enumerate(df['raw_idx']):\n\n query_idx_start = query_indices[query_idx]\n query_idx_end = query_indices[query_idx + 1]\n query_frag = query_frags[query_idx_start:query_idx_end]\n query_int = query_ints[query_idx_start:query_idx_end]\n\n query_frag_idx = query_frag < label.masses[-1]+1\n query_frag = query_frag[query_frag_idx]\n query_int = query_int[query_frag_idx]\n\n if labeled[idx]:\n label_int, off_mass = label_search(query_frag, query_int, label.masses, reporter_frag_tol, ppm)\n label_intensities[idx, :] = label_int\n off_masses[idx, :] = off_mass\n \n df[label.channels] = label_intensities\n df[[_+'_off_ppm' for _ in label.channels]] = off_masses\n \n ms_file.write(df, dataset_name=\"peptide_fdr\", overwrite=True) #Overwrite dataframe with label information\n",
"_____no_output_____"
],
[
"#export\nimport logging\nimport os\n\nfrom alphapept.constants import label_dict\n\ndef find_labels(\n to_process: dict,\n callback: callable = None,\n parallel:bool = False\n) -> bool:\n \"\"\"Wrapper function to search for labels.\n\n Args:\n to_process (dict): A dictionary with settings indicating which files are to be processed and how.\n callback (callable): A function that accepts a float between 0 and 1 as progress. Defaults to None.\n parallel (bool): If True, process multiple files in parallel.\n This is not implemented yet!\n Defaults to False.\n\n Returns:\n bool: True if and only if the label finding was succesful.\n\n \"\"\"\n index, settings = to_process\n raw_file = settings['experiment']['file_paths'][index]\n try:\n base, ext = os.path.splitext(raw_file)\n file_name = base+'.ms_data.hdf'\n label = label_dict[settings['isobaric_label']['label']]\n \n reporter_frag_tol = settings['isobaric_label']['reporter_frag_tolerance']\n ppm = settings['isobaric_label']['reporter_frag_tolerance_ppm']\n \n search_label_on_ms_file(file_name, label, reporter_frag_tol, ppm)\n \n logging.info(f'Tag finding of file {file_name} complete.')\n return True\n except Exception as e:\n logging.error(f'Tag finding of file {file_name} failed. Exception {e}')\n return f\"{e}\" #Can't return exception object, cast as string\n return True",
"_____no_output_____"
],
[
"#hide\nfrom nbdev.export import *\nnotebook2script()",
"Converted 00_settings.ipynb.\nConverted 01_chem.ipynb.\nConverted 02_io.ipynb.\nConverted 03_fasta.ipynb.\nConverted 04_feature_finding.ipynb.\nConverted 05_search.ipynb.\nConverted 06_score.ipynb.\nConverted 07_recalibration.ipynb.\nConverted 08_quantification.ipynb.\nConverted 09_matching.ipynb.\nConverted 10_constants.ipynb.\nConverted 11_interface.ipynb.\nConverted 12_performance.ipynb.\nConverted 13_export.ipynb.\nConverted 14_display.ipynb.\nConverted 15_label.ipynb.\nConverted additional_code.ipynb.\nConverted contributing.ipynb.\nConverted file_formats.ipynb.\nConverted index.ipynb.\nConverted minimal_stuff.ipynb.\nConverted Untitled.ipynb.\n"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
]
] |
cb920c0beab47a5f7a1fdf04c67a242405c7df70
| 10,034 |
ipynb
|
Jupyter Notebook
|
Tuesday's_Challenge.ipynb
|
coopwilliams/Neural_network_foundations_code_challenges
|
d3b5a75aa7e1559fd7436b964818e56d103a9d03
|
[
"MIT"
] | null | null | null |
Tuesday's_Challenge.ipynb
|
coopwilliams/Neural_network_foundations_code_challenges
|
d3b5a75aa7e1559fd7436b964818e56d103a9d03
|
[
"MIT"
] | null | null | null |
Tuesday's_Challenge.ipynb
|
coopwilliams/Neural_network_foundations_code_challenges
|
d3b5a75aa7e1559fd7436b964818e56d103a9d03
|
[
"MIT"
] | null | null | null | 24.413625 | 614 | 0.535978 |
[
[
[
"# For today's code challenge you will be reviewing yesterdays lecture material. Have fun!\n\n### if you get done early check out [these videos](https://www.3blue1brown.com/neural-networks).",
"_____no_output_____"
],
[
"# The Perceptron\n\nThe first and simplest kind of neural network that we could talk about is the perceptron. A perceptron is just a single node or neuron of a neural network with nothing else. It can take any number of inputs and spit out an output. What a neuron does is it takes each of the input values, multplies each of them by a weight, sums all of these products up, and then passes the sum through what is called an \"activation function\" the result of which is the final value.\n\nI really like figure 2.1 found in this [pdf](http://www.uta.fi/sis/tie/neuro/index/Neurocomputing2.pdf) even though it doesn't have bias term represented there.\n\n\n\nIf we were to write what is happening in some verbose mathematical notation, it might look something like this:\n\n\\begin{align}\n y = sigmoid(\\sum(weight_{1}input_{1} + weight_{2}input_{2} + weight_{3}input_{3}) + bias)\n\\end{align}\n\nUnderstanding what happens with a single neuron is important because this is the same pattern that will take place for all of our networks. \n\nWhen imagining a neural network I like to think about the arrows as representing the weights, like a wire that has a certain amount of resistance and only lets a certain amount of current through. And I like to think about the node itselef as containing the prescribed activation function that neuron will use to decide how much signal to pass onto the next layer.",
"_____no_output_____"
],
[
"# Activation Functions (transfer functions)\n\nIn Neural Networks, each node has an activation function. Each node in a given layer typically has the same activation function. These activation functions are the biggest piece of neural networks that have been inspired by actual biology. The activation function decides whether a cell \"fires\" or not. Sometimes it is said that the cell is \"activated\" or not. In Artificial Neural Networks activation functions decide how much signal to pass onto the next layer. This is why they are sometimes referred to as transfer functions because they determine how much signal is transferred to the next layer.\n\n## Common Activation Functions:\n\n",
"_____no_output_____"
],
[
"# Implementing a Perceptron from scratch in Python",
"_____no_output_____"
],
[
"### Establish training data",
"_____no_output_____"
]
],
[
[
"import numpy as np\n\nnp.random.seed(812)\n\ninputs = np.array([\n [0, 0, 1],\n [1, 1, 1],\n [1, 0, 1],\n [0, 1, 1]\n])\n\ncorrect_outputs = [[0], [1], [1], [0]]",
"_____no_output_____"
]
],
[
[
"### Sigmoid activation function and its derivative for updating weights",
"_____no_output_____"
]
],
[
[
"def sigmoid(x):\n return 1 / (1 + np.exp(-x))\n\ndef sigmoid_derivative(x):\n sx = sigmoid(x)\n return sx * (1 - sx)",
"_____no_output_____"
]
],
[
[
"## Updating weights with derivative of sigmoid function:\n\n",
"_____no_output_____"
],
[
"### Initialize random weights for our three inputs",
"_____no_output_____"
]
],
[
[
"weights = 2 * np.random.random((3, 1)) - 1",
"_____no_output_____"
],
[
"weights.shape",
"_____no_output_____"
],
[
"inputs.shape",
"_____no_output_____"
]
],
[
[
"### Calculate weighted sum of inputs and weights",
"_____no_output_____"
]
],
[
[
"weighted_sum = np.dot(inputs, weights)",
"_____no_output_____"
],
[
"weighted_sum",
"_____no_output_____"
]
],
[
[
"\n### Output the activated value for the end of 1 training epoch",
"_____no_output_____"
]
],
[
[
"activated_value = sigmoid(weighted_sum)\nactivated_value",
"_____no_output_____"
]
],
[
[
"### take difference of output and true values to calculate error",
"_____no_output_____"
]
],
[
[
"error = correct_outputs - activated_value\nerror\n",
"_____no_output_____"
]
],
[
[
"### Put it all together",
"_____no_output_____"
]
],
[
[
"adjustments = error * sigmoid_derivative(activated_value)\nadjustments, inputs.T",
"_____no_output_____"
],
[
"for i in range(10000):\n weighted_sum = np.dot(inputs, weights)\n activated_value = sigmoid(weighted_sum)\n error = correct_outputs - activated_value\n adjustments = error * sigmoid_derivative(activated_value)\n weights += np.dot(inputs.T, adjustments)\n\nprint(weights)\nprint(\"\\n-----\\n\", activated_value)",
"[[15.03804491]\n [-0.40666422]\n [-7.23278107]]\n\n-----\n [[7.22059439e-04]\n [9.99388204e-01]\n [9.99592541e-01]\n [4.80912067e-04]]\n"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
]
] |
cb9219c5e28c3cc7217ab02c1b4e6a8e2ed73db1
| 48,449 |
ipynb
|
Jupyter Notebook
|
201228-ThompsonSampling.ipynb
|
MartinTschendel/Reinforcement-Learning
|
b2a5fe00c692a27040e32888b05de1ebde4df7c5
|
[
"MIT"
] | null | null | null |
201228-ThompsonSampling.ipynb
|
MartinTschendel/Reinforcement-Learning
|
b2a5fe00c692a27040e32888b05de1ebde4df7c5
|
[
"MIT"
] | null | null | null |
201228-ThompsonSampling.ipynb
|
MartinTschendel/Reinforcement-Learning
|
b2a5fe00c692a27040e32888b05de1ebde4df7c5
|
[
"MIT"
] | null | null | null | 236.336585 | 11,464 | 0.925158 |
[
[
[
"import numpy as np\nimport matplotlib.pyplot as plt\nimport pandas as pd",
"_____no_output_____"
],
[
"dataset = pd.read_csv('Ads_CTR_Optimisation.csv')",
"_____no_output_____"
]
],
[
[
"<img src=\"Thompson_Sampling_Slide.png\" width=600 align=\"left\">",
"_____no_output_____"
]
],
[
[
"import random\ndef get_best_ad(N):\n d = 10\n ads_selected = []\n numbers_of_rewards_1 = [0] * d\n numbers_of_rewards_0 = [0] * d\n total_reward = 0\n #'n' represents the rounds\n for n in range(0, N):\n ad = 0\n #max_random refers to theta, the variable in step 2 and 3 of the algorithm\n max_random = 0\n for i in range(0, d):\n #step 2: take a random draw from the beta distribution\n random_beta = random.betavariate(numbers_of_rewards_1[i] + 1, numbers_of_rewards_0[i] + 1)\n #step 3: select the ad with the highest theta\n if random_beta > max_random:\n max_random = random_beta\n #if the case above applies, we should also make the 'ad' to become the index 'i'\n ad = i\n #we update the list 'ads_selected' with that ad, that has the highest theta\n ads_selected.append(ad)\n #the 'reward' picks out the respective values from the dataset (from round 'n' and specific 'ad')\n reward = dataset.values[n, ad]\n if reward == 1:\n numbers_of_rewards_1[ad] = numbers_of_rewards_1[ad] + 1\n else:\n numbers_of_rewards_0[ad] = numbers_of_rewards_0[ad] + 1\n total_reward = total_reward + reward\n return ads_selected",
"_____no_output_____"
],
[
"plt.hist(get_best_ad(10000))\nplt.title('Histogram of ads selections')\nplt.xlabel('Ads')\nplt.ylabel('Number of times each ad was selected')\nplt.show()",
"_____no_output_____"
],
[
"#check if algorithm can already identify best ad after fewer rounds (1000)\nplt.hist(get_best_ad(1000))\nplt.title('Histogram of ads selections')\nplt.xlabel('Ads')\nplt.ylabel('Number of times each ad was selected')\nplt.show()",
"_____no_output_____"
],
[
"#check if algorithm can already identify best ad after fewer rounds (500)\nplt.hist(get_best_ad(500))\nplt.title('Histogram of ads selections')\nplt.xlabel('Ads')\nplt.ylabel('Number of times each ad was selected')\nplt.show()",
"_____no_output_____"
],
[
"#check if algorithm can already identify best ad after fewer rounds (300)\nplt.hist(get_best_ad(300))\nplt.title('Histogram of ads selections')\nplt.xlabel('Ads')\nplt.ylabel('Number of times each ad was selected')\nplt.show()",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code"
] |
[
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
]
] |
cb921d679b2e608395a10b605af77dbd4174b356
| 1,445 |
ipynb
|
Jupyter Notebook
|
Generators.ipynb
|
profhall/python_jupyter
|
fce06a5f9ed8b884bfb4dcec88a368c59f5518e8
|
[
"MIT"
] | null | null | null |
Generators.ipynb
|
profhall/python_jupyter
|
fce06a5f9ed8b884bfb4dcec88a368c59f5518e8
|
[
"MIT"
] | null | null | null |
Generators.ipynb
|
profhall/python_jupyter
|
fce06a5f9ed8b884bfb4dcec88a368c59f5518e8
|
[
"MIT"
] | null | null | null | 16.802326 | 74 | 0.483045 |
[
[
[
"# Generators\nStrings are..\n\n",
"_____no_output_____"
]
],
[
[
"print(\"This Line is a python statement that produces an output\")\n",
"_____no_output_____"
]
],
[
[
"## String Methods\n\n",
"_____no_output_____"
]
],
[
[
"print(\"This Line is a python statement that produces an output\")",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb922d6f18c188cb669bb942af84effdde4437e2
| 57,844 |
ipynb
|
Jupyter Notebook
|
Random_Forest_titanic.ipynb
|
mehdijafarzadeh/Classification_Titanic
|
20cf2caa3cc1fda1edddec9a136b0324ea827375
|
[
"MIT"
] | null | null | null |
Random_Forest_titanic.ipynb
|
mehdijafarzadeh/Classification_Titanic
|
20cf2caa3cc1fda1edddec9a136b0324ea827375
|
[
"MIT"
] | null | null | null |
Random_Forest_titanic.ipynb
|
mehdijafarzadeh/Classification_Titanic
|
20cf2caa3cc1fda1edddec9a136b0324ea827375
|
[
"MIT"
] | null | null | null | 50.607174 | 27,580 | 0.647932 |
[
[
[
"# Random Forest\n\non the penguin dataset",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport pandas as pd\nimport seaborn as sns\nfrom matplotlib import pyplot as plt",
"_____no_output_____"
],
[
"from sklearn.compose import ColumnTransformer\nfrom sklearn.pipeline import make_pipeline\nfrom sklearn.impute import SimpleImputer\nfrom sklearn.preprocessing import OneHotEncoder\nfrom sklearn.preprocessing import KBinsDiscretizer\nfrom sklearn.preprocessing import MinMaxScaler",
"_____no_output_____"
],
[
"from sklearn.linear_model import LogisticRegression\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import accuracy_score\nfrom sklearn.ensemble import RandomForestClassifier\n",
"_____no_output_____"
]
],
[
[
"### Preparations",
"_____no_output_____"
]
],
[
[
"# let's bring in the data and get rid of the Nans\n\ndf = pd.read_csv('./data/train.csv')\n# df.dropna(inplace=True)",
"_____no_output_____"
]
],
[
[
"#### 1. Inspect the size of the dataset",
"_____no_output_____"
]
],
[
[
"df.head()",
"_____no_output_____"
],
[
"df.isna().sum()",
"_____no_output_____"
],
[
"# sns.scatterplot(x=df['Age'], y=df['Pclass'], hue=df['Sex'])Second ways to do the same!\n\nsns.scatterplot(x='Age', y='Pclass', hue='Sex', data= df) \n",
"_____no_output_____"
],
[
"y = df['Survived']\nX = df[['Pclass', 'Age', 'Sex' ]]",
"_____no_output_____"
],
[
"X_train, X_test, y_train, y_test = train_test_split(X, y)",
"_____no_output_____"
],
[
"X_train.shape, X_test.shape",
"_____no_output_____"
],
[
"X_train.head()",
"_____no_output_____"
]
],
[
[
"## pipline",
"_____no_output_____"
]
],
[
[
"impute_and_MinMaxScaler = make_pipeline(\n SimpleImputer(strategy='most_frequent'), \n MinMaxScaler()\n)",
"_____no_output_____"
]
],
[
[
"## Column Transformer",
"_____no_output_____"
]
],
[
[
"fe = ColumnTransformer([\n # (name, transformer, column-names)\n # ('do-nothing', 'passthrough', ['culmen_length_mm', 'body_mass_g']),\n \n ('imputation, scaling', impute_and_MinMaxScaler, ['Age']),\n# ('imputation', SimpleImputer(), ['Age']),\n# ('binning', KBinsDiscretizer(n_bins=3, encode='onehot-dense'), ['Sex']),\n ('one-hot-encode', OneHotEncoder(), ['Sex', 'Pclass']),\n# ('one-hot', OneHotEncoder, ['Age']),\n \n # DON'T DO THIS:\n # ('one-hot-encode', OneHotEncoder(sparse=False, handle_unknown='ignore'), ['body_mass_g'])\n])\n\nfe",
"_____no_output_____"
],
[
"fe.fit(X_train)\n\n# transform the training data\nX_train_trans = fe.transform(X_train)\npd.DataFrame(X_train_trans)",
"_____no_output_____"
]
],
[
[
"## Find the optimal separation with Scikit",
"_____no_output_____"
],
[
"#### 7. Train the model",
"_____no_output_____"
]
],
[
[
"from sklearn.tree import DecisionTreeClassifier, plot_tree",
"_____no_output_____"
],
[
"\n# initialize and train model\n# m = DecisionTreeClassifier(max_depth=2)\nm= RandomForestClassifier(\n n_estimators=100, # number of decision trees in the forest\n max_depth=4 # depth of each tree\n)\n\nm.fit(X_train_trans, y_train)",
"_____no_output_____"
]
],
[
[
"## predictions",
"_____no_output_____"
]
],
[
[
"y_pred_train = m.predict(X_train_trans)\naccuracy_score(y_train, y_pred_train)",
"_____no_output_____"
]
],
[
[
"#### Calculate the accuracy",
"_____no_output_____"
]
],
[
[
"m.classes_",
"_____no_output_____"
],
[
"titanic_kaggel = pd.read_csv('./data/test.csv', index_col = 0)\ntitanic_kaggel.head()",
"_____no_output_____"
],
[
"X_kaggel = titanic_kaggel[['Pclass', 'Age', 'Sex']]\nX_kaggel_trans = fe.transform(X_kaggel)",
"_____no_output_____"
],
[
"y_pred_kaggel = m.predict(X_kaggel_trans)\ntype(y_pred_kaggel)",
"_____no_output_____"
],
[
"\nsubmission = pd.DataFrame(y_pred_kaggel, index=X_kaggel.index, columns = ['Survived'])\n\n\nsubmission.to_csv('submission3.csv', index=True)\n\nsubmission",
"_____no_output_____"
],
[
"# plt.figure(figsize=(10, 12))\n# t = plot(m, feature_names=['Age','female', 'male', 'class1', 'class2','class3'], class_names=['Non_survived','Survived'])",
"_____no_output_____"
],
[
"fe.named_transformers_['one-hot-encode'].get_feature_names()",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb923003a47d310531da9e915c4d199ccca9d519
| 4,919 |
ipynb
|
Jupyter Notebook
|
Examples/PE_Production_2.ipynb
|
ChampionApe/GamsPythonModels
|
aaa234b2627cda2b92e478e8e8503bf9778aebeb
|
[
"MIT"
] | null | null | null |
Examples/PE_Production_2.ipynb
|
ChampionApe/GamsPythonModels
|
aaa234b2627cda2b92e478e8e8503bf9778aebeb
|
[
"MIT"
] | null | null | null |
Examples/PE_Production_2.ipynb
|
ChampionApe/GamsPythonModels
|
aaa234b2627cda2b92e478e8e8503bf9778aebeb
|
[
"MIT"
] | null | null | null | 23.312796 | 392 | 0.541167 |
[
[
[
"# **Static, partial equilibrium model**\n### **Production module, example 2:**\n\n*In some cases, it can be usefull to allow for a sector to use the same type of inputs more than once in the nesting tree (see the AbatementProject for an example). In such a case, the prices and quantities of the sector are defined over different, but overlapping sets. This example shows how to adjust the nesting trees, and subsequently the production module to account for this.*\n\n*In this example, a sector uses $E_i$-types of input with $i\\in\\lbrace 1,2\\rbrace$, and $B_j$-types with $j\\in \\lbrace 1,2\\rbrace$. However, $B_j$ are utilized under both $E_i$ branches. Thus there are components of the form 'Ei_Bj' that mesaures the use of $B_j$ under branch $E_i$.*",
"_____no_output_____"
],
[
"Import standard collection of packages:",
"_____no_output_____"
]
],
[
[
"clean_up=True # removes gams-related files in work-folder if true\n%run StdPackages.ipynb",
"The file_gams_py_gdb1.gdx is still active and was not deleted.\n"
]
],
[
[
"## **Nesting structure:**",
"_____no_output_____"
],
[
"*Nesting trees:*",
"_____no_output_____"
]
],
[
[
"tree_in = {'Y': ['E1', 'E2'],\n 'E1': ['E1_B1','E1_B2','K'],\n 'E2': ['E2_B1','E2_B2']}\ntree_out = {'Y': ['Y1','Y2']}",
"_____no_output_____"
]
],
[
[
"*Initialize empty class with some name:*",
"_____no_output_____"
]
],
[
[
"nt = nesting_tree.nesting_tree(name='test')",
"_____no_output_____"
]
],
[
[
"*Add trees to the class. Note that the 'input' tree is now of version Q2P, whereas the output-type is standard ('std'):*",
"_____no_output_____"
]
],
[
[
"nt.add_tree(tree_in,'in',**{'tree_name': 'test_in','version': 'Q2P'})\nnt.add_tree(tree_out,'out',**{'tree_name': 'test_out','type_io': 'output','version': 'std'})",
"_____no_output_____"
]
],
[
[
"*Define the Q2P settings. This includes a multiindex (mapping) from Ei_Bj types to Bj types, included in a dictionary for the trees:*",
"_____no_output_____"
]
],
[
[
"q2p = pd.MultiIndex.from_tuples([('E1_B1', 'B1'), ('E1_B2', 'B2'),('E2_B1', 'B1'), ('E2_B2','B2')],names=['n','nn'])\nQ2Ps = {'in': q2p}",
"_____no_output_____"
]
],
[
[
"*Run all, with Q2Ps settings:*",
"_____no_output_____"
]
],
[
[
"nt.run_all(Q2Ps=Q2Ps)",
"_____no_output_____"
]
],
[
[
"## **Create production module:**",
"_____no_output_____"
],
[
"*Partial equilibrium model, production module from nesting tree:*",
"_____no_output_____"
]
],
[
[
"pm = PE.GPM_STA_PE.production(nt,**{'work_folder': work_folder})",
"_____no_output_____"
]
],
[
[
"*Write gams files to a repo called TESTS/example_1, and export the 'settings' as well:*",
"_____no_output_____"
]
],
[
[
"pm.df_write(repo=os.getcwd()+'\\\\PE_P\\\\Example_2')",
"_____no_output_____"
]
],
[
[
"*Create model instance and run:*",
"_____no_output_____"
]
],
[
[
"pm.create_model_instance()\npm.df_run()",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb926dd9b94a1ad2a22fb7efe0bf721542ddf0f6
| 4,489 |
ipynb
|
Jupyter Notebook
|
code/Data prep - download eras5.ipynb
|
goldford/NEMO-Salish-Sea-2021
|
c6a78956293f1741ff9537747dd02ce37c20a4d3
|
[
"MIT"
] | null | null | null |
code/Data prep - download eras5.ipynb
|
goldford/NEMO-Salish-Sea-2021
|
c6a78956293f1741ff9537747dd02ce37c20a4d3
|
[
"MIT"
] | null | null | null |
code/Data prep - download eras5.ipynb
|
goldford/NEMO-Salish-Sea-2021
|
c6a78956293f1741ff9537747dd02ce37c20a4d3
|
[
"MIT"
] | null | null | null | 34.530769 | 255 | 0.484072 |
[
[
[
"# follow instructions here: \n#insert these two lines into .cdsapirc file, likely into c:/users/username/\n#url: https://cds.climate.copernicus.eu/api/v2\n#key: 51075:b42dad06-8bbd-4987-855f-c3a4a851d26a\n",
"_____no_output_____"
],
[
"import cdsapi\n\nc = cdsapi.Client()\n\nc.retrieve(\n 'reanalysis-era5-single-levels',\n {\n 'product_type': 'reanalysis',\n 'variable': [\n '10m_u_component_of_wind', '10m_v_component_of_wind', '2m_dewpoint_temperature',\n '2m_temperature', 'mean_snowfall_rate', 'mean_surface_downward_long_wave_radiation_flux',\n 'mean_surface_downward_short_wave_radiation_flux', 'mean_total_precipitation_rate', 'surface_pressure',\n ],\n 'year': '2015',\n 'month': [\n '01', '02', '03',\n '04', '05', '06',\n '07', '08', '09',\n '10', '11', '12',\n ],\n 'day': [\n '01', '02', '03',\n '04', '05', '06',\n '07', '08', '09',\n '10', '11', '12',\n '13', '14', '15',\n '16', '17', '18',\n '19', '20', '21',\n '22', '23', '24',\n '25', '26', '27',\n '28', '29', '30',\n '31',\n ],\n 'time': [\n '00:00', '01:00', '02:00',\n '03:00', '04:00', '05:00',\n '06:00', '07:00', '08:00',\n '09:00', '10:00', '11:00',\n '12:00', '13:00', '14:00',\n '15:00', '16:00', '17:00',\n '18:00', '19:00', '20:00',\n '21:00', '22:00', '23:00',\n ],\n 'area': [\n 51.07, -127, 46.73,\n -120.7,\n ],\n 'format': 'netcdf',\n },\n 'ERA5_SalishSea_9vars_2015.nc')",
"2021-05-27 22:37:50,543 INFO Welcome to the CDS\n2021-05-27 22:37:50,552 INFO Sending request to https://cds.climate.copernicus.eu/api/v2/resources/reanalysis-era5-single-levels\n2021-05-27 22:37:50,725 INFO Request is queued\n2021-05-27 23:08:29,075 INFO Request is running\n2021-05-28 00:35:23,756 WARNING Recovering from connection error [('Connection aborted.', OSError(\"(10054, 'WSAECONNRESET')\"))], attemps 0 of 500\n2021-05-28 00:35:23,757 WARNING Retrying in 120 seconds\n2021-05-28 00:37:23,770 INFO Retrying now...\n2021-05-28 00:37:25,070 INFO Request is completed\n2021-05-28 00:37:25,082 INFO Downloading https://download-0005.copernicus-climate.eu/cache-compute-0005/cache/data3/adaptor.mars.internal-1622181999.5896902-24906-3-c3e97795-7d25-4a73-91af-9fda140cc341.nc to ERA5_SalishSea_9vars_2015.nc (70.4M)\n2021-05-28 00:37:56,378 INFO Download rate 2.3M/s \n"
]
]
] |
[
"code"
] |
[
[
"code",
"code"
]
] |
cb92942d47489c5102cb455dc88ca95865aac13e
| 367,148 |
ipynb
|
Jupyter Notebook
|
archives/baseline_linear.ipynb
|
maks-p/final_project_bi
|
c5e1d38682541c028ca15ef7e3650035309a582d
|
[
"MIT"
] | 8 |
2021-01-15T17:14:24.000Z
|
2021-12-30T06:27:10.000Z
|
archives/baseline_linear.ipynb
|
maks-p/final_project_bi
|
c5e1d38682541c028ca15ef7e3650035309a582d
|
[
"MIT"
] | 1 |
2021-03-23T12:32:23.000Z
|
2021-03-23T12:32:23.000Z
|
archives/baseline_linear.ipynb
|
maks-p/final_project_bi
|
c5e1d38682541c028ca15ef7e3650035309a582d
|
[
"MIT"
] | 8 |
2020-03-05T01:18:18.000Z
|
2021-12-14T17:08:50.000Z
| 93.875735 | 94,356 | 0.778095 |
[
[
[
"import numpy as np \nimport pandas as pd \nimport warnings\nwarnings.filterwarnings('ignore')",
"_____no_output_____"
],
[
"import seaborn as sns\nsns.set_palette('Set2')\n\nimport matplotlib.pyplot as plt\n%matplotlib inline",
"_____no_output_____"
],
[
"from sklearn.metrics import confusion_matrix, mean_squared_error\nfrom sklearn.preprocessing import LabelEncoder, MinMaxScaler, StandardScaler\nfrom sklearn.model_selection import train_test_split, cross_val_score\nfrom sklearn.linear_model import LinearRegression, Lasso, Ridge, SGDRegressor\nfrom sklearn.preprocessing import KBinsDiscretizer\nfrom sklearn.ensemble import RandomForestRegressor\nfrom sklearn.svm import LinearSVC, SVC, LinearSVR, SVR\nfrom sklearn.tree import DecisionTreeRegressor\nfrom sklearn.model_selection import GridSearchCV\nfrom scipy.stats import zscore\nfrom sklearn.metrics import mean_squared_error",
"_____no_output_____"
],
[
"import requests\nimport json\nfrom datetime import datetime\nimport time",
"_____no_output_____"
],
[
"import os\nfrom pandas.tseries.holiday import USFederalHolidayCalendar as calendar",
"_____no_output_____"
],
[
"from config import yelp_api_key\nfrom config import darksky_api_key",
"_____no_output_____"
]
],
[
[
"## Set Up",
"_____no_output_____"
]
],
[
[
"# Analysis Dates\nstart_date = '2017-01-01' # Start Date Inclusive\nend_date = '2019-06-19' # End Date Exclusive",
"_____no_output_____"
],
[
"search_business = 'Jupiter Disco'\nlocation = 'Brooklyn, NY'",
"_____no_output_____"
]
],
[
[
"## Pull Weather Data",
"_____no_output_____"
],
[
"### Latitude + Longitude from Yelp API",
"_____no_output_____"
]
],
[
[
"host = 'https://api.yelp.com'\npath = '/v3/businesses/search'\n\nsearch_limit = 10\n\n# Yelp Authorization Header with API Key\nheaders = {\n 'Authorization': 'Bearer {}'.format(yelp_api_key) \n }\n\n# Build Requests Syntax with Yelp Host and Path and URL Paramaters\n# Return JSON response\ndef request(host, path, url_params=None):\n \n url_params = url_params or {}\n url = '{}{}'.format(host, path)\n\n response = requests.get(url, headers=headers, params=url_params)\n \n return response.json()\n\n# Build URL Params for the Request and provide the host and path\ndef search(term, location):\n \n url_params = {\n 'term': term.replace(' ', '+'),\n 'location': location.replace(' ', '+'),\n 'limit': search_limit\n }\n \n return request(host, path, url_params=url_params)\n\n# Return Coordinates if Exact Match Found\ndef yelp_lat_long(business, location):\n \n # Call search function here with business name and location\n response = search(business, location)\n \n # Set state to 'No Match' in case no Yelp match found\n state = 'No Match'\n possible_matches = []\n \n # Check search returns for match wtith business\n for i in range(len(response['businesses'])):\n\n # If match found:\n if response['businesses'][i]['name'] == business:\n\n # Local variables to help navigate JSON return\n response_ = response['businesses'][0]\n name_ = response_['name']\n\n print(f'Weather Location: {name_}')\n state = 'Match Found'\n #print(response['businesses'][0])\n return response_['coordinates']['latitude'], response_['coordinates']['longitude']\n\n else:\n \n # If no exact match, append all search returns to list\n possible_matches.append(response['businesses'][i]['name'])\n \n # If no match, show user potential matches\n if state == 'No Match':\n \n print('Exact match not found, did you mean one of the following? \\n')\n \n for possible_match in possible_matches:\n print(possible_match)\n \n return None, None\n\nlat, long = yelp_lat_long(search_business, location)\n#print(f'Latitude: {lat}\\nLongitude: {long}')",
"Weather Location: Jupiter Disco\n"
]
],
[
[
"### Darksky API Call",
"_____no_output_____"
]
],
[
[
"# Create List of Dates of target Weather Data\ndef find_dates(start_date, end_date):\n \n list_of_days = []\n daterange = pd.date_range(start_date, end_date)\n for single_date in daterange:\n list_of_days.append(single_date.strftime(\"%Y-%m-%d\"))\n \n return list_of_days",
"_____no_output_____"
],
[
"# Concatenate URL to make API Call\ndef build_url(api_key, lat, long, day):\n \n _base_url = 'https://api.darksky.net/forecast/'\n _time = 'T20:00:00'\n _url = f'{_base_url}{api_key}/{lat},{long},{day + _time}?America/New_York&exclude=flags'\n return _url\n\ndef make_api_call(url):\n \n r = requests.get(url)\n \n return r.json()",
"_____no_output_____"
],
[
"# Try / Except Helper Function for Handling JSON API Output\ndef find_val(dictionary, *keys):\n\n level = dictionary\n \n for key in keys:\n \n try:\n level = level[key]\n \n except:\n return np.NAN\n \n return level\n\n# Parse API Call Data using Try / Except Helper Function\ndef parse_data(data):\n\n time = datetime.fromtimestamp(data['currently']['time']).strftime('%Y-%m-%d')\n \n try:\n precip_max_time = datetime.fromtimestamp(find_val(data, 'daily', 'data', 0, 'precipIntensityMaxTime')).strftime('%I:%M%p')\n \n except:\n precip_max_time = datetime(1900,1,1,5,1).strftime('%I:%M%p')\n \n entry = {'date': time,\n 'temperature': float(find_val(data, 'currently', 'temperature')),\n 'apparent_temperature': float(find_val(data, 'currently', 'apparentTemperature')),\n 'humidity': float(find_val(data, 'currently', 'humidity')),\n 'precip_intensity_max': float(find_val(data,'daily','data', 0, 'precipIntensityMax')),\n 'precip_type': find_val(data, 'daily', 'data', 0, 'precipType'),\n 'precip_prob': float(find_val(data, 'currently', 'precipProbability')),\n 'pressure': float(find_val(data, 'currently', 'pressure')),\n 'summary': find_val(data, 'currently', 'icon'),\n 'precip_max_time': precip_max_time}\n \n return entry",
"_____no_output_____"
],
[
"# Create List of Weather Data Dictionaries & Input Target Dates\ndef weather_call(start_date, end_date, _lat, _long):\n \n weather = []\n list_of_days = find_dates(start_date, end_date)\n \n for day in list_of_days:\n \n data = make_api_call(build_url(darksky_api_key, _lat, _long, day))\n \n weather.append(parse_data(data))\n \n return weather\n\nresult = weather_call(start_date, end_date, lat, long)",
"_____no_output_____"
],
[
"# Build DataFrame from List of Dictionaries\ndef build_weather_df(api_call_results):\n\n df = pd.DataFrame(api_call_results)\n\n # Add day of week to DataFrame + Set Index as date\n df['date'] = pd.to_datetime(df['date'])\n df['day_of_week'] = df['date'].dt.weekday\n df['month'] = df['date'].dt.month\n \n df.set_index('date', inplace=True)\n \n df['apparent_temperature'].fillna(method='ffill',inplace=True)\n df['temperature'].fillna(method='ffill',inplace=True)\n df['humidity'].fillna(method='ffill',inplace=True)\n df['precip_prob'].fillna(method='ffill', inplace=True)\n df['pressure'].fillna(method='ffill', inplace=True)\n df['precip_type'].fillna(value='none', inplace=True)\n \n return df\n\nweather_df = build_weather_df(result);",
"_____no_output_____"
],
[
"weather_df.to_csv(f'weather_{start_date}_to_{end_date}.csv')\nweather_csv_file = f'weather_{start_date}_to_{end_date}.csv'",
"_____no_output_____"
]
],
[
[
"## Import / Clean / Prep File",
"_____no_output_____"
]
],
[
[
"# Import Sales Data\nbar_sales_file = 'bar_x_sales_export.csv'\nrest_1_file = 'rest_1_dinner_sales_w_covers_061819.csv'",
"_____no_output_____"
],
[
"# Set File\ncurrent_file = rest_1_file\nweather_csv_file = 'weather_2017-01-01_to_2019-06-19.csv'",
"_____no_output_____"
],
[
"# HELPER FUNCTION\ndef filter_df(df, start_date, end_date):\n \n return df[(df.index > start_date) & (df.index < end_date)]",
"_____no_output_____"
],
[
"# HELPER FUNCTION\ndef import_parse(file):\n\n data = pd.read_csv(file, index_col = 'date', parse_dates=True)\n df = pd.DataFrame(data)\n \n # Rename Column to 'sales'\n df = df.rename(columns={df.columns[0]: 'sales',\n 'dinner_covers': 'covers'})\n \n # Drop NaN\n #df = df.query('sales > 0').copy()\n df.fillna(0, inplace=True)\n \n print(f'\"{file}\" has been imported + parsed. The file has {len(df)} rows.')\n \n return df",
"_____no_output_____"
],
[
"# HELPER FUNCTION\ndef prepare_data(current_file, weather_file):\n \n df = filter_df(import_parse(current_file), start_date, end_date)\n weather_df_csv = pd.read_csv(weather_csv_file, parse_dates=True, index_col='date')\n weather_df_csv['summary'].fillna(value='none', inplace=True)\n \n df = pd.merge(df, weather_df_csv, how='left', on='date')\n \n return df",
"_____no_output_____"
]
],
[
[
"### Encode Closed Days",
"_____no_output_____"
]
],
[
[
"# Set Closed Dates using Sales\n\n## REST 1 CLOSED DATES\nadditional_closed_dates = ['2018-12-24', '2017-12-24', '2017-02-05', '2017-03-14', '2018-01-01', '2018-02-04', '2019-02-03']\n\n## BAR CLOSED DATES\n#additional_closed_dates = ['2018-12-24', '2017-12-24', '2017-10-22']\n\nclosed_dates = [pd.to_datetime(date) for date in additional_closed_dates]\n\n\n# Drop or Encode Closed Days\ndef encode_closed_days(df):\n\n # CLOSED FEATURE\n cal = calendar()\n\n # Local list of days with zero sales\n potential_closed_dates = df[df['sales'] == 0].index\n\n # Enocodes closed days with 1\n df['closed'] = np.where((((df.index.isin(potential_closed_dates)) & \\\n (df.index.isin(cal.holidays(start_date, end_date)))) | df.index.isin(closed_dates)), 1, 0)\n\n df['sales'] = np.where(df['closed'] == 1, 0, df['sales'])\n \n return df",
"_____no_output_____"
],
[
"baseline_df = encode_closed_days(prepare_data(current_file, weather_csv_file))",
"\"rest_1_dinner_sales_w_covers_061819.csv\" has been imported + parsed. The file has 899 rows.\n"
],
[
"baseline_df = baseline_df[['sales', 'outside', 'day_of_week', 'month', 'closed']]\nbaseline_df = add_dummies(add_clusters(baseline_df))\n",
"_____no_output_____"
],
[
"mod_baseline = target_trend_engineering(add_cal_features(impute_outliers(baseline_df)))",
"_____no_output_____"
],
[
"mod_baseline = mod_baseline.drop(['month'], axis=1)",
"_____no_output_____"
]
],
[
[
"### Replace Outliers in Training Data",
"_____no_output_____"
]
],
[
[
"# Replace Outliers with Medians\n## Targets for Outliers\n\nz_thresh = 3\n\ndef impute_outliers(df, *col):\n \n # Check for Outliers in Sales + Covers\n for c in col:\n \n # Impute Median for Sales & Covers Based on Day of Week Outiers\n for d in df['day_of_week'].unique():\n \n # Median / Mean / STD for each day of the week\n daily_median = np.median(df[df['day_of_week'] == d][c])\n daily_mean = np.mean(df[df['day_of_week'] == d][c])\n daily_std = np.std(df[df['day_of_week'] ==d ][c])\n \n # Temporary column encoded if Target Columns have an Outlier\n df['temp_col'] = np.where((df['day_of_week'] == d) & (df['closed'] == 0) & ((np.abs(df[c] - daily_mean)) > (daily_std * z_thresh)), 1, 0)\n \n # Replace Outlier with Median\n df[c] = np.where(df['temp_col'] == 1, daily_median, df[c])\n df = df.drop(['temp_col'], axis=1)\n \n \n return df",
"_____no_output_____"
],
[
"def add_ppa(df):\n \n df['ppa'] = np.where(df['covers'] > 0, df['sales'] / df['covers'], 0)\n \n return df",
"_____no_output_____"
]
],
[
[
"## Clean File Here",
"_____no_output_____"
]
],
[
[
"data = add_ppa(impute_outliers(encode_closed_days(prepare_data(current_file, weather_csv_file)), 'sales', 'covers'))",
"\"rest_1_dinner_sales_w_covers_061819.csv\" has been imported + parsed. The file has 899 rows.\n"
]
],
[
[
"### Download CSV for EDA",
"_____no_output_____"
]
],
[
[
"data.to_csv('CSV_for_EDA.csv')",
"_____no_output_____"
],
[
"df_outside = data['outside']",
"_____no_output_____"
]
],
[
[
"## CHOOSE TARGET --> SALES OR COVERS",
"_____no_output_____"
]
],
[
[
"target = 'sales'",
"_____no_output_____"
],
[
"def daily_average_matrix_ann(df, target):\n \n matrix = df.groupby([df.index.dayofweek, df.index.month, df.index.year]).agg({target: 'mean'})\n matrix = matrix.rename_axis(['day', 'month', 'year'])\n return matrix.unstack(level=1)\n\ndaily_average_matrix_ann(data, target)",
"_____no_output_____"
]
],
[
[
"### Create Month Clusters",
"_____no_output_____"
]
],
[
[
"from sklearn.cluster import KMeans\n\nday_k = 7\nmo_k = 3\n\ndef create_clusters(df, target, col, k):\n \n # MAKE DATAFRAME USING CENTRAL TENDENCIES AS FEATURES\n describe = df.groupby(col)[target].aggregate(['median', 'std', 'max'])\n df = describe.reset_index()\n \n # SCALE TEMPORARY DF\n scaler = MinMaxScaler()\n f = scaler.fit_transform(df)\n \n # INSTANTIATE MODEL\n km = KMeans(n_clusters=k, random_state=0).fit(f)\n \n # GET KMEANS CLUSTER PREDICTIONS\n labels = km.predict(f)\n \n # MAKE SERIES FROM PREDICTIONS\n temp = pd.DataFrame(labels, columns = ['cluster'], index=df.index)\n \n # CONCAT CLUSTERS TO DATAFRAME\n df = pd.concat([df, temp], axis=1)\n \n # CREATE CLUSTER DICTIONARY\n temp_dict = {}\n for i in list(df[col]):\n \n temp_dict[i] = df.loc[df[col] == i, 'cluster'].iloc[0]\n \n return temp_dict\n\n# Create Global Dictionaries to Categorize Day / Month\n#day_dict = create_clusters(data, 'day_of_week', day_k)\nmonth_dict = create_clusters(data, target, 'month', mo_k)",
"_____no_output_____"
],
[
"# Print Clusters\n#print('Day Clusters: ', day_dict, '\\n', 'Total Clusters: ', len(set(day_dict.values())), '\\n')\nprint('Month Clusters: ', month_dict, '\\n', 'Total Clusters: ', len(set(month_dict.values())))",
"Month Clusters: {1: 2, 2: 2, 3: 2, 4: 2, 5: 0, 6: 0, 7: 0, 8: 0, 9: 0, 10: 1, 11: 1, 12: 1} \n Total Clusters: 3\n"
]
],
[
[
"### Add Temperature Onehot Categories",
"_____no_output_____"
]
],
[
[
"def encode_temp(df):\n\n temp_enc = KBinsDiscretizer(n_bins=5, encode='onehot', strategy='kmeans')\n temp_enc.fit(df[['apparent_temperature']])\n \n return temp_enc\n \ndef one_hot_temp(df, temp_enc):\n \n binned_transform = temp_enc.transform(df[['apparent_temperature']])\n binned_df = pd.DataFrame(binned_transform.toarray(), index=df.index, columns=['temp_very_cold', 'temp_cold', 'temp_warm', 'temp_hot', 'temp_very_hot'])\n df = df.merge(binned_df, how='left', on='date')\n df.drop(['apparent_temperature', 'temperature'], axis=1, inplace=True)\n\n return df, temp_enc",
"_____no_output_____"
]
],
[
[
"## Feature Engineering",
"_____no_output_____"
]
],
[
[
"# Add Clusters to DataFrame to use as Features\ndef add_clusters(df):\n \n #df['day_cluster'] = df['day_of_week'].apply(lambda x: day_dict[x]).astype('category')\n df['month_cluster'] = df['month'].apply(lambda x: month_dict[x]).astype('category')\n \n return df",
"_____no_output_____"
]
],
[
[
"### Add Weather Features",
"_____no_output_____"
]
],
[
[
"hours_start = '05:00PM'\nhours_end = '11:59PM'\n\nhs_dt = datetime.strptime(hours_start, \"%I:%M%p\")\nhe_dt = datetime.strptime(hours_end, \"%I:%M%p\")\n\ndef between_time(check_time):\n \n if hs_dt <= datetime.strptime(check_time, \"%I:%M%p\") <= he_dt:\n\n return 1\n\n else:\n\n return 0",
"_____no_output_____"
],
[
"add_weather = True\n\ntemp_delta_window = 1\n\ndef add_weather_features(df):\n \n if add_weather:\n \n # POOR WEATHER FEATURES\n df['precip_while_open'] = df['precip_max_time'].apply(lambda x: between_time(x))\n \n # DROP FEATURES\n features_to_drop = ['precip_max_time']\n df.drop(features_to_drop, axis=1, inplace=True)\n \n return df",
"_____no_output_____"
]
],
[
[
"### Add Calendar Features",
"_____no_output_____"
]
],
[
[
"def add_cal_features(df):\n \n cal = calendar()\n\n # THREE DAY WEEKEND FEATURE\n sunday_three_days = [date + pd.DateOffset(-1) for date in cal.holidays(start_date, end_date) if date.dayofweek == 0]\n df['sunday_three_day'] = np.where(df.index.isin(sunday_three_days), 1, 0)\n \n return df",
"_____no_output_____"
]
],
[
[
"### Add Dummies",
"_____no_output_____"
]
],
[
[
"def add_dummies(df):\n \n df['day_of_week'] = df['day_of_week'].astype('category')\n \n df = pd.get_dummies(data=df, columns=['day_of_week', 'month_cluster'])\n \n return df",
"_____no_output_____"
]
],
[
[
"### Add Interactions",
"_____no_output_____"
]
],
[
[
"def add_interactions(df):\n \n apply_this_interaction = False\n \n if apply_this_interaction:\n \n for d in [col for col in df.columns if col.startswith('day_cluster')]:\n \n for m in [col for col in df.columns if col.startswith('month_cluster')]:\n \n col_name = d + '_X_' + m\n \n df[col_name] = df[d] * df[m]\n \n df.drop([d], axis=1, inplace=True)\n \n df.drop([col for col in df.columns if col.startswith('month_cluster')], axis=1, inplace=True)\n \n return df\n \n else:\n \n return df",
"_____no_output_____"
],
[
"def add_weather_interactions(df):\n \n apply_this_interaction = True\n \n if apply_this_interaction:\n \n try:\n \n df['outside_X_precip_open'] = df['outside'] * df['precip_while_open']\n \n \n for w in [col for col in df.columns if col.startswith('temp_')]:\n\n col_name = w + '_X_' + 'outside'\n df[col_name] = df[w] * df['outside']\n\n df.drop(['outside'], axis=1, inplace=True)\n \n except:\n \n pass\n \n return df\n \n else:\n \n return df",
"_____no_output_____"
]
],
[
[
"### Feature Selection",
"_____no_output_____"
]
],
[
[
"def feature_selection(df):\n \n try:\n target_list = ['sales', 'covers', 'ppa']\n\n target_to_drop = [t for t in target_list if t != target]\n\n df = df.drop(target_to_drop, axis=1)\n \n except:\n pass\n \n # Feature Selection / Drop unnecessary or correlated columns\n cols_to_drop = ['month', 'precip_type', 'summary', 'pressure', 'precip_intensity_max', 'day_of_week_0']\n \n df = df.drop(cols_to_drop, axis=1)\n \n return df",
"_____no_output_____"
]
],
[
[
"### Add Target Trend Feature Engineering",
"_____no_output_____"
]
],
[
[
"trend_days_rolling = 31\ntrend_days_shift = 7\ndays_fwd = trend_days_rolling + trend_days_shift + 1\n\ndef target_trend_engineering(df):\n \n df['target_trend'] = df[target].rolling(trend_days_rolling).mean() / df[target].shift(trend_days_shift).rolling(trend_days_rolling).mean()\n #df['target_delta'] = df[target].shift(7) + df[target].shift(14) - df[target].shift(21) - df[target].shift(28)\n \n return df",
"_____no_output_____"
]
],
[
[
"## Start Here",
"_____no_output_____"
]
],
[
[
"# IMPORT & PARSE CLEAN TRAINING SET\ndata = add_ppa(impute_outliers(encode_closed_days(prepare_data(current_file, weather_csv_file)), 'sales', 'covers'));",
"\"rest_1_dinner_sales_w_covers_061819.csv\" has been imported + parsed. The file has 899 rows.\n"
],
[
"# One Hot Encode Temperature Data\ndata, temp_enc = one_hot_temp(data, encode_temp(data))",
"_____no_output_____"
],
[
"# Create CSV\ndata.to_csv('csv_before_features.csv')",
"_____no_output_____"
],
[
"def feature_engineering(df):\n \n df.columns = df.columns.map(str)\n \n # Add day & Month Clusters // Dicts with data held in Global Variable\n df = add_clusters(df)\n \n # Add Engineered Features for Weather & Calendar\n df = add_weather_features(df)\n df = add_cal_features(df)\n \n # Create Dummies\n df = add_dummies(df)\n \n # Add Interactions\n df = add_interactions(df)\n df = add_weather_interactions(df)\n \n # Drop Selected Columns\n df = feature_selection(df)\n \n return df\n\ndfx = feature_engineering(data)",
"_____no_output_____"
],
[
"dfx = target_trend_engineering(dfx)",
"_____no_output_____"
],
[
"def corr_chart(df):\n \n corr = df.corr()\n\n mask = np.zeros_like(corr, dtype=np.bool)\n mask[np.triu_indices_from(mask)] = True\n\n # Set up the matplotlib figure\n sns.set_style('whitegrid')\n f, ax = plt.subplots(figsize=(16, 12))\n\n # Generate a custom diverging colormap\n cmap = sns.diverging_palette(220, 10, as_cmap=True)\n\n # Draw the heatmap with the mask and correct aspect ratio\n sns.heatmap(corr, mask=mask, cmap=cmap, vmax=1, vmin=-1, center=0,\n square=True, linewidths=.75, annot=False, cbar_kws={\"shrink\": .75});\n \ncorr_chart(dfx)",
"_____no_output_____"
]
],
[
[
"## Update Sales",
"_____no_output_____"
]
],
[
[
"# # File from Start based on Target Variable\n# current_sales_df = import_parse(rest_1_file)",
"_____no_output_____"
],
[
"# date = '2019-06-13'\n# sales = 15209.75\n# covers = 207\n# outside = 1\n# closed = 0\n\n# def add_sales_row(date, sales):\n \n# df = pd.DataFrame({'sales': sales,\n# 'covers': covers,\n# 'outside': outside,\n# 'closed': closed},\n# index=[date])\n \n# return df\n\n# temp = add_sales_row(date, sales)",
"_____no_output_____"
],
[
"# def build_sales_df(df, temp):\n \n# df = df.append(temp)\n \n# return df\n \n# current_sales_df = build_sales_df(current_sales_df, temp)",
"_____no_output_____"
],
[
"# # Download Current DataFrame to CSV\n# current_sales_df.to_csv(f'rest_1_clean_updated_{start_date}_to_{end_date}.csv')",
"_____no_output_____"
],
[
"# df_import = pd.read_csv('rest_1_clean_updated_2017-01-01_to_2019-06-17.csv', parse_dates=True, index_col='Unnamed: 0')\n\n# def import_current(df):\n \n# df.index = pd.to_datetime(df.index)\n \n# df = add_ppa(df)\n \n# target_list = ['sales', 'covers', 'ppa']\n\n# target_to_drop = [t for t in target_list if t != target]\n\n# df = df.drop(target_to_drop, axis=1)\n \n# return df\n\n# current_df = import_current(df_import)",
"_____no_output_____"
]
],
[
[
"### Add Recent Sales Data",
"_____no_output_____"
]
],
[
[
"# # Import Most Recent DataFrame\n# df_before_features = pd.read_csv('csv_before_features.csv', index_col='date', parse_dates=True)\n\n# # Create New Weather DataFrame with Updated Data\n# new_date_start = '2019-06-15'\n# new_date_end = '2019-06-17'\n\n# def update_current_df(sales_df, df_before_features, new_date_start, new_end_date):\n \n# sales_df = sales_df[new_date_start:]\n \n# sales_df = sales_df.rename_axis(index = 'date')\n# sales_df.index = pd.to_datetime(sales_df.index)\n\n# ## Find Lat Long for Business\n# lat, long = yelp_lat_long(search_business, location)\n\n# ## Pull Weather Data / Forecast\n# weather_df = build_weather_df(weather_call(new_date_start, new_date_end, lat, long))\n \n# ## Parse, Clean, Engineer\n# df = pd.merge(sales_df, weather_df, how='left', on='date')\n# df, _ = one_hot_temp(df, temp_enc)\n# df = pd.concat([df_before_features, df])\n# df = target_trend_engineering(feature_engineering(df))\n \n# return df\n \n# current_df = update_current_df(current_df, df_before_features, new_date_start, new_date_end)",
"_____no_output_____"
]
],
[
[
"## Test / Train / Split",
"_____no_output_____"
],
[
"### Drop Closed Days?",
"_____no_output_____"
]
],
[
[
"drop_all_closed = False\n\nif drop_all_closed:\n \n current_df = current_df[current_df['closed'] == 0]",
"_____no_output_____"
],
[
"dfx.columns",
"_____no_output_____"
],
[
"def drop_weather(df):\n \n no_weather = False\n\n if no_weather:\n\n df = df.drop(['humidity', 'precip_prob', 'temp_very_cold', 'temp_cold', 'temp_hot', 'temp_very_hot', 'precip_while_open', \\\n 'temp_very_cold_X_outside', 'temp_cold_X_outside', 'temp_hot_X_outside','temp_very_hot_X_outside', 'outside_X_precip_open'], axis=1)\n \n df = df.merge(df_outside, on='date', how='left')\n \n return df\n \n else:\n \n return df\n \ndfx = drop_weather(dfx)\ndfx.head()\n ",
"_____no_output_____"
],
[
"def cv_split(df):\n \n features = dfx.drop([target], axis=1)[days_fwd:]\n y = dfx[target][days_fwd:]\n \n return features, y\n\ncv_features, cv_y = cv_split(dfx)\nbaseline_cv_x, baseline_cv_y = cv_split(mod_baseline)",
"_____no_output_____"
],
[
"def train_test_split(df):\n \n # Separate Target & Features\n y = df[target]\n features = df.drop([target], axis=1)\n \n # Test / Train / Split\n train_date_start = '2017-01-01'\n train_date_end = '2018-12-31'\n \n X_train = features[pd.to_datetime(train_date_start) + pd.DateOffset(days_fwd):train_date_end]\n X_test = features[pd.to_datetime(train_date_end) + pd.DateOffset(1): ]\n \n y_train = y[pd.to_datetime(train_date_start) + pd.DateOffset(days_fwd):train_date_end]\n y_test = y[pd.to_datetime(train_date_end) + pd.DateOffset(1): ]\n \n # Scale\n scaler = MinMaxScaler()\n X_train_scaled = scaler.fit_transform(X_train)\n X_test_scaled = scaler.transform(X_test)\n \n X_train = pd.DataFrame(X_train_scaled, columns=X_train.columns)\n X_test = pd.DataFrame(X_test_scaled, columns=X_train.columns)\n \n print('Train set: ', len(X_train))\n print('Test set: ', len(X_test))\n \n return X_train, X_test, y_train, y_test, scaler\n \nX_train, X_test, y_train, y_test, scaler = train_test_split(dfx)\nbaseline_X_train, baseline_X_test, baseline_y_train, baseline_y_test, baseline_scaler = train_test_split(mod_baseline)",
"Train set: 691\nTest set: 169\nTrain set: 691\nTest set: 169\n"
]
],
[
[
"### Linear Regression",
"_____no_output_____"
]
],
[
[
"def linear_regression_model(X_train, y_train):\n \n lr = LinearRegression(fit_intercept=True)\n lr_rgr = lr.fit(X_train, y_train)\n \n return lr_rgr",
"_____no_output_____"
],
[
"lr_rgr = linear_regression_model(X_train, y_train)\nbaseline_lr_rgr = linear_regression_model(baseline_X_train, baseline_y_train)",
"_____no_output_____"
],
[
"def rgr_score(rgr, X_train, y_train, X_test, y_test, cv_features, cv_y):\n \n y_hat = rgr.predict(X_test)\n sum_squares_residual = sum((y_test - y_hat)**2)\n sum_squares_total = sum((y_test - np.mean(y_test))**2)\n r_squared = 1 - (float(sum_squares_residual))/sum_squares_total\n adjusted_r_squared = 1 - (1-r_squared)*(len(y_test)-1)/(len(y_test)-X_test.shape[1]-1)\n print('Formula Scores - R-Squared: ', r_squared, 'Adjusted R-Squared: ', adjusted_r_squared, '\\n')\n \n train_score = rgr.score(X_train, y_train)\n test_score = rgr.score(X_test, y_test)\n \n y_pred = rgr.predict(X_test)\n rmse = np.sqrt(mean_squared_error(y_test, y_pred))\n \n pred_df = pd.DataFrame(y_pred, index=y_test.index)\n pred_df = pred_df.rename(columns={0: target})\n \n print('Train R-Squared: ', train_score)\n print('Test R-Squared: ', test_score, '\\n')\n \n print('Root Mean Squared Error: ', rmse, '\\n')\n \n print('Cross Val Avg R-Squared: ', \\\n np.mean(cross_val_score(rgr, cv_features, cv_y, cv=10, scoring='r2')), '\\n')\n \n print('Intercept: ', rgr.intercept_, '\\n')\n print('Coefficients: \\n')\n \n for index, col_name in enumerate(X_test.columns):\n print(col_name, ' --> ', rgr.coef_[index])\n \n return pred_df",
"_____no_output_____"
],
[
"pred_df = rgr_score(lr_rgr, X_train, y_train, X_test, y_test, cv_features, cv_y)\n\nbaseline_preds = rgr_score(baseline_lr_rgr, baseline_X_train, baseline_y_train, baseline_X_test, baseline_y_test, baseline_cv_x, baseline_cv_y)",
"Formula Scores - R-Squared: 0.8314590736057758 Adjusted R-Squared: 0.8005994673645798 \n\nTrain R-Squared: 0.781664938228909\nTest R-Squared: 0.8314590736057759 \n\nRoot Mean Squared Error: 1294.243986043061 \n\nCross Val Avg R-Squared: 0.6784487693497173 \n\nIntercept: 13151.944683753753 \n\nCoefficients: \n\nhumidity --> -1430.5968014821779\nprecip_prob --> -285.63684206335836\nclosed --> -13810.080481616358\ntemp_very_cold --> -257.0253619843726\ntemp_cold --> -34.53827320300603\ntemp_warm --> 520.331150761081\ntemp_hot --> 516.9458825323867\ntemp_very_hot --> -745.7133981060797\nprecip_while_open --> 329.17268853749164\nsunday_three_day --> 1459.4669674345812\nday_of_week_1 --> -131.00317778178092\nday_of_week_2 --> 263.2031013715046\nday_of_week_3 --> 896.9247158051236\nday_of_week_4 --> 2701.813099501862\nday_of_week_5 --> 3747.196055875591\nday_of_week_6 --> 1130.87365120462\nmonth_cluster_0 --> 30.30786650568939\nmonth_cluster_1 --> 261.03403614494533\nmonth_cluster_2 --> -291.3419026506342\noutside_X_precip_open --> -1152.7558581560686\ntemp_very_cold_X_outside --> -548.6628878490104\ntemp_cold_X_outside --> -15.384005348701548\ntemp_warm_X_outside --> 1250.2483159129388\ntemp_hot_X_outside --> 3192.161249597661\ntemp_very_hot_X_outside --> 4730.427249167204\ntarget_trend --> 2191.55917774509\nFormula Scores - R-Squared: 0.5547269411073399 Adjusted R-Squared: 0.514247572117098 \n\nTrain R-Squared: 0.6532346004008801\nTest R-Squared: 0.5547269411073397 \n\nRoot Mean Squared Error: 2629.9341607918627 \n\nCross Val Avg R-Squared: 0.6784487693497173 \n\nIntercept: 1.787689908062782e+16 \n\nCoefficients: \n\noutside --> 2924.183365954575\nclosed --> -13932.679232421144\nday_of_week_0 --> -53270307486902.28\nday_of_week_1 --> -53270307487428.79\nday_of_week_2 --> -53270307486850.15\nday_of_week_3 --> -53270307486337.086\nday_of_week_4 --> -53270307484571.22\nday_of_week_5 --> -53270307483496.83\nday_of_week_6 --> -53270307486096.63\nmonth_cluster_0 --> -1.7823628773128064e+16\nmonth_cluster_1 --> -1.782362877312821e+16\nmonth_cluster_2 --> -1.7823628773129166e+16\nsunday_three_day --> 912.0294684571295\ntarget_trend --> 3327.242985861117\n"
]
],
[
[
"### Prediction Function",
"_____no_output_____"
]
],
[
[
"outside = 1\n\ndef predict_df(clf, scaler, X_train, current_df, date_1, date_2):\n \n # Find Lat Long for Business\n lat, long = yelp_lat_long(search_business, location)\n \n # Pull Weather Data / Forecast\n weather_df = build_weather_df(weather_call(date_1, date_2, lat, long))\n \n day_of_week, apparent_temperature = weather_df['day_of_week'], weather_df['apparent_temperature']\n weather_df['outside'] = outside\n \n # One Hot Encode Temperature Using Fitted Encoder\n df, _ = one_hot_temp(weather_df, temp_enc)\n \n df['closed'] = 0\n \n # Add Feature Engineering\n df = feature_engineering(df)\n \n # Add Sales Data for Sales Trend Engineering\n current_df = current_df[target]\n df = pd.merge(df, current_df, on='date', how='left')\n df[target] = df[target].fillna(method='ffill')\n \n df = target_trend_engineering(df)\n df = df.drop([target], axis=1)\n \n # Ensure Column Parity\n missing_cols = set(X_train.columns) - set(df.columns)\n \n for c in missing_cols:\n df[c] = 0\n \n df = df[X_train.columns][-2:]\n \n # Scale Transform\n df_scaled = scaler.transform(df)\n df = pd.DataFrame(df_scaled, columns=df.columns, index=df.index)\n \n # Predict and Build Prediction DataFrame for Review\n pred_array = pd.DataFrame(clf.predict(df), index=df.index, columns=[target])\n pred_df = df[df.columns[(df != 0).any()]]\n pred_df = pd.concat([pred_df, day_of_week, apparent_temperature], axis=1)\n \n final_predict = pd.concat([pred_array, pred_df], axis=1)\n \n return final_predict",
"_____no_output_____"
],
[
"tonight = predict_df(lr_rgr, scaler, X_train, dfx, pd.datetime.now().date() + pd.DateOffset(-days_fwd), pd.datetime.now().date())\ntonight[-2:]",
"Weather Location: Jupiter Disco\n"
]
],
[
[
"## Lasso",
"_____no_output_____"
]
],
[
[
"def lasso_model(X_train, y_train):\n\n lassoReg = Lasso(fit_intercept=True, alpha=.05)\n lasso_rgr = lassoReg.fit(X_train,y_train)\n\n return lasso_rgr\n\nlasso_rgr = lasso_model(X_train, y_train)\n\nbaseline_lasso = lasso_model(baseline_X_train, baseline_y_train)",
"_____no_output_____"
],
[
"pred_df_ppa_lasso = rgr_score(lasso_rgr, X_train, y_train, X_test, y_test, cv_features, cv_y)\n\nbaseline_lasso = rgr_score(baseline_lasso, baseline_X_train, baseline_y_train, baseline_X_test, baseline_y_test, baseline_cv_x, baseline_cv_y)",
"Formula Scores - R-Squared: 0.8314323274764113 Adjusted R-Squared: 0.8005678240565992 \n\nTrain R-Squared: 0.7816645985160647\nTest R-Squared: 0.8314323274764113 \n\nRoot Mean Squared Error: 1294.3466751779988 \n\nCross Val Avg R-Squared: 0.678457985108043 \n\nIntercept: 13148.727273355167 \n\nCoefficients: \n\nhumidity --> -1430.1528362812999\nprecip_prob --> -283.34453313562307\nclosed --> -13805.88566023607\ntemp_very_cold --> -221.80990823827818\ntemp_cold --> 0.42931203698637216\ntemp_warm --> 555.9345371314138\ntemp_hot --> 552.6446950518861\ntemp_very_hot --> -695.2977967077175\nprecip_while_open --> 325.0905311124925\nsunday_three_day --> 1455.5221565455486\nday_of_week_1 --> -131.766259785456\nday_of_week_2 --> 261.5629839884926\nday_of_week_3 --> 895.483365148691\nday_of_week_4 --> 2700.2704367121014\nday_of_week_5 --> 3745.7127220957136\nday_of_week_6 --> 1130.0307377729198\nmonth_cluster_0 --> 0.0\nmonth_cluster_1 --> 230.44835585239383\nmonth_cluster_2 --> -321.18001583943624\noutside_X_precip_open --> -1147.8559947757474\ntemp_very_cold_X_outside --> -541.4106343569797\ntemp_cold_X_outside --> -0.0\ntemp_warm_X_outside --> 1249.5522516363753\ntemp_hot_X_outside --> 3191.6350462776413\ntemp_very_hot_X_outside --> 4715.139045684163\ntarget_trend --> 2189.4283021727774\nFormula Scores - R-Squared: 0.5555023119791402 Adjusted R-Squared: 0.5150934312499712 \n\nTrain R-Squared: 0.6532604874181484\nTest R-Squared: 0.5555023119791402 \n\nRoot Mean Squared Error: 2627.6433610242184 \n\nCross Val Avg R-Squared: 0.678457985108043 \n\nIntercept: 13463.334368752663 \n\nCoefficients: \n\noutside --> 2893.046989647196\nclosed --> -13917.965394269797\nday_of_week_0 --> -1197.2232967267432\nday_of_week_1 --> -1725.5463796595861\nday_of_week_2 --> -1143.797856114545\nday_of_week_3 --> -629.3400946702291\nday_of_week_4 --> 1131.6386560644487\nday_of_week_5 --> 2207.4641670584297\nday_of_week_6 --> -393.37566833264526\nmonth_cluster_0 --> 611.9120990124821\nmonth_cluster_1 --> 449.03603863194314\nmonth_cluster_2 --> -499.1021943255487\nsunday_three_day --> 912.804845886708\ntarget_trend --> 3321.1131545660255\n"
],
[
"tonight = predict_df(lasso_rgr, scaler, X_train, dfx, pd.datetime.now().date() + pd.DateOffset(-days_fwd), pd.datetime.now().date())\ntonight[-2:]",
"Weather Location: Jupiter Disco\n"
],
[
"from yellowbrick.regressor import ResidualsPlot\nfrom yellowbrick.features.importances import FeatureImportances\n\nplt.figure(figsize=(12,8))\n\nvisualizer = ResidualsPlot(lasso_rgr, hist=False)\n\nvisualizer.fit(X_train, y_train)\nvisualizer.score(X_test, y_test)\nvisualizer.poof()",
"_____no_output_____"
],
[
"features = list(X_train.columns)\n\nfig = plt.figure(figsize=(12,8))\nax = fig.add_subplot()\n\nlabels = list(map(lambda x: x.title(), features))\nvisualizer = FeatureImportances(lasso_rgr, ax=ax, labels=labels, relative=False)\n\nvisualizer.fit(X_train, y_train)\nvisualizer.poof()",
"_____no_output_____"
]
],
[
[
"### Random Forest Regression",
"_____no_output_____"
]
],
[
[
"(1/6)**20",
"_____no_output_____"
],
[
"8543 * 8543",
"_____no_output_____"
],
[
"def rf_regression_model(X_train, y_train):\n \n rfr = RandomForestRegressor(max_depth= 11, \n max_features= 0.60, \n min_impurity_decrease= 0.005, \n n_estimators= 300,\n min_samples_leaf = 2,\n min_samples_split = 2,\n random_state = 0)\n \n rfr_rgr = rfr.fit(X_train, y_train)\n \n return rfr_rgr",
"_____no_output_____"
],
[
"rfr_rgr = rf_regression_model(X_train, y_train)",
"_____no_output_____"
],
[
"def rfr_score(rgr, X_test, y_test, cv_features, cv_y):\n \n y_hat = rgr.predict(X_test)\n sum_squares_residual = sum((y_test - y_hat)**2)\n sum_squares_total = sum((y_test - np.mean(y_test))**2)\n r_squared = 1 - (float(sum_squares_residual))/sum_squares_total\n adjusted_r_squared = 1 - (1-r_squared)*(len(y_test)-1)/(len(y_test)-X_test.shape[1]-1)\n print('Formula Scores - R-Squared: ', r_squared, 'Adjusted R-Squared: ', adjusted_r_squared, '\\n')\n \n train_score = rgr.score(X_train, y_train)\n test_score = rgr.score(X_test, y_test)\n \n y_pred = rgr.predict(X_test)\n rmse = np.sqrt(mean_squared_error(y_test, y_pred))\n \n pred_df = pd.DataFrame(y_pred, index=y_test.index)\n pred_df = pred_df.rename(columns={0: target})\n \n print('Train R-Squared: ', train_score)\n print('Test R-Squared: ', test_score, '\\n')\n \n print('Root Mean Squared Error: ', rmse, '\\n')\n \n print('Cross Val Avg R-Squared: ', \\\n np.mean(cross_val_score(rgr, cv_features, cv_y, cv=10, scoring='r2')), '\\n')\n \n return pred_df\n\npred_df = rfr_score(rfr_rgr, X_test, y_test, cv_features, cv_y) ",
"Formula Scores - R-Squared: 0.8196000605472993 Adjusted R-Squared: 0.7865690857179315 \n\nTrain R-Squared: 0.8914251745921165\nTest R-Squared: 0.8196000605472994 \n\nRoot Mean Squared Error: 1339.0033420819104 \n\nCross Val Avg R-Squared: 0.6470395821127155 \n\n"
]
],
[
[
"# Random Forest Regression Prediction",
"_____no_output_____"
]
],
[
[
"tonight = predict_df(rfr_rgr, scaler, X_train, dfx, pd.datetime.now().date() + pd.DateOffset(-days_fwd), pd.datetime.now().date())\ntonight[-2:]",
"Weather Location: Jupiter Disco\n"
]
],
[
[
"### Grid Search Helper Function",
"_____no_output_____"
]
],
[
[
"def run_grid_search(rgr, params, X_train, y_train):\n \n cv = 5\n n_jobs = -1\n scoring = 'neg_mean_squared_error'\n \n grid = GridSearchCV(rgr, params, cv=cv, n_jobs=n_jobs, scoring=scoring, verbose=10)\n grid = grid.fit(X_train, y_train)\n \n best_grid_rgr = grid.best_estimator_\n \n print('Grid Search: ', rgr.__class__.__name__, '\\n')\n print('Grid Search Best Score: ', grid.best_score_)\n print('Grid Search Best Params: ', grid.best_params_)\n print('Grid Search Best Estimator: ', grid.best_estimator_)\n\n return best_grid_rgr\n",
"_____no_output_____"
],
[
"params = {\n 'n_estimators': [250, 275, 300, 350, 400,500],\n 'max_depth': [5, 7, 9, 11, 13, 15],\n 'min_impurity_decrease': [0.005, 0.001, 0.0001],\n 'max_features': ['auto', 0.65, 0.75, 0.85, 0.95]\n }\n\nbest_grid_rgr = run_grid_search(rfr_rgr, params, X_train, y_train)",
"Fitting 5 folds for each of 540 candidates, totalling 2700 fits\n"
]
],
[
[
"### OLS Model",
"_____no_output_____"
]
],
[
[
"import statsmodels.api as sm\nfrom statsmodels.formula.api import ols",
"_____no_output_____"
],
[
"f = ''\n\nfor c in dfx.columns:\n \n f += c + '+'\n \nx = f[6:-1]",
"_____no_output_____"
],
[
"f= target + '~' + x\n\nmodel = ols(formula=f, data=dfx).fit()\nmodel.summary()",
"_____no_output_____"
]
],
[
[
"## XGB Regressor",
"_____no_output_____"
]
],
[
[
"from xgboost import XGBRegressor",
"_____no_output_____"
],
[
"def xgb_model(X_train, y_train):\n \n objective = 'reg:linear'\n booster = 'gbtree'\n nthread = 4\n learning_rate = 0.02\n max_depth = 3\n colsample_bytree = 0.75\n n_estimators = 450\n min_child_weight = 2\n \n xgb_rgr= XGBRegressor(booster=booster, objective=objective, colsample_bytree=colsample_bytree, learning_rate=learning_rate, \\\n max_depth=max_depth, nthread=nthread, n_estimators=n_estimators, min_child_weight=min_child_weight, random_state = 0)\n \n xgb_rgr = xgb_rgr.fit(X_train, y_train)\n \n return xgb_rgr",
"_____no_output_____"
],
[
"xgb_rgr = xgb_model(X_train, y_train)",
"_____no_output_____"
],
[
"# Convert column names back to original\nX_test = X_test[X_train.columns]",
"_____no_output_____"
],
[
"pred_df_covers_xgb = rfr_score(xgb_rgr, X_test, y_test, cv_features, cv_y) ",
"Formula Scores - R-Squared: 0.835448257897047 Adjusted R-Squared: 0.8053190656810134 \n\nTrain R-Squared: 0.8272349873106326\nTest R-Squared: 0.8354482578970471 \n\nRoot Mean Squared Error: 1278.8355762750668 \n\nCross Val Avg R-Squared: 0.6668620522192908 \n\n"
]
],
[
[
"## HYBRID",
"_____no_output_____"
]
],
[
[
"filex = 'predicted_ppa_timeseries.csv'\ndf_ts = pd.read_csv(filex,index_col='date',parse_dates=True)",
"_____no_output_____"
],
[
"pred_df = pred_df_covers_xgb.merge(df_ts, on='date', how='left')\npred_df.head()",
"_____no_output_____"
],
[
"pred_df['sales'] = pred_df['covers'] * pred_df['pred_ppa']",
"_____no_output_____"
],
[
"pred_df = pred_df[['sales']]",
"_____no_output_____"
],
[
"r2 = metrics.r2_score(y_test, pred_df)",
"_____no_output_____"
],
[
"r2",
"_____no_output_____"
],
[
"rmse = np.sqrt(mean_squared_error(y_test, pred_df))\nrmse",
"_____no_output_____"
],
[
"tonight = predict_df(xgb_rgr, scaler, X_train, dfx, pd.datetime.now().date() + pd.DateOffset(-days_fwd), pd.datetime.now().date())\ntonight[-2:]",
"_____no_output_____"
],
[
"xgb1 = XGBRegressor()\nparameters = {'nthread':[4], #when use hyperthread, xgboost may become slower\n 'objective':['reg:linear'],\n 'learning_rate': [.015, 0.02, .025], #so called `eta` value\n 'max_depth': [3, 4, 5],\n 'min_child_weight': [1, 2, 3],\n 'silent': [1],\n 'subsample': [0.7],\n 'colsample_bytree': [0.55, 0.60, 0.65],\n 'n_estimators': [400, 500, 600]}\n\nxgb_grid = GridSearchCV(xgb1,\n parameters,\n cv = 3,\n n_jobs = 5,\n verbose=True,\n scoring = 'neg_mean_squared_error')\n\nxgb_grid.fit(X_train,\n y_train)\n\nprint(xgb_grid.best_score_)\nprint(xgb_grid.best_params_)",
"Fitting 3 folds for each of 243 candidates, totalling 729 fits\n"
]
],
[
[
"## CLEAN RUN",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
]
] |
cb929602368d6e40b75d600d00688cdbf6eb6d88
| 1,705 |
ipynb
|
Jupyter Notebook
|
project_euler/097_Large non-Mersenne prime.ipynb
|
Sabihxh/projectEuler
|
4b0d4bc56598dd2a32c1f00b41bfcadbb495a88b
|
[
"MIT"
] | 1 |
2018-03-20T12:04:06.000Z
|
2018-03-20T12:04:06.000Z
|
project_euler/097_Large non-Mersenne prime.ipynb
|
Sabihxh/projectEuler
|
4b0d4bc56598dd2a32c1f00b41bfcadbb495a88b
|
[
"MIT"
] | null | null | null |
project_euler/097_Large non-Mersenne prime.ipynb
|
Sabihxh/projectEuler
|
4b0d4bc56598dd2a32c1f00b41bfcadbb495a88b
|
[
"MIT"
] | null | null | null | 22.733333 | 308 | 0.553666 |
[
[
[
"### Problem 97: Large non-Mersenne prime\n\n<p>The first known prime found to exceed one million digits was discovered in 1999, and is a Mersenne prime of the form 2<sup>6972593</sup>−1; it contains exactly 2,098,960 digits. Subsequently other Mersenne primes, of the form 2<sup><i>p</i></sup>−1, have been found which contain more digits.</p>\n<p>However, in 2004 there was found a massive non-Mersenne prime which contains 2,357,207 digits: 28433×2<sup>7830457</sup>+1.</p>\n<p>Find the last ten digits of this prime number.</p>\n",
"_____no_output_____"
]
],
[
[
"print((28433 * 2**7830457 + 1) % 10**10)",
"8739992577\n"
]
]
] |
[
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
]
] |
cb929bb5468cddfa574271030e0a7521af0cd718
| 26,992 |
ipynb
|
Jupyter Notebook
|
02_Intro_Numpy.ipynb
|
jorgemauricio/curso_itesm
|
2bcdce57c852387e0365c7706cfe1519878ab1e0
|
[
"MIT"
] | 1 |
2020-05-16T00:57:06.000Z
|
2020-05-16T00:57:06.000Z
|
02_Intro_Numpy.ipynb
|
jorgemauricio/curso_itesm
|
2bcdce57c852387e0365c7706cfe1519878ab1e0
|
[
"MIT"
] | null | null | null |
02_Intro_Numpy.ipynb
|
jorgemauricio/curso_itesm
|
2bcdce57c852387e0365c7706cfe1519878ab1e0
|
[
"MIT"
] | null | null | null | 18.487671 | 100 | 0.433128 |
[
[
[
"# Numpy\n\n### GitHub repository: https://github.com/jorgemauricio/curso_itesm\n\n### Instructor: Jorge Mauricio",
"_____no_output_____"
]
],
[
[
"# librerías\nimport numpy as np",
"_____no_output_____"
]
],
[
[
"# Crear Numpy Arrays\n## De una lista de python\nCreamos el arreglo directamente de una lista o listas de python",
"_____no_output_____"
]
],
[
[
"my_list = [1,2,3]\nmy_list",
"_____no_output_____"
],
[
"np.array(my_list)",
"_____no_output_____"
],
[
"my_matrix = [[1,2,3],[4,5,6],[7,8,9]]\nmy_matrix",
"_____no_output_____"
]
],
[
[
"## Métodos\n\n### arange",
"_____no_output_____"
]
],
[
[
"np.arange(0,10)",
"_____no_output_____"
],
[
"np.arange(0,11,2)",
"_____no_output_____"
]
],
[
[
"### ceros y unos\nGenerar arreglos de ceros y unos",
"_____no_output_____"
]
],
[
[
"np.zeros(3)",
"_____no_output_____"
],
[
"np.zeros((5,5))",
"_____no_output_____"
],
[
"np.ones(3)",
"_____no_output_____"
],
[
"np.ones((3,3))",
"_____no_output_____"
]
],
[
[
"### linspace\nGenerar un arreglo especificando un intervalo",
"_____no_output_____"
]
],
[
[
"np.linspace(0,10,3)",
"_____no_output_____"
],
[
"np.linspace(0,10,50)",
"_____no_output_____"
]
],
[
[
"### eye\nGenerar matrices de identidad",
"_____no_output_____"
]
],
[
[
"np.eye(4)",
"_____no_output_____"
]
],
[
[
"### Random\n### rand\nGenerar un arreglo con una forma determinada de numeros con una distribución uniforme [0,1]",
"_____no_output_____"
]
],
[
[
"np.random.rand(2)",
"_____no_output_____"
],
[
"np.random.rand(5,5)",
"_____no_output_____"
]
],
[
[
"### randn\nGenerar un arreglo con una distribucion estandar a diferencia de rand que es uniforme",
"_____no_output_____"
]
],
[
[
"np.random.randn(2)",
"_____no_output_____"
],
[
"np.random.randn(5,5)",
"_____no_output_____"
]
],
[
[
"### randint\nGenerar numeros aleatorios en un rango determinado",
"_____no_output_____"
]
],
[
[
"np.random.randint(1,100)",
"_____no_output_____"
],
[
"np.random.randint(1,100,10)",
"_____no_output_____"
]
],
[
[
"### Métodos y atributos de arreglos\n",
"_____no_output_____"
]
],
[
[
"arr = np.arange(25)\nranarr = np.random.randint(0,50,10)",
"_____no_output_____"
],
[
"arr",
"_____no_output_____"
],
[
"ranarr",
"_____no_output_____"
]
],
[
[
"### Reshape\nRegresa el mismo arreglo pero en diferente forma",
"_____no_output_____"
]
],
[
[
"arr.reshape(5,5)",
"_____no_output_____"
]
],
[
[
"### max, min, argmax, argmin\nMetodos para encontrar máximos y mínimos de los valores y sus indices",
"_____no_output_____"
]
],
[
[
"ranarr",
"_____no_output_____"
],
[
"ranarr.max()",
"_____no_output_____"
],
[
"ranarr.argmax()",
"_____no_output_____"
],
[
"ranarr.min()",
"_____no_output_____"
],
[
"ranarr.argmin()",
"_____no_output_____"
]
],
[
[
"### Shape\nAtributo para desplegar la forma que tienen el arreglo",
"_____no_output_____"
]
],
[
[
"# Vector\narr.shape",
"_____no_output_____"
],
[
"# Tomar en cuenta que se implementan dos corchetes\narr.reshape(1,25)",
"_____no_output_____"
],
[
"arr.reshape(1,25).shape",
"_____no_output_____"
],
[
"arr.reshape(25,1)",
"_____no_output_____"
],
[
"arr.reshape(25,1).shape",
"_____no_output_____"
]
],
[
[
"### dtype\nDespliega el tipo de dato de los objetos del arreglo",
"_____no_output_____"
]
],
[
[
"arr.dtype",
"_____no_output_____"
]
],
[
[
"# Selección e indices en Numpy\n",
"_____no_output_____"
]
],
[
[
"# crear un arreglo\narr = np.arange(0,11)",
"_____no_output_____"
],
[
"# desplegar el arreglo\narr",
"_____no_output_____"
]
],
[
[
"# Selección utilizando corchetes",
"_____no_output_____"
]
],
[
[
"# obtener el valor del indice 8\narr[8]",
"_____no_output_____"
],
[
"# obtener los valores de un rango\narr[1:5]",
"_____no_output_____"
],
[
"#obtener los valores de otro rango\narr[2:6]",
"_____no_output_____"
]
],
[
[
"# Reemplazar valores",
"_____no_output_____"
]
],
[
[
"# reemplazar valores en un rango determinado\narr[0:5]=100\n\n# desplegar el arreglo\n\narr",
"_____no_output_____"
],
[
"# Generar nuevamente el arreglo\narr = np.arange(0,11)\n\n# desplegar\narr",
"_____no_output_____"
],
[
"# corte de un arreglo\nslice_of_arr = arr[0:6]\n\n# desplegar el corte\nslice_of_arr",
"_____no_output_____"
],
[
"# cambiar valores del corte\nslice_of_arr[:]=99\n\n# desplegar los valores del corte\nslice_of_arr",
"_____no_output_____"
],
[
"# desplegar arreglo\narr",
"_____no_output_____"
],
[
"# para obtener una copia se debe hacer explicitamente\narr_copy = arr.copy()\n\n# desplegar el arreglo copia\narr_copy",
"_____no_output_____"
]
],
[
[
"## Indices en un arreglo 2D (matrices)\n\nLa forma general de un arreglo 2d es la siguiente **arr_2d[row][col]** o **arr_2d[row,col]**\n",
"_____no_output_____"
]
],
[
[
"# generar un arreglo 2D\narr_2d = np.array(([5,10,15],[20,25,30],[35,40,45]))\n\n#Show\narr_2d",
"_____no_output_____"
],
[
"# indices de filas\narr_2d[1]\n",
"_____no_output_____"
],
[
"# Formato es arr_2d[row][col] o arr_2d[row,col]\n\n# Seleccionar un solo elemento\narr_2d[1][0]",
"_____no_output_____"
],
[
"# Seleccionar un solo elemento\narr_2d[1,0]",
"_____no_output_____"
],
[
"# Cortes en 2D\n\n# forma (2,2) desde la esquina superior derecha\narr_2d[:2,1:]",
"_____no_output_____"
],
[
"#forma desde la ultima fila\narr_2d[2]",
"_____no_output_____"
],
[
"# forma desde la ultima fila\narr_2d[2,:]",
"_____no_output_____"
],
[
"# longitud de un arreglo\narr_length = arr_2d.shape[1]\n\narr_length",
"_____no_output_____"
]
],
[
[
"# Selección",
"_____no_output_____"
]
],
[
[
"arr = np.arange(1,11)\narr",
"_____no_output_____"
],
[
"arr > 4",
"_____no_output_____"
],
[
"bool_arr = arr>4",
"_____no_output_____"
],
[
"bool_arr",
"_____no_output_____"
],
[
"arr[bool_arr]",
"_____no_output_____"
],
[
"arr[arr>2]",
"_____no_output_____"
],
[
"x = 2\narr[arr>x]",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb92a16a6efeb87b9256ecacbc5d0b936608d942
| 24,564 |
ipynb
|
Jupyter Notebook
|
tops_sample_qc.ipynb
|
tlefsad/esninja
|
346449f90490e50daa700d928badc8c1a99f7abe
|
[
"Apache-2.0"
] | 15 |
2020-04-26T07:13:53.000Z
|
2021-12-16T08:41:33.000Z
|
tops_sample_qc.ipynb
|
tlefsad/esninja
|
346449f90490e50daa700d928badc8c1a99f7abe
|
[
"Apache-2.0"
] | null | null | null |
tops_sample_qc.ipynb
|
tlefsad/esninja
|
346449f90490e50daa700d928badc8c1a99f7abe
|
[
"Apache-2.0"
] | 6 |
2020-04-27T10:05:48.000Z
|
2021-03-15T02:39:07.000Z
| 32.708389 | 120 | 0.422081 |
[
[
[
"# TOPS Sample Quality Check",
"_____no_output_____"
]
],
[
[
"import qgrid\nimport pandas as pd\npd.options.display.max_columns=999\npd.options.display.max_rows=999\nimport numpy as np\n%reload_ext autoreload\n%autoreload 2\nfrom tqdm import tqdm_notebook, trange\nfrom IPython.core.interactiveshell import InteractiveShell\nInteractiveShell.ast_node_interactivity = \"all\"\nimport warnings\nwarnings.filterwarnings(\"ignore\")\ndef check_missing(df):\n per_missing = df.isnull().mean()\n missing_df = pd.DataFrame({'col': df.columns, 'per_missing': per_missing})\n missing_df = missing_df.sort_values('per_missing',ascending=False).reset_index(drop=True)\n missing_df['col'] = pd.Categorical(missing_df.col, categories=missing_df.col, ordered=True)\n return missing_df",
"_____no_output_____"
],
[
"df = pd.read_csv('data/tops_sample.csv')\nqgrid.show_grid(df[[c for c in df.columns if '_en' in c]])",
"_____no_output_____"
],
[
"qgrid.show_grid(df[[c for c in df.columns if '_th' in c]])",
"_____no_output_____"
],
[
"#missing values\ncheck_missing(df)",
"_____no_output_____"
],
[
"#thai characters in english names\nimport re\ndef char_percent(pattern,text):\n return len(re.findall(pattern,text)) / (len(text)+0.01)\n\ndf[df.name_en.map(lambda x: char_percent('[ก-๙]',str(x)))>0].iloc[:,:4]",
"_____no_output_____"
],
[
"#thai characters in english brands\nimport re\ndef char_percent(pattern,text):\n return len(re.findall(pattern,text)) / (len(text)+0.01)\n\ndf[df.brand_en.map(lambda x: char_percent('[ก-๙]',str(x)))>0].iloc[:,:6]",
"_____no_output_____"
],
[
"#thai characters in english descriptions\nimport re\ndef char_percent(pattern,text):\n return len(re.findall(pattern,text)) / (len(text)+0.01)\n\ndf[df.long_desc_en.map(lambda x: char_percent('[ก-๙]',str(x)))>0].iloc[:,-2:]",
"_____no_output_____"
],
[
"#be careful about the translation\ndf.category_en.value_counts()",
"_____no_output_____"
],
[
"df.category_th.value_counts()",
"_____no_output_____"
],
[
"#WTF!?\ndf[df.category_en=='Promotion Crazy Price 12.12'][['category_en','subcategory_en','category_th','subcategory_th']]",
"_____no_output_____"
],
[
"df.columns",
"_____no_output_____"
]
]
] |
[
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb92b1ed001c0aae6b584c457cc8fba790ce08cf
| 273,215 |
ipynb
|
Jupyter Notebook
|
nkasmanoff/burstdoc/Example.ipynb
|
lepalmer/Users
|
275147c651abb2ebb61cc9f44c566a8f40b0dcf3
|
[
"MIT"
] | null | null | null |
nkasmanoff/burstdoc/Example.ipynb
|
lepalmer/Users
|
275147c651abb2ebb61cc9f44c566a8f40b0dcf3
|
[
"MIT"
] | null | null | null |
nkasmanoff/burstdoc/Example.ipynb
|
lepalmer/Users
|
275147c651abb2ebb61cc9f44c566a8f40b0dcf3
|
[
"MIT"
] | 1 |
2018-07-23T19:18:05.000Z
|
2018-07-23T19:18:05.000Z
| 1,729.208861 | 47,894 | 0.947184 |
[
[
[
"%matplotlib inline",
"_____no_output_____"
],
[
"from bcSim import simFiles\nfrom plotSim import plotAeff",
"_____no_output_____"
],
[
"sfs = simFiles('../test/config.yaml')",
"Loading $BURSTCUBE/Simulation/MEGAlib/test//test_100.000keV_Cos0.500.inc1.id1.sim\nLoading $BURSTCUBE/Simulation/MEGAlib/test//test_173.205keV_Cos0.500.inc1.id1.sim\nLoading $BURSTCUBE/Simulation/MEGAlib/test//test_300.000keV_Cos0.500.inc1.id1.sim\nLoading $BURSTCUBE/Simulation/MEGAlib/test//test_100.000keV_Cos0.750.inc1.id1.sim\nLoading $BURSTCUBE/Simulation/MEGAlib/test//test_173.205keV_Cos0.750.inc1.id1.sim\nLoading $BURSTCUBE/Simulation/MEGAlib/test//test_300.000keV_Cos0.750.inc1.id1.sim\nLoading $BURSTCUBE/Simulation/MEGAlib/test//test_100.000keV_Cos1.000.inc1.id1.sim\nLoading $BURSTCUBE/Simulation/MEGAlib/test//test_173.205keV_Cos1.000.inc1.id1.sim\nLoading $BURSTCUBE/Simulation/MEGAlib/test//test_300.000keV_Cos1.000.inc1.id1.sim\n"
],
[
"plotAeff(sfs)",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code"
]
] |
cb92b3a8f505a8d77d4956b015a69b9ab5fb4400
| 96,992 |
ipynb
|
Jupyter Notebook
|
.ipynb_checkpoints/Test-checkpoint.ipynb
|
wlof-2/Text2Relation
|
a1321e3627fee4714d2c39c964d93d12d0802467
|
[
"MIT"
] | null | null | null |
.ipynb_checkpoints/Test-checkpoint.ipynb
|
wlof-2/Text2Relation
|
a1321e3627fee4714d2c39c964d93d12d0802467
|
[
"MIT"
] | null | null | null |
.ipynb_checkpoints/Test-checkpoint.ipynb
|
wlof-2/Text2Relation
|
a1321e3627fee4714d2c39c964d93d12d0802467
|
[
"MIT"
] | null | null | null | 53.854525 | 7,616 | 0.618804 |
[
[
[
"from transformers import (\n AutoConfig,\n AutoModelForSeq2SeqLM,\n AutoTokenizer,\n DataCollatorForSeq2Seq,\n HfArgumentParser,\n MBartTokenizer,\n default_data_collator,\n AutoModelWithLMHead,\n set_seed\n)\nmodel_name = \"./models/First/\"\nmodel = AutoModelWithLMHead.from_pretrained(model_name)\ntokenizer = AutoTokenizer.from_pretrained(model_name)\n\nif tokenizer.encode(\"<extra_id_0> <extra_id_1>\") != [32099, 32098, 32097, 1]:\n # For non-t5 tokenizer\n tokenizer.add_special_tokens(\n {\"additional_special_tokens\": [\"<Temp_S>\", \"<Temp_E>\", \"<Relation_S>\", \"<Relation_E>\", \\\n \"<ORG>\", \"<VEH>\", \"<WEA>\", \"<LOC>\",\"<FAC>\",\"<End>\" ,\"<PER>\",\"<GPE>\"]})\n",
"/home/pesante/anaconda3/envs/text2event/lib/python3.8/site-packages/transformers/models/auto/modeling_auto.py:1006: FutureWarning: The class `AutoModelWithLMHead` is deprecated and will be removed in a future version. Please use `AutoModelForCausalLM` for causal language models, `AutoModelForMaskedLM` for masked language models and `AutoModelForSeq2SeqLM` for encoder-decoder models.\n warnings.warn(\n"
],
[
"b = tokenizer.convert_tokens_to_ids([\"<Temp_S>\", \"<Temp_E>\", \"<Relation_S>\", \"<Relation_E>\"])\nprint(b)",
"[32100, 32110, 32101, 32102]\n"
],
[
"tokenizer.add_special_tokens(\n {\"additional_special_tokens\": [\"<Temp_S>\", \"<Temp_E>\", \"<Relation_S>\", \"<Relation_E>\", \\\n \"<ORG>\", \"<VEH>\", \"<WEA>\", \"<LOC>\",\"<FAC>\",\"<End>\" ,\"<PER>\",\"<GPE>\"]})\nb = tokenizer.convert_tokens_to_ids([\"<Temp_S>\", \"<Temp_E>\", \"<Relation_S>\", \"<Relation_E>\"])\nprint(b)\nprint(tokenizer.convert_tokens_to_ids([\"<Temp_S>\"])[0])\nif tokenizer.convert_tokens_to_ids([\"<Temp_S>\"])[0] == 32100:\n print(\"kakd\")",
"[32100, 32110, 32101, 32102]\n32100\nkakd\n"
],
[
"def add_space(text):\n \"\"\"\n add space between special token\n :param text:\n :return:\n \"\"\"\n new_text_list = list()\n for item in zip(split_bracket.findall(text), split_bracket.split(text)[1:]):\n new_text_list += item\n return ' '.join(new_text_list)",
"_____no_output_____"
]
],
[
[
"# Read Data",
"_____no_output_____"
]
],
[
[
"from datasets import load_dataset, load_metric\ndata_files = {}\ndata_files[\"train\"] = './data/new_text2tree/one_ie_ace2005_subtype/train.json'\nextension = 'json'\ndatasets = load_dataset(extension, data_files=data_files)",
"Using custom data configuration default-521fb87c8c936007\n"
],
[
"column_names = datasets[\"train\"].column_names\nprint(column_names)",
"['text', 'relation']\n"
],
[
"train_dataset = datasets[\"train\"]\ntext_column = \"text\"\nsummary_column = \"event\"\ninputs = train_dataset[text_column]\ntargets = train_dataset[summary_column]\npadding = \"max_length\"\nmodel_inputs = tokenizer(inputs, max_length= 256, padding=padding, truncation=True)",
"_____no_output_____"
],
[
"print(type(model_inputs))",
"<class 'transformers.tokenization_utils_base.BatchEncoding'>\n"
]
],
[
[
"# New data format",
"_____no_output_____"
]
],
[
[
"import os\nimport json\nfrom collections import Counter, defaultdict\nfrom data_convert.format.text2tree import Text2Tree\nfrom data_convert.task_format.event_extraction import Event, DyIEPP\nfrom data_convert.utils import read_file, check_output, data_counter_to_table, get_schema, output_schema\nfrom nltk.corpus import stopwords\n\ntarget_class=Text2Tree\ntype_format='subtype'",
"_____no_output_____"
],
[
"import os\nin_filename = \"data/raw_data/ace05-EN/train.oneie.json\"\noutput_filename = \"data/new_text2tree/ace2005_event/dev\"\nif not os.path.exists(output_filename):\n os.makedirs(output_filename)\nevent_output = open(output_filename + '.json', 'w')\n\ncount = 0\nnumber = 0\nfor line in read_file(in_filename):\n document = Event(json.loads(line.strip()))\n if len(document.entities) > 3:\n break\n for sentence in document.generate_relations():\n number += 1\n if(len(sentence['relations']) == 0):\n count += 1\nprint(count)\nprint(number)",
" 0%| | 6/17172 [00:00<00:00, 37560.93it/s]"
],
[
"import os\nin_filename = \"data/raw_data/ace05-EN/train.oneie.json\"\noutput_filename = \"data/new_text2tree/ace2005_event/dev\"\nif not os.path.exists(output_filename):\n os.makedirs(output_filename)\nevent_output = open(output_filename + '.json', 'w')\n\ncount = 0\nnumber = 0\nfor line in read_file(in_filename):\n document = Event(json.loads(line.strip()))\n sentences = document.generate_relations()\n for sentence in sentences:\n if(len(sentence['relations']) > 0):\n print(sentence['relations'])\n break\n break",
" 0%| | 0/17172 [00:00<?, ?it/s]\n"
],
[
"for sentence in document.generate_relations():\n print(sentence['relations'])\n print(\" \")",
"[]\n \n"
],
[
"import os\nin_filename = \"data/datasets/conll04/conll04_train.json\"\ncount = 0\nnumber = 0\nentity_set = set()\nrelation_type_set = set()\nfor line in read_file(in_filename):\n for line_ in json.loads(line.strip()):\n for entity in line_['entities']\n break",
" 0%| | 0/1 [00:00<?, ?it/s]"
],
[
"print(entity_set)\nprint(relation_type_set)",
"{'Metric', 'Task', 'OtherScientificTerm', 'Generic', 'Material', 'Method'}\n{'EVALUATE-FOR', 'PART-OF', 'CONJUNCTION', 'COMPARE', 'FEATURE-OF', 'HYPONYM-OF', 'USED-FOR'}\n"
],
[
"for relations_in_sentence, sentence_start, sentence, entity in zip(document.relations, document.sentence_start, document.sentences, document.ner):\n print(sentence)\n print(relations_in_sentence)\n print(entity)\n print(\" \")",
"['English', 'is', 'shown', 'to', 'be', 'trans-context-free', 'on', 'the', 'basis', 'of', 'coordinations', 'of', 'the', 'respectively', 'type', 'that', 'involve', 'strictly', 'syntactic', 'cross-serial', 'agreement', '.']\n[]\n[[0, 0, 'Material'], [10, 10, 'OtherScientificTerm'], [17, 20, 'OtherScientificTerm']]\n \n['The', 'agreement', 'in', 'question', 'involves', 'number', 'in', 'nouns', 'and', 'reflexive', 'pronouns', 'and', 'is', 'syntactic', 'rather', 'than', 'semantic', 'in', 'nature', 'because', 'grammatical', 'number', 'in', 'English', ',', 'like', 'grammatical', 'gender', 'in', 'languages', 'such', 'as', 'French', ',', 'is', 'partly', 'arbitrary', '.']\n[[29, 29, 31, 32, 'CONJUNCTION'], [48, 49, 51, 51, 'FEATURE-OF'], [54, 54, 51, 51, 'HYPONYM-OF']]\n[[23, 23, 'Generic'], [29, 29, 'OtherScientificTerm'], [31, 32, 'OtherScientificTerm'], [42, 43, 'OtherScientificTerm'], [45, 45, 'Material'], [48, 49, 'OtherScientificTerm'], [51, 51, 'Material'], [54, 54, 'Material']]\n \n['The', 'formal', 'proof', ',', 'which', 'makes', 'crucial', 'use', 'of', 'the', 'Interchange', 'Lemma', 'of', 'Ogden', 'et', 'al.', ',', 'is', 'so', 'constructed', 'as', 'to', 'be', 'valid', 'even', 'if', 'English', 'is', 'presumed', 'to', 'contain', 'grammatical', 'sentences', 'in', 'which', 'respectively', 'operates', 'across', 'a', 'pair', 'of', 'coordinate', 'phrases', 'one', 'of', 'whose', 'members', 'has', 'fewer', 'conjuncts', 'than', 'the', 'other', ';', 'it', 'thus', 'goes', 'through', 'whatever', 'the', 'facts', 'may', 'be', 'regarding', 'constructions', 'with', 'unequal', 'numbers', 'of', 'conjuncts', 'in', 'the', 'scope', 'of', 'respectively', ',', 'whereas', 'other', 'arguments', 'have', 'foundered', 'on', 'this', 'problem', '.']\n[]\n[[70, 71, 'Method'], [86, 86, 'Material']]\n \n"
],
[
"def generate_relations(document):\n relations = list()\n for relation in document.relations:\n arguments = list()\n relation_type = relation['relation_type']\n for argument in relation['arguments']:\n argument_entity = document.entities[argument['entity_id']]\n arguments += [list(range(argument_entity['start'], argument_entity['end']))]\n for old_relation in relations:\n if relation_type == old_relation['type']:\n old_relation['arguments'].append(arguments)\n continue\n relations += [{'type': relation_type, 'arguments': [arguments]}]\n return relations",
"_____no_output_____"
],
[
"relations = generate_relations(document)\nprint(relations)",
"_____no_output_____"
],
[
"def generate_relation(document):\n for relations_in_sentence, sentence_start in zip(document.relations, document.sentence_start):\n relations = list()\n type_set = set()\n for relation in relations_in_sentence:\n# 'arguments': [['Arg-1', [9]], ['Arg-2', [14]]]\n arguments = [list(range(relation[0]-sentence_start, relation[1]+1-sentence_start)),\n list(range(relation[2]-sentence_start,relation[3]+1-sentence_start))]\n relation_type = relation[4].split('.')[0]\n if relation_type in type_set:\n for old_relation in relations:\n if relation_type == old_relation['type']:\n old_relation['arguments'].append(arguments)\n else:\n type_set.add(relation_type)\n relations += [{'type': relation_type, 'arguments': [arguments]}]\n print(relations)",
"_____no_output_____"
],
[
"generate_relation(document)",
"_____no_output_____"
],
[
"Name_Entity_Type = set()\nfor ner, sentence, events_in_sentence, sentence_start in zip(document.ner , document.sentences, document.events, document.sentence_start):\n events = list()\n for event in events_in_sentence:\n trigger, event_type = event[0]\n trigger_ner = ner\n trigger -= sentence_start\n\n suptype, subtype = event_type.split('.')\n\n if type_format == 'subtype':\n event_type = subtype\n elif type_format == 'suptype':\n event_type = suptype\n else:\n event_type = suptype + type_format + subtype\n\n arguments = list()\n for start, end, role in event[1:]:\n start -= sentence_start\n end -= sentence_start\n arguments += [[role, list(range(start, end + 1))]]\n\n for argument in arguments:\n for ner_pos in ner:\n Name_Entity_Type.add(ner_pos[2])\n if((ner_pos[0]-sentence_start) == argument[1][0]):\n argument.insert(1, ner_pos[2])\n if(len(argument) != 3):\n print(\"Wrong\")\n\n event = {'type': event_type, 'tokens': [trigger], 'arguments': arguments}\n\n events += [event]\n# print(events)\n if(len(event['arguments']) >1):\n A_predict = {'tokens': sentence, 'events': events}\n# print(A_predict)\nprint(Name_Entity_Type)",
"_____no_output_____"
],
[
"list_A = ['ORG', 'VEH', 'WEA', 'LOC', 'FAC', 'PER', 'GPE']\nfor a in list_A:\n entity_token = tokenizer.encode(a)\n print(entity_token)\nprint(tokenizer.encode('<extra_id_0>'))",
"[4674, 517, 1]\n[3, 8575, 566, 1]\n[9664, 188, 1]\n[301, 5618, 1]\n[377, 5173, 1]\n[3, 8742, 1]\n[350, 5668, 1]\n[32099, 1]\n"
],
[
"entity_dic = {'ORG', 'VEH', 'WEA', 'LOC', 'FAC', 'PER', 'GPE'}\ntokenizer.add_special_tokens({\"additional_special_tokens\": list_A})\nmodel.resize_token_embeddings(len(tokenizer))\nprint(tokenizer.convert_tokens_to_ids(list_A))\nif 'ORG' in list_A:\n print(\"Wrong\")\n \nprint(tokenizer.decode(32103))\nprint(tokenizer.pad_token_id)\na = 1.213232\nb = round(a, 4)\nprint(b)",
"[32100, 32101, 32102, 32103, 32104, 8742, 32105]\nWrong\nLOC\n0\n1.2132\n"
]
],
[
[
"# OneIE data processing",
"_____no_output_____"
]
],
[
[
"events = list()\n# print(document.events)\n# print(document.entities)\nfor event, entity in zip(document.events, document.entities):\n# print(event)\n# print(entity)\n arguments = list()\n for argument in event['arguments']:\n argument_entity = document.entities[argument['entity_id']]\n# print(\"argument_entity\", argument_entity)\n arguments += [[argument['role'], argument_entity['entity_type'] ,list(range(argument_entity['start'], argument_entity['end']))]]\n\n suptype, subtype = event['event_type'].split(':')\n\n if type_format == 'subtype':\n event_type = subtype\n elif type_format == 'suptype':\n event_type = suptype\n else:\n event_type = suptype + type_format + subtype\n\n events += [{\n 'type': event_type,\n 'tokens': list(range(event['trigger']['start'], event['trigger']['end'])),\n 'arguments': arguments\n }]\n print(events)",
"[{'type': 'Attack', 'tokens': [8], 'arguments': []}]\n[{'type': 'Attack', 'tokens': [8], 'arguments': []}, {'type': 'End-Position', 'tokens': [10], 'arguments': [['Person', 'PER', [12, 13]], ['Entity', 'GPE', [17]]]}]\n"
],
[
"for ner, sentence, events_in_sentence, sentence_start in zip(document.ner , document.sentences, document.events, document.sentence_start):\n if(len(ner) > 0):\n print((ner))",
"_____no_output_____"
]
],
[
[
"# Annotated from event text to tree",
"_____no_output_____"
]
],
[
[
"from data_convert.utils import read_file, check_output, data_counter_to_table, get_schema, output_schema\nfrom nltk.corpus import stopwords\n\ntype_start = '<extra_id_0>'\ntype_end = '<extra_id_1>'\nrole_start = '<extra_id_2>'\nrole_end = '<extra_id_3>'\n\ndef get_str_from_tokens(tokens, sentence, separator=' '):\n start, end_exclude = tokens[0], tokens[-1] + 1\n return separator.join(sentence[start:end_exclude])",
"_____no_output_____"
],
[
"event_schema_set = set()\n\nfor event in A_predict['events']:\n print(event)\n event_schema_set = event_schema_set | get_schema(event)\n sep = ' '\n predicate = sep.join([A_predict['tokens'][index]\n for index in event['tokens']])\n# counter['pred'].update([predicate])\n# counter['type'].update([event['type']])\n# data_counter[in_filename].update(['event'])\n# for argument in event['arguments']:\n# data_counter[in_filename].update(['argument'])\n# counter['role'].update([argument[0]])\n\nprint(predicate)\nprint(event_schema_set)",
"{'type': 'Attack', 'tokens': [7], 'arguments': [['Attacker', 'PER', [5]], ['Attacker', 'PER', [14]], ['Attacker', 'PER', [19]]]}\nfight\n{('Attack', 'Attacker')}\n"
],
[
"tokens=A_predict['tokens']\npredicate_arguments=A_predict['events']\n\ntoken_separator = ' '\n\nevent_str_rep_list = list()\n\nfor predicate_argument in predicate_arguments:\n event_type = predicate_argument['type']\n\n # predicate_argument['tokens'] is the trigger index\n # tokens is the sentence tokens, we get the trigger text span here\n predicate_text = get_str_from_tokens(predicate_argument['tokens'], tokens, separator=token_separator)\n\n # prefix_tokens[predicate_argument['tokens'][0]] = ['[ ']\n # suffix_tokens[predicate_argument['tokens'][-1]] = [' ]']\n\n role_str_list = list()\n # role_name is the argument role, role_tokens are corresponding text span index\n for role_name, role_entity, role_tokens in predicate_argument['arguments']:\n # if role_name == 'Place' or role_name.startswith('Time'):\n if role_name == event_type:\n continue\n # get the role text span from role tokens index\n role_text = get_str_from_tokens(role_tokens, tokens, separator=token_separator)\n# print(role_text)\n# print(role_entity)\n if False:\n role_str = ' '.join([role_start, role_name, role_entity ,role_text, role_end])\n else:\n role_str = ' '.join([type_start, role_name,role_entity ,role_text, type_end])\n # All arguments in the sentence\n# print(role_str)\n role_str_list += [role_str]\n role_str_list_str = ' '.join(role_str_list)\n event_str_rep = f\"{type_start} {event_type} {predicate_text} {role_str_list_str} {type_end}\"\n event_str_rep_list += [event_str_rep]\n\n# print(tokens) \nsource_text = token_separator.join(tokens)\ntarget_text = ' '.join(event_str_rep_list)\n\nif not False:\n target_text = f'{type_start} ' + \\\n ' '.join(event_str_rep_list) + f' {type_end}'\n\nprint(source_text) \nprint(\" \")\nprint(target_text)",
"Britain has deployed some 45,000 troops to fight with the more than 250,000 US soldiers lined up against Iraqi troops .\n \n<extra_id_0> <extra_id_0> Attack fight <extra_id_0> Attacker PER troops <extra_id_1> <extra_id_0> Attacker PER soldiers <extra_id_1> <extra_id_0> Attacker PER troops <extra_id_1> <extra_id_1> <extra_id_1>\n"
],
[
"a = [\"%s%s\" % (\"type_start \", \"type_end\")] * 2\nprint(a)",
"['type_start type_end', 'type_start type_end']\n"
]
],
[
[
"# Get Label Name",
"_____no_output_____"
]
],
[
[
"def get_label_name_tree(label_name_list, tokenizer, end_symbol='<end>'):\n # Change recurring into non-recurring labels, \n sub_token_tree = dict()\n\n # this is label_name token ids\n label_tree = dict()\n for typename in label_name_list:\n# print(typename)\n after_tokenized = tokenizer.encode(typename)\n label_tree[typename] = after_tokenized\n print(label_tree)\n for _, sub_label_seq in label_tree.items():\n # sub_label_seq is the tokenize_ids of typename\n parent = sub_token_tree\n for value in sub_label_seq:\n if value not in parent:\n parent[value] = dict()\n parent = parent[value]\n parent[end_symbol] = None\n\n return sub_token_tree",
"_____no_output_____"
],
[
"from extraction.event_schema import EventSchema\nevent_schema = './data/text2tree/dyiepp_ace2005_subtype/event.schema'\ndecoding_type_schema = EventSchema.read_from_file(event_schema)\n# print(decoding_type_schema.type_list)\n# print(decoding_type_schema.role_list)\n# print(decoding_type_schema.type_role_dict)\ntype_tree = get_label_name_tree(decoding_type_schema.type_list, tokenizer, end_symbol='<tree-end>')",
"{'Extradite': [8505, 10700, 1], 'Sentence': [4892, 17, 1433, 1], 'Trial-Hearing': [20660, 18, 3845, 9, 1007, 1], 'Die': [316, 1], 'Elect': [3, 21543, 1], 'Declare-Bankruptcy': [28596, 60, 18, 21347, 9433, 75, 63, 1], 'Appeal': [25024, 1], 'Attack': [24655, 1], 'Merge-Org': [4039, 397, 18, 7395, 122, 1], 'Charge-Indict': [15907, 18, 1570, 12194, 1], 'Release-Parole': [13048, 18, 13212, 32, 109, 1], 'Divorce': [2043, 1967, 565, 1], 'Pardon': [2180, 2029, 1], 'Sue': [17564, 1], 'Start-Position': [3273, 18, 345, 32, 7, 4749, 1], 'Meet': [12325, 1], 'Demonstrate': [15782, 29, 7, 17, 2206, 1], 'Execute': [25183, 15, 1], 'Convict': [1193, 7287, 17, 1], 'Transport': [7608, 1], 'Transfer-Ownership': [9900, 18, 667, 210, 687, 2009, 1], 'Arrest-Jail': [1533, 6216, 18, 683, 9, 173, 1], 'Start-Org': [3273, 18, 7395, 122, 1], 'Transfer-Money': [9900, 18, 9168, 15, 63, 1], 'Phone-Write': [8924, 18, 24965, 15, 1], 'End-Position': [3720, 18, 345, 32, 7, 4749, 1], 'Fine': [11456, 1], 'End-Org': [3720, 18, 7395, 122, 1], 'Acquit': [4292, 10073, 1], 'Nominate': [465, 51, 8660, 1], 'Be-Born': [493, 18, 279, 127, 29, 1], 'Injure': [86, 10609, 15, 1], 'Marry': [1571, 651, 1]}\n"
],
[
"# print(list(type_tree.keys()))\n# print(\" \")\nsubtree = type_tree[9900][18][667]\nprint(len(subtree))\nprint(subtree)\nif '<tree-end>' in subtree:\n print(\"end_tree\")",
"1\n{210: {687: {2009: {1: {'<tree-end>': None}}}}}\n"
],
[
"role_tree = get_label_name_tree(decoding_type_schema.role_list, tokenizer, end_symbol='<tree-end>')\nprint(role_tree.keys())\nprint(role_tree)",
"{'Plaintiff': [30837, 1], 'Instrument': [13507, 1], 'Person': [5780, 1], 'Entity': [4443, 485, 1], 'Beneficiary': [30570, 1208, 1], 'Adjudicator': [1980, 14312, 447, 1016, 1], 'Seller': [19980, 1], 'Prosecutor': [749, 7, 15, 3044, 127, 1], 'Artifact': [1261, 23, 8717, 1], 'Vehicle': [15095, 1], 'Attacker': [24655, 49, 1], 'Origin': [19477, 1], 'Victim': [12060, 2998, 1], 'Defendant': [3, 16196, 989, 288, 1], 'Agent': [8628, 1], 'Org': [955, 122, 1], 'Buyer': [19099, 1], 'Place': [3399, 1], 'Destination': [19344, 257, 1], 'Target': [12615, 1], 'Recipient': [419, 3389, 4741, 1], 'Giver': [6434, 52, 1]}\ndict_keys([30837, 13507, 5780, 4443, 30570, 1980, 19980, 749, 1261, 15095, 24655, 19477, 12060, 3, 8628, 955, 19099, 3399, 19344, 12615, 419, 6434])\n{30837: {1: {'<tree-end>': None}}, 13507: {1: {'<tree-end>': None}}, 5780: {1: {'<tree-end>': None}}, 4443: {485: {1: {'<tree-end>': None}}}, 30570: {1208: {1: {'<tree-end>': None}}}, 1980: {14312: {447: {1016: {1: {'<tree-end>': None}}}}}, 19980: {1: {'<tree-end>': None}}, 749: {7: {15: {3044: {127: {1: {'<tree-end>': None}}}}}}, 1261: {23: {8717: {1: {'<tree-end>': None}}}}, 15095: {1: {'<tree-end>': None}}, 24655: {49: {1: {'<tree-end>': None}}}, 19477: {1: {'<tree-end>': None}}, 12060: {2998: {1: {'<tree-end>': None}}}, 3: {16196: {989: {288: {1: {'<tree-end>': None}}}}}, 8628: {1: {'<tree-end>': None}}, 955: {122: {1: {'<tree-end>': None}}}, 19099: {1: {'<tree-end>': None}}, 3399: {1: {'<tree-end>': None}}, 19344: {257: {1: {'<tree-end>': None}}}, 12615: {1: {'<tree-end>': None}}, 419: {3389: {4741: {1: {'<tree-end>': None}}}}, 6434: {52: {1: {'<tree-end>': None}}}}\n"
],
[
"print(role_tree)\nif('<tree-end>' in role_tree):\n print(\"Wrong\")",
"{30837: {1: {'<tree-end>': None}}, 13507: {1: {'<tree-end>': None}}, 5780: {1: {'<tree-end>': None}}, 4443: {485: {1: {'<tree-end>': None}}}, 30570: {1208: {1: {'<tree-end>': None}}}, 1980: {14312: {447: {1016: {1: {'<tree-end>': None}}}}}, 19980: {1: {'<tree-end>': None}}, 749: {7: {15: {3044: {127: {1: {'<tree-end>': None}}}}}}, 1261: {23: {8717: {1: {'<tree-end>': None}}}}, 15095: {1: {'<tree-end>': None}}, 24655: {49: {1: {'<tree-end>': None}}}, 19477: {1: {'<tree-end>': None}}, 12060: {2998: {1: {'<tree-end>': None}}}, 3: {16196: {989: {288: {1: {'<tree-end>': None}}}}}, 8628: {1: {'<tree-end>': None}}, 955: {122: {1: {'<tree-end>': None}}}, 19099: {1: {'<tree-end>': None}}, 3399: {1: {'<tree-end>': None}}, 19344: {257: {1: {'<tree-end>': None}}}, 12615: {1: {'<tree-end>': None}}, 419: {3389: {4741: {1: {'<tree-end>': None}}}}, 6434: {52: {1: {'<tree-end>': None}}}}\n"
],
[
"first_tree = role_tree[30837]\nprint(first_tree[1])\nif '<tree-end>' in first_tree[1]:\n print(\"Wrong\")",
"{'<tree-end>': None}\nWrong\n"
]
],
[
[
"# Transform output to Tree",
"_____no_output_____"
]
],
[
[
"def convert_bracket(_text):\n # replace the special token labels to Formal statement\n _text = add_space(_text)\n for start in [role_start, type_start]:\n _text = _text.replace(start, left_bracket)\n for end in [role_end, type_end]:\n _text = _text.replace(end, right_bracket)\n return _text\n\ndef add_space(text):\n \"\"\"\n add space between special token\n :param text:\n :return:\n \"\"\"\n new_text_list = list()\n for item in zip(split_bracket.findall(text), split_bracket.split(text)[1:]):\n new_text_list += item\n return ' '.join(new_text_list)\n\ndef get_tree_str(tree):\n \"\"\"\n get str from event tree\n :param tree:\n :return:\n \"\"\"\n str_list = list()\n for element in tree:\n if isinstance(element, str):\n str_list += [element]\n return ' '.join(str_list)",
"_____no_output_____"
],
[
"import re\nleft_bracket = '【'\nright_bracket = '】'\ntext = \"<extra_id_0> <extra_id_0> ORG-AFF <extra_id_0> Minister <extra_id_0> British <extra_id_1> <extra_id_1> ORG-AFF <extra_id_1> <extra_id_0> PART-WHOLE <extra_id_0> Grand Hotel Europe <extra_id_0> Saint Petersburg <extra_id_1> <extra_id_1> PART-WHOLE <extra_id_1> <extra_id_1>\"\nfrom nltk.tree import ParentedTree\nbrackets = left_bracket + right_bracket\nprint(text)",
"<extra_id_0> <extra_id_0> ORG-AFF <extra_id_0> Minister <extra_id_0> British <extra_id_1> <extra_id_1> ORG-AFF <extra_id_1> <extra_id_0> PART-WHOLE <extra_id_0> Grand Hotel Europe <extra_id_0> Saint Petersburg <extra_id_1> <extra_id_1> PART-WHOLE <extra_id_1> <extra_id_1>\n"
],
[
"split_bracket = re.compile(r\"<extra_id_\\d>\")\nnew_text_list = list()\nfor item in zip(split_bracket.findall(text), split_bracket.split(text)[1:]):\n new_text_list += item\nprint(' '.join(new_text_list))\n",
"<extra_id_0> <extra_id_0> ORG-AFF <extra_id_0> Minister <extra_id_0> British <extra_id_1> <extra_id_1> ORG-AFF <extra_id_1> <extra_id_0> PART-WHOLE <extra_id_0> Grand Hotel Europe <extra_id_0> Saint Petersburg <extra_id_1> <extra_id_1> PART-WHOLE <extra_id_1> <extra_id_1> \n"
],
[
"new_text = convert_bracket(text)\nprint(new_text)",
"【 【 ORG-AFF 【 Minister 【 British 】 】 ORG-AFF 】 【 PART-WHOLE 【 Grand Hotel Europe 【 Saint Petersburg 】 】 PART-WHOLE 】 】 \n"
],
[
"gold_tree = ParentedTree.fromstring(new_text, brackets=brackets)",
"_____no_output_____"
],
[
"print(gold_tree)",
"(\n (ORG-AFF (Minister (British )) ORG-AFF)\n (PART-WHOLE (Grand Hotel Europe (Saint Petersburg)) PART-WHOLE))\n"
],
[
"str_list = list()\nfor relation_tree in gold_tree:\n print(relation_tree.label())\n for role_tree in relation_tree:\n if isinstance(role_tree, str):\n print(\"end\", role_tree)\n else:\n role1_text = role_tree.label() + ' ' + get_tree_str(role_tree)\n role2_text = role_tree[-1].label() + ' '+get_tree_str(role_tree[-1])\n print(role1_text + \" aa \"+ role2_text)\n# print(role_tree.label()+ ' ' +a)\n# print(role_tree[-1].label() + ' '+get_tree_str(role_tree[-1]))\n# print(role_tree[0])",
"ORG-AFF\nMinister aa British \nend ORG-AFF\nPART-WHOLE\nGrand Hotel Europe aa Saint Petersburg\nend PART-WHOLE\n"
],
[
"kind = ' '.join(str_list)\nprint(kind)\nnew_kind = ' '.join(kind.split(' ')[1:])\nprint(new_kind)",
"\n\n"
],
[
"b = []\na = [\"LOC\", \"ORG\", \"VEH\", \"FAC\", \"PER\", \"WEA\", \"GPE\"]\na.append(\" \")\nb =a\nprint(b)",
"['LOC', 'ORG', 'VEH', 'FAC', 'PER', 'WEA', 'GPE', ' ']\n"
],
[
"c = b+[1]\nprint(c)",
"['LOC', 'ORG', 'VEH', 'FAC', 'PER', 'WEA', 'GPE', ' ', 1]\n"
],
[
"print(tokenizer.encode(\" \"))",
"[1]\n"
],
[
"a = tokenizer.decode(3)\nb = [1,2,3]\nb.append(a)\nprint(b)",
"[1, 2, 3, '']\n"
],
[
"c = [0,1]\nprint(c+a)",
"[0, 1, 1548, 3, 6, 38, 17952, 3859, 3292, 3457, 3, 6, 79, 33, 6326, 53, 3, 9, 775, 682, 13, 4719, 3, 6, 652, 5413, 13, 8, 7749, 9534, 45, 8, 10101, 3, 5]\n"
],
[
"a = [1548, 3, 6, 38, 17952, 3859, 3292, 3457, 3, 6, 79, 33, 6326, 53, 3, 9, 775, 682, 13, 4719, 3, 6, 652, 5413, 13, 8, 7749, 9534, 45, 8, 10101, 3, 5]\nb = list()\nfor x in a:\n b.append(tokenizer.decode(x))\nprint(' '.join(b))",
"Well , as coalition forces push north , they are encounter ing a unique problem of combat , getting rid of the weapons captured from the enemy .\n"
],
[
"import torch\nlabels = torch.tensor(a)\ndecoded_labels = tokenizer.batch_decode(labels, skip_special_tokens=False)\nprint(decoded_labels)\nprint(' '.join(decoded_labels))",
"['Well', '', ',', 'as', 'coalition', 'forces', 'push', 'north', '', ',', 'they', 'are', 'encounter', 'ing', '', 'a', 'unique', 'problem', 'of', 'combat', '', ',', 'getting', 'rid', 'of', 'the', 'weapons', 'captured', 'from', 'the', 'enemy', '', '.']\nWell , as coalition forces push north , they are encounter ing a unique problem of combat , getting rid of the weapons captured from the enemy .\n"
],
[
"print(type([3]))",
"<class 'list'>\n"
],
[
"special_token_set = {-1,-2}\ntgt_generated = [1,2,3,-1,2,3,2,-1,4,3,-2,4,-2]\nspecial_index_token = list(filter(lambda x: x[1] in special_token_set, list(enumerate(tgt_generated))))\nprint(special_index_token)",
"[(3, -1), (7, -1), (10, -2), (12, -2)]\n"
],
[
"Target = [0, 32099, 32099, 4674, 517, 18, 188, 9089, 32099, 7471, 32099, 13143, 32099, 1]\nTarget = [0, 32099, 32099, 3, 19846, 18, 518, 6299, 3765, 32099, 3, 19600, 989, 23, 12489, 32099, 783, 563, 32099, 1]\ntext = tokenizer.decode(Target)\nprint(text)",
"<pad><extra_id_0><extra_id_0> PART-WHOLE<extra_id_0> Vivendi Universal<extra_id_0> media group<extra_id_0></s>\n"
],
[
"Source = [2, 0, 7, 7, 7, 5]\ntext2 = tokenizer.decode(Source)\nprint(text2)",
"_____no_output_____"
],
[
"add_special_tokens\nprint(tokenizer.sep_token)",
"Using sep_token, but it is not set yet.\n"
]
],
[
[
"# Test List in Python",
"_____no_output_____"
]
],
[
[
"record_list = list()\nrecord = {'relation': int}\nrecord['relation'] = 1\nrecord_list += [record]\nprint(record_list)",
"[{'relation': 1}]\n"
],
[
"record['relation'] = 2\nrecord_list += [record]\nprint(record_list)",
"[{'relation': 2}, {'relation': 2}]\n"
],
[
"tokenizer.decode(2)",
"_____no_output_____"
],
[
"from transformers import AutoModelForSeq2SeqLM, AutoTokenizer\n\ndef extract_triplets(text):\n triplets = []\n relation, subject, relation, object_ = '', '', '', ''\n text = text.strip()\n current = 'x'\n for token in text.replace(\"<s>\", \"\").replace(\"<pad>\", \"\").replace(\"</s>\", \"\").split():\n if token == \"<triplet>\":\n current = 't'\n if relation != '':\n triplets.append({'head': subject.strip(), 'type': relation.strip(),'tail': object_.strip()})\n relation = ''\n subject = ''\n elif token == \"<subj>\":\n current = 's'\n if relation != '':\n triplets.append({'head': subject.strip(), 'type': relation.strip(),'tail': object_.strip()})\n object_ = ''\n elif token == \"<obj>\":\n current = 'o'\n relation = ''\n else:\n if current == 't':\n subject += ' ' + token\n elif current == 's':\n object_ += ' ' + token\n elif current == 'o':\n relation += ' ' + token\n if subject != '' and relation != '' and object_ != '':\n triplets.append({'head': subject.strip(), 'type': relation.strip(),'tail': object_.strip()})\n return triplets\n\n# Load model and tokenizer\ntokenizer = AutoTokenizer.from_pretrained(\"Babelscape/rebel-large\")\nmodel = AutoModelForSeq2SeqLM.from_pretrained(\"Babelscape/rebel-large\")\ngen_kwargs = {\n \"max_length\": 256,\n \"length_penalty\": 0,\n \"num_beams\": 3,\n \"num_return_sequences\": 3,\n}\n\n# Text to extract triplets from\ntext = 'Punta Cana is a resort town in the municipality of Higüey, in La Altagracia Province, the easternmost province of the Dominican Republic.'\n\n# Tokenizer text\nmodel_inputs = tokenizer(text, max_length=256, padding=True, truncation=True, return_tensors = 'pt')\n\n# Generate\ngenerated_tokens = model.generate(\n model_inputs[\"input_ids\"].to(model.device),\n attention_mask=model_inputs[\"attention_mask\"].to(model.device),\n **gen_kwargs,\n)\n\n# Extract text\ndecoded_preds = tokenizer.batch_decode(generated_tokens, skip_special_tokens=False)\n\n# Extract triplets\nfor idx, sentence in enumerate(decoded_preds):\n print(f'Prediction triplets sentence {idx}')\n print(extract_triplets(sentence))",
"_____no_output_____"
],
[
"tokenizer = AutoTokenizer.from_pretrained(\"Babelscape/rebel-large\")\nmodel_name_or_path = \"Babelscape/rebel-large\"\nmodel = AutoModelForSeq2SeqLM.from_pretrained(model_name_or_path)",
"_____no_output_____"
]
],
[
[
"# Get the label data",
"_____no_output_____"
]
],
[
[
"output_filename = './data/new_text2tree/ace_label/'\nif not os.path.exists(output_filename):\n os.makedirs(output_filename)\nimport json\nwith open('./data/new_text2tree/one_ie_ace2005_subtype/test.json', 'r', encoding=\"utf-8\") as f:\n # 读取所有行 每行会是一个字符串\n for jsonstr in f.readlines():\n # 将josn字符串转化为dict字典\n jsonstr = json.loads(jsonstr)\n with open(output_filename + 'test_end_before.json', 'a+', encoding='utf-8') as f2:\n line = json.dumps(jsonstr[\"relation\"], ensure_ascii=False)\n f2.write(line+'\\n')\n",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb92c06a733617c824808fbeb198e414ada7ca09
| 10,078 |
ipynb
|
Jupyter Notebook
|
Notebooks/1-Merging Data.ipynb
|
AtefEl-Hennawy/Arabic-Dialect-Identification
|
499129613d2917390d18b164723f4c5843382c3b
|
[
"MIT"
] | 1 |
2022-03-16T13:12:02.000Z
|
2022-03-16T13:12:02.000Z
|
Notebooks/1-Merging Data.ipynb
|
AtefEl-Hennawy/Arabic-Dialect-Identification
|
499129613d2917390d18b164723f4c5843382c3b
|
[
"MIT"
] | null | null | null |
Notebooks/1-Merging Data.ipynb
|
AtefEl-Hennawy/Arabic-Dialect-Identification
|
499129613d2917390d18b164723f4c5843382c3b
|
[
"MIT"
] | null | null | null | 31.009231 | 90 | 0.426474 |
[
[
[
"import pandas as pd ",
"_____no_output_____"
],
[
"# original Dataset without text\noriginal = pd.read_csv('dialect_dataset.csv', index_col=0)\noriginal.size",
"_____no_output_____"
],
[
"# Dataset after fetching Data\napi_re = pd.read_csv('Data.csv', index_col=0)\napi_re.size",
"_____no_output_____"
],
[
"api_re",
"_____no_output_____"
],
[
"# merging in ID to make sure nothing is missing\nto_work = original.join(api_re)\nto_work",
"_____no_output_____"
],
[
"to_work.to_csv('Work.csv')",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb92c941eff0217ae05cdfefe1090bf2db8361df
| 6,999 |
ipynb
|
Jupyter Notebook
|
fairness_benchmark/notebooks/hyperparameter_table.ipynb
|
som-shahlab/fairness_benchmark
|
7d9e2619bf636acb2a9261ed7b4fdf59c105b8a1
|
[
"MIT"
] | 9 |
2020-07-22T09:01:40.000Z
|
2022-02-03T18:44:15.000Z
|
fairness_benchmark/notebooks/hyperparameter_table.ipynb
|
som-shahlab/fairness_benchmark
|
7d9e2619bf636acb2a9261ed7b4fdf59c105b8a1
|
[
"MIT"
] | null | null | null |
fairness_benchmark/notebooks/hyperparameter_table.ipynb
|
som-shahlab/fairness_benchmark
|
7d9e2619bf636acb2a9261ed7b4fdf59c105b8a1
|
[
"MIT"
] | 2 |
2020-07-26T01:28:05.000Z
|
2021-12-07T13:16:52.000Z
| 30.697368 | 151 | 0.496928 |
[
[
[
"import os\nimport numpy as np\nimport pandas as pd\nimport glob\nfrom prediction_utils.util import yaml_read, df_dict_concat",
"_____no_output_____"
],
[
"table_path = '../figures/hyperparameters/'\nos.makedirs(table_path, exist_ok = True)",
"_____no_output_____"
],
[
"param_grid_base = {\n \"lr\": [1e-3, 1e-4, 1e-5],\n \"batch_size\": [128, 256, 512],\n \"drop_prob\": [0.0, 0.25, 0.5, 0.75],\n \"num_hidden\": [1, 2, 3],\n \"hidden_dim\": [128, 256],\n}\n\nthe_dict = {'hyperparameter': [], 'Grid': []}\nfor key, value in param_grid_base.items():\n the_dict['hyperparameter'].append(key)\n the_dict['Grid'].append(value)\nthe_df = pd.DataFrame(the_dict)\nrename_grid = {\n 'hyperparameter': ['lr', 'batch_size', 'drop_prob', 'num_hidden', 'hidden_dim'],\n 'Hyperparameter': ['Learning Rate', 'Batch Size', 'Dropout Probability', 'Number of Hidden Layers', 'Hidden Dimension']\n}\nrename_df = pd.DataFrame(rename_grid)\nthe_df = the_df.merge(rename_df)[['Hyperparameter', 'Grid']].sort_values('Hyperparameter')",
"_____no_output_____"
],
[
"the_df",
"_____no_output_____"
],
[
"the_df.to_latex(os.path.join(table_path, 'param_grid.txt'), index=False)",
"_____no_output_____"
],
[
"selected_models_path = '/share/pi/nigam/projects/spfohl/cohorts/admissions/optum/experiments/baseline_tuning_fold_1/config/selected_models'",
"_____no_output_____"
],
[
"selected_models_path_dict = {\n 'starr': '/share/pi/nigam/projects/spfohl/cohorts/admissions/starr_20200523/experiments/baseline_tuning_fold_1_10/config/selected_models',\n 'mimic': '/share/pi/nigam/projects/spfohl/cohorts/admissions/mimic_omop/experiments/baseline_tuning_fold_1_10/config/selected_models',\n 'optum': '/share/pi/nigam/projects/spfohl/cohorts/admissions/optum/experiments/baseline_tuning_fold_1/config/selected_models',\n}\nselected_param_dict = {\n db: {\n task: yaml_read(glob.glob(os.path.join(db_path, task, '*.yaml'), recursive=True)[0]) for task in os.listdir(db_path)\n }\n for db, db_path in selected_models_path_dict.items()\n}",
"_____no_output_____"
],
[
"col_order = {\n 'starr': ['hospital_mortality', 'LOS_7', 'readmission_30'],\n 'mimic': ['los_icu_3days', 'los_icu_7days', 'mortality_hospital', 'mortality_icu'],\n 'optum': ['readmission_30', 'LOS_7'],\n}",
"_____no_output_____"
],
[
"for db in selected_param_dict.keys():\n db_params = selected_param_dict[db]\n db_df = (\n pd.concat({key: pd.DataFrame(value, index= [0]) for key, value in db_params.items()})\n .reset_index(level=1,drop=True)\n .rename_axis('task')\n .transpose()\n .rename_axis('hyperparameter')\n .reset_index()\n .merge(rename_df, how ='right')\n )\n db_df = db_df[['Hyperparameter'] + list(set(db_df.columns) - set(['hyperparameter', 'Hyperparameter']))].sort_values('Hyperparameter')\n db_df = db_df[['Hyperparameter'] + col_order[db]].sort_values('Hyperparameter')\n selected_param_path = os.path.join(table_path, db)\n os.makedirs(selected_param_path, exist_ok=True)\n db_df.to_latex(os.path.join(selected_param_path, 'selected_param_table.txt'), index=False)",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb92d4e2019f32a8a1056416d41a841bde9d9929
| 32,407 |
ipynb
|
Jupyter Notebook
|
Time_Series_Forecasting/Energy_Consumption_Exercise.ipynb
|
tianyiran02/udacity_CN-ML_SageMaker_Studies
|
6a63445f4c5391a932b137aa29ac97f91a0e9700
|
[
"MIT"
] | 4 |
2020-05-29T02:48:21.000Z
|
2020-10-25T03:31:37.000Z
|
Time_Series_Forecasting/Energy_Consumption_Exercise.ipynb
|
tianyiran02/udacity_CN-ML_SageMaker_Studies
|
6a63445f4c5391a932b137aa29ac97f91a0e9700
|
[
"MIT"
] | null | null | null |
Time_Series_Forecasting/Energy_Consumption_Exercise.ipynb
|
tianyiran02/udacity_CN-ML_SageMaker_Studies
|
6a63445f4c5391a932b137aa29ac97f91a0e9700
|
[
"MIT"
] | 9 |
2020-02-16T08:10:17.000Z
|
2020-10-03T11:36:21.000Z
| 28.402279 | 228 | 0.562595 |
[
[
[
"# 时间序列预测\n\n时间序列是随着时间的推移定期收集的数据。时间序列预测是指根据历史数据预测未来数据点的任务。时间序列预测用途很广泛，包括天气预报、零售和销量预测、股市预测，以及行为预测（例如预测一天的车流量）。时间序列数据有很多，识别此类数据中的模式是很活跃的机器学习研究领域。\n\n<img src='notebook_ims/time_series_examples.png' width=80% />\n\n在此 notebook 中，我们将学习寻找时间规律的一种方法，即使用 SageMaker 的监督式学习模型 [DeepAR](https://docs.aws.amazon.com/sagemaker/latest/dg/deepar.html)。\n\n\n### DeepAR\n\nDeepAR 使用循环神经网络 (RNN)，它会接受序列数据点作为历史输入，并生成预测序列数据点。这种模型如何学习？\n\n在训练过程中，你需要向 DeepAR estimator 提供训练数据集（由多个时间序列组成）。该 estimator 会查看所有的训练时间序列并尝试发现它们之间的相似性。它通过从训练时间序列中随机抽取**训练样本**进行训练。\n* 每个训练样本都由相邻的**上下文**和**预测**窗口（长度已提前固定好）对组成。\n * `context_length` 参数会控制模型能看到过去多久的数据。\n * `prediction_length` 参数会控制模型可以对未来多久做出预测。\n * 详情请参阅[此文档](https://docs.aws.amazon.com/sagemaker/latest/dg/deepar_how-it-works.html)。\n\n<img src='notebook_ims/context_prediction_windows.png' width=50% />\n\n> 因为 DeepAR 用多个时间序列进行训练，所以很适合有**循环规律**的数据。\n\n在任何预测任务中，选择的上下文窗口都应该能向模型提供足够的**相关**信息，这样才能生成准确的预测。通常，最接近预测时间帧的数据包含的信息对确定预测结果的影响最大。在很多预测应用中，例如预测月销量，上下文和预测窗口大小一样，但有时候有必要设置更大的上下文窗口，从而发现数据中的更长期规律。\n\n### 能耗数据\n\n在此 notebook 中，我们将使用的数据是全球的家庭耗电量数据。数据集来自 [Kaggle](https://www.kaggle.com/uciml/electric-power-consumption-data-set)，表示从 2006 年到 2010 年的耗电量数据。对于这么庞大的数据集，我们可以预测很长时间的耗电量，例如几天、几周或几个月。预测能耗有很多用途，例如确定耗电量的季节性价格，以及根据预测用量有效地向居民供电。\n\n**趣味阅读**：Google DeepMind 最近展开了一项逆相关项目，他们使用机器学习预测风力发电机产生的电量，并有效地将电力输送给电网。你可以在[这篇帖子](https://deepmind.com/blog/machine-learning-can-boost-value-wind-energy/)中详细了解这项研究。\n\n### 机器学习工作流程\n\n此 notebook 将时间序列预测分成了以下几个步骤：\n* 加载和探索数据\n* 创建时间序列训练集和测试集\n* 将数据变成 JSON 文件并上传到 S3\n* 实例化和训练 DeepAR estimator\n* 部署模型并创建预测器\n* 评估预测器\n\n---\n\n首先加载常规资源。",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\n%matplotlib inline",
"_____no_output_____"
]
],
[
[
"# 加载和探索数据\n\n我们将收集在几年内收集的全球能耗数据。以下单元格将加载并解压缩此数据，并为你提供一个文本数据文件 `household_power_consumption.txt`。",
"_____no_output_____"
]
],
[
[
"! wget https://s3.amazonaws.com/video.udacity-data.com/topher/2019/March/5c88a3f1_household-electric-power-consumption/household-electric-power-consumption.zip\n! unzip household-electric-power-consumption",
"_____no_output_____"
]
],
[
[
"### 读取 `.txt` 文件\n\n下个单元格显示了文本文件里的前几行，我们可以看看数据格式。",
"_____no_output_____"
]
],
[
[
"# display first ten lines of text data\nn_lines = 10\n\nwith open('household_power_consumption.txt') as file:\n head = [next(file) for line in range(n_lines)]\n \ndisplay(head)",
"_____no_output_____"
]
],
[
[
"## 预处理数据\n\nhousehold_power_consumption.txt 文件具有以下属性：\n * 每个数据点都具有日期和时间记录 (时:分:秒)\n * 各个数据特征用分号 (;) 分隔\n * 某些数据为“nan”或“?”，我们将它们都当做 `NaN` 值\n\n### 处理 `NaN` 值\n\n此 DataFrame 包含一些缺失值的数据点。到目前为止，我们只是丢弃这些值，但是还有其他处理 `NaN` 值的方式。一种技巧是用缺失值所在列的**均值**填充；这样填充的值可能比较符合实际。\n\n我在 `txt_preprocessing.py` 中提供了一些辅助函数，可以帮助你将原始文本文件加载为 DataFrame，并且用各列的平均特征值填充 `NaN` 值。这种技巧对于长期预测来说是可行的，如果是按小时分析和预测，则最好丢弃这些 `NaN` 值或对很小的滑动窗口求平均值，而不是采用整个数据列的平均值。\n\n**在下面的单元格中，我将文件读取为 DataFrame 并用特征级平均值填充 `NaN` 值。**",
"_____no_output_____"
]
],
[
[
"import txt_preprocessing as pprocess\n\n# create df from text file\ninitial_df = pprocess.create_df('household_power_consumption.txt', sep=';')\n\n# fill NaN column values with *average* column value\ndf = pprocess.fill_nan_with_mean(initial_df)\n\n# print some stats about the data\nprint('Data shape: ', df.shape)\ndf.head()",
"_____no_output_____"
]
],
[
[
"## 全球有效能耗\n\n在此示例中，我们想要预测全球有效能耗，即全球的家庭每分钟平均有效能耗（千瓦）。在下面获取这列数据并显示生成的图形。",
"_____no_output_____"
]
],
[
[
"# Select Global active power data\npower_df = df['Global_active_power'].copy()\nprint(power_df.shape)",
"_____no_output_____"
],
[
"# display the data \nplt.figure(figsize=(12,6))\n# all data points\npower_df.plot(title='Global active power', color='blue') \nplt.show()",
"_____no_output_____"
]
],
[
[
"因为数据是每分钟记录的，上图包含很多值。所以我只在下面显示了一小部分数据。",
"_____no_output_____"
]
],
[
[
"# can plot a slice of hourly data\nend_mins = 1440 # 1440 mins = 1 day\n\nplt.figure(figsize=(12,6))\npower_df[0:end_mins].plot(title='Global active power, over one day', color='blue') \nplt.show()",
"_____no_output_____"
]
],
[
[
"### 每小时与每天\n\n每分钟收集了很多数据，我可以通过以下两种方式之一分析数据：\n1. 创建很多简短的时间序列，例如一周左右，并且每小时都记录一次能耗，尝试预测接下来的几小时或几天的能耗。\n2. 创建更少的很长时间序列，数据每天记录一次，并使用这些数据预测未来几周或几个月的能耗。\n\n两种任务都很有意思。具体取决于你是要预测一天/一周还是更长时间（例如一个月）的规律。鉴于我拥有的数据量，我认为可以查看在多个月或一年内发生的更长重复性趋势。所以我将重采样“全球有效能耗”，将**每日**数据点记录为 24 小时的平均值。\n\n> 我们可以使用 pandas [时间序列工具](https://pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html)根据特定的频率重采样数据，例如按照每小时 ('H') 或每天 ('D') 重采样数据点。\n\n",
"_____no_output_____"
]
],
[
[
"# resample over day (D)\nfreq = 'D'\n# calculate the mean active power for a day\nmean_power_df = power_df.resample(freq).mean()\n\n# display the mean values\nplt.figure(figsize=(15,8))\nmean_power_df.plot(title='Global active power, mean per day', color='blue') \nplt.tight_layout()\nplt.show()",
"_____no_output_____"
]
],
[
[
"在此图形中，可以看到每年都出现了有趣的趋势。每年初和每年末都会出现能耗高峰，这时候是冬季，供暖和照明使用量都更高。在 8 月份左右也出现小高峰，这时候全球的温度通常更高。\n\n数据依然不够平滑，但是展现了明显的趋势，所以适合用机器学习模型识别这些规律。",
"_____no_output_____"
],
[
"---\n## 创建时间序列\n\n我的目标是看看能否根据从 2007-2009 的整年数据，准确地预测 2010 年多个月的平均全球有效能耗。\n\n接下来为每个完整的年份数据创建一个时间序列。这只是一种设计决策，我决定使用一整年的数据，从 2007 年 1 月开始，因为 2006 年的数据点不太多，并且这种划分更容易处理闰年。我还可以从第一个收集的数据点开始构建时间序列，只需在下面的函数中更改 `t_start` 和 `t_end` 即可。\n\n函数 `make_time_series` 将为传入的每个年份 `['2007', '2008', '2009']` 创建 pandas `Series`。\n* 所有的时间序列将从相同的时间点 `t_start`（或 t0）开始。\n * 在准备数据时，需要为每个时间序列使用一致的起始点；DeepAR 将此时间点作为参考帧，从而学习循环规律，例如工作日的行为与周末不一样，或者夏天与冬天不一样。\n * 你可以更改起始和结束索引，并定义你创建的任何时间序列。\n* 在创建时间序列时，我们应该考虑到闰年，例如 2008 年。\n* 通常，我们通过从 DataFrame 获取相关的全球能耗数据和日期索引创建 `Series`。\n\n```\n# get global consumption data\ndata = mean_power_df[start_idx:end_idx]\n\n# create time series for the year\nindex = pd.DatetimeIndex(start=t_start, end=t_end, freq='D')\ntime_series.append(pd.Series(data=data, index=index))\n```",
"_____no_output_____"
]
],
[
[
"def make_time_series(mean_power_df, years, freq='D', start_idx=16):\n '''Creates as many time series as there are complete years. This code\n accounts for the leap year, 2008.\n :param mean_power_df: A dataframe of global power consumption, averaged by day.\n This dataframe should also be indexed by a datetime.\n :param years: A list of years to make time series out of, ex. ['2007', '2008'].\n :param freq: The frequency of data recording (D = daily)\n :param start_idx: The starting dataframe index of the first point in the first time series.\n The default, 16, points to '2017-01-01'. \n :return: A list of pd.Series(), time series data.\n '''\n \n # store time series\n time_series = []\n \n # store leap year in this dataset\n leap = '2008'\n\n # create time series for each year in years\n for i in range(len(years)):\n\n year = years[i]\n if(year == leap):\n end_idx = start_idx+366\n else:\n end_idx = start_idx+365\n\n # create start and end datetimes\n t_start = year + '-01-01' # Jan 1st of each year = t_start\n t_end = year + '-12-31' # Dec 31st = t_end\n\n # get global consumption data\n data = mean_power_df[start_idx:end_idx]\n\n # create time series for the year\n index = pd.DatetimeIndex(start=t_start, end=t_end, freq=freq)\n time_series.append(pd.Series(data=data, index=index))\n \n start_idx = end_idx\n \n # return list of time series\n return time_series\n ",
"_____no_output_____"
]
],
[
[
"## 测试结果\n\n下面为每个完整的年份创建一个时间序列，并显示结果。",
"_____no_output_____"
]
],
[
[
"# test out the code above\n\n# yearly time series for our three complete years\nfull_years = ['2007', '2008', '2009']\nfreq='D' # daily recordings\n\n# make time series\ntime_series = make_time_series(mean_power_df, full_years, freq=freq)",
"_____no_output_____"
],
[
"# display first time series\ntime_series_idx = 0\n\nplt.figure(figsize=(12,6))\ntime_series[time_series_idx].plot()\nplt.show()",
"_____no_output_____"
]
],
[
[
"---\n# 按时间拆分数据\n\n我们将用测试数据集评估模型。对于分类等机器学习任务，我们通常随机将样本拆分成不同的数据集，创建训练/测试数据。对于未来预测来说，一定要按照**时间**拆分训练/测试数据，不能按照单个数据点拆分。\n> 通常，在创建训练数据时，我们从每个完整的时间序列中去除最后 `prediction_length` 个数据点，并创建训练时间序列。\n\n### 练习：创建训练时间序列\n\n请完成 `create_training_series` 函数，它应该接受完整时间序列数据列表，并返回截断的训练时间序列列表。\n\n* 在此例中，我们想预测一个月的数据，将 `prediction_length` 设为 30 天。\n* 为了创建训练数据集，我们将从生成的每个时间序列中去除最后 30 个数据点，所以仅使用第一部分作为训练数据。\n* **测试集包含每个时间序列的完整范围**。\n",
"_____no_output_____"
]
],
[
[
"# create truncated, training time series\ndef create_training_series(complete_time_series, prediction_length):\n '''Given a complete list of time series data, create training time series.\n :param complete_time_series: A list of all complete time series.\n :param prediction_length: The number of points we want to predict.\n :return: A list of training time series.\n '''\n # your code here\n \n pass\n ",
"_____no_output_____"
],
[
"# test your code!\n\n# set prediction length\nprediction_length = 30 # 30 days ~ a month\n\ntime_series_training = create_training_series(time_series, prediction_length)\n",
"_____no_output_____"
]
],
[
[
"### 训练和测试序列\n\n我们可以将训练/测试序列绘制到同一个坐标轴上，可视化这些序列。我们应该看到测试序列包含一年的所有数据，训练序列包含最后 `prediction_length` 个数据点之外的数据。",
"_____no_output_____"
]
],
[
[
"# display train/test time series\ntime_series_idx = 0\n\nplt.figure(figsize=(15,8))\n# test data is the whole time series\ntime_series[time_series_idx].plot(label='test', lw=3)\n# train data is all but the last prediction pts\ntime_series_training[time_series_idx].plot(label='train', ls=':', lw=3)\n\nplt.legend()\nplt.show()",
"_____no_output_____"
]
],
[
[
"## 转换为 JSON \n\n根据 [DeepAR 文档](https://docs.aws.amazon.com/sagemaker/latest/dg/deepar.html)，DeepAR 要求输入训练数据是 JSON 格式，并包含以下字段：\n\n* **start**：定义时间序列开始日期的字符串，格式为“YYYY-MM-DD HH:MM:SS”。\n* **target**：表示时间序列的数值数组。\n* **cat**（可选）：类别特征数值数组，可以用于表示记录所属的组。这个字段适合按照项目类别寻找模型，例如对于零售销量，可以将 {'shoes', 'jackets', 'pants'} 表示成类别 {0, 1, 2}。\n\n输入数据的格式应该为，在 JSON 文件中每行一个时间序列。每行看起来像字典，例如：\n```\n{\"start\":'2007-01-01 00:00:00', \"target\": [2.54, 6.3, ...], \"cat\": [1]}\n{\"start\": \"2012-01-30 00:00:00\", \"target\": [1.0, -5.0, ...], \"cat\": [0]} \n...\n```\n在上述示例中，每个时间序列都有一个相关的类别特征和一个时间序列特征。\n\n### 练习：格式化能耗数据\n\n对于我们的数据来说：\n* 开始日期“start”将为时间序列中第一行的索引，即这一年的 1 月 1 日。\n* “target”将为时间序列存储的所有能耗值。\n* 我们将不使用可选“cat”字段。\n\n请完成以下实用函数，它应该将 `pandas.Series` 对象转换成 DeepAR 可以使用的相应 JSON 字符串。",
"_____no_output_____"
]
],
[
[
"def series_to_json_obj(ts):\n '''Returns a dictionary of values in DeepAR, JSON format.\n :param ts: A single time series.\n :return: A dictionary of values with \"start\" and \"target\" keys.\n '''\n # your code here\n \n pass\n",
"_____no_output_____"
],
[
"# test out the code\nts = time_series[0]\n\njson_obj = series_to_json_obj(ts)\n\nprint(json_obj)",
"_____no_output_____"
]
],
[
[
"### 将数据保存到本地\n\n下面的辅助函数会将一个序列放入 JSON 文件的一行中，并使用换行符“\\n”分隔。数据还会编码并写入我们指定的文件名中。",
"_____no_output_____"
]
],
[
[
"# import json for formatting data\nimport json\nimport os # and os for saving\n\ndef write_json_dataset(time_series, filename): \n with open(filename, 'wb') as f:\n # for each of our times series, there is one JSON line\n for ts in time_series:\n json_line = json.dumps(series_to_json_obj(ts)) + '\\n'\n json_line = json_line.encode('utf-8')\n f.write(json_line)\n print(filename + ' saved.')",
"_____no_output_____"
],
[
"# save this data to a local directory\ndata_dir = 'json_energy_data'\n\n# make data dir, if it does not exist\nif not os.path.exists(data_dir):\n os.makedirs(data_dir)",
"_____no_output_____"
],
[
"# directories to save train/test data\ntrain_key = os.path.join(data_dir, 'train.json')\ntest_key = os.path.join(data_dir, 'test.json')\n\n# write train/test JSON files\nwrite_json_dataset(time_series_training, train_key) \nwrite_json_dataset(time_series, test_key)",
"_____no_output_____"
]
],
[
[
"---\n## 将数据上传到 S3\n\n接下来，为了使 estimator 能够访问此数据，我将数据上传到 S3。\n\n### Sagemaker 资源\n\n首先指定：\n* 训练模型用到的 sagemaker 角色和会话。\n* 默认 S3 存储桶，我们可以在其中存储训练、测试和模型数据。",
"_____no_output_____"
]
],
[
[
"import boto3\nimport sagemaker\nfrom sagemaker import get_execution_role",
"_____no_output_____"
],
[
"# session, role, bucket\nsagemaker_session = sagemaker.Session()\nrole = get_execution_role()\n\nbucket = sagemaker_session.default_bucket()\n",
"_____no_output_____"
]
],
[
[
"### 练习：将训练和测试 JSON 文件上传到 S3\n\n指定唯一的训练和测试 prefix，它们定义了数据在 S3 中的位置。\n* 将训练数据上传到 S3 中的某个位置，并将该位置保存到 `train_path`\n* 将测试数据上传到 S3 中的某个位置，并将该位置保存到 `test_path`",
"_____no_output_____"
]
],
[
[
"# suggested that you set prefixes for directories in S3\n\n# upload data to S3, and save unique locations\ntrain_path = None\ntest_path = None",
"_____no_output_____"
],
[
"# check locations\nprint('Training data is stored in: '+ train_path)\nprint('Test data is stored in: '+ test_path)",
"_____no_output_____"
]
],
[
[
"---\n# 训练 DeepAR Estimator\n\n某些 estimator 具有特定的 SageMaker 构造函数，但是并非都有。你可以创建一个基本 `Estimator` 并传入保存特定模型的特定镜像（或容器）。\n\n接下来，配置要在我们运行模型所在的区域使用的容器镜像。",
"_____no_output_____"
]
],
[
[
"from sagemaker.amazon.amazon_estimator import get_image_uri\n\nimage_name = get_image_uri(boto3.Session().region_name, # get the region\n 'forecasting-deepar') # specify image\n",
"_____no_output_____"
]
],
[
[
"### 练习：实例化 Estimator \n\n现在可以定义将启动训练作业的 estimator 了。一般的 Estimator 将由普通的构造函数参数和 `image_name` 进行定义。\n> 你可以查看 [estimator 源代码](https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/estimator.py#L595)了解详情。\n",
"_____no_output_____"
]
],
[
[
"from sagemaker.estimator import Estimator\n\n# instantiate a DeepAR estimator\nestimator = None\n",
"_____no_output_____"
]
],
[
[
"## 设置超参数\n\n接下来，我们需要定义一些 DeepAR 超参数，这些超参数定义了模型大小和训练行为。需要定义周期数评论、预测时长和上下文时长。\n\n* **epochs**：在训练时遍历数据的最大次数。\n* **time_freq**：数据集中的时间序列频率（“D”表示每天）。\n* **prediction_length**：一个字符串，表示训练模型预测的时间步数量（基于频率单位）。\n* **context_length**：模型在做出预测之前可以看到的时间点数量。\n\n### 上下文长度\n\n通常，建议从 `context_length`=`prediction_length` 开始。这是因为 DeepAR 模型还会从目标时间序列那接收“延迟”输入，使模型能够捕获长期依赖关系。例如，每日时间序列可以具有每年季节效应，DeepAR 会自动包含一年延迟。所以上下文长度可以短于一年，模型依然能够捕获这种季节效应。\n\n模型选择的延迟值取决于时间序列的频率。例如，每日频率的延迟值是上一周、上两周、上三周、上四周和一年。详情请参阅 [DeepAR 工作原理文档](https://docs.aws.amazon.com/sagemaker/latest/dg/deepar_how-it-works.html)。\n\n### 可选超参数\n\n你还可以配置可选超参数，以进一步优化模型。包括 RNN 模型的层数、每层的单元格数量、似然率函数，以及训练选项，例如批次大小和学习速率。\n\n要了解所有不同 DeepAR 超参数的详尽列表，请参阅 DeepAR [超参数文档](https://docs.aws.amazon.com/sagemaker/latest/dg/deepar_hyperparameters.html)。",
"_____no_output_____"
]
],
[
[
"freq='D'\ncontext_length=30 # same as prediction_length\n\nhyperparameters = {\n \"epochs\": \"50\",\n \"time_freq\": freq,\n \"prediction_length\": str(prediction_length),\n \"context_length\": str(context_length),\n \"num_cells\": \"50\",\n \"num_layers\": \"2\",\n \"mini_batch_size\": \"128\",\n \"learning_rate\": \"0.001\",\n \"early_stopping_patience\": \"10\"\n}",
"_____no_output_____"
],
[
"# set the hyperparams\nestimator.set_hyperparameters(**hyperparameters)",
"_____no_output_____"
]
],
[
[
"## 训练作业\n\n现在我们可以启动训练作业了。SageMaker 将启动 EC2 实例、从 S3 下载数据、开始训练模型并保存训练过的模型。\n\n如果你提供了 `test` 数据通道（就像在示例中一样），DeepAR 还会计算训练过的模型在此测试数据集上的准确率指标。计算方法是预测测试集中每个时间序列的最后 `prediction_length` 个点，并将它们与时间序列的实际值进行比较。计算的误差指标将包含在日志输出中。\n\n下个单元格可能需要几分钟才能完成，取决于数据大小、模型复杂度和训练选项。",
"_____no_output_____"
]
],
[
[
"%%time\n# train and test channels\ndata_channels = {\n \"train\": train_path,\n \"test\": test_path\n}\n\n# fit the estimator\nestimator.fit(inputs=data_channels)",
"_____no_output_____"
]
],
[
[
"## 部署和创建预测器\n\n训练模型后，我们可以将模型部署到预测器端点上，并使用模型做出预测。\n\n在此 notebook 结束时，记得**删除端点**。我们将在此 notebook 的最后提供一个删除单元格，但是建议提前记住。",
"_____no_output_____"
]
],
[
[
"%%time\n\n# create a predictor\npredictor = estimator.deploy(\n initial_instance_count=1,\n instance_type='ml.t2.medium',\n content_type=\"application/json\" # specify that it will accept/produce JSON\n)",
"_____no_output_____"
]
],
[
[
"---\n# 生成预测\n\n根据 DeepAR 的[推理格式](https://docs.aws.amazon.com/sagemaker/latest/dg/deepar-in-formats.html)，`predictor` 要求输入数据是 JSON 格式，并具有以下键：\n* **instances**：一个 JSON 格式的时间序列列表，模型应该预测这些时间序列。\n* **configuration**（可选）：一个配置信息字典，定义了请求希望的响应类型。\n\n在 configuration 中，可以配置以下键：\n* **num_samples**：一个整数，指定了模型在做出概率预测时，生成的样本数。\n* **output_types**：一个指定响应类型的列表。我们需要 **quantiles**，它会查看模型生成的 num_samples 列表，并根据这些值为每个时间点生成[分位数估值](https://en.wikipedia.org/wiki/Quantile)。\n* **quantiles**：一个列表，指定生成哪些分位数估值并在响应中返回这些估值。\n\n\n下面是向 DeepAR 模型端点发出 JSON 查询的示例。\n\n```\n{\n \"instances\": [\n { \"start\": \"2009-11-01 00:00:00\", \"target\": [4.0, 10.0, 50.0, 100.0, 113.0] },\n { \"start\": \"1999-01-30\", \"target\": [2.0, 1.0] }\n ],\n \"configuration\": {\n \"num_samples\": 50,\n \"output_types\": [\"quantiles\"],\n \"quantiles\": [\"0.5\", \"0.9\"]\n }\n}\n```\n\n\n## JSON 预测请求\n\n以下代码接受时间序列**列表**作为输入并接受一些配置参数。然后将该序列变成 JSON 实例格式，并将输入转换成相应格式的 JSON_input。",
"_____no_output_____"
]
],
[
[
"def json_predictor_input(input_ts, num_samples=50, quantiles=['0.1', '0.5', '0.9']):\n '''Accepts a list of input time series and produces a formatted input.\n :input_ts: An list of input time series.\n :num_samples: Number of samples to calculate metrics with.\n :quantiles: A list of quantiles to return in the predicted output.\n :return: The JSON-formatted input.\n '''\n # request data is made of JSON objects (instances)\n # and an output configuration that details the type of data/quantiles we want\n \n instances = []\n for k in range(len(input_ts)):\n # get JSON objects for input time series\n instances.append(series_to_json_obj(input_ts[k]))\n\n # specify the output quantiles and samples\n configuration = {\"num_samples\": num_samples, \n \"output_types\": [\"quantiles\"], \n \"quantiles\": quantiles}\n\n request_data = {\"instances\": instances, \n \"configuration\": configuration}\n\n json_request = json.dumps(request_data).encode('utf-8')\n \n return json_request",
"_____no_output_____"
]
],
[
[
"### 获得预测\n\n然后，我们可以使用该函数获得格式化时间序列的预测。\n\n在下个单元格中，我获得了时间序列输入和已知目标，并将格式化输入传入预测器端点中，以获得预测。",
"_____no_output_____"
]
],
[
[
"# get all input and target (test) time series\ninput_ts = time_series_training\ntarget_ts = time_series\n\n# get formatted input time series\njson_input_ts = json_predictor_input(input_ts)\n\n# get the prediction from the predictor\njson_prediction = predictor.predict(json_input_ts)\n\nprint(json_prediction)",
"_____no_output_____"
]
],
[
[
"## 解码预测\n\n预测器返回 predictor returns JSON 格式的预测，所以我们需要提取可视化结果所需的预测和分位数数据。以下函数会读取 JSON 格式的预测并生成每个分位数中的预测列表。",
"_____no_output_____"
]
],
[
[
"# helper function to decode JSON prediction\ndef decode_prediction(prediction, encoding='utf-8'):\n '''Accepts a JSON prediction and returns a list of prediction data.\n '''\n prediction_data = json.loads(prediction.decode(encoding))\n prediction_list = []\n for k in range(len(prediction_data['predictions'])):\n prediction_list.append(pd.DataFrame(data=prediction_data['predictions'][k]['quantiles']))\n return prediction_list\n",
"_____no_output_____"
],
[
"# get quantiles/predictions\nprediction_list = decode_prediction(json_prediction)\n\n# should get a list of 30 predictions \n# with corresponding quantile values\nprint(prediction_list[0])",
"_____no_output_____"
]
],
[
[
"## 显示结果\n\n分位数数据可以提供查看预测结果所需的所有信息。\n* 分位数 0.1 和 0.9 表示预测值的上下限。\n* 分位数 0.5 表示所有样本预测的中位数。\n",
"_____no_output_____"
]
],
[
[
"# display the prediction median against the actual data\ndef display_quantiles(prediction_list, target_ts=None):\n # show predictions for all input ts\n for k in range(len(prediction_list)):\n plt.figure(figsize=(12,6))\n # get the target month of data\n if target_ts is not None:\n target = target_ts[k][-prediction_length:]\n plt.plot(range(len(target)), target, label='target')\n # get the quantile values at 10 and 90%\n p10 = prediction_list[k]['0.1']\n p90 = prediction_list[k]['0.9']\n # fill the 80% confidence interval\n plt.fill_between(p10.index, p10, p90, color='y', alpha=0.5, label='80% confidence interval')\n # plot the median prediction line\n prediction_list[k]['0.5'].plot(label='prediction median')\n plt.legend()\n plt.show()",
"_____no_output_____"
],
[
"# display predictions\ndisplay_quantiles(prediction_list, target_ts)",
"_____no_output_____"
]
],
[
[
"## 预测未来\n\n我们没有向模型提供任何 2010 年数据，但是我们看看如果只有已知开始日期，**没有目标**，模型能否预测能耗。\n\n### 练习：为“未来”预测设定请求\n\n请创建一个格式化输入并传入部署的 `predictor`，同时传入常规“configuration”参数。这里的“instances”只有 1 个实例，定义如下：\n* **start**：开始时间将为你指定的时间戳。要预测 2010 年的前 30 天，从 1 月 1 日“2010-01-01”开始。\n* **target**：目标将为空列表，因为这一年没有完整的相关时间序列。我们特意从模型中去除了该信息，以便测试模型。\n```\n{\"start\": start_time, \"target\": []} # empty target\n```",
"_____no_output_____"
]
],
[
[
"# Starting my prediction at the beginning of 2010\nstart_date = '2010-01-01'\ntimestamp = '00:00:00'\n\n# formatting start_date\nstart_time = start_date +' '+ timestamp\n\n# format the request_data\n# with \"instances\" and \"configuration\"\nrequest_data = None\n\n\n# create JSON input\njson_input = json.dumps(request_data).encode('utf-8')\n\nprint('Requesting prediction for '+start_time)",
"_____no_output_____"
]
],
[
[
"然后正常地获取和解码预测响应。",
"_____no_output_____"
]
],
[
[
"# get prediction response\njson_prediction = predictor.predict(json_input)\n\nprediction_2010 = decode_prediction(json_prediction)\n",
"_____no_output_____"
]
],
[
[
"最后，将预测与已知目标序列进行比较。此目标将来自 2010 年的时间序列，我在下面创建了该序列。",
"_____no_output_____"
]
],
[
[
"# create 2010 time series\nts_2010 = []\n# get global consumption data\n# index 1112 is where the 2011 data starts\ndata_2010 = mean_power_df.values[1112:]\n\nindex = pd.DatetimeIndex(start=start_date, periods=len(data_2010), freq='D')\nts_2010.append(pd.Series(data=data_2010, index=index))\n",
"_____no_output_____"
],
[
"# range of actual data to compare\nstart_idx=0 # days since Jan 1st 2010\nend_idx=start_idx+prediction_length\n\n# get target data\ntarget_2010_ts = [ts_2010[0][start_idx:end_idx]]\n\n# display predictions\ndisplay_quantiles(prediction_2010, target_2010_ts)",
"_____no_output_____"
]
],
[
[
"## 删除端点\n\n请用不同的时间序列尝试你的代码。建议调节 DeepAR 超参数，看看能否改进此预测器的性能。\n\n评估完预测器（任何预测器）后，记得删除端点。",
"_____no_output_____"
]
],
[
[
"## TODO: delete the endpoint\npredictor.delete_endpoint()",
"_____no_output_____"
]
],
[
[
"## 总结\n\n你已经见过一个复杂但是应用广泛的时间序列预测方法，并且掌握了将 DeepAR 模型应用到你感兴趣的数据上所需的技能。",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
]
] |
cb92f496020703997fb9f13c2553ee2054f64a26
| 74,647 |
ipynb
|
Jupyter Notebook
|
intro-to-pytorch/Part 2 - Neural Networks in PyTorch (Exercises).ipynb
|
gghknb/deep-learning-pytorch
|
a8a049947506c82df5200216d8396e71bb23e60a
|
[
"MIT"
] | null | null | null |
intro-to-pytorch/Part 2 - Neural Networks in PyTorch (Exercises).ipynb
|
gghknb/deep-learning-pytorch
|
a8a049947506c82df5200216d8396e71bb23e60a
|
[
"MIT"
] | null | null | null |
intro-to-pytorch/Part 2 - Neural Networks in PyTorch (Exercises).ipynb
|
gghknb/deep-learning-pytorch
|
a8a049947506c82df5200216d8396e71bb23e60a
|
[
"MIT"
] | null | null | null | 84.442308 | 16,152 | 0.785711 |
[
[
[
"# Neural networks with PyTorch\n\nDeep learning networks tend to be massive with dozens or hundreds of layers, that's where the term \"deep\" comes from. You can build one of these deep networks using only weight matrices as we did in the previous notebook, but in general it's very cumbersome and difficult to implement. PyTorch has a nice module `nn` that provides a nice way to efficiently build large neural networks.",
"_____no_output_____"
]
],
[
[
"# Import necessary packages\n\n%matplotlib inline\n%config InlineBackend.figure_format = 'retina'\n\nimport numpy as np\nimport torch\n\nimport helper\n\nimport matplotlib.pyplot as plt",
"_____no_output_____"
]
],
[
[
"\nNow we're going to build a larger network that can solve a (formerly) difficult problem, identifying text in an image. Here we'll use the MNIST dataset which consists of greyscale handwritten digits. Each image is 28x28 pixels, you can see a sample below\n\n<img src='assets/mnist.png'>\n\nOur goal is to build a neural network that can take one of these images and predict the digit in the image.\n\nFirst up, we need to get our dataset. This is provided through the `torchvision` package. The code below will download the MNIST dataset, then create training and test datasets for us. Don't worry too much about the details here, you'll learn more about this later.",
"_____no_output_____"
]
],
[
[
"### Run this cell\n\nfrom torchvision import datasets, transforms\n\n# Define a transform to normalize the data\ntransform = transforms.Compose([transforms.ToTensor(),\n transforms.Normalize((0.5,), (0.5,)),\n ])\n\n# Download and load the training data\ntrainset = datasets.MNIST('~/.pytorch/MNIST_data/', download=True, train=True, transform=transform)\ntrainloader = torch.utils.data.DataLoader(trainset, batch_size=64, shuffle=True)",
"Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz\nDownloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz\nDownloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz\nDownloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz\nProcessing...\nDone!\n"
]
],
[
[
"We have the training data loaded into `trainloader` and we make that an iterator with `iter(trainloader)`. Later, we'll use this to loop through the dataset for training, like\n\n```python\nfor image, label in trainloader:\n ## do things with images and labels\n```\n\nYou'll notice I created the `trainloader` with a batch size of 64, and `shuffle=True`. The batch size is the number of images we get in one iteration from the data loader and pass through our network, often called a *batch*. And `shuffle=True` tells it to shuffle the dataset every time we start going through the data loader again. But here I'm just grabbing the first batch so we can check out the data. We can see below that `images` is just a tensor with size `(64, 1, 28, 28)`. So, 64 images per batch, 1 color channel, and 28x28 images.",
"_____no_output_____"
]
],
[
[
"dataiter = iter(trainloader)\nimages, labels = dataiter.next()\nprint(type(images))\nprint(images.shape)\nprint(labels.shape)",
"<class 'torch.Tensor'>\ntorch.Size([64, 1, 28, 28])\ntorch.Size([64])\n"
]
],
[
[
"This is what one of the images looks like. ",
"_____no_output_____"
]
],
[
[
"plt.imshow(images[1].numpy().squeeze(), cmap='Greys_r');",
"_____no_output_____"
]
],
[
[
"First, let's try to build a simple network for this dataset using weight matrices and matrix multiplications. Then, we'll see how to do it using PyTorch's `nn` module which provides a much more convenient and powerful method for defining network architectures.\n\nThe networks you've seen so far are called *fully-connected* or *dense* networks. Each unit in one layer is connected to each unit in the next layer. In fully-connected networks, the input to each layer must be a one-dimensional vector (which can be stacked into a 2D tensor as a batch of multiple examples). However, our images are 28x28 2D tensors, so we need to convert them into 1D vectors. Thinking about sizes, we need to convert the batch of images with shape `(64, 1, 28, 28)` to a have a shape of `(64, 784)`, 784 is 28 times 28. This is typically called *flattening*, we flattened the 2D images into 1D vectors.\n\nPreviously you built a network with one output unit. Here we need 10 output units, one for each digit. We want our network to predict the digit shown in an image, so what we'll do is calculate probabilities that the image is of any one digit or class. This ends up being a discrete probability distribution over the classes (digits) that tells us the most likely class for the image. That means we need 10 output units for the 10 classes (digits). We'll see how to convert the network output into a probability distribution next.\n\n> **Exercise:** Flatten the batch of images `images`. Then build a multi-layer network with 784 input units, 256 hidden units, and 10 output units using random tensors for the weights and biases. For now, use a sigmoid activation for the hidden layer. Leave the output layer without an activation, we'll add one that gives us a probability distribution next.",
"_____no_output_____"
]
],
[
[
"## Your solution\ndef sigmoid(x) :\n return 1/(1+torch.exp(-x))\n\ninput_unit = images.shape[2]*images.shape[3]\nhidden_unit = 256\noutput_unit = 10\n\nimg = images.view(images.shape[0],-1) # img = images.view(-1,input_unit) \n\nW1 = torch.randn(input_unit,hidden_unit)\nW2 = torch.randn(hidden_unit,output_unit)\n\nB1 = torch.randn(hidden_unit)\nB2 = torch.randn(output_unit)\n\nout = torch.mm(sigmoid(torch.mm(img,W1)+B1),W2)+B2 # output of your network, should have shape (64,10)\nprint(out.shape)",
"torch.Size([64, 10])\n"
]
],
[
[
"Now we have 10 outputs for our network. We want to pass in an image to our network and get out a probability distribution over the classes that tells us the likely class(es) the image belongs to. Something that looks like this:\n<img src='assets/image_distribution.png' width=500px>\n\nHere we see that the probability for each class is roughly the same. This is representing an untrained network, it hasn't seen any data yet so it just returns a uniform distribution with equal probabilities for each class.\n\nTo calculate this probability distribution, we often use the [**softmax** function](https://en.wikipedia.org/wiki/Softmax_function). Mathematically this looks like\n\n$$\n\\Large \\sigma(x_i) = \\cfrac{e^{x_i}}{\\sum_k^K{e^{x_k}}}\n$$\n\nWhat this does is squish each input $x_i$ between 0 and 1 and normalizes the values to give you a proper probability distribution where the probabilites sum up to one.\n\n> **Exercise:** Implement a function `softmax` that performs the softmax calculation and returns probability distributions for each example in the batch. Note that you'll need to pay attention to the shapes when doing this. If you have a tensor `a` with shape `(64, 10)` and a tensor `b` with shape `(64,)`, doing `a/b` will give you an error because PyTorch will try to do the division across the columns (called broadcasting) but you'll get a size mismatch. The way to think about this is for each of the 64 examples, you only want to divide by one value, the sum in the denominator. So you need `b` to have a shape of `(64, 1)`. This way PyTorch will divide the 10 values in each row of `a` by the one value in each row of `b`. Pay attention to how you take the sum as well. You'll need to define the `dim` keyword in `torch.sum`. Setting `dim=0` takes the sum across the rows while `dim=1` takes the sum across the columns.",
"_____no_output_____"
]
],
[
[
"def softmax(x):\n ## TODO: Implement the softmax function here\n return torch.exp(x) / torch.sum(torch.exp(x),dim=1).view(-1,1) # matrix / vector => error (broadcasting)\n# Here, out should be the output of the network in the previous excercise with shape (64,10)\nprint(torch.sum(torch.exp(out),dim=1).shape)\nprobabilities = softmax(out)\n# Does it have the right shape? Should be (64, 10)\nprint(probabilities.shape)\n# Does it sum to 1?\nprint(probabilities.sum(dim=1))",
"torch.Size([64])\ntorch.Size([64, 10])\ntensor([1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000,\n 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000,\n 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000,\n 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000,\n 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000,\n 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000,\n 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000, 1.0000,\n 1.0000])\n"
]
],
[
[
"## Building networks with PyTorch\n\nPyTorch provides a module `nn` that makes building networks much simpler. Here I'll show you how to build the same one as above with 784 inputs, 256 hidden units, 10 output units and a softmax output.",
"_____no_output_____"
]
],
[
[
"from torch import nn",
"_____no_output_____"
],
[
"class Network(nn.Module):\n def __init__(self):\n super().__init__() #register layers \n \n # Inputs to hidden layer linear transformation\n self.hidden = nn.Linear(784, 256)\n # Output layer, 10 units - one for each digit\n self.output = nn.Linear(256, 10)\n \n # Define sigmoid activation and softmax output \n self.sigmoid = nn.Sigmoid()\n self.softmax = nn.Softmax(dim=1)\n \n def forward(self, x):\n # Pass the input tensor through each of our operations\n x = self.hidden(x)\n x = self.sigmoid(x)\n x = self.output(x)\n x = self.softmax(x)\n \n return x",
"_____no_output_____"
]
],
[
[
"Let's go through this bit by bit.\n\n```python\nclass Network(nn.Module):\n```\n\nHere we're inheriting from `nn.Module`. Combined with `super().__init__()` this creates a class that tracks the architecture and provides a lot of useful methods and attributes. It is mandatory to inherit from `nn.Module` when you're creating a class for your network. The name of the class itself can be anything.\n\n```python\nself.hidden = nn.Linear(784, 256)\n```\n\nThis line creates a module for a linear transformation, $x\\mathbf{W} + b$, with 784 inputs and 256 outputs and assigns it to `self.hidden`. The module automatically creates the weight and bias tensors which we'll use in the `forward` method. You can access the weight and bias tensors once the network (`net`) is created with `net.hidden.weight` and `net.hidden.bias`.\n\n```python\nself.output = nn.Linear(256, 10)\n```\n\nSimilarly, this creates another linear transformation with 256 inputs and 10 outputs.\n\n```python\nself.sigmoid = nn.Sigmoid()\nself.softmax = nn.Softmax(dim=1)\n```\n\nHere I defined operations for the sigmoid activation and softmax output. Setting `dim=1` in `nn.Softmax(dim=1)` calculates softmax across the columns.\n\n```python\ndef forward(self, x):\n```\n\nPyTorch networks created with `nn.Module` must have a `forward` method defined. It takes in a tensor `x` and passes it through the operations you defined in the `__init__` method.\n\n```python\nx = self.hidden(x)\nx = self.sigmoid(x)\nx = self.output(x)\nx = self.softmax(x)\n```\n\nHere the input tensor `x` is passed through each operation and reassigned to `x`. We can see that the input tensor goes through the hidden layer, then a sigmoid function, then the output layer, and finally the softmax function. It doesn't matter what you name the variables here, as long as the inputs and outputs of the operations match the network architecture you want to build. The order in which you define things in the `__init__` method doesn't matter, but you'll need to sequence the operations correctly in the `forward` method.\n\nNow we can create a `Network` object.",
"_____no_output_____"
]
],
[
[
"# Create the network and look at it's text representation\nmodel = Network()\nmodel",
"_____no_output_____"
]
],
[
[
"You can define the network somewhat more concisely and clearly using the `torch.nn.functional` module. This is the most common way you'll see networks defined as many operations are simple element-wise functions. We normally import this module as `F`, `import torch.nn.functional as F`.",
"_____no_output_____"
]
],
[
[
"import torch.nn.functional as F\n\nclass Network(nn.Module):\n def __init__(self):\n super().__init__()\n # Inputs to hidden layer linear transformation\n self.hidden = nn.Linear(784, 256)\n # Output layer, 10 units - one for each digit\n self.output = nn.Linear(256, 10)\n \n def forward(self, x):\n # Hidden layer with sigmoid activation\n x = F.sigmoid(self.hidden(x)) # functional 함수는 init할 필요없이 그냥 가져다 쓰면 되서 편함 extra parameter를 만들지 않아도 되서 좋기도 함 \n # Output layer with softmax activation\n x = F.softmax(self.output(x), dim=1)\n \n return x",
"_____no_output_____"
]
],
[
[
"### Activation functions\n\nSo far we've only been looking at the sigmoid activation function, but in general any function can be used as an activation function. The only requirement is that for a network to approximate a non-linear function, the activation functions must be non-linear. Here are a few more examples of common activation functions: Tanh (hyperbolic tangent), and ReLU (rectified linear unit).\n\n<img src=\"assets/activation.png\" width=700px>\n\nIn practice, the ReLU function is used almost exclusively as the activation function for hidden layers.",
"_____no_output_____"
],
[
"### Your Turn to Build a Network\n\n<img src=\"assets/mlp_mnist.png\" width=600px>\n\n> **Exercise:** Create a network with 784 input units, a hidden layer with 128 units and a ReLU activation, then a hidden layer with 64 units and a ReLU activation, and finally an output layer with a softmax activation as shown above. You can use a ReLU activation with the `nn.ReLU` module or `F.relu` function.\n\nIt's good practice to name your layers by their type of network, for instance 'fc' to represent a fully-connected layer. As you code your solution, use `fc1`, `fc2`, and `fc3` as your layer names.",
"_____no_output_____"
]
],
[
[
"import torch.nn.functional as F\n\n## Your solution here\nclass fc(nn.Module) :\n def __init__(self) :\n super().__init__()\n \n self.fc1 = nn.Linear(784,128)\n self.fc2 = nn.Linear(128,64)\n self.fc3 = nn.Linear(64,10)\n def forward(self,x) :\n x = self.fc1(x)\n x = F.relu(x)\n x = self.fc2(x)\n x = F.relu(x)\n x = self.fc3(x)\n x = F.softmax(x)\n \n return x",
"_____no_output_____"
]
],
[
[
"### Initializing weights and biases\n\nThe weights and such are automatically initialized for you, but it's possible to customize how they are initialized. The weights and biases are tensors attached to the layer you defined, you can get them with `model.fc1.weight` for instance.",
"_____no_output_____"
]
],
[
[
"model = fc()\nprint(model.fc1.weight)\nprint(model.fc1.bias)",
"Parameter containing:\ntensor([[-0.0325, 0.0320, 0.0227, ..., -0.0082, 0.0117, 0.0127],\n [ 0.0208, -0.0003, -0.0114, ..., 0.0331, 0.0309, -0.0074],\n [-0.0132, -0.0222, 0.0262, ..., 0.0053, 0.0251, -0.0117],\n ...,\n [ 0.0013, -0.0214, 0.0262, ..., -0.0126, -0.0129, -0.0237],\n [ 0.0340, 0.0028, -0.0216, ..., 0.0024, 0.0036, 0.0213],\n [ 0.0079, 0.0088, 0.0344, ..., 0.0173, -0.0086, -0.0228]],\n requires_grad=True)\nParameter containing:\ntensor([-0.0108, -0.0198, 0.0002, -0.0349, -0.0105, -0.0004, -0.0049, 0.0256,\n -0.0312, 0.0059, -0.0326, 0.0041, 0.0073, 0.0079, -0.0153, -0.0121,\n -0.0167, 0.0286, 0.0301, 0.0281, 0.0309, -0.0051, -0.0276, -0.0251,\n 0.0312, 0.0143, -0.0025, -0.0319, 0.0120, 0.0102, 0.0134, -0.0356,\n -0.0303, 0.0204, 0.0010, -0.0245, -0.0262, 0.0316, -0.0336, -0.0341,\n 0.0262, -0.0206, 0.0222, -0.0157, 0.0127, -0.0093, 0.0067, -0.0092,\n 0.0261, -0.0113, 0.0328, -0.0150, 0.0246, -0.0267, -0.0179, 0.0201,\n 0.0246, 0.0309, -0.0310, -0.0353, -0.0328, -0.0017, -0.0069, -0.0084,\n 0.0107, -0.0275, -0.0352, -0.0059, 0.0009, -0.0168, 0.0019, 0.0355,\n 0.0243, 0.0148, 0.0216, 0.0309, 0.0090, 0.0056, 0.0321, -0.0015,\n 0.0340, 0.0145, -0.0181, -0.0105, 0.0194, -0.0230, 0.0314, 0.0332,\n 0.0016, 0.0287, -0.0342, -0.0259, -0.0034, 0.0241, 0.0153, 0.0221,\n -0.0184, -0.0046, 0.0153, -0.0287, 0.0015, -0.0042, -0.0069, 0.0356,\n 0.0119, 0.0078, 0.0187, -0.0212, 0.0004, -0.0193, -0.0149, -0.0131,\n 0.0336, -0.0304, -0.0076, 0.0276, -0.0055, -0.0183, -0.0256, 0.0246,\n -0.0031, 0.0112, -0.0146, -0.0326, 0.0098, 0.0075, -0.0120, 0.0162],\n requires_grad=True)\n"
]
],
[
[
"For custom initialization, we want to modify these tensors in place. These are actually autograd *Variables*, so we need to get back the actual tensors with `model.fc1.weight.data`. Once we have the tensors, we can fill them with zeros (for biases) or random normal values.",
"_____no_output_____"
]
],
[
[
"# Set biases to all zeros\nmodel.fc1.bias.data.fill_(0)",
"_____no_output_____"
],
[
"# sample from random normal with standard dev = 0.01\nmodel.fc1.weight.data.normal_(std=0.01)",
"_____no_output_____"
]
],
[
[
"### Forward pass\n\nNow that we have a network, let's see what happens when we pass in an image.",
"_____no_output_____"
]
],
[
[
"# Grab some data \ndataiter = iter(trainloader)\nimages, labels = dataiter.next()\n\n# Resize images into a 1D vector, new shape is (batch size, color channels, image pixels) \nimages.resize_(64, 1, 784)\n# or images.resize_(images.shape[0], 1, 784) to automatically get batch size\n\n# Forward pass through the network\nimg_idx = 0\nps = model.forward(images[img_idx,:])\n\nimg = images[img_idx]\nhelper.view_classify(img.view(1, 28, 28), ps)",
"/root/anaconda3/envs/py36/lib/python3.6/site-packages/ipykernel_launcher.py:17: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n"
]
],
[
[
"As you can see above, our network has basically no idea what this digit is. It's because we haven't trained it yet, all the weights are random!\n\n### Using `nn.Sequential`\n\nPyTorch provides a convenient way to build networks like this where a tensor is passed sequentially through operations, `nn.Sequential` ([documentation](https://pytorch.org/docs/master/nn.html#torch.nn.Sequential)). Using this to build the equivalent network:",
"_____no_output_____"
]
],
[
[
"# Hyperparameters for our network\ninput_size = 784\nhidden_sizes = [128, 64]\noutput_size = 10\n\n# Build a feed-forward network\nmodel = nn.Sequential(nn.Linear(input_size, hidden_sizes[0]),\n nn.ReLU(),\n nn.Linear(hidden_sizes[0], hidden_sizes[1]),\n nn.ReLU(),\n nn.Linear(hidden_sizes[1], output_size),\n nn.Softmax(dim=1))\nprint(model)\n\n# Forward pass through the network and display output\nimages, labels = next(iter(trainloader))\nimages.resize_(images.shape[0], 1, 784)\nps = model.forward(images[0,:])\nhelper.view_classify(images[0].view(1, 28, 28), ps)",
"Sequential(\n (0): Linear(in_features=784, out_features=128, bias=True)\n (1): ReLU()\n (2): Linear(in_features=128, out_features=64, bias=True)\n (3): ReLU()\n (4): Linear(in_features=64, out_features=10, bias=True)\n (5): Softmax()\n)\n"
]
],
[
[
"Here our model is the same as before: 784 input units, a hidden layer with 128 units, ReLU activation, 64 unit hidden layer, another ReLU, then the output layer with 10 units, and the softmax output.\n\nThe operations are available by passing in the appropriate index. For example, if you want to get first Linear operation and look at the weights, you'd use `model[0]`.",
"_____no_output_____"
]
],
[
[
"print(model[0])\nmodel[0].weight",
"Linear(in_features=784, out_features=128, bias=True)\n"
]
],
[
[
"You can also pass in an `OrderedDict` to name the individual layers and operations, instead of using incremental integers. Note that dictionary keys must be unique, so _each operation must have a different name_.",
"_____no_output_____"
]
],
[
[
"from collections import OrderedDict\nmodel = nn.Sequential(OrderedDict([\n ('fc1', nn.Linear(input_size, hidden_sizes[0])),\n ('relu1', nn.ReLU()),\n ('fc2', nn.Linear(hidden_sizes[0], hidden_sizes[1])),\n ('relu2', nn.ReLU()),\n ('output', nn.Linear(hidden_sizes[1], output_size)),\n ('softmax', nn.Softmax(dim=1))]))\nmodel",
"_____no_output_____"
]
],
[
[
"Now you can access layers either by integer or the name",
"_____no_output_____"
]
],
[
[
"print(model[0])\nprint(model.fc1)",
"Linear(in_features=784, out_features=128, bias=True)\nLinear(in_features=784, out_features=128, bias=True)\n"
]
],
[
[
"In the next notebook, we'll see how we can train a neural network to accuractly predict the numbers appearing in the MNIST images.",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
]
] |
cb92fbb366d19f0448ec793b6f82b3736f42fd76
| 4,372 |
ipynb
|
Jupyter Notebook
|
Automaton/Netflix_Scrapper.ipynb
|
tehliang01/Jupyter
|
e779188d68d35ecc8edf781eec6cb7715a5b6020
|
[
"MIT"
] | 394 |
2019-10-21T00:47:25.000Z
|
2022-03-31T21:06:03.000Z
|
Automaton/Netflix_Scrapper.ipynb
|
Zl970409/Jupyter
|
244f99aaae953fd04d44732198325a658c2c1630
|
[
"MIT"
] | 52 |
2021-08-09T22:40:20.000Z
|
2022-03-07T16:56:36.000Z
|
Automaton/Netflix_Scrapper.ipynb
|
Zl970409/Jupyter
|
244f99aaae953fd04d44732198325a658c2c1630
|
[
"MIT"
] | 205 |
2019-10-08T17:28:46.000Z
|
2022-03-31T18:26:23.000Z
| 31.228571 | 168 | 0.513266 |
[
[
[
"### Netflix Scrapper\n\nThe purpose of the code is to get details of all the Categories on Netflix and then to gather information about Sub-Categories and movies under each Sub-Category.",
"_____no_output_____"
]
],
[
[
"from bs4 import BeautifulSoup\nimport requests\nimport pandas as pd\nimport numpy as np",
"_____no_output_____"
],
[
"def make_soup(url):\n return BeautifulSoup(requests.get(url).text, 'html.parser')",
"_____no_output_____"
],
[
"def browseCategory(category, data):\n category_url = data[category-1][2]\n category = data[category-1][1]\n subCategory_details = []\n count = 1\n subCategories = []\n soup = make_soup(category_url)\n cards_list = soup.find_all('section',{'class':'nm-collections-row'})\n for card in cards_list:\n try:\n subCategory = card.find('h1').text\n movie_list = []\n movies = card.find_all('li')\n movie_count = 1\n for movie in movies:\n try:\n movie_title = movie.find('span',{'class':'nm-collections-title-name'}).text\n movie_link = movie.find('a').get('href')\n movie_list.append([movie_count, movie_title , movie_link])\n movie_count += 1\n except AttributeError:\n pass\n subCategories.append(subCategory)\n subCategory_details.append(movie_list)\n count += 1\n except AttributeError:\n pass\n return subCategories, subCategory_details, count-1",
"_____no_output_____"
],
[
"def getCategories(base_url):\n category_soup = make_soup(base_url)\n categories = category_soup.find_all('section',{'class':'nm-collections-row'})\n result=[]\n count = 1\n for category in categories:\n try:\n Title = category.find('span', {'class':'nm-collections-row-name'}).text\n url = category.find('a').get('href')\n result.append([count, Title, url])\n count += 1\n except AttributeError:\n pass\n #print(result)\n return result",
"_____no_output_____"
],
[
"def main():\n netflix_url = \"https://www.netflix.com/in/browse/genre/839338\"\n categories = getCategories(netflix_url)\n print(\"Please select one of the category\")\n df = pd.DataFrame(np.array(categories), columns=['Sr.No', 'Title', 'link'])\n print(df.to_string(index=False))\n choice = int(input('\\n\\n Please Enter your Choice: \\n'))\n subCategories, movieList, count = browseCategory(choice, categories)\n for i in range(0, count):\n print(subCategories[i],'\\n\\n')\n subCategory_df = pd.DataFrame(np.array(movieList[i]), columns=['Sr.No', 'Title', 'link'])\n print(subCategory_df.to_string(index=False))\n print(\"\\n\\n\\n\")\n \nif __name__ == '__main__':\n main()",
"_____no_output_____"
]
]
] |
[
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
]
] |
cb93128adf28d93e3dd60ce11a6a673cfbbea34d
| 232,694 |
ipynb
|
Jupyter Notebook
|
tidyingMessySongDataframe.ipynb
|
mghanava/dataWrangling
|
395fc110d9e25af82c25c612f1a4d99f2a002dda
|
[
"MIT"
] | null | null | null |
tidyingMessySongDataframe.ipynb
|
mghanava/dataWrangling
|
395fc110d9e25af82c25c612f1a4d99f2a002dda
|
[
"MIT"
] | null | null | null |
tidyingMessySongDataframe.ipynb
|
mghanava/dataWrangling
|
395fc110d9e25af82c25c612f1a4d99f2a002dda
|
[
"MIT"
] | null | null | null | 37.966063 | 1,421 | 0.326613 |
[
[
[
"# Tidy datasets are easy to manipulate, model and visualise, and have a specific structure: each variable is a column, each observation is a row, and each type of observational unit is a table (3rd level normalization of relational database).\n\n### Tidy Data, Hadley Wickham (2014)",
"_____no_output_____"
]
],
[
[
"# This statement widens the notebook page to the window size.\nfrom IPython.core.display import display, HTML\ndisplay(HTML(\"<style>.container { width:100% !important; }</style>\"))",
"_____no_output_____"
],
[
"import numpy as np\nimport pandas as pd",
"_____no_output_____"
],
[
"# error message in reading as pandas cannot decode the encoding\nbillboard = pd.read_csv('billboard.csv')",
"_____no_output_____"
],
[
"# Latin_1 encoding accept any possible byte as input and converts it to the unicode character of same code.\nbillboard = pd.read_csv('billboard.csv', encoding='latin_1')\nbillboard",
"_____no_output_____"
]
],
[
[
"## This table is messy in two ways:\n1. Table combines two observational unit: the song and its rank in each week. \n2. Variable of week is repreated in over 76 columns!",
"_____no_output_____"
]
],
[
[
"# rename columns to be more readable\nnew_cols = [\n 'wk1', 'wk2', 'wk3', 'wk4', 'wk5', 'wk6', 'wk7', 'wk8', 'wk9', 'wk10',\n 'wk11', 'wk12', 'wk13', 'wk14', 'wk15', 'wk16', 'wk17', 'wk18', 'wk19',\n 'wk20', 'wk21', 'wk22', 'wk23', 'wk24', 'wk25', 'wk26', 'wk27', 'wk28',\n 'wk29', 'wk30', 'wk31', 'wk32', 'wk33', 'wk34', 'wk35', 'wk36', 'wk37',\n 'wk38', 'wk39', 'wk40', 'wk41', 'wk42', 'wk43', 'wk44', 'wk45', 'wk46',\n 'wk47', 'wk48', 'wk49', 'wk50', 'wk51', 'wk52', 'wk53', 'wk54', 'wk55',\n 'wk56', 'wk57', 'wk58', 'wk59', 'wk60', 'wk61', 'wk62', 'wk63', 'wk64',\n 'wk65', 'wk66', 'wk67', 'wk68', 'wk69', 'wk70', 'wk71', 'wk72', 'wk73',\n 'wk74', 'wk75', 'wk76'\n]\nbillboard.rename(columns=dict(zip(billboard.columns[7:], new_cols)),\n inplace=True)\n\n# If there is no need to make the changes inplace in the original read datafile (for example for a later computation),\n# following method is more efficient as just the indices are read not the whole DataFrame.\n\n# billboard.columns.values[range(7,83)] = new_cols\n\nbillboard.rename(columns={'artist.inverted': 'artist'}, inplace=True)\n\nbillboard.head()",
"_____no_output_____"
],
[
"# `id_vars` or identifier variables are columns not to be changed while `value_vars` or measured varaible are \"unpivoted\" to the row axis leaving just two non-identifier columns, 'variable' and 'value'.\n# Note how readable and coincise the table becomes compared to the original table.\n# Note that 'year' and 'date.peaked' extra columns are eliminated as they can be computed them from the 'date.entered' and 'week' columns.\n\nbillboardmelt = billboard.melt(\n id_vars=['artist', 'track', 'time', 'genre', 'date.entered'],\n value_vars=new_cols,\n var_name='week',\n value_name='rank')\nbillboardmelt",
"_____no_output_____"
],
[
"# Covnvert the string-integer format of week column values to integers to be more readable and also be computable.\nbillboardmelt['week'] = billboardmelt['week'].apply(lambda s: int(s[2:]))\n# Sort the table to note the redundancy; multiple columns of artist, track, time, genre have repretitive records.\nbillboardmelt.sort_values(['artist', 'track'], inplace=True)\nbillboardmelt",
"_____no_output_____"
]
],
[
[
"## First let's remove date redundancy\n\n### \"date.entered\" column was usable for original dataset with one date row for all the multiple weeks columns. However, now that mutiple columns are rows each consequtive week needs its separate corresponding date otherwise for the rows of all weeks from 1 to 76, the \"date.entered\" of the first week would be repeated (date redundancy). To remove this redundancy, date of each subsequent week is computed by adding multiples of 7 to the original \"date.entered\": for example; date for week 3 is \"date.entered\" + 14 while date of week 76 is date.entered + 75*7.",
"_____no_output_____"
]
],
[
[
"# converts date.entered strings to proper panda date format to make is computable\nbillboardmelt['date.entered'] = pd.to_datetime(billboardmelt['date.entered'])\n\n# Compute date column from date.entered (the date of the first week that a song enters the market) and week columns.\nbillboardmelt['date'] = billboardmelt['date.entered'] + pd.Timedelta(\n '7 days') * (billboardmelt['week'] - 1)\n\n# Drop the original date.entered column.\nbillboardmelt.drop(['date.entered'], axis=1, inplace=True)",
"_____no_output_____"
],
[
"billboardmelt",
"_____no_output_____"
]
],
[
[
"## To remove artist, track, time and genre redundancies we move them into a SEPARATE table as distinct observational unit of song details and move week, rank and date column to another SEPARATE table as distinct observational unit of ranking and then link them together through unique indices.",
"_____no_output_____"
]
],
[
[
"# Create a separate table with track data and unique index (like primary key in database) using drop_duplictaes() method\ntracks = billboardmelt[['artist', 'track', 'time', 'genre']].drop_duplicates()\n\n# Basically, unique indices of tracks dataframe is inserted into the tracksid datafarme as a separate column. To carry the unique index \"id\" explicitly as a separate column, a new index is made by\n# reseting the index to default. tracks dataframe is assigned to tracksid dataframe with a default index and a column of unique indices.\ntracks.index.name = 'id'\ntracksid = tracks.reset_index()\ntracksid",
"_____no_output_____"
],
[
"# Link tracksid table to the ranking using unique indices by joining tracksid dataframe with billboardmelt and dropping common columns.\n# This makes the second table with unique observational unit of ranking with the relevant date and week columns.\n# Note that unique ids are repeated accoring to week number but date, week and rank are unique.\n\nranking = pd.merge(tracksid,\n billboardmelt,\n on=['artist', 'track', 'time',\n 'genre']).drop(['artist', 'track', 'time', 'genre'],\n axis=1)\nranking",
"_____no_output_____"
]
],
[
[
"## In summary, messy billboard dataframe is split into two tidy tables of tracksid and ranking.\n -Should we need rank of a song, we query on the distinct observational ranking table.\n -Should we need details of song, we query on the distinct observational song table usinf the song unique index.",
"_____no_output_____"
],
[
"## Example: Find the details for the song with maximum ranking in the first week",
"_____no_output_____"
]
],
[
[
"# First find the unique song index\nranking.loc[ranking[ranking.week == 1]['rank'].idxmax()]",
"_____no_output_____"
],
[
"# Then find the track details\ntracksid.query('id ==248')",
"_____no_output_____"
],
[
"%load_ext version_information\n%version_information numpy, pandas, matplotlib, version_information",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code"
]
] |
cb93129dfa377235cc801653367fb58b02da970f
| 27,230 |
ipynb
|
Jupyter Notebook
|
SANDBOX.ipynb
|
VCG/gp
|
cd106b604f8670a70add469d41180e34df3b1068
|
[
"MIT"
] | null | null | null |
SANDBOX.ipynb
|
VCG/gp
|
cd106b604f8670a70add469d41180e34df3b1068
|
[
"MIT"
] | null | null | null |
SANDBOX.ipynb
|
VCG/gp
|
cd106b604f8670a70add469d41180e34df3b1068
|
[
"MIT"
] | null | null | null | 158.313953 | 23,690 | 0.891443 |
[
[
[
"%load_ext autoreload\n%autoreload 2\n\nimport os; import sys; sys.path.append('..')\nimport gp\nimport gp.nets as nets\n\nfrom matplotlib.pyplot import imshow\nimport matplotlib.pyplot as plt\n%matplotlib inline",
"The autoreload extension is already loaded. To reload it, use:\n %reload_ext autoreload\n"
],
[
"X_train, y_train, X_test, y_test = gp.Patch.load_rgb('iplb')",
"Loaded /home/d/patches//iplb/ in 0.0600819587708 seconds.\n"
],
[
"aaa = X_train[:,:-1,:,:]",
"_____no_output_____"
],
[
"aaa[0][0]",
"_____no_output_____"
],
[
"gp.Util.view_rgba(aaa[0])",
"_____no_output_____"
],
[
"len(range(5,80,5))",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb931e8249e008b989372157c2bb789388ea988d
| 10,652 |
ipynb
|
Jupyter Notebook
|
test_xarray_gfs_bucket.ipynb
|
tito-am/noaa_collect
|
18f38a4d0237f4f0385d8218a06da3c5edb81769
|
[
"MIT"
] | null | null | null |
test_xarray_gfs_bucket.ipynb
|
tito-am/noaa_collect
|
18f38a4d0237f4f0385d8218a06da3c5edb81769
|
[
"MIT"
] | null | null | null |
test_xarray_gfs_bucket.ipynb
|
tito-am/noaa_collect
|
18f38a4d0237f4f0385d8218a06da3c5edb81769
|
[
"MIT"
] | null | null | null | 31.329412 | 167 | 0.498498 |
[
[
[
"import xarray as xr\nimport fsspec \nimport s3fs\nclient_kwargs={'region_name':'us-east-1'}\nconfig_kwargs = {'max_pool_connections': 30}\n\nfs = s3fs.S3FileSystem(client_kwargs=client_kwargs, \n config_kwargs=config_kwargs, \n anon=True) # public read\n\nfs.ls('noaa-gfs-bdp-pds/gfs.20200122/00/')\n\nfs.download('noaa-gfs-bdp-pds/gfs.20200122/00/gfs.t00z.pgrb2full.0p50.f027','gfs.t00z.pgrb2full.0p50.f027.grib2')\n\n\n#https://docs.opendata.aws/noaa-gefs-pds/readme.html for gefs ensemble data information\n#https://www.nco.ncep.noaa.gov/pmb/products/gens/ inventary of GFS contents",
"_____no_output_____"
],
[
"ds = xr.open_dataset('gfs.t00z.pgrb2full.0p50.f027.grib2', engine='cfgrib',backend_kwargs={'filter_by_keys': {'cfVarName': 'acpcp','typeOfLevel': 'surface'}})\n\n",
"Ignoring index file '/Users/caramelo/Downloads/03OC003/gfs.t00z.pgrb2full.0p50.f027.grib2.4cc40.idx' older than GRIB file\n"
],
[
"ds",
"_____no_output_____"
],
[
"acpcp=ds.acpcp",
"_____no_output_____"
],
[
"acpcp.attrs = ds.acpcp.attrs",
"_____no_output_____"
],
[
"acpcp.attrs['units'] = 'mm'",
"_____no_output_____"
],
[
"acpcp1d = acpcp.isel(latitude=10, longitude=10)",
"_____no_output_____"
],
[
"acpcp1d.plot()",
"_____no_output_____"
],
[
"acpcp1d.attrs",
"_____no_output_____"
],
[
"ds",
"_____no_output_____"
],
[
"ds.mean(dim=['latitude','longitude'])",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb9335480e8dbd30f0e27fbe8da6889868171c07
| 174,842 |
ipynb
|
Jupyter Notebook
|
docs/2notebook/Example_5_Explain_BERT.ipynb
|
lijianwen96/TextAttack
|
cd8a7f0c908e3ed2910e4f6360d861843d6358f9
|
[
"MIT"
] | 7 |
2021-10-30T03:44:53.000Z
|
2021-11-24T08:18:38.000Z
|
docs/2notebook/Example_5_Explain_BERT.ipynb
|
lijianwen96/TextAttack
|
cd8a7f0c908e3ed2910e4f6360d861843d6358f9
|
[
"MIT"
] | null | null | null |
docs/2notebook/Example_5_Explain_BERT.ipynb
|
lijianwen96/TextAttack
|
cd8a7f0c908e3ed2910e4f6360d861843d6358f9
|
[
"MIT"
] | 3 |
2021-03-28T09:02:57.000Z
|
2021-12-28T01:17:20.000Z
| 404.726852 | 153,339 | 0.510598 |
[
[
[
"# Explain Attacking BERT models using CAptum\n\nCaptum is a PyTorch library to explain neural networks\nHere we show a minimal example using Captum to explain BERT models from TextAttack",
"_____no_output_____"
],
[
"[](https://colab.research.google.com/github/QData/TextAttack/blob/master/docs/2notebook/Example_5_Explain_BERT.ipynb)\n\n[](https://github.com/QData/TextAttack/blob/master/docs/2notebook/Example_5_Explain_BERT.ipynb)",
"_____no_output_____"
]
],
[
[
"import torch\nfrom copy import deepcopy",
"_____no_output_____"
],
[
"from textattack.datasets import HuggingFaceDataset\nfrom textattack.models.tokenizers import AutoTokenizer\nfrom textattack.models.wrappers import HuggingFaceModelWrapper\nfrom textattack.models.wrappers import ModelWrapper\nfrom transformers import AutoModelForSequenceClassification",
"\u001b[34m\u001b[1mwandb\u001b[0m: \u001b[33mWARNING\u001b[0m W&B installed but not logged in. Run `wandb login` or set the WANDB_API_KEY env variable.\n"
],
[
"from captum.attr import IntegratedGradients, LayerConductance, LayerIntegratedGradients, LayerDeepLiftShap, InternalInfluence, LayerGradientXActivation\nfrom captum.attr import visualization as viz",
"_____no_output_____"
],
[
"device = torch.device(\"cuda:2\" if torch.cuda.is_available() else \"cpu\")\nprint(device)\ntorch.cuda.set_device(device) ",
"cuda:2\n"
],
[
"dataset = HuggingFaceDataset(\"ag_news\", None, \"train\")\noriginal_model = AutoModelForSequenceClassification.from_pretrained(\"textattack/bert-base-uncased-ag-news\")\noriginal_tokenizer = AutoTokenizer(\"textattack/bert-base-uncased-ag-news\")\nmodel = HuggingFaceModelWrapper(original_model,original_tokenizer)",
"Using custom data configuration default\nReusing dataset ag_news (/u/ss7mu/.cache/huggingface/datasets/ag_news/default/0.0.0/fb5c5e74a110037311ef5e904583ce9f8b9fbc1354290f97b4929f01b3f48b1a)\n\u001b[34;1mtextattack\u001b[0m: Loading \u001b[94mdatasets\u001b[0m dataset \u001b[94mag_news\u001b[0m, split \u001b[94mtrain\u001b[0m.\n"
],
[
"def captum_form(encoded):\n input_dict = {k: [_dict[k] for _dict in encoded] for k in encoded[0]}\n batch_encoded = { k: torch.tensor(v).to(device) for k, v in input_dict.items()}\n return batch_encoded\n\ndef get_text(tokenizer,input_ids,token_type_ids,attention_mask):\n list_of_text = []\n number = input_ids.size()[0]\n for i in range(number):\n ii = input_ids[i,].cpu().numpy()\n tt = token_type_ids[i,]\n am = attention_mask[i,]\n txt = tokenizer.decode(ii, skip_special_tokens=True)\n list_of_text.append(txt)\n return list_of_text\n \nsel =2\nencoded = model.tokenizer.batch_encode([dataset[i][0]['text'] for i in range(sel)])\nlabels = [dataset[i][1] for i in range(sel)]\n\nbatch_encoded = captum_form(encoded)\n\nclone = deepcopy(model)\nclone.model.to(device)\n\ndef calculate(input_ids,token_type_ids,attention_mask):\n #convert back to list of text\n return clone.model(input_ids,token_type_ids,attention_mask)[0]\n \n# x = calculate(**batch_encoded) \n\nlig = LayerIntegratedGradients(calculate, clone.model.bert.embeddings)\n# lig = InternalInfluence(calculate, clone.model.bert.embeddings)\n# lig = LayerGradientXActivation(calculate, clone.model.bert.embeddings)\n\nbsl = torch.zeros(batch_encoded['input_ids'].size()).type(torch.LongTensor).to(device)\nlabels = torch.tensor(labels).to(device)\n\nattributions,delta = lig.attribute(inputs=batch_encoded['input_ids'],\n baselines=bsl,\n additional_forward_args=(batch_encoded['token_type_ids'], batch_encoded['attention_mask']),\n n_steps = 10,\n target = labels,\n return_convergence_delta=True\n )\natts = attributions.sum(dim=-1).squeeze(0)\natts = atts / torch.norm(atts)",
"_____no_output_____"
],
[
"# print(attributions.size())\natts = attributions.sum(dim=-1).squeeze(0)\natts = atts / torch.norm(atts)",
"_____no_output_____"
],
[
"from textattack.attack_recipes import PWWSRen2019\nattack = PWWSRen2019.build(model)",
"\u001b[34;1mtextattack\u001b[0m: Unknown if model of class <class 'textattack.models.wrappers.huggingface_model_wrapper.HuggingFaceModelWrapper'> compatible with goal function <class 'textattack.goal_functions.classification.untargeted_classification.UntargetedClassification'>.\n"
],
[
"results_iterable = attack.attack_dataset(dataset, indices=range(10))\n\nviz_list = []\n\nfor n,result in enumerate(results_iterable):\n orig = result.original_text()\n pert = result.perturbed_text()\n \n encoded = model.tokenizer.batch_encode([orig])\n batch_encoded = captum_form(encoded)\n x = calculate(**batch_encoded) \n print(x)\n print(dataset[n][1])\n \n pert_encoded = model.tokenizer.batch_encode([pert])\n pert_batch_encoded = captum_form(pert_encoded)\n x_pert = calculate(**pert_batch_encoded) \n\n\n attributions,delta = lig.attribute(inputs=batch_encoded['input_ids'],\n# baselines=bsl,\n additional_forward_args=(batch_encoded['token_type_ids'], batch_encoded['attention_mask']),\n n_steps = 10,\n target = torch.argmax(calculate(**batch_encoded)).item(),\n return_convergence_delta=True\n )\n attributions_pert,delta_pert = lig.attribute(inputs=pert_batch_encoded['input_ids'],\n# baselines=bsl,\n additional_forward_args=(pert_batch_encoded['token_type_ids'], pert_batch_encoded['attention_mask']),\n n_steps = 10,\n target = torch.argmax(calculate(**pert_batch_encoded)).item(),\n return_convergence_delta=True\n )\n \n orig = original_tokenizer.tokenizer.tokenize(orig)\n pert = original_tokenizer.tokenizer.tokenize(pert)\n \n atts = attributions.sum(dim=-1).squeeze(0)\n atts = atts / torch.norm(atts)\n\n \n \n atts_pert = attributions_pert.sum(dim=-1).squeeze(0)\n atts_pert = atts_pert / torch.norm(atts)\n \n \n \n all_tokens = original_tokenizer.tokenizer.convert_ids_to_tokens(batch_encoded['input_ids'][0])\n all_tokens_pert = original_tokenizer.tokenizer.convert_ids_to_tokens(pert_batch_encoded['input_ids'][0])\n \n \n v = viz.VisualizationDataRecord(\n atts[:45].detach().cpu(),\n torch.max(x).item(),\n torch.argmax(x,dim=1).item(),\n dataset[n][1],\n 2,\n atts.sum().detach(), \n all_tokens[:45],\n delta)\n \n v_pert = viz.VisualizationDataRecord(\n atts_pert[:45].detach().cpu(),\n torch.max(x_pert).item(),\n torch.argmax(x_pert,dim=1).item(),\n dataset[n][1],\n 2,\n atts_pert.sum().detach(), \n all_tokens_pert[:45],\n delta_pert)\n viz_list.append(v)\n viz_list.append(v_pert)\n\n# print(result.perturbed_text())\n print(result.__str__(color_method='ansi'))\n\n",
"tensor([[-3.3016, -2.1467, 2.0953, 2.7272]], device='cuda:2',\n grad_fn=<AddmmBackward>)\n2\n\u001b[94mBusiness (96%)\u001b[0m --> \u001b[35mSci/tech (68%)\u001b[0m\n\nWall St. \u001b[94mBears\u001b[0m Claw Back Into the \u001b[94mBlack\u001b[0m (Reuters) Reuters - Short-sellers, Wall Street's \u001b[94mdwindling\u001b[0m\\\u001b[94mband\u001b[0m of ultra-cynics, are \u001b[94mseeing\u001b[0m \u001b[94mgreen\u001b[0m again.\n\nWall St. \u001b[35msuffer\u001b[0m Claw Back Into the \u001b[35mlightlessness\u001b[0m (Reuters) Reuters - Short-sellers, Wall Street's \u001b[35mdwindle\u001b[0m\\\u001b[35misthmus\u001b[0m of ultra-cynics, are \u001b[35mexamine\u001b[0m \u001b[35mgreenish\u001b[0m again.\ntensor([[-4.8619, -0.8199, 3.9263, 1.1104]], device='cuda:2',\n grad_fn=<AddmmBackward>)\n2\n\u001b[94mBusiness (100%)\u001b[0m --> \u001b[35mSci/tech (50%)\u001b[0m\n\nCarlyle Looks Toward Commercial Aerospace (Reuters) Reuters - Private \u001b[94minvestment\u001b[0m firm Carlyle Group,\\which has a reputation for \u001b[94mmaking\u001b[0m well-timed and occasionally\\controversial plays in the \u001b[94mdefense\u001b[0m industry, has quietly \u001b[94mplaced\u001b[0m\\its bets on another part of the market.\n\nCarlyle Looks Toward Commercial Aerospace (Reuters) Reuters - Private \u001b[35minvestiture\u001b[0m firm Carlyle Group,\\which has a reputation for \u001b[35mca-ca\u001b[0m well-timed and occasionally\\controversial plays in the \u001b[35mdenial\u001b[0m industry, has quietly \u001b[35msite\u001b[0m\\its bets on another part of the market.\ntensor([[-4.3489, -2.6201, 4.6131, 1.2504]], device='cuda:2',\n grad_fn=<AddmmBackward>)\n2\n\u001b[94mBusiness (100%)\u001b[0m --> \u001b[35mSci/tech (100%)\u001b[0m\n\nOil and Economy \u001b[94mCloud\u001b[0m Stocks' Outlook (Reuters) Reuters - Soaring crude prices plus worries\\about the economy and the outlook for \u001b[94mearnings\u001b[0m are expected to\\hang over the \u001b[94mstock\u001b[0m market next week during the depth of the\\summer doldrums.\n\nOil and Economy \u001b[35mswarm\u001b[0m Stocks' Outlook (Reuters) Reuters - Soaring crude prices plus worries\\about the economy and the outlook for \u001b[35mlucre\u001b[0m are expected to\\hang over the \u001b[35mgillyflower\u001b[0m market next week during the depth of the\\summer doldrums.\ntensor([[ 0.0232, -3.9863, 0.2248, 3.1301]], device='cuda:2',\n grad_fn=<AddmmBackward>)\n2\n\u001b[94mBusiness (78%)\u001b[0m --> \u001b[91mWorld (58%)\u001b[0m\n\nIraq \u001b[94mHalts\u001b[0m Oil Exports from Main Southern Pipeline (Reuters) Reuters - Authorities have halted oil export\\\u001b[94mflows\u001b[0m from the main pipeline in southern \u001b[94mIraq\u001b[0m after\\intelligence showed a rebel militia could \u001b[94mstrike\u001b[0m\\infrastructure, an oil official said on Saturday.\n\nIraq \u001b[91mkibosh\u001b[0m Oil Exports from Main Southern Pipeline (Reuters) Reuters - Authorities have halted oil export\\\u001b[91mhang\u001b[0m from the main pipeline in southern \u001b[91mIrak\u001b[0m after\\intelligence showed a rebel militia could \u001b[91mfall\u001b[0m\\infrastructure, an oil official said on Saturday.\ntensor([[-3.2209, -3.5844, 5.4475, -0.4632]], device='cuda:2',\n grad_fn=<AddmmBackward>)\n2\n\u001b[94mBusiness (99%)\u001b[0m --> \u001b[91mWorld (82%)\u001b[0m\n\nOil prices soar to all-time record, posing new menace to \u001b[94mUS\u001b[0m \u001b[94meconomy\u001b[0m (AFP) AFP - Tearaway world oil prices, toppling records and straining wallets, present a new economic menace barely three months before the US presidential elections.\n\nOil prices soar to all-time record, posing new menace to \u001b[91muranium\u001b[0m \u001b[91mthriftiness\u001b[0m (AFP) AFP - Tearaway world oil prices, toppling records and straining wallets, present a new economic menace barely three months before the US presidential elections.\ntensor([[-4.5941, -2.0886, 5.1424, 0.3504]], device='cuda:2',\n grad_fn=<AddmmBackward>)\n2\n\u001b[94mBusiness (100%)\u001b[0m --> \u001b[35mSci/tech (55%)\u001b[0m\n\n\u001b[94mStocks\u001b[0m \u001b[94mEnd\u001b[0m \u001b[94mUp\u001b[0m, But Near Year Lows (Reuters) Reuters - \u001b[94mStocks\u001b[0m \u001b[94mended\u001b[0m slightly higher on Friday\\but \u001b[94mstayed\u001b[0m \u001b[94mnear\u001b[0m \u001b[94mlows\u001b[0m for the year as \u001b[94moil\u001b[0m prices surged past #36;\u001b[94m46\u001b[0m\\a \u001b[94mbarrel\u001b[0m, \u001b[94moffsetting\u001b[0m a \u001b[94mpositive\u001b[0m \u001b[94moutlook\u001b[0m from computer \u001b[94mmaker\u001b[0m\\Dell Inc. (DELL.O)\n\n\u001b[35mstock\u001b[0m \u001b[35mterminate\u001b[0m \u001b[35mup\u001b[0m, But Near Year Lows (Reuters) Reuters - \u001b[35minventory\u001b[0m \u001b[35mterminate\u001b[0m slightly higher on Friday\\but \u001b[35mcontinue\u001b[0m \u001b[35mvirtually\u001b[0m \u001b[35mmoo\u001b[0m for the year as \u001b[35membrocate\u001b[0m prices surged past #36;\u001b[35mxlvi\u001b[0m\\a \u001b[35mdrum\u001b[0m, \u001b[35mcancel\u001b[0m a \u001b[35mincontrovertible\u001b[0m \u001b[35mmindset\u001b[0m from computer \u001b[35mCreator\u001b[0m\\Dell Inc. (DELL.O)\ntensor([[-3.4198, -3.5160, 6.3658, -1.3811]], device='cuda:2',\n grad_fn=<AddmmBackward>)\n2\n\u001b[94mBusiness (100%)\u001b[0m --> \u001b[91mWorld (76%)\u001b[0m\n\nMoney Funds \u001b[94mFell\u001b[0m in \u001b[94mLatest\u001b[0m \u001b[94mWeek\u001b[0m (AP) AP - \u001b[94mAssets\u001b[0m of the nation's retail money \u001b[94mmarket\u001b[0m \u001b[94mmutual\u001b[0m funds \u001b[94mfell\u001b[0m by #36;1.17 \u001b[94mbillion\u001b[0m in the latest week to #36;849.98 \u001b[94mtrillion\u001b[0m, the Investment Company \u001b[94mInstitute\u001b[0m \u001b[94msaid\u001b[0m Thursday.\n\nMoney Funds \u001b[91mhide\u001b[0m in \u001b[91mup-to-the-minute\u001b[0m \u001b[91mhebdomad\u001b[0m (AP) AP - \u001b[91masset\u001b[0m of the nation's retail money \u001b[91mcommercialise\u001b[0m \u001b[91mcommon\u001b[0m funds \u001b[91mcruel\u001b[0m by #36;1.17 \u001b[91mgazillion\u001b[0m in the latest week to #36;849.98 \u001b[91m1000000000000\u001b[0m, the Investment Company \u001b[91mconstitute\u001b[0m \u001b[91mstate\u001b[0m Thursday.\ntensor([[-3.9547, -3.9751, 5.8308, 0.0899]], device='cuda:2',\n grad_fn=<AddmmBackward>)\n2\n\u001b[94mBusiness (100%)\u001b[0m --> \u001b[91mWorld (76%)\u001b[0m\n\n\u001b[94mFed\u001b[0m \u001b[94mminutes\u001b[0m \u001b[94mshow\u001b[0m \u001b[94mdissent\u001b[0m over inflation (USATODAY.com) USATODAY.com - Retail sales \u001b[94mbounced\u001b[0m \u001b[94mback\u001b[0m a bit in July, and \u001b[94mnew\u001b[0m claims for jobless \u001b[94mbenefits\u001b[0m fell \u001b[94mlast\u001b[0m \u001b[94mweek\u001b[0m, the government said Thursday, indicating the economy is \u001b[94mimproving\u001b[0m from a midsummer slump.\n\n\u001b[91meat\u001b[0m \u001b[91mhour\u001b[0m \u001b[91mtestify\u001b[0m \u001b[91mprotest\u001b[0m over inflation (USATODAY.com) USATODAY.com - Retail sales \u001b[91mbound\u001b[0m \u001b[91mhind\u001b[0m a bit in July, and \u001b[91myoung\u001b[0m claims for jobless \u001b[91mwelfare\u001b[0m fell \u001b[91mdeath\u001b[0m \u001b[91mworkweek\u001b[0m, the government said Thursday, indicating the economy is \u001b[91mamend\u001b[0m from a midsummer slump.\ntensor([[-3.9061, -3.7002, 5.6281, 0.1403]], device='cuda:2',\n grad_fn=<AddmmBackward>)\n2\n\u001b[94mBusiness (100%)\u001b[0m --> \u001b[35mSci/tech (100%)\u001b[0m\n\nSafety \u001b[94mNet\u001b[0m (Forbes.com) Forbes.com - After earning a PH.D. in Sociology, Danny Bazil Riley started to work as the general manager at a commercial real estate firm at an annual base salary of #36;70,000. Soon after, a financial planner stopped by his desk to drop off brochures about insurance benefits available through his employer. But, at 32, \"buying insurance was the furthest thing from my mind,\" says Riley.\n\nSafety \u001b[35mcyberspace\u001b[0m (Forbes.com) Forbes.com - After earning a PH.D. in Sociology, Danny Bazil Riley started to work as the general manager at a commercial real estate firm at an annual base salary of #36;70,000. Soon after, a financial planner stopped by his desk to drop off brochures about insurance benefits available through his employer. But, at 32, \"buying insurance was the furthest thing from my mind,\" says Riley.\ntensor([[-5.1085, -1.4586, 3.8248, 1.9658]], device='cuda:2',\n grad_fn=<AddmmBackward>)\n2\n\u001b[94mBusiness (100%)\u001b[0m --> \u001b[91mWorld (71%)\u001b[0m\n\nWall St. Bears \u001b[94mClaw\u001b[0m Back Into the Black \u001b[94mNEW\u001b[0m YORK (Reuters) - Short-sellers, Wall Street's dwindling band of ultra-cynics, are seeing green again.\n\nWall St. Bears \u001b[91mchela\u001b[0m Back Into the Black \u001b[91mnovel\u001b[0m YORK (Reuters) - Short-sellers, Wall Street's dwindling band of ultra-cynics, are seeing green again.\n"
],
[
"print('\\033[1m', 'Visualizations For AG NEWS', '\\033[0m')\nviz.visualize_text(viz_list)",
"\u001b[1m Visualizations For AG NEWS \u001b[0m\n"
],
[
"# reference for viz datarecord\n# def __init__(\n# self,\n# word_attributions,\n# pred_prob,\n# pred_class,\n# true_class,\n# attr_class,\n# attr_score,\n# raw_input,\n# convergence_score,\n# ):",
"_____no_output_____"
]
]
] |
[
"markdown",
"code"
] |
[
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb9361ff29eedb6caaeba4f0e159ed093ed1f5d1
| 665,605 |
ipynb
|
Jupyter Notebook
|
project_main/project_advanced_line_dect.ipynb
|
govinsprabhu/Advanced_line_dectection
|
bea5d58fe7fc495cd5c292ab19f2f087e5911bbf
|
[
"MIT"
] | null | null | null |
project_main/project_advanced_line_dect.ipynb
|
govinsprabhu/Advanced_line_dectection
|
bea5d58fe7fc495cd5c292ab19f2f087e5911bbf
|
[
"MIT"
] | null | null | null |
project_main/project_advanced_line_dect.ipynb
|
govinsprabhu/Advanced_line_dectection
|
bea5d58fe7fc495cd5c292ab19f2f087e5911bbf
|
[
"MIT"
] | null | null | null | 898.252362 | 129,068 | 0.94885 |
[
[
[
"## Advanced Lane Finding Project\n\nThe goals / steps of this project are the following:\n\n* Compute the camera calibration matrix and distortion coefficients given a set of chessboard images.\n* Apply a distortion correction to raw images.\n* Use color transforms, gradients, etc., to create a thresholded binary image.\n* Apply a perspective transform to rectify binary image (\"birds-eye view\").\n* Detect lane pixels and fit to find the lane boundary.\n* Determine the curvature of the lane and vehicle position with respect to center.\n* Warp the detected lane boundaries back onto the original image.\n* Output visual display of the lane boundaries and numerical estimation of lane curvature and vehicle position.\n\n---\n## First, I'll compute the camera calibration using chessboard images",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport cv2\nimport glob\nimport matplotlib.pyplot as plt\nimport matplotlib.image as mpimg\n%matplotlib inline\n\n# prepare object points, like (0,0,0), (1,0,0), (2,0,0) ....,(6,5,0)\nobjp = np.zeros((6*9,3), np.float32)\nobjp[:,:2] = np.mgrid[0:9,0:6].T.reshape(-1,2)\n\n# Arrays to store object points and image points from all the images.\nobj_points = [] # 3d points in real world space\nimg_points = [] # 2d points in image plane.\n\n# Make a list of calibration images\nimages = glob.glob('../camera_cal/calibration*.jpg')\n\n# Step through the list and search for chessboard corners\nfor fname in images:\n img = cv2.imread(fname)\n gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)\n\n # Find the chessboard corners\n ret, corners = cv2.findChessboardCorners(gray, (9,6),None)\n\n # If found, add object points, image points\n if ret:\n #print('found the corners')\n obj_points.append(objp)\n img_points.append(corners)\n\n # Draw and display the corners\n img = cv2.drawChessboardCorners(img, (9,6), corners, ret)\n plt.imshow(img)\n #cv2.waitKey(500)\n else:\n print('not found')\n\ncv2.destroyAllWindows()",
"not found\nnot found\nnot found\n"
]
],
[
[
"## Now applying distortion correction to the row image",
"_____no_output_____"
]
],
[
[
"def remove_distortion(img, obj_points, img_points):\n #print('inside remove distotion')\n gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)\n ret, mtx, dist, rvecs, tvecs = cv2.calibrateCamera(obj_points, img_points, gray.shape[::-1], None, None)\n #ret, corners = cv2.findChessboardCorners(gray, (nx, ny), None)\n un_dst = cv2.undistort(img, mtx, dist, None, mtx)\n #f, (ax1, ax2) = plt.subplots(1, 2, figsize=(24, 9))\n #f.tight_layout()\n #ax1.imshow(img)\n #ax1.set_title('Original Image', fontsize=50)\n #ax2.imshow(un_dst)\n #ax2.set_title('Undistorted Image', fontsize=50)\n #plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)\n #plt.savefig('un_distorted_real.png')\n #plt.show()\n return dist, mtx, un_dst\n\n#org_img = cv2.imread('../test_images/straight_lines1.jpg')\n#img = cv2.imread('../camera_cal/calibration1.jpg')\n#org_img = cv2.imread('../test_images/test5.jpg')\norg_img = cv2.imread('not_working.jpg')\nplt.imshow(org_img)\ndist, mtx, un_dst = remove_distortion(org_img, obj_points, img_points)\n",
"_____no_output_____"
]
],
[
[
"## Applying color and gradient transform to the image",
"_____no_output_____"
]
],
[
[
"def gradient_color_transform(img, s_thresh=(170, 255), sx_thresh=(30, 80)):\n hls = cv2.cvtColor(img, cv2.COLOR_RGB2HLS)\n s_channel = hls[:,:,2]\n gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)\n\n sobelx = cv2.Sobel(gray, cv2.CV_64F, 1, 0) # Take the derivative in x\n abs_sobelx = np.absolute(sobelx) # Absolute x derivative to accentuate lines away from horizontal\n scaled_sobel = np.uint8(255*abs_sobelx/np.max(abs_sobelx))\n thresh_min = 20\n thresh_max = 100\n sxbinary = np.zeros_like(scaled_sobel)\n sxbinary[(scaled_sobel >= thresh_min) & (scaled_sobel <= thresh_max)] = 1\n s_thresh_min = 170\n s_thresh_max = 255\n s_binary = np.zeros_like(s_channel)\n s_binary[(s_channel >= s_thresh_min) & (s_channel <= s_thresh_max)] = 1\n color_binary = np.dstack(( np.zeros_like(sxbinary), sxbinary, s_binary)) * 255\n\n combined_binary = np.zeros_like(sxbinary)\n combined_binary[(s_binary == 1) | (sxbinary == 1)] = 1\n\n\n #Plotting thresholded images\n# f, (ax1, ax2) = plt.subplots(1, 2, figsize=(20, 10))\n# ax1.set_title('Stacked thresholds')\n# ax1.imshow(color_binary)\n\n# ax2.set_title('Combined S channel and gradient thresholds')\n# ax2.imshow(combined_binary, cmap='gray')\n# plt.imshow(combined_binary, cmap='gray')\n# plt.savefig('binary_combo_real.png')\n# plt.savefig('../combined_image.jpg')\n return combined_binary\n\n\nimage = un_dst\n#image = cv2.imread('../test_images/test1.jpg')\nplt.imshow(image)\nplt.show()\ncombined_image = gradient_color_transform(image)\n",
"_____no_output_____"
]
],
[
[
"## Applying perspective transform",
"_____no_output_____"
]
],
[
[
"def transform(cur, dist, mtx):\n #img = np.copy(cur)\n img = cur\n #print(img.shape)\n un_dst = cv2.undistort(img, mtx, dist, None, mtx)\n #plt.imshow(un_dst)\n #plt.show()\n #print(un_dst.shape)\n img_size = (img.shape[1], img.shape[0])\n\n src = np.float32([[200, img_size[1] - 1], [595, 450], [685, 450], [1100, img_size[1] - 1]])\n dst = np.float32([[350, img_size[1] - 1],\n [350, 1],\n [970, 1],\n [970, img_size[1] - 1]])\n \n M = cv2.getPerspectiveTransform(src, dst)\n Minv = cv2.getPerspectiveTransform(dst, src)\n warped = cv2.warpPerspective(un_dst, M, img_size)\n \n color=[255, 0, 0] \n thickness=5\n# cv2.line(img, (200, img_size[1] - 1), (590, 450), color, thickness)\n# cv2.line(img, (590, 450), (690, 450), color, thickness)\n# cv2.line(img, (690, 450), (1100, img_size[1] - 1), color, thickness)\n# cv2.line(img, (1100, img_size[1] - 1), (300, img_size[1] - 1), color, thickness)\n \n# cv2.line(warped, (350, img_size[1] - 1), (350, 1), color, thickness)\n# cv2.line(warped, (350, 1), (970, 1), color, thickness)\n# cv2.line(warped, (970, 1), (970, img_size[1] - 1), color, thickness)\n# cv2.line(warped, (970, img_size[1] - 1), (350, img_size[1] - 1), color, thickness)\n \n \n \n# f, (ax1, ax2) = plt.subplots(1, 2, figsize=(30,9))\n# f.tight_layout()\n# ax1.imshow(img)\n# ax1.set_title('Un-distorted Image with source points', fontsize=50)\n# ax2.imshow(warped)\n# ax2.set_title('Warped Image with destination', fontsize=50)\n# plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)\n# plt.savefig('warped_straight_lines_real.png')\n# plt.show()\n return warped, Minv\n\n#combined_image = cv2.cvtColor(mpimg.imread('../test_images/straight_lines1.jpg'), cv2.COLOR_RGB2GRAY)\n#combined_image = mpimg.imread('../binary-combo-img.jpg')\nplt.imshow(combined_image, cmap='gray')\nbinary_warped, Minv = transform(combined_image, dist, mtx)\nplt.imshow(binary_warped, cmap='gray')",
"_____no_output_____"
]
],
[
[
"## Detecting lane And Finding the curvature",
"_____no_output_____"
]
],
[
[
"def getCurvature_and_center_of_line(leftx, lefty, rightx, righty, y_eval, shape):\n ym_per_pix = 30 / 720 # meters per pixel in y dimension\n xm_per_pix = 3.7 / 700 # meters per pixel in x dimension\n left_fit = np.polyfit(lefty * ym_per_pix, leftx * xm_per_pix, 2)\n right_fit = np.polyfit(righty * ym_per_pix, rightx * xm_per_pix, 2)\n \n # Generate x and y values for plotting\n #print(y_eval, left_fit, right_fit)\n \n y_eval_meters = y_eval * ym_per_pix\n \n left_bottom = left_fit[0] * y_eval_meters ** 2 + left_fit[1] * y_eval_meters + left_fit[2]\n right_bottom = right_fit[0] * y_eval_meters ** 2 + right_fit[1] * y_eval_meters + right_fit[2]\n line_center = (left_bottom + right_bottom) / 2\n img_center = shape[1] /2 * xm_per_pix\n dif = img_center - line_center\n #print('curvature left ', left_bottom, 'right',right_bottom, 'line_center ', line_center, img_center, dif)\n \n \n left_curvature = ((1 + (2 * left_fit[0] * y_eval *ym_per_pix + left_fit[1]) ** 2) ** 1.5) / np.absolute(2 * left_fit[0])\n right_curvature = ((1 + (2 * right_fit[0] * y_eval *ym_per_pix + right_fit[1]) ** 2) ** 1.5) / np.absolute(2 * right_fit[0])\n \n mean_curvature = (left_curvature + right_curvature) /2\n \n return mean_curvature, dif\n\n\ndef find_lane_pixels(binary_warped):\n # Take a histogram of the bottom half of the image\n histogram = np.sum(binary_warped[binary_warped.shape[0] // 2:, :], axis=0)\n #plt.plot(histogram)\n # Create an output image to draw on and visualize the result\n out_img = np.dstack((binary_warped, binary_warped, binary_warped))\n # Find the peak of the left and right halves of the histogram\n # These will be the starting point for the left and right lines\n midpoint = np.int(histogram.shape[0] // 2)\n left_margin = 300 #to avoid wrong identification of dividers as lines\n leftx_base = left_margin + np.argmax(histogram[left_margin:midpoint])\n #print('leftx_base', leftx_base, midpoint)\n rightx_base = np.argmax(histogram[midpoint:]) + midpoint\n #print(leftx_base,rightx_base)\n # HYPERPARAMETERS\n # Choose the number of sliding windows\n nwindows = 9\n # Set the width of the windows +/- margin\n margin = 100\n # Set minimum number of pixels found to recenter window\n minpix = 50\n\n # Set height of windows - based on nwindows above and image shape\n window_height = np.int(binary_warped.shape[0] // nwindows)\n # Identify the x and y positions of all nonzero pixels in the image\n nonzero = binary_warped.nonzero()\n # print('non-zero', len(nonzero[0]), len(nonzero[1]))\n nonzeroy = np.array(nonzero[0])\n nonzerox = np.array(nonzero[1])\n # Current positions to be updated later for each window in nwindows\n leftx_current = leftx_base\n rightx_current = rightx_base\n\n # Create empty lists to receive left and right lane pixel indices\n left_lane_inds = []\n right_lane_inds = []\n\n # Step through the windows one by one\n for window in range(nwindows):\n # Identify window boundaries in x and y (and right and left)\n win_y_low = binary_warped.shape[0] - (window + 1) * window_height\n win_y_high = binary_warped.shape[0] - window * window_height\n ### TO-DO: Find the four below boundaries of the window ###\n\n win_xleft_low = leftx_current - margin # Update this\n win_xleft_high = leftx_current + margin # Update this\n win_xright_low = rightx_current - margin # Update this\n win_xright_high = rightx_current + margin # Update this\n\n # Draw the windows on the visualization image\n# cv2.rectangle(out_img, (win_xleft_low, win_y_low),\n# (win_xleft_high, win_y_high), (0, 255, 0), 2)\n# cv2.rectangle(out_img, (win_xright_low, win_y_low),\n# (win_xright_high, win_y_high), (0, 255, 0), 2)\n\n # Identify the nonzero pixels in x and y within the window \n good_left_inds = ((nonzerox > win_xleft_low) & (nonzerox < win_xleft_high) & (nonzeroy < win_y_high) & (\n nonzeroy > win_y_low)).nonzero()[0]\n good_right_inds = ((nonzerox > win_xright_low) & (nonzerox < win_xright_high) & (nonzeroy < win_y_high) & (\n nonzeroy > win_y_low)).nonzero()[0]\n\n # Append these indices to the lists\n if (good_left_inds.shape[0] != 0):\n left_lane_inds.append(good_left_inds)\n if (good_right_inds.shape[0] != 0):\n right_lane_inds.append(good_right_inds)\n #print(win_y_low,good_left_inds.shape)\n #(`right` or `leftx_current`) on their mean position #\n try:\n if (len(good_left_inds) > minpix):\n leftx_current = np.int(np.mean(nonzerox[good_left_inds]))\n rightx_current = np.int(np.mean(nonzerox[good_right_inds]))\n except:\n pass\n #print(len(left_lane_inds),'l eft ', len(right_lane_inds))\n # Concatenate the arrays of indices (previously was a list of lists of pixels)\n try:\n left_lane_inds = np.concatenate(left_lane_inds)\n right_lane_inds = np.concatenate(right_lane_inds)\n except ValueError:\n \n pass\n\n # Extract left and right line pixel positions\n leftx = nonzerox[left_lane_inds]\n lefty = nonzeroy[left_lane_inds]\n rightx = nonzerox[right_lane_inds]\n righty = nonzeroy[right_lane_inds]\n #print(leftx.shape, lefty.shape, rightx.shape, righty.shape)\n return leftx, lefty, rightx, righty, out_img\n\n\ndef fit_polynomial(binary_warped):\n # Find our lane pixels first\n leftx, lefty, rightx, righty, out_img = find_lane_pixels(binary_warped)\n\n ### TO-DO: Fit a second order polynomial to each using `np.polyfit` ###\n left_fit = np.polyfit(lefty, leftx, 2)\n right_fit = np.polyfit(righty, rightx, 2)\n #print('left fit', left_fit, right_fit)\n # Generate x and y values for plotting\n ploty = np.linspace(0, binary_warped.shape[0] - 1, binary_warped.shape[0])\n try:\n left_fitx = left_fit[0] * ploty ** 2 + left_fit[1] * ploty + left_fit[2]\n right_fitx = right_fit[0] * ploty ** 2 + right_fit[1] * ploty + right_fit[2]\n except TypeError:\n # Avoids an error if `left` and `right_fit` are still none or incorrect\n print('The function failed to fit a line!')\n left_fitx = 1 * ploty ** 2 + 1 * ploty\n right_fitx = 1 * ploty ** 2 + 1 * ploty\n\n ## Visualization ##\n # Colors in the left and right lane regions\n out_img[lefty, leftx] = [255, 0, 0]\n out_img[righty, rightx] = [0, 0, 255]\n \n y_eval = np.max(ploty)\n mean_curvature, center_to_image = getCurvature_and_center_of_line(leftx, lefty, rightx, righty, y_eval, out_img.shape)\n #Plots the left and right polynomials on the lane lines\n #plt.plot(left_fitx, ploty, color='yellow')\n #plt.plot(right_fitx, ploty, color='yellow')\n lines_mid = (left_fitx[-1] + right_fitx[-1]) //2\n \n return out_img, left_fitx, right_fitx, ploty, mean_curvature, center_to_image\n\n\nplt.imshow(binary_warped)\nplt.show()\nwarped_line_detected_box_method, left_fitx, right_fitx, ploty, mean_curvature, center_to_image = fit_polynomial(binary_warped)\nplt.imshow(warped_line_detected_box_method)\nplt.show()\n#plt.savefig('warped_line_detected_box_method.png')\n",
"_____no_output_____"
]
],
[
[
"## Warp line back into the original image",
"_____no_output_____"
]
],
[
[
"\ndef transform_back_add_text(binary_warped, left_fitx, right_fitx, ploty, org_img, mean_curvature, center_to_image ):\n warp_zero = np.zeros_like(binary_warped).astype(np.uint8)\n color_warp = np.dstack((warp_zero, warp_zero, warp_zero))\n # Recast the x and y points into usable format for cv2.fillPoly()\n pts_left = np.array([np.transpose(np.vstack([left_fitx, ploty]))])\n pts_right = np.array([np.flipud(np.transpose(np.vstack([right_fitx, ploty])))])\n pts = np.hstack((pts_left, pts_right))\n #print(np.int_(pts).shape)\n # Draw the lane onto the warped blank image\n cv2.fillPoly(color_warp, np.int_([pts]), (0, 255, 0))\n\n # Warp the blank back to original image space using inverse perspective matrix (Minv)\n\n newwarp = cv2.warpPerspective(color_warp, Minv, (org_img.shape[1], org_img.shape[0]))\n # Combine the result with the original image\n result = cv2.addWeighted(org_img, 1, newwarp, 0.3, 0)\n \n \n direction = 'left' if center_to_image < 0 else 'right'\n string_direction = 'Vehicle is %.2fm' %abs(center_to_image)+' '+direction+' to the center'\n string_mean_curvature = 'Radius of Curvature %.2f(m)' %mean_curvature \n \n cv2.putText(result,string_mean_curvature, (100,50), cv2.FONT_HERSHEY_SIMPLEX, 2,(255,255,255),2)\n cv2.putText(result,string_direction, (100,100), cv2.FONT_HERSHEY_SIMPLEX, 2,(255,255,255),2)\n #cv2.imshow(result)\n #plt.imshow(result)\n return result\n \nfinal_image = transform_back_add_text(binary_warped, left_fitx, right_fitx, ploty,org_img, mean_curvature, center_to_image) \n#plt.imshow(final_image)\n",
"_____no_output_____"
]
],
[
[
"## Pipeline for video",
"_____no_output_____"
]
],
[
[
"def pipeline(image):\n dist, mtx, un_dst = remove_distortion(image, obj_points, img_points)\n combined_image = gradient_color_transform(un_dst)\n binary_warped, Minv = transform(combined_image, dist, mtx)\n warped_line_detected_box_method, left_fitx, right_fitx, ploty, mean_curvature, center_to_image = fit_polynomial(binary_warped)\n final_image = transform_back_add_text(binary_warped, left_fitx, right_fitx, ploty,image, mean_curvature, center_to_image)\n #plt.imshow(final_image)\n #print(final_image.shape)\n return final_image\n\norg_img = cv2.imread('not_working.jpg')\nres = pipeline(org_img) \nplt.imshow(res)\nplt.show()\n#org_img = cv2.imread('../test_images/test1.jpg')\n#res = pipeline(org_img) \n#plt.imshow(res)\n#plt.show()\n ",
"_____no_output_____"
],
[
"def process_image(image):\n return pipeline(image)",
"_____no_output_____"
],
[
"from moviepy.editor import VideoFileClip\nfrom IPython.display import HTML",
"_____no_output_____"
],
[
"video_output = '../harder_challenge_video_output_copy1.mp4'\nvideo_input = \"../harder_challenge_video.mp4\"\nclip1 = VideoFileClip(video_input)\nprint(clip1.duration)\nwhite_clip = clip1.fl_image(process_image) #NOTE: this function expects color images!!\n#clip1.reader.close()\n#clip1.audio.reader.close_proc()\n\n%time white_clip.write_videofile(video_output, audio=False)",
"47.96\n[MoviePy] >>>> Building video ../harder_challenge_video_output_copy1.mp4\n[MoviePy] Writing video ../harder_challenge_video_output_copy1.mp4\n"
],
[
"#print(white_clip)\n\nHTML(\"\"\"\n<video width=\"960\" height=\"540\" controls>\n <source src=\"{0}\">\n</video>\n\"\"\".format(video_output))",
"_____no_output_____"
],
[
"video = VideoFileClip('../project_video.mp4')\nnp_frame = video.get_frame(23)\nplt.imshow(np_frame)\n#plt.savefig('not_working.jpg')\nvideo.save_frame('not_working.jpg', t=23) ",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb936f813014b8dc4445efe76224a97b43cb8d39
| 5,364 |
ipynb
|
Jupyter Notebook
|
testing_tokenizers.ipynb
|
uunal/playground-ws
|
eadfd07092684069f9dcb3ff3b731d33ab76c529
|
[
"MIT"
] | null | null | null |
testing_tokenizers.ipynb
|
uunal/playground-ws
|
eadfd07092684069f9dcb3ff3b731d33ab76c529
|
[
"MIT"
] | null | null | null |
testing_tokenizers.ipynb
|
uunal/playground-ws
|
eadfd07092684069f9dcb3ff3b731d33ab76c529
|
[
"MIT"
] | null | null | null | 25.065421 | 137 | 0.563758 |
[
[
[
"import torch",
"_____no_output_____"
],
[
"tokenizer = torch.hub.load('huggingface/pytorch-transformers', 'tokenizer', 'bert-base-cased')\n\ntext_1 = \"session opened for user root by zero\"\ntext_2 = \"session closed for user root\"",
"Using cache found in C:\\Users\\uunal/.cache\\torch\\hub\\huggingface_pytorch-transformers_master\n"
],
[
"# Tokenized input with special tokens around it (for BERT: [CLS] at the beginning and [SEP] at the end)\nindexed_tokens = tokenizer.encode(text_1, text_2, add_special_tokens=True)",
"_____no_output_____"
],
[
"indexed_tokens",
"_____no_output_____"
],
[
"# Define sentence A and B indices associated to 1st and 2nd sentences (see paper)\nsegments_ids = [0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1]\n\n# Convert inputs to PyTorch tensors\nsegments_tensors = torch.tensor([segments_ids])\ntokens_tensor = torch.tensor([indexed_tokens])\n\nmodel = torch.hub.load('huggingface/pytorch-transformers', 'model', 'bert-base-cased')\n\nwith torch.no_grad():\n encoded_layers, _ = model(tokens_tensor, token_type_ids=segments_tensors)",
"Using cache found in C:\\Users\\uunal/.cache\\torch\\hub\\huggingface_pytorch-transformers_master\n"
],
[
"# Mask a token that we will try to predict back with `BertForMaskedLM`\nmasked_index = 7\nindexed_tokens[masked_index] = tokenizer.mask_token_id\ntokens_tensor = torch.tensor([indexed_tokens])\n\nmasked_lm__model = torch.hub.load('huggingface/pytorch-transformers', 'modelWithLMHead', 'bert-base-cased')\n\nwith torch.no_grad():\n predictions = masked_lm__model(tokens_tensor, token_type_ids=segments_tensors)\n\n# Get the predicted token\npredicted_index = torch.argmax(predictions[0][0], dim=1)[masked_index].item()\npredicted_token = tokenizer.convert_ids_to_tokens([predicted_index])[0]\n# assert predicted_token == 'Jim'",
"Using cache found in C:\\Users\\uunal/.cache\\torch\\hub\\huggingface_pytorch-transformers_master\n"
],
[
"tokenizer.convert_ids_to_tokens([torch.argmax(predictions[0][0], dim=1)[7].item()])[0]",
"_____no_output_____"
],
[
"torch.argmax(predictions[0][0], dim=1)[9].item()",
"_____no_output_____"
],
[
"torch.argmax(predictions[0][0], dim=1)",
"_____no_output_____"
],
[
"predicted_token",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb9373d267e237ccf6ad2836ba387857a49c028e
| 81,759 |
ipynb
|
Jupyter Notebook
|
K-Nearest Neighbors.ipynb
|
oyuka1112/Stocks
|
40e9b7c163ec512d5a4c03aca62b7554d1f2d4c3
|
[
"Apache-2.0"
] | null | null | null |
K-Nearest Neighbors.ipynb
|
oyuka1112/Stocks
|
40e9b7c163ec512d5a4c03aca62b7554d1f2d4c3
|
[
"Apache-2.0"
] | null | null | null |
K-Nearest Neighbors.ipynb
|
oyuka1112/Stocks
|
40e9b7c163ec512d5a4c03aca62b7554d1f2d4c3
|
[
"Apache-2.0"
] | null | null | null | 190.137209 | 70,488 | 0.897247 |
[
[
[
"import pandas as pd\nfrom diff_cap_packages import Xy\nfrom diff_cap_packages import ta\nfrom sklearn import neighbors\nfrom sklearn.metrics import mean_squared_error \nfrom math import sqrt\nimport matplotlib.pyplot as plt\n%matplotlib inline",
"_____no_output_____"
],
[
"X,Y = Xy.get_market_Xy(target_id=\"930060\")",
"_____no_output_____"
],
[
"Y = ta.add_ATR(Y, timeperiod=2)\nY = ta.add_RSI(Y, timeperiod=2)\nY = ta.add_SMA(Y, timeperiod=2)\nY = ta.add_EMA(Y, timeperiod=2)\n#Y = ta.add_OBV(Y)\n",
"_____no_output_____"
],
[
"Y = Y.dropna()",
"_____no_output_____"
],
[
"from sklearn.model_selection import train_test_split\ntrain , test = train_test_split(Y, test_size = 0.2)\n\nx_train = train.drop('930060 +0 day', axis=1)\ny_train = train['930060 +0 day']\n\nx_test = test.drop('930060 +0 day', axis = 1)\ny_test = test['930060 +0 day']",
"_____no_output_____"
],
[
"from sklearn.preprocessing import MinMaxScaler\nscaler = MinMaxScaler(feature_range=(0, 1))\n\nx_train_scaled = scaler.fit_transform(x_train)\nx_train = pd.DataFrame(x_train_scaled)\n\nx_test_scaled = scaler.fit_transform(x_test)\nx_test = pd.DataFrame(x_test_scaled)",
"_____no_output_____"
],
[
"rmse_val = [] #to store rmse values for different k\nfor K in range(20):\n K = K+1\n model = neighbors.KNeighborsRegressor(n_neighbors = K)\n\n model.fit(x_train, y_train) #fit the model\n pred=model.predict(x_test) #make prediction on test set\n error = sqrt(mean_squared_error(y_test,pred)) #calculate rmse\n rmse_val.append(error) #store rmse values\n print('RMSE value for k= ' , K , 'is:', error)",
"RMSE value for k= 1 is: 1.9223466236226452\nRMSE value for k= 2 is: 1.7038884574416227\nRMSE value for k= 3 is: 1.6524329331022074\nRMSE value for k= 4 is: 1.595332374565986\nRMSE value for k= 5 is: 1.6392686155676144\nRMSE value for k= 6 is: 1.610559140768324\nRMSE value for k= 7 is: 1.6019378335478591\nRMSE value for k= 8 is: 1.5941212789666424\nRMSE value for k= 9 is: 1.5720617307934204\nRMSE value for k= 10 is: 1.5419628026123628\nRMSE value for k= 11 is: 1.53033088069718\nRMSE value for k= 12 is: 1.536963104031556\nRMSE value for k= 13 is: 1.5397277217445093\nRMSE value for k= 14 is: 1.5395596017691389\nRMSE value for k= 15 is: 1.5456249167696865\nRMSE value for k= 16 is: 1.5514959652042217\nRMSE value for k= 17 is: 1.5684474475894885\nRMSE value for k= 18 is: 1.5456603511398561\nRMSE value for k= 19 is: 1.5359642681049774\nRMSE value for k= 20 is: 1.5508763506984513\n"
],
[
"from sklearn.model_selection import GridSearchCV\nparams = {'n_neighbors':[2,3,4,5,6,7,8,9]}\n\nknn = neighbors.KNeighborsRegressor()\n\nmodel = GridSearchCV(knn, params, cv=5)\nmodel.fit(x_train,y_train)\nmodel.best_params_",
"_____no_output_____"
],
[
"#predicting on the test set and creating submission file\nfrom sklearn.neighbors import KNeighborsRegressor\nknn_model = KNeighborsRegressor(n_neighbors=9)",
"_____no_output_____"
],
[
"knn_model.fit(x_train, y_train)",
"_____no_output_____"
],
[
"from sklearn.metrics import mean_squared_error\nfrom math import sqrt\ntrain_preds = knn_model.predict(x_train)\nmse = mean_squared_error(y_train, train_preds)\nrmse = sqrt(mse)\nrmse",
"_____no_output_____"
],
[
"test_preds = knn_model.predict(x_test)\nmse = mean_squared_error(y_test, test_preds)\nrmse = sqrt(mse)\nrmse",
"_____no_output_____"
],
[
"cmap = sns.cubehelix_palette(as_cmap=True)\nf, ax = plt.subplots()\npoints = ax.scatter(x_test[0], x_test[1], c=test_preds, s=50, cmap=cmap)\nf.colorbar(points)\nplt.show()",
"_____no_output_____"
],
[
"x_test",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb9376e9ab67962eeb4bedff527d2bd87f5d5859
| 16,526 |
ipynb
|
Jupyter Notebook
|
Python/Lesson_7/KC_lesson_7.ipynb
|
LubovFedoseeva/Learning
|
2d0e87c31298186b4e8f0b7301207aea574e0d03
|
[
"MIT"
] | 1 |
2021-12-08T19:41:57.000Z
|
2021-12-08T19:41:57.000Z
|
Python/Lesson_7/KC_lesson_7.ipynb
|
LubovFedoseeva/Learning
|
2d0e87c31298186b4e8f0b7301207aea574e0d03
|
[
"MIT"
] | null | null | null |
Python/Lesson_7/KC_lesson_7.ipynb
|
LubovFedoseeva/Learning
|
2d0e87c31298186b4e8f0b7301207aea574e0d03
|
[
"MIT"
] | null | null | null | 21.888742 | 200 | 0.474222 |
[
[
[
"import pandas as pd\nimport datetime\nimport vk_api\nimport os\nimport requests\nimport json\nimport random\n\n%matplotlib inline\nimport matplotlib.pyplot as plt\nimport seaborn as sns\nimport sys",
"_____no_output_____"
],
[
"token = '4e6e771d37dbcbcfcc3b53d291a274d3ae21560a2e81f058a7c177aff044b5141941e89aff1fead50be4f'",
"_____no_output_____"
],
[
"vk_session = vk_api.VkApi(token=token)\nvk = vk_session.get_api()",
"_____no_output_____"
],
[
"vk.messages.send(\n chat_id=1,\n random_id=2,\n message='Matrix has you ...')",
"_____no_output_____"
],
[
"df = pd.read_csv('/home/jupyter-an.karpov/shared/ads_data.csv.zip', compression='zip')",
"_____no_output_____"
],
[
"ad_data = df.groupby(['ad_id', 'event'], as_index=False) \\\n .agg({'user_id': 'count'})",
"_____no_output_____"
],
[
"ad_data = ad_data.pivot(index='ad_id', columns='event', values='user_id').reset_index()",
"_____no_output_____"
],
[
"ad_data = ad_data.fillna(0).assign(ctr=ad_data.click / ad_data.view)",
"_____no_output_____"
],
[
"top_ctr = ad_data.query('click > 20 & view > 100').sort_values('ctr').tail(10)",
"_____no_output_____"
],
[
"top_ctr.to_csv('top_ctr_data.csv', index=False)",
"_____no_output_____"
],
[
"path = '/home/jupyter-an.karpov/lesson_7/top_ctr_data.csv'\nfile_name = 'top_ctr_data.csv'\n\npath_to_file = path\nupload_url = vk.docs.getMessagesUploadServer(peer_id=2000000001)[\"upload_url\"]\nfile = {'file': (file_name, open(path_to_file, 'rb'))}\n\nresponse = requests.post(upload_url, files=file)\n\njson_data = json.loads(response.text)\njson_data",
"_____no_output_____"
],
[
"saved_file = vk.docs.save(file=json_data['file'], title=file_name)",
"_____no_output_____"
],
[
"saved_file",
"_____no_output_____"
],
[
"attachment = 'doc{}_{}'.format(saved_file['doc']['owner_id'], saved_file['doc']['id'])",
"_____no_output_____"
],
[
"attachment",
"_____no_output_____"
],
[
"vk.messages.send(\n chat_id=1,\n random_id=3,\n message='Топ объявлений по CTR', \n attachment = attachment\n)",
"_____no_output_____"
],
[
"import gspread\nfrom df2gspread import df2gspread as d2g\nfrom oauth2client.service_account import ServiceAccountCredentials",
"_____no_output_____"
],
[
"scope = ['https://spreadsheets.google.com/feeds',\n 'https://www.googleapis.com/auth/drive']\n\nmy_mail = '[email protected]'\n\n# Authorization\ncredentials = ServiceAccountCredentials.from_json_keyfile_name('heroic-venture-268009-1dfbcc34e5fa.json', scope)\ngs = gspread.authorize(credentials)",
"_____no_output_____"
],
[
"# Name of the table in google sheets, \n# can be url for open_by_url\n# or id (key) part for open_by_key\ntable_name = 'to_sequence' # Your table\n# Get this table\nwork_sheet = gs.open(table_name)",
"_____no_output_____"
],
[
"# Select 1st sheet\nsheet1 = work_sheet.sheet1\n\n# Get data in python lists format\ndata = sheet1.get_all_values()",
"_____no_output_____"
],
[
"data",
"_____no_output_____"
],
[
"headers = data.pop(0)",
"_____no_output_____"
],
[
"df = pd.DataFrame(data, columns=headers)",
"_____no_output_____"
],
[
"df",
"_____no_output_____"
],
[
"df.sort_values('price', ascending=False)",
"_____no_output_____"
],
[
"sheet1.append_row([500, 'group_4'])",
"_____no_output_____"
],
[
"# Looks like spreadsheet should be already present at the dist (so, run code in create table section)\nspreadsheet_name = 'A new spreadsheet'\nsheet = 'KarpovCorses2'\nd2g.upload(df, spreadsheet_name, sheet, credentials=credentials, row_names=True)",
"_____no_output_____"
],
[
"url = 'https://api-metrika.yandex.net/stat/v1/data?'\nvisits = 'metrics=ym:s:visits&dimensions=ym:s:date&id=44147844'",
"_____no_output_____"
],
[
"vistis_url = url + visits",
"_____no_output_____"
],
[
"vistis_request = requests.get(vistis_url)",
"_____no_output_____"
],
[
"vistis_request",
"_____no_output_____"
],
[
"json_data = json.loads(vistis_request.text)",
"_____no_output_____"
],
[
"json_data['data']",
"_____no_output_____"
],
[
"y_df = pd.DataFrame([(i['dimensions'][0]['name'], \n i['metrics'][0]) for i in json_data['data']], columns=['date', \n 'visits'])",
"_____no_output_____"
],
[
"spreadsheet_name = 'A new spreadsheet'\nsheet = 'Yandex_visits'\nd2g.upload(y_df, spreadsheet_name, sheet, credentials=credentials, row_names=True)",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb9381e83627ae8af52eb789ce9ec15a4da0fa44
| 204,202 |
ipynb
|
Jupyter Notebook
|
SlicerPlayground/02_Hello_Slicer.ipynb
|
piiq/SlicerNotebooks
|
c02ccab403723aa48fa205a171b54cbdf08ed9a6
|
[
"BSD-3-Clause"
] | null | null | null |
SlicerPlayground/02_Hello_Slicer.ipynb
|
piiq/SlicerNotebooks
|
c02ccab403723aa48fa205a171b54cbdf08ed9a6
|
[
"BSD-3-Clause"
] | null | null | null |
SlicerPlayground/02_Hello_Slicer.ipynb
|
piiq/SlicerNotebooks
|
c02ccab403723aa48fa205a171b54cbdf08ed9a6
|
[
"BSD-3-Clause"
] | null | null | null | 578.475921 | 144,878 | 0.943585 |
[
[
[
"# Hello, Slicer!\n\nLet's test that the setup was successful by checking that:\n\n1. We're running the python kernel bundled with Slicer\n2. The python kernel actually works\n3. We have access to Slicer functionality\n",
"_____no_output_____"
]
],
[
[
"import sys\nprint(f'We\\'re running {sys.executable}\\n')\nprint(\"Hello, Slicer!\")",
"We're running /Applications/MedicalImaging/Slicer.app/Contents/bin/PythonSlicer\n\nHello, Slicer!\n"
],
[
"try:\n import vtk\n import numpy as np\n from emoji import UNICODE_EMOJI\n import JupyterNotebooksLib as slicernb\n print('Imports succeeded. Slicer environment is setup correctly.')\nexcept ImportError as e:\n print(f'Imports failed with the following message: {e}.\\nSlicer environment was not setup correctly.')",
"Imports succeeded. Slicer environment is setup correctly.\n"
]
],
[
[
"---\n\n## Great!\n\nNow that we have everything setup and python is working let's try to create and view some objects in Slicer.\n\nLet's create a function that generates a cube:",
"_____no_output_____"
]
],
[
[
"def create_cube(size_xyz: list, center_xyz: list) -> vtk.vtkCubeSource:\n \"\"\"\n Creates a cube of a given size at a given point.\n\n :param size_xyz: List specifying size [x,y,z].\n :type size_xyz: list\n :param center_xyz: List specifying centroid coordinated [x,y,z].\n :type center_xyz: list\n\n :returns: A VKT box.\n :rtype: vtkCubeSource\n \"\"\"\n box = vtk.vtkCubeSource()\n box.SetXLength(size_xyz[0])\n box.SetYLength(size_xyz[1])\n box.SetZLength(size_xyz[2])\n box.SetCenter(center_xyz)\n return box",
"_____no_output_____"
]
],
[
[
"Now let's generate some cubes",
"_____no_output_____"
]
],
[
[
"some = 5 # why not?\n\nfor i in range(some):\n box = create_cube(np.random.randint(1, 42, 3).tolist(),\n np.random.randint(-50, 50, 3).tolist())\n\n # Create a model node that displays our box\n boxNode = slicer.modules.models.logic().AddModel(box.GetOutputPort())\n boxNode.SetName(f'Box{str(i+1)}')\n\n # Adjust display properties (color and opacity)\n boxNode.GetDisplayNode().SetColor(np.random.sample(3))\n boxNode.GetDisplayNode().SetOpacity(np.random.uniform(0, 1))",
"_____no_output_____"
]
],
[
[
"And let's adjust the 3D viewport camera position to render our cubes in the notebook output.",
"_____no_output_____"
]
],
[
[
"lm = slicer.app.layoutManager()\nview = lm.threeDWidget(0).threeDView()\nthreeDViewNode = view.mrmlViewNode()\ncameraNode = slicer.modules.cameras.logic().GetViewActiveCameraNode(threeDViewNode)\ncameraNode.SetPosition([342,242,242])\nslicernb.View3DDisplay()",
"_____no_output_____"
]
],
[
[
"## Nice!\n\nSo we've tested that VTK is in place and that the 3D view works. What about the slice views?\n\n---\n\n## Let's create some images to display them in slice views.\n\nWe'll create a \"💀 Hello Slicer\" sign spanned across the Red, Green and Yellow slice views.",
"_____no_output_____"
],
[
"Let's start with creating images represented as numpy arrays",
"_____no_output_____"
]
],
[
[
"def create_np_text_img(text: str, size: tuple = (128, 128)) -> np.ndarray:\n \"\"\"\n Creates a numpy text image.\n\n Creates a text-on-background image and returns it as a flat 3D numpy array.\n \n Check font paths when copying this function.\n The font paths should point to actual true-type font files on the disk.\n\n :param text: Input unicode text.\n :type text: str\n :param size: Target image size (optional).\n :type size: tuple\n\n :returns: Flat 3D numpy array containing pixel values.\n :rtype: np.ndarray\n \"\"\"\n from PIL import Image, ImageDraw, ImageFont\n if bool(set(text).intersection(UNICODE_EMOJI)):\n font_path = \"/System/Library/Fonts/Apple Color Emoji.ttc\"\n font = ImageFont.truetype(font_path, 64)\n else:\n font_path = \"/System/Library/Fonts/Microsoft/Arial Black.ttf\"\n font = ImageFont.truetype(font_path, 24)\n\n text_width, text_height = font.getsize(text)\n\n text_image = Image.new('I', size, \"black\")\n draw = ImageDraw.Draw(text_image)\n draw.text((text_width/2, text_height/2), text, 'white', font)\n return np.asarray(text_image).reshape(*size, 1)",
"_____no_output_____"
],
[
"skull_image_array = create_np_text_img('💀')\nhello_image_array = create_np_text_img('Hello')\nslicer_image_array = create_np_text_img('Slicer')",
"_____no_output_____"
]
],
[
[
"The coordinate space of the images, the arrays and slice views is usually not the same. \nThere is a good wiki article about it here:\n\n[https://www.slicer.org/wiki/Coordinate_systems](https://www.slicer.org/wiki/Coordinate_systems)\n\nThat means that displaying our sign in the slice views might require some rotations and flipping.\n\nLet's create an empty (128, 128, 128) array and paste our images onto it's sides.",
"_____no_output_____"
]
],
[
[
"volume = np.zeros((128, 128, 128))\n\nvolume[:1, :, :] = np.rot90(skull_image_array, 2, (0, 1)).transpose(2, 0, 1)\nvolume[:, :1, :] = np.rot90(hello_image_array[::-1], 1, (0, 1)).transpose(1, 2, 0)\nvolume[:, :, :1] = np.rot90(slicer_image_array, 2, (0, 1))",
"_____no_output_____"
]
],
[
[
"Finally we can create a volume node and render it's slices in the slice views",
"_____no_output_____"
]
],
[
[
"volumeNode = slicer.util.addVolumeFromArray(volume)\nvolumeNode.SetName('Hello Slicer!')",
"_____no_output_____"
],
[
"def show_slice_in_slice_view(volumeNode: slicer.vtkMRMLScalarVolumeNode,\n sliceNum: int = 0,\n sliceView: str = 'Red'):\n \"\"\"\n Render a numpy image on slice view.\n\n :param volumeNode: The volume node\n :type volumeNode: vtkMRMLScalarVolumeNode\n :param sliceNum: The number of the slice that we want to show. Optional. Defaults to 0.\n :type sliceNum: int\n :param sliceView: One of default slice views ('Red', 'Green', 'Yellow')\n :type sliceView: str\n \"\"\"\n sliceViewWidget = slicer.app.layoutManager().sliceWidget(sliceView)\n sliceWidgetLogic = sliceViewWidget.sliceLogic()\n sliceWidgetLogic.GetSliceCompositeNode().SetForegroundVolumeID(volumeNode.GetID())\n sliceWidgetLogic.FitSliceToAll()\n sliceWidgetLogic.SetSliceOffset(sliceNum)\n pass",
"_____no_output_____"
],
[
"for sliceView in ['Red', 'Green', 'Yellow']:\n show_slice_in_slice_view(volumeNode, 0, sliceView)\nslicernb.ViewDisplay('FourUp', False)",
"_____no_output_____"
]
],
[
[
"# All right! This was fun!\n\nLet's play with volumes a bit more in the next notebook.",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
]
] |
cb939caf97b5fa25f80ac262c4ddeb7282fe826d
| 263,561 |
ipynb
|
Jupyter Notebook
|
ContinuousDataAcquisitionPython/Old/data/.ipynb_checkpoints/visualization-checkpoint.ipynb
|
atick-faisal/Hand-Gesture-Recognition-v2
|
f26e53b68e3cddd26c40edc9b056d82b5e887ff3
|
[
"MIT"
] | 2 |
2021-02-22T19:09:30.000Z
|
2021-07-01T04:11:00.000Z
|
ContinuousDataAcquisitionPython/Old/data/.ipynb_checkpoints/visualization-checkpoint.ipynb
|
atick-faisal/Hand-Gesture-Recognition-v2
|
f26e53b68e3cddd26c40edc9b056d82b5e887ff3
|
[
"MIT"
] | null | null | null |
ContinuousDataAcquisitionPython/Old/data/.ipynb_checkpoints/visualization-checkpoint.ipynb
|
atick-faisal/Hand-Gesture-Recognition-v2
|
f26e53b68e3cddd26c40edc9b056d82b5e887ff3
|
[
"MIT"
] | 2 |
2020-11-20T14:29:36.000Z
|
2021-04-07T08:00:45.000Z
| 1,166.199115 | 62,260 | 0.955866 |
[
[
[
"import numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nfrom scipy.signal import medfilt",
"_____no_output_____"
],
[
"dataset = pd.read_csv('train_data.csv')\ndataset.columns = ['accX', 'accY', 'accZ']",
"_____no_output_____"
],
[
"accX = np.array(dataset['accX'])\naccY = np.array(dataset['accY'])\naccZ = np.array(dataset['accZ'])",
"_____no_output_____"
],
[
"plt.plot(accX + 150)\nplt.plot(accY)\nplt.plot(accZ - 150)\nplt.title('Raw Data')\nplt.legend(['AccX', 'AccY', 'AccZ'], loc = 1)\nplt.show()",
"_____no_output_____"
],
[
"plt.plot(medfilt(accX) + 150)\nplt.plot(medfilt(accY))\nplt.plot(medfilt(accZ) - 150)\nplt.title('Filtered Data')\nplt.legend(['AccX', 'AccY', 'AccZ'], loc = 1)\nplt.show()",
"_____no_output_____"
],
[
"plt.plot(accX[100:300] + 200)\nplt.plot(accY[100:300])\nplt.plot(accZ[100:300] - 200)\nplt.title('Acceleration for \\'Painting\\' Gesture')\nplt.legend(['Acc-X', 'Acc-Y', 'Acc-Z'], loc = 1)\nplt.show()",
"_____no_output_____"
],
[
"plt.plot(accX[2100:2300] + 200)\nplt.plot(accY[2100:2300])\nplt.plot(accZ[2100:2300] - 200)\nplt.title('Acceleration for \\'Thank You\\' Gesture')\nplt.legend(['Acc-X', 'Acc-Y', 'Acc-Z'], loc = 1)\nplt.show()",
"_____no_output_____"
],
[
"plt.plot(accX[3100:3300] + 200)\nplt.plot(accY[3100:3300])\nplt.plot(accZ[3100:3300] - 200)\nplt.title('Acceleration for \\'Sorry\\' Gesture')\nplt.legend(['Acc-X', 'Acc-Y', 'Acc-Z'], loc = 1)\nplt.show()",
"_____no_output_____"
],
[
"features = pd.read_csv('feature_vector.csv')\nplt.plot(features)\nplt.title('Feature')\nplt.show()",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb939d4fa9d516cf6bacac72ad02c3d760633452
| 2,887 |
ipynb
|
Jupyter Notebook
|
docs/machine_learning/basics/loading_features_from_dictionaries.ipynb
|
TheJKFever/notes
|
5afa769301f6f3efbe45332e9457df130a1ec8a8
|
[
"CC0-1.0"
] | 1 |
2018-01-09T19:06:03.000Z
|
2018-01-09T19:06:03.000Z
|
docs/machine_learning/basics/loading_features_from_dictionaries.ipynb
|
vigneshprajapati/notes
|
69b52689bec2185eb8ff532932617b665c3ecb95
|
[
"CC0-1.0"
] | null | null | null |
docs/machine_learning/basics/loading_features_from_dictionaries.ipynb
|
vigneshprajapati/notes
|
69b52689bec2185eb8ff532932617b665c3ecb95
|
[
"CC0-1.0"
] | 1 |
2020-10-17T22:00:42.000Z
|
2020-10-17T22:00:42.000Z
| 19.639456 | 81 | 0.488396 |
[
[
[
"---\ntitle: \"Loading Features From Dictionaries\"\nauthor: \"Chris Albon\"\ndate: 2017-12-20T11:53:49-07:00\ndescription: \"Loading Features From Dictionaries Using Python.\"\ntype: technical_note\ndraft: false\n---",
"_____no_output_____"
]
],
[
[
"## Preliminaries",
"_____no_output_____"
]
],
[
[
"from sklearn.feature_extraction import DictVectorizer",
"_____no_output_____"
]
],
[
[
"## Create A Dictionary",
"_____no_output_____"
]
],
[
[
"staff = [{'name': 'Steve Miller', 'age': 33.},\n {'name': 'Lyndon Jones', 'age': 12.},\n {'name': 'Baxter Morth', 'age': 18.}]",
"_____no_output_____"
]
],
[
[
"## Convert Dictionary To Feature Matrix",
"_____no_output_____"
]
],
[
[
"# Create an object for our dictionary vectorizer\nvec = DictVectorizer()",
"_____no_output_____"
],
[
"# Fit then transform the staff dictionary with vec, then output an array\nvec.fit_transform(staff).toarray()",
"_____no_output_____"
]
],
[
[
"## View Feature Names",
"_____no_output_____"
]
],
[
[
"# Get Feature Names\nvec.get_feature_names()",
"_____no_output_____"
]
]
] |
[
"raw",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"raw"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb939dabd741d60374e87bc1903c8f7be280e2e9
| 1,274 |
ipynb
|
Jupyter Notebook
|
search for values in a list.ipynb
|
AntonioRobG/programming-fundamentals-for-data-science
|
ae943238dd064d884f808ba182be4f6845f7ad20
|
[
"MIT"
] | null | null | null |
search for values in a list.ipynb
|
AntonioRobG/programming-fundamentals-for-data-science
|
ae943238dd064d884f808ba182be4f6845f7ad20
|
[
"MIT"
] | null | null | null |
search for values in a list.ipynb
|
AntonioRobG/programming-fundamentals-for-data-science
|
ae943238dd064d884f808ba182be4f6845f7ad20
|
[
"MIT"
] | null | null | null | 21.965517 | 114 | 0.550235 |
[
[
[
"\"\"\" \nCreate a list containing 67, 12, 46, 43 and 13, then use list method .index() to search for a 43 and 44. \nEnsure that no ValueError occurs when searching for 44.\n\"\"\" ",
"_____no_output_____"
],
[
"mylist = [67, 12, 46, 43, 13]\nif 43 in mylist: print(mylist.index(43))\nif 44 in mylist: print(mylist.index(44))",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code"
]
] |
cb93c8e4cd2ce9b57c8ccdaa54878a96f5be875d
| 18,652 |
ipynb
|
Jupyter Notebook
|
ml-zb/lec08.ipynb
|
JasonWayne/course-notes
|
feff7a0636e7f8f2353c1dea24fe25296a8b33c3
|
[
"MIT"
] | null | null | null |
ml-zb/lec08.ipynb
|
JasonWayne/course-notes
|
feff7a0636e7f8f2353c1dea24fe25296a8b33c3
|
[
"MIT"
] | null | null | null |
ml-zb/lec08.ipynb
|
JasonWayne/course-notes
|
feff7a0636e7f8f2353c1dea24fe25296a8b33c3
|
[
"MIT"
] | null | null | null | 16.971793 | 70 | 0.467725 |
[
[
[
"##### 1",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 2",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 3",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 4",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 5",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 6",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 7",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 8",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 9",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 10",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 11",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 12",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 13",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 14",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 15",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 16",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 17",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 18",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 19",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 20",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 21",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 22",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 23",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 24",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 25",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 26",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 27",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 28",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 29",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 30",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 31",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 32",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 33",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 34",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 35",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 36",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 37",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 38",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 39",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 40",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 41",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 42",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 43",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 44",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 45",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 46",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 47",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 48",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 49",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 50",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 51",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 52",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 53",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 54",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 55",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 56",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 57",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 58",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 59",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 60",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 61",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 62",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 63",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 64",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 65",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 66",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 67",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 68",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 69",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 70",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 71",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 72",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 73",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 74",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 75",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"##### 76",
"_____no_output_____"
],
[
"",
"_____no_output_____"
]
]
] |
[
"markdown"
] |
[
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
]
] |
cb93d3921bc6fc606d51df2ad1f34dba0b01be65
| 21,567 |
ipynb
|
Jupyter Notebook
|
TransferLearn/facialRecogn/Preprocessing4_Final.ipynb
|
zcemycl/DNNexamples-KerasTFPyTorch
|
ca5b7abc385c975b05940701ecc06a73755e1d10
|
[
"MIT"
] | null | null | null |
TransferLearn/facialRecogn/Preprocessing4_Final.ipynb
|
zcemycl/DNNexamples-KerasTFPyTorch
|
ca5b7abc385c975b05940701ecc06a73755e1d10
|
[
"MIT"
] | null | null | null |
TransferLearn/facialRecogn/Preprocessing4_Final.ipynb
|
zcemycl/DNNexamples-KerasTFPyTorch
|
ca5b7abc385c975b05940701ecc06a73755e1d10
|
[
"MIT"
] | null | null | null | 26.593095 | 140 | 0.375852 |
[
[
[
"import numpy as np\nimport matplotlib.pyplot as plt\nimport math\nimport cv2\nimport os\nimport tqdm\nimport glob\nfrom statistics import mode\nfrom sklearn.cluster import KMeans\nfrom sklearn.cluster import MeanShift",
"_____no_output_____"
],
[
"def loadFileName(directory):\n dealerIDir = glob.glob(directory+'/*')\n allVidir = []\n for i in range(len(dealerIDir)):\n allVidir.append(glob.glob(dealerIDir[i]+'/*mp4'))\n return np.array(allVidir)\ndef eulidist(rectlist,center):\n dx1_2 = (rectlist[0]-center[0])**2\n dx2_2 = (rectlist[2]-center[0])**2\n dy1_2 = (rectlist[1]-center[1])**2\n dy2_2 = (rectlist[3]-center[1])**2\n return dx1_2+dx2_2+dy1_2+dy2_2\ndef selectFaces(faces,center):\n disList = []\n for i in range(len(faces)):\n (x,y,w,h) = faces[i]\n loc = np.array((x,y,x+w,y+h))\n disList.append(eulidist(loc,center))\n disList = np.array(disList)\n# print(faces)\n return faces[np.argmin(disList)]",
"_____no_output_____"
],
[
"def extractFrame(mp4Dir):\n vid = cv2.VideoCapture(mp4Dir)\n allFrame = []\n while vid.isOpened():\n ret,frame = vid.read()\n if ret:\n allFrame.append(frame)\n else:\n break\n return np.array(allFrame)\ndef extractXYWH(allFrame):\n face_cascade = cv2.CascadeClassifier('C:/Users/44754/Anaconda3/Lib/site-packages/cv2/data/haarcascade_frontalface_default.xml')\n imgList = allFrame.copy()\n for i in tqdm.tqdm(range(len(allFrame))):\n img = imgList[i].copy()\n gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)\n faces = face_cascade.detectMultiScale(gray)\n if not len(faces):\n face = np.array((0,0,0,0))\n else:\n center = np.array(imgList[i][:,:,::-1].shape)[:2][::-1]/2\n (x,y,w,h) = selectFaces(faces,center)\n face = np.array((x,y,w,h))\n if i == 0:\n coorList = face\n else:\n coorList = np.vstack((coorList,face))\n return coorList\ndef meanShiftCluster(coorList):\n ms = MeanShift(bandwidth=20)\n ms.fit(coorList[:,:2])\n labels = ms.labels_\n remainIdx = mode(labels)\n w = int(np.mean(coorList[labels==remainIdx,2]))\n h = int(np.mean(coorList[labels==remainIdx,3]))\n return labels,remainIdx,w,h\ndef plotCluster(coorList,labels):\n xmin,xmax = np.min(coorList[:,0]),np.max(coorList[:,0])\n ymin,ymax = np.min(coorList[:,1]),np.max(coorList[:,1])\n numCol = int(np.ceil((len(np.unique(labels))+1)/2))\n plt.figure(figsize=(20,10))\n plt.subplot(2,numCol,1)\n plt.scatter(coorList[:,0],coorList[:,1],c=labels)\n plt.xlim([xmin,xmax])\n plt.ylim([ymin,ymax])\n for i in range(len(np.unique(labels))):\n plt.subplot(2,numCol,i+2)\n plt.scatter(coorList[labels==i,0],coorList[labels==i,1])\n plt.xlim([xmin,xmax])\n plt.ylim([ymin,ymax])\n plt.title(np.sum(labels==i))\ndef modifyFrames(allFrame,labels,remainIdx,w,h):\n imgList = allFrame.copy()\n modList = []\n for i in tqdm.tqdm(range(len(allFrame))):\n img = imgList[i].copy()\n if labels[i] == remainIdx:\n x,y = coorList[i,0],coorList[i,1]\n img = cv2.rectangle(img,(x,y),(x+w,y+h),(255,0,0),2)\n modList.append(img)\n modList = np.array(modList)\n return modList",
"_____no_output_____"
],
[
"def extractFaces(allFrame,coorList,labels,remainIdx,w,h):\n imgList = allFrame.copy()\n facesList = []\n for i in range(len(allFrame)):\n if labels[i] == remainIdx:\n img = imgList[i].copy()\n x,y = coorList[i,0],coorList[i,1]\n tmp = img[y:y+h,x:x+w]\n facesList.append(tmp)\n facesList = np.array(facesList)\n return facesList\ndef dir2processedarr(mp4Dir):\n allFrame = extractFrame(mp4Dir)\n print(allFrame.shape)\n coorArr = extractXYWH(allFrame)\n labels,remainIdx,w,h = meanShiftCluster(coorArr)\n facesArr = extractFaces(allFrame,coorArr,labels,remainIdx,w,h)\n return facesArr\ndef storeProcessedImgs(directory):\n currentdir = os.getcwd()\n # create train dir\n path = os.path.join(currentdir,'train')\n if not os.path.exists(path):\n os.makedirs(path)\n \n # list all IDs\n dealersID = os.listdir(directory)\n for i in range(len(dealersID)):\n print('{} dealers'.format(i))\n tmppath = os.path.join(path,dealersID[i])\n if not os.path.exists(tmppath):\n os.makedirs(tmppath)\n dealDir = os.path.join(directory,dealersID[i])\n vidsID = os.listdir(dealDir)\n j = 0\n # Each video for a specific dealer\n for vid in vidsID:\n mp4Dir = os.path.join(dealDir,vid)\n tmparr = dir2processedarr(mp4Dir)\n # saving\n for k in range(tmparr.shape[0]):\n savepath = os.path.join(tmppath,'{}.jpg'.format(j))\n cv2.imwrite(savepath,tmparr[k])\n j+=1",
"_____no_output_____"
],
[
"vidir = 'D:/DreamAI/videosubset'\nstoreProcessedImgs(vidir)",
"0 dealers\n(2085, 540, 1024, 3)\n"
],
[
"currentdir = os.getcwd()\n# create train dir\npath = os.path.join(currentdir,'train')\n# list all IDs\ndealersID = os.listdir(path)\nfor ids in dealersID:\n tmp = os.path.join(path,ids)\n print('Dealer {}: '.format(ids)+str(len(glob.glob(tmp+'/*.jpg'))))",
"Dealer 231: 15876\nDealer 634: 8803\nDealer 717: 16645\nDealer 818: 5979\n"
]
],
[
[
"### There are differences among the numbers of images for different dealers.",
"_____no_output_____"
]
]
] |
[
"code",
"markdown"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
]
] |
cb93e1c5a2c4515dfa1bd095d343c04784c5d963
| 11,753 |
ipynb
|
Jupyter Notebook
|
tutorial/00 Jupyter Notebook and Python Basics.ipynb
|
feststelltaste/software-analytics-workshop
|
2ea5622a544c5238ca0d4044a66f13d2fb549a18
|
[
"MIT"
] | 16 |
2019-09-18T15:05:01.000Z
|
2021-04-18T14:50:56.000Z
|
tutorial/00 Jupyter Notebook and Python Basics.ipynb
|
feststelltaste/software-analytics-workshop
|
2ea5622a544c5238ca0d4044a66f13d2fb549a18
|
[
"MIT"
] | null | null | null |
tutorial/00 Jupyter Notebook and Python Basics.ipynb
|
feststelltaste/software-analytics-workshop
|
2ea5622a544c5238ca0d4044a66f13d2fb549a18
|
[
"MIT"
] | 2 |
2020-03-13T00:26:26.000Z
|
2021-05-02T02:37:15.000Z
| 37.912903 | 425 | 0.434187 |
[
[
[
"# Software Analytics Mini Tutorial Part I: Jupyter Notebook and Python basics",
"_____no_output_____"
],
[
"## Introduction\nThis series of notebooks are a simple mini tutorial to introduce you to the basic functionality of Jupyter, Python, pandas and matplotlib. The comprehensive explanations should guide you to be able to analyze software data on your own. Therefore, the examples is chosen in such a way that we come across the typical methods in a data analysis. Have fun!\n\n*This is part I: The basics of Jupyter Notebook and Python. For the data analysis framework pandas and the visualization library matplotlib, go to [the other tutorial](10%20pandas%20and%20matplotlib%20basics.ipynb).*",
"_____no_output_____"
],
[
"## The Jupyter Notebook System\nFirst, we'll take a closer look at Jupyter Notebook. What you see here is Jupyter, the interactive notebook environment for programming. Jupyter Notebook allows us us to write code and documentation in executable **cells**.\n\nWe see below a cell in which we can enter Python code.\n\n#### Execute a cell\n\n1. select the next cell (mouse click or arrow keys).\n1. type in for example `\"Hello World!`.\n1. execute the cell with a `Ctrl`+`Enter`. \n1. click on the cell again\n1. execute the cell with `Shift`+`Enter`.",
"_____no_output_____"
],
[
"##### Discussion\n\n* What's the difference between the two different ways of executing cells?",
"_____no_output_____"
],
[
"#### Create a new cell\n\nWe will use the built-in keyboard shortcuts for this:\n\n1. if not happened yet, click on this cell.\n1. enter **command mode**, selectable with `Esc` key.\n1. create a **new cell** after this text by pressing the key `b`.\n1. change the **cell type** to **Markdown** with key `m`.\n1. switch to **edit mode** with `Enter`\n1. write a text\n1. execute cell with `Ctrl` + `Enter`.",
"_____no_output_____"
],
[
"*Additional information:*\n\n* We've seen an important feature of Jupyter Notebook: The distinction between **command mode** (accessible via the `Esc` key) and **edit mode** (accessible via the `Enter` key). Note the differences:\n * In command mode, the border of the current cell is blue. This mode allows you to manipulate the **notebook**'s content.\n * In edit mode, the border turns green. This allows you to manipulate the **cell**'s content.\n\n* **Markdown** is a simple markup language that can be used to write and format texts. This allows us to directly document the steps we have taken.",
"_____no_output_____"
],
[
"## Python Basics\nLet's take a look at some basic Python programming constructs that we will need later when working with the pandas data analysis framework.\n\nWe look at very basic functions:\n\n* variable assignments\n* value range accesses\n* method calls",
"_____no_output_____"
],
[
"#### Assign text to a variable\n1. **assign** the text **value** \"Hello World\" to the **variable** `text` by using the syntax `<variable> = <value>`. \n1. type the variable `text` in the next line and execute the cell.\n1. execute the cell (this will be necessary for each upcoming cell, so we won't mention it from now on)",
"_____no_output_____"
],
[
"#### Accessing slices of information\n\nBy using the array notation with the square brackets `[` and `]` (the slice operators), we can access the first letter of our text with a 0-based index (this also works for other types like lists).\n\n1. access the first letter in `text` with `[0]`.",
"_____no_output_____"
],
[
"#### Select last character\n1. access the last letter in `text` with `[-1]`.",
"_____no_output_____"
],
[
"#### Select ranges\n1. access a range of `text` with the **slice** `[2:5]`.",
"_____no_output_____"
],
[
"#### Select open ranges\n1. access an open range with the slice `[:5]` (which is an abbreviation for a 0-based slice `[0:5]`)",
"_____no_output_____"
],
[
"#### Reverse a list of values\n1. reverse the text (or a list) by using the `::` notation with an following `-1`.",
"_____no_output_____"
],
[
"#### Use auto completion and execute a method\n1. append a `.` to `text` and look at the functions with the `Tab` key.\n1. find and execute the **method** `upper()` of the `text` object (Tip: Type a `u` when using auto completion).",
"_____no_output_____"
],
[
"#### Execute a method with parameters\n\n...and find out how this works by using the integrated, interactive documentation:\n1. select the `split` method of `text`.\n1. press `Shift`+`Tab`.\n1. press `Shift`+`Tab` twice in quick succession.\n1. press `Shift`+`Tab` three times in quick succession (and then `ESC` to hide).\n1. read the documentation of `split`\n1. split the text in `text` with `split` exactly once (**parameter** `maxsplit`) apart by using an `l` (a lower case \"L\") as separator (parameter `sep`).\n\n*Note: What we are using here additionally is Python's feature to swap the methods parameter. This works, if you assign the inputs directly to the method argument's name (e.g. `maxsplit=1`).",
"_____no_output_____"
],
[
"## Summary\nOK, we were just warming up! Proceed to [the next section](10%20pandas%20and%20matplotlib%20basics.ipynb)\n\nIf you want to dive deeper into this topic, take a look at my [blog posts on that topic](http://www.feststelltaste.de/category/software-analytics/) or my microsite [softwareanalytics.de](https://softwareanalytics.de/). I'm looking forward to your comments and feedback on [GitHub](https://github.com/feststelltaste/software-analytics-workshop/issues) or on [Twitter](https://www.twitter.com/feststelltaste)!",
"_____no_output_____"
]
]
] |
[
"markdown"
] |
[
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
]
] |
cb93e7b304d4d6160c856386d644c38eab24eb80
| 637,144 |
ipynb
|
Jupyter Notebook
|
image_cluster/code_cafe_color_map/code_folium_map/code_folium_soryeong_ver05.ipynb
|
dataitgirls2/m3-B
|
75430132e98a14b41feed0da958c4147e22e4f1b
|
[
"MIT"
] | 2 |
2018-08-30T13:04:19.000Z
|
2018-09-04T14:07:35.000Z
|
image_cluster/code_cafe_color_map/code_folium_map/code_folium_soryeong_ver05.ipynb
|
dataitgirls2/m3-B
|
75430132e98a14b41feed0da958c4147e22e4f1b
|
[
"MIT"
] | null | null | null |
image_cluster/code_cafe_color_map/code_folium_map/code_folium_soryeong_ver05.ipynb
|
dataitgirls2/m3-B
|
75430132e98a14b41feed0da958c4147e22e4f1b
|
[
"MIT"
] | 5 |
2018-08-11T12:25:25.000Z
|
2018-08-16T09:55:37.000Z
| 2,806.801762 | 630,792 | 0.993647 |
[
[
[
"import cv2\nimport matplotlib.pyplot as plt\nimport matplotlib.image as mpimg\nimport pandas as pd",
"_____no_output_____"
],
[
"df_cafe = pd.read_csv('final_cafe_info_with_path.csv')\ndf_cafe.head()",
"_____no_output_____"
],
[
"# df_shop = pd.read_csv('open_data_shop.csv', encoding='cp949')\n# df_seoul = df_shop.loc[df_shop['도로명주소'].str.contains('서울')]\n# df_seoul['상권업종중분류명'].value_counts()",
"_____no_output_____"
],
[
"import googlemaps\ngmaps_key = 'AIzaSyAejv70nUFd8T1Gli59yiKEE3-ECkpDpU8' # 자신의 key를 사용합니다.\ngmaps = googlemaps.Client(key=gmaps_key)",
"_____no_output_____"
],
[
"import base64\nimport folium\n\n\nmap = folium.Map(location=[df_cafe['위도'].mean(), df_cafe['경도'].mean()], zoom_start=13)\n\nfor n in df_cafe.index:\n png = './cafe_color_result/' + df_cafe['파일명'][n]\n encoded = base64.b64encode(open(png, 'rb').read()).decode('utf-8')\n cafe_name = df_cafe['카페명'][n] + ' - ' + df_cafe['주소'][n]\n html = f'<p>{cafe_name}</p> <img src=\"data:image/png;base64,{encoded}\">'\n iframe = folium.IFrame(html, width=700, height=130)\n popup = folium.Popup(iframe, max_width=300)\n folium.Marker([df_cafe['위도'][n], df_cafe['경도'][n]], popup=popup).add_to(map)\n \nmap",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code"
]
] |
cb93eaf86bbec852c3512c562d93a695dc0cf602
| 771,980 |
ipynb
|
Jupyter Notebook
|
scripts/histogram.ipynb
|
geohackweek/ghw2019_wiggles
|
9b636db8d97986e038a301e36b808e820ccc525f
|
[
"BSD-3-Clause"
] | 3 |
2019-10-09T19:42:12.000Z
|
2021-05-28T00:10:54.000Z
|
scripts/histogram.ipynb
|
geohackweek/ghw2019_wiggles
|
9b636db8d97986e038a301e36b808e820ccc525f
|
[
"BSD-3-Clause"
] | 1 |
2019-09-11T16:37:59.000Z
|
2019-09-11T16:37:59.000Z
|
scripts/histogram.ipynb
|
geohackweek/ghw2019_wiggles
|
9b636db8d97986e038a301e36b808e820ccc525f
|
[
"BSD-3-Clause"
] | 3 |
2019-09-10T20:41:59.000Z
|
2019-09-10T20:42:57.000Z
| 787.734694 | 21,436 | 0.949269 |
[
[
[
"## plot the histogram showing the modeled and labeled result \nimport numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\n%matplotlib inline\n",
"_____no_output_____"
],
[
"# for loop version\ndef read_comp(file):\n Pwave = {}\n Pwave['correct'] = []\n Pwave['wrongphase'] = []\n Pwave['miss'] = 0\n Pwave['multiphase'] = [] \n \n with open(file, 'r') as comp:\n for line in comp:\n pickline = line.split()\n if pickline[0].strip()[-1] == pickline[1].strip() and len(pickline)==3:\n Pwave['correct'].append(float(pickline[2][:-2]))\n if pickline[0].strip()[-1] != pickline[1].strip() and pickline[1].strip() != 'N':\n Pwave['wrongphase'].append (float(pickline[2][:-2]))\n if pickline[0].strip()[-1] == pickline[1].strip() and len(pickline)>3: \n Pwave['multiphase'].append(len(pickline)-2)\n if pickline[1].strip() == 'N':\n Pwave['miss'] +=1\n \n return Pwave",
"_____no_output_____"
],
[
"# run all the output file \nty = ['EQS','EQP','SUS','SUP','THS','THP','SNS','SNP','PXS','PXP']\nwave = {}\nfor name in ty:\n wave[name] = read_comp('../comparison_out/comp.'+name+'.out')",
"_____no_output_____"
],
[
"# plot histogram of the correct plot\ndef plotfig(name):\n fig, ax = plt.subplots(figsize = (10,6))\n# filename = wave['EQP']['correct']\n fig = plt.hist(name,bins= 10)\n ax.set_ylabel('number', fontsize=15)\n ax.set_xlabel('time difference (s)', fontsize=15)\n ax.set_title('Phase with time difference')\n plt.xticks(fontsize=15)\n #plt.xticks( rotation='vertical')\n plt.xticks(np.arange(-10, 10, step=1))\n plt.yticks(fontsize=15)\n plt.legend(fontsize=15)\n #plt.savefig('test.jpg')",
"_____no_output_____"
],
[
"for k in wave.keys():\n for t in wave[k].keys():\n# print (k, t)\n fig, ax = plt.subplots(figsize = (10,6))\n# filename = wave['EQP']['correct']\n fig = plt.hist(wave[k][t],bins= 10)\n ax.set_ylabel('number', fontsize=15)\n ax.set_xlabel('time difference (s)', fontsize=15)\n ax.set_title('{} Phase with time difference for {}'.format(t,k))\n plt.xticks(fontsize=15)\n #plt.xticks( rotation='vertical')\n plt.xticks(np.arange(-10, 10, step=1))\n plt.yticks(fontsize=15)\n #plt.legend(fontsize=15)\n #plt.savefig('{}_{}.jpg'.format(k,t))",
"/srv/conda/envs/notebook/lib/python3.7/site-packages/ipykernel_launcher.py:4: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\n after removing the cwd from sys.path.\n"
],
[
"plt.hist(wave['EQS']['correct'])\n#plt.xlim(0, 50)\nplt.savefig('time_diff.jpg')",
"_____no_output_____"
],
[
"plotfig(wave['EQS']['wrongphase'])",
"_____no_output_____"
],
[
"plotfig(wave['EQS']['multiphase'])",
"_____no_output_____"
],
[
"plotfig(wave['EQS']['correct'])",
"_____no_output_____"
],
[
"# plot histogram of the wrongphase plot \nfig, ax = plt.subplots(figsize = (10,6))\n\n#k = np.random.normal(float(test['EQS'][ID]['time']), 3, 1000)\nfig = plt.hist(wave['EQP']['wrongphase'])\nax.set_ylabel('number', fontsize=15)\nax.set_xlabel('time difference (S)', fontsize=15)\nax.set_title('Correct Phase with time difference')\nplt.xticks(fontsize=15)\nplt.yticks(fontsize=15)\nplt.xticks(np.arange(-10, 10, step=1))\nplt.legend(fontsize=15)",
"_____no_output_____"
],
[
"# plot histogram of the wrongphase plot \nfig, ax = plt.subplots(figsize = (10,6))\n#k = np.random.normal(float(test['EQS'][ID]['time']), 3, 1000)\nfig = plt.hist(test['multiphase'])\nax.set_ylabel('number', fontsize=15)\nax.set_xlabel('time difference (S)', fontsize=15)\nax.set_title('Correct Phase with time difference')\nplt.xticks(fontsize=15)\nplt.yticks(fontsize=15)\n#plt.xticks(np.arange(-10, 10, step=1))\nplt.legend(fontsize=15)",
"_____no_output_____"
],
[
"comp = pd. read_csv('../comparison.out', names=['type','mod_type','time'],sep = ' ')\ncomp.head() ",
"_____no_output_____"
],
[
"comp['mod_type'].value_counts()",
"_____no_output_____"
],
[
"comp['time'][comp['type']=='THP'].describe()",
"_____no_output_____"
],
[
"comp = pd. read_csv('../comparison_out/comp.EQP.out', sep = '/s+')\ncomp.head() ",
"/srv/conda/envs/notebook/lib/python3.7/site-packages/ipykernel_launcher.py:1: ParserWarning: Falling back to the 'python' engine because the 'c' engine does not support regex separators (separators > 1 char and different from '\\s+' are interpreted as regex); you can avoid this warning by specifying engine='python'.\n \"\"\"Entry point for launching an IPython kernel.\n"
],
[
"with open('../comparison_out/comp.EQP.out', 'r') as comp:\n leng = []\n for line in comp:\n pickline = line.split(' ')\n leng.append(len(pickline))",
"_____no_output_____"
],
[
"max(leng)",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb93ecbf2619cd1da7592ab5065595caf70fa8ad
| 334,609 |
ipynb
|
Jupyter Notebook
|
.ipynb_checkpoints/The_Annotated__Attention_is_All_You_Need_-checkpoint.ipynb
|
yapdianang/fast_transformer
|
3c47e67dd98a3c5642304d5c7d8343618ada1f24
|
[
"MIT"
] | null | null | null |
.ipynb_checkpoints/The_Annotated__Attention_is_All_You_Need_-checkpoint.ipynb
|
yapdianang/fast_transformer
|
3c47e67dd98a3c5642304d5c7d8343618ada1f24
|
[
"MIT"
] | null | null | null |
.ipynb_checkpoints/The_Annotated__Attention_is_All_You_Need_-checkpoint.ipynb
|
yapdianang/fast_transformer
|
3c47e67dd98a3c5642304d5c7d8343618ada1f24
|
[
"MIT"
] | null | null | null | 91.348348 | 87,048 | 0.743082 |
[
[
[
"<img src=\"aiayn.png\">",
"_____no_output_____"
],
[
"> When teaching, I emphasize implementation as a way to understand recent developments in ML. This post is an attempt to keep myself honest along this goal. The recent [\"Attention is All You Need\"]\n(https://arxiv.org/abs/1706.03762) paper from NIPS 2017 has been instantly impactful paper as a new method for machine translation and potentiall NLP generally. The paper is very clearly written, but the conventional wisdom has been that it is quite difficult to implement correctly. \n>\n> In this post I follow the paper through from start to finish and try to implement each component in code. \n(I have done some minor reordering and skipping from the original paper). This document itself is a working notebook, and should be a completely usable and efficient implementation. To follow along you will first need to install [PyTorch](http://pytorch.org/) and [torchtext](https://github.com/pytorch/text). The complete code is available on [github](https://github.com/harvardnlp/annotated-transformer).\n>- Alexander \"Sasha\" Rush ([@harvardnlp](https://twitter.com/harvardnlp))\n",
"_____no_output_____"
]
],
[
[
"# Standard PyTorch imports\nimport numpy as np\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nimport math, copy\nfrom torch.autograd import Variable\n\n# For plots\n%matplotlib inline\nimport matplotlib.pyplot as plt\n",
"_____no_output_____"
]
],
[
[
"# Background",
"_____no_output_____"
],
[
"The goal of reducing sequential computation also forms the foundation of the Extended Neural GPU\n[16], ByteNet [18] and ConvS2S [9], all of which use convolutional neural networks as basic building\nblock, computing hidden representations in parallel for all input and output positions. In these models,\nthe number of operations required to relate signals from two arbitrary input or output positions grows\nin the distance between positions, linearly for ConvS2S and logarithmically for ByteNet. This makes\nit more difficult to learn dependencies between distant positions [12]. In the Transformer this is\nreduced to a constant number of operations, albeit at the cost of reduced effective resolution due\nto averaging attention-weighted positions, an effect we counteract with Multi-Head Attention as\ndescribed in section 3.2.\n\nSelf-attention, sometimes called intra-attention is an attention mechanism relating different positions\nof a single sequence in order to compute a representation of the sequence. Self-attention has been\nused successfully in a variety of tasks including reading comprehension, abstractive summarization,\ntextual entailment and learning task-independent sentence representations [4, 27, 28, 22].\nEnd-to-end memory networks are based on a recurrent attention mechanism instead of sequencealigned\nrecurrence and have been shown to perform well on simple-language question answering and\nlanguage modeling tasks [34].\n\nTo the best of our knowledge, however, the Transformer is the first transduction model relying\nentirely on self-attention to compute representations of its input and output without using sequencealigned\nRNNs or convolution. In the following sections, we will describe the Transformer, motivate\nself-attention and discuss its advantages over models such as [17, 18] and [9].",
"_____no_output_____"
],
[
"# Model Architecture",
"_____no_output_____"
],
[
"Most competitive neural sequence transduction models have an encoder-decoder structure [(cite)](cho2014learning,bahdanau2014neural,sutskever14). Here, the encoder maps an input sequence of symbol representations $(x_1, ..., x_n)$ to a sequence of continuous representations $\\mathbf{z} = (z_1, ..., z_n)$. Given $\\mathbf{z}$, the decoder then generates an output sequence $(y_1,...,y_m)$ of symbols one element at a time. At each step the model is auto-regressive [(cite)](graves2013generating), consuming the previously generated symbols as additional input when generating the next. ",
"_____no_output_____"
]
],
[
[
"class EncoderDecoder(nn.Module):\n \"\"\"\n A standard Encoder-Decoder architecture. Base model for this and many \n other models.\n \"\"\"\n def __init__(self, encoder, decoder, src_embed, tgt_embed, generator):\n super(EncoderDecoder, self).__init__()\n self.encoder = encoder\n self.decoder = decoder\n self.src_embed = src_embed\n self.tgt_embed = tgt_embed\n self.generator = generator\n \n def forward(self, src, tgt, src_mask, tgt_mask):\n \"Take in and process masked src and target sequences.\"\n memory = self.encoder(self.src_embed(src), src_mask)\n output = self.decoder(self.tgt_embed(tgt), memory, src_mask, tgt_mask)\n return output",
"_____no_output_____"
]
],
[
[
"The Transformer follows this overall architecture using stacked self-attention and point-wise, fully connected layers for both the encoder and decoder, shown in the left and right halves of Figure 1, respectively. ",
"_____no_output_____"
],
[
"<img src=\"ModalNet-21.png\" width=400px>",
"_____no_output_____"
],
[
"## Encoder and Decoder Stacks \n\n### Encoder: \n\nThe encoder is composed of a stack of $N=6$ identical layers. ",
"_____no_output_____"
]
],
[
[
"def clones(module, N):\n \"Produce N identical layers.\"\n return nn.ModuleList([copy.deepcopy(module) for _ in range(N)])",
"_____no_output_____"
],
[
"class Encoder(nn.Module):\n \"Core encoder is a stack of N layers\"\n def __init__(self, layer, N):\n super(Encoder, self).__init__()\n self.layers = clones(layer, N)\n self.norm = LayerNorm(layer.size)\n \n def forward(self, x, mask):\n \"Pass the input (and mask) through each layer in turn.\"\n for layer in self.layers:\n x = layer(x, mask)\n return self.norm(x)",
"_____no_output_____"
]
],
[
[
"We employ a residual connection [(cite)](he2016deep) around each of the two sub-layers, followed by layer normalization [(cite)](layernorm2016). ",
"_____no_output_____"
]
],
[
[
"class LayerNorm(nn.Module):\n \"Construct a layernorm module (See citation for details).\"\n def __init__(self, features, eps=1e-6):\n super(LayerNorm, self).__init__()\n self.a_2 = nn.Parameter(torch.ones(features))\n self.b_2 = nn.Parameter(torch.zeros(features))\n self.eps = eps\n\n def forward(self, x):\n mean = x.mean(-1, keepdim=True)\n std = x.std(-1, keepdim=True)\n return self.a_2 * (x - mean) / (std + self.eps) + self.b_2",
"_____no_output_____"
]
],
[
[
"That is, the output of each sub-layer is $\\mathrm{LayerNorm}(x + \\mathrm{Sublayer}(x))$, where $\\mathrm{Sublayer}(x)$ is the function implemented by the sub-layer itself. We apply dropout [(cite)](srivastava2014dropout) to the output of each sub-layer, before it is added to the sub-layer input and normalized. \n\nTo facilitate these residual connections, all sub-layers in the model, as well as the embedding layers, produce outputs of dimension $d_{\\text{model}}=512$. ",
"_____no_output_____"
]
],
[
[
"class SublayerConnection(nn.Module):\n \"\"\"\n A residual connection followed by a layer norm.\n Note for code simplicity we apply the norm first as opposed to last.\n \"\"\"\n def __init__(self, size, dropout):\n super(SublayerConnection, self).__init__()\n self.norm = LayerNorm(size)\n self.dropout = nn.Dropout(dropout)\n\n def forward(self, x, sublayer):\n \"Apply residual connection to any sublayer function that maintains the same size.\"\n return x + self.dropout(sublayer(self.norm(x)))",
"_____no_output_____"
]
],
[
[
"Each layer has two sub-layers. The first is a multi-head self-attention mechanism, and the second is a simple, position-wise fully connected feed-forward network.",
"_____no_output_____"
]
],
[
[
"class EncoderLayer(nn.Module):\n \"Encoder is made up of two sublayers, self-attn and feed forward (defined below)\"\n def __init__(self, size, self_attn, feed_forward, dropout):\n super(EncoderLayer, self).__init__()\n self.self_attn = self_attn\n self.feed_forward = feed_forward\n self.sublayer = clones(SublayerConnection(size, dropout), 2)\n self.size = size\n\n def forward(self, x, mask):\n \"Follow Figure 1 (left) for connections.\"\n x = self.sublayer[0](x, lambda x: self.self_attn(x, x, x, mask))\n return self.sublayer[1](x, self.feed_forward)",
"_____no_output_____"
]
],
[
[
"### Decoder:\n\nThe decoder is also composed of a stack of $N=6$ identical layers. \n",
"_____no_output_____"
]
],
[
[
"class Decoder(nn.Module):\n \"Generic N layer decoder with masking.\"\n def __init__(self, layer, N):\n super(Decoder, self).__init__()\n self.layers = clones(layer, N)\n self.norm = LayerNorm(layer.size)\n \n def forward(self, x, memory, src_mask, tgt_mask):\n for layer in self.layers:\n x = layer(x, memory, src_mask, tgt_mask)\n return self.norm(x)",
"_____no_output_____"
]
],
[
[
"In addition to the two sub-layers in each encoder layer, the decoder inserts a third sub-layer, which performs multi-head attention over the output of the encoder stack. Similar to the encoder, we employ residual connections around each of the sub-layers, followed by layer normalization. ",
"_____no_output_____"
]
],
[
[
"class DecoderLayer(nn.Module):\n \"Decoder is made up of three sublayers, self-attn, src-attn, and feed forward (defined below)\"\n def __init__(self, size, self_attn, src_attn, feed_forward, dropout):\n super(DecoderLayer, self).__init__()\n self.size = size\n self.self_attn = self_attn\n self.src_attn = src_attn\n self.feed_forward = feed_forward\n self.sublayer = clones(SublayerConnection(size, dropout), 3)\n \n def forward(self, x, memory, src_mask, tgt_mask):\n \"Follow Figure 1 (right) for connections.\"\n m = memory\n x = self.sublayer[0](x, lambda x: self.self_attn(x, x, x, tgt_mask))\n x = self.sublayer[1](x, lambda x: self.src_attn(x, m, m, src_mask))\n return self.sublayer[2](x, self.feed_forward)",
"_____no_output_____"
]
],
[
[
"We also modify the self-attention sub-layer in the decoder stack to prevent positions from attending to subsequent positions. This masking, combined with fact that the output embeddings are offset by one position, ensures that the predictions for position $i$ can depend only on the known outputs at positions less than $i$.",
"_____no_output_____"
]
],
[
[
"def subsequent_mask(size):\n \"Mask out subsequent positions.\"\n attn_shape = (1, size, size)\n subsequent_mask = np.triu(np.ones(attn_shape), k=1).astype('uint8')\n return torch.from_numpy(subsequent_mask) == 0",
"_____no_output_____"
],
[
"# The attention mask shows the position each tgt word (row) is allowed to look at (column).\n# Words are blocked for attending to future words during training. \nplt.figure(figsize=(5,5))\nplt.imshow(subsequent_mask(20)[0])",
"_____no_output_____"
]
],
[
[
"### Attention: \nAn attention function can be described as mapping a query and a set of key-value pairs to an output, where the query, keys, values, and output are all vectors. The output is computed as a weighted sum of the values, where the weight assigned to each value is computed by a compatibility function of the query with the corresponding key. \n\nWe call our particular attention \"Scaled Dot-Product Attention\". The input consists of queries and keys of dimension $d_k$, and values of dimension $d_v$. We compute the dot products of the query with all keys, divide each by $\\sqrt{d_k}$, and apply a softmax function to obtain the weights on the values. \n<img width=\"220px\" src=\"ModalNet-19.png\">\n\nIn practice, we compute the attention function on a set of queries simultaneously, packed together into a matrix $Q$. The keys and values are also packed together into matrices $K$ and $V$. We compute the matrix of outputs as: \n \n$$ \n \\mathrm{Attention}(Q, K, V) = \\mathrm{softmax}(\\frac{QK^T}{\\sqrt{d_k}})V \n$$ \n ",
"_____no_output_____"
]
],
[
[
"def attention(query, key, value, mask=None, dropout=0.0):\n \"Compute 'Scaled Dot Product Attention'\"\n d_k = query.size(-1)\n scores = torch.matmul(query, key.transpose(-2, -1)) \\\n / math.sqrt(d_k)\n if mask is not None:\n scores = scores.masked_fill(mask == 0, -1e9)\n p_attn = F.softmax(scores, dim = -1)\n # (Dropout described below)\n p_attn = F.dropout(p_attn, p=dropout)\n return torch.matmul(p_attn, value), p_attn",
"_____no_output_____"
]
],
[
[
"The two most commonly used attention functions are additive attention [(cite)](bahdanau2014neural), and dot-product (multiplicative) attention. Dot-product attention is identical to our algorithm, except for the scaling factor of $\\frac{1}{\\sqrt{d_k}}$. Additive attention computes the compatibility function using a feed-forward network with a single hidden layer. While the two are similar in theoretical complexity, dot-product attention is much faster and more space-efficient in practice, since it can be implemented using highly optimized matrix multiplication code. \n\n \nWhile for small values of $d_k$ the two mechanisms perform similarly, additive attention outperforms dot product attention without scaling for larger values of $d_k$ [(cite)](DBLP:journals/corr/BritzGLL17). We suspect that for large values of $d_k$, the dot products grow large in magnitude, pushing the softmax function into regions where it has extremely small gradients (To illustrate why the dot products get large, assume that the components of $q$ and $k$ are independent random variables with mean $0$ and variance $1$. Then their dot product, $q \\cdot k = \\sum_{i=1}^{d_k} q_ik_i$, has mean $0$ and variance $d_k$.). To counteract this effect, we scale the dot products by $\\frac{1}{\\sqrt{d_k}}$. ",
"_____no_output_____"
],
[
" ",
"_____no_output_____"
],
[
"### Multi-Head Attention \n \nInstead of performing a single attention function with $d_{\\text{model}}$-dimensional keys, values and queries, we found it beneficial to linearly project the queries, keys and values $h$ times with different, learned linear projections to $d_k$, $d_k$ and $d_v$ dimensions, respectively. \nOn each of these projected versions of queries, keys and values we then perform the attention function in parallel, yielding $d_v$-dimensional output values. These are concatenated and once again projected, resulting in the final values:\n\n<img width=\"270px\" src=\"ModalNet-20.png\">\n \nMulti-head attention allows the model to jointly attend to information from different representation subspaces at different positions. With a single attention head, averaging inhibits this. \n \n \n \n$$ \n\\mathrm{MultiHead}(Q, K, V) = \\mathrm{Concat}(\\mathrm{head_1}, ..., \\mathrm{head_h})W^O \\\\ \n \\text{where}~\\mathrm{head_i} = \\mathrm{Attention}(QW^Q_i, KW^K_i, VW^V_i) \n$$ \n \nWhere the projections are parameter matrices $W^Q_i \\in \\mathbb{R}^{d_{\\text{model}} \\times d_k}$, $W^K_i \\in \\mathbb{R}^{d_{\\text{model}} \\times d_k}$, $W^V_i \\in \\mathbb{R}^{d_{\\text{model}} \\times d_v}$ and $W^O \\in \\mathbb{R}^{hd_v \\times d_{\\text{model}}}$. \n \n\n \nIn this work we employ $h=8$ parallel attention layers, or heads. For each of these we use $d_k=d_v=d_{\\text{model}}/h=64$. Due to the reduced dimension of each head, the total computational cost is similar to that of single-head attention with full dimensionality. ",
"_____no_output_____"
]
],
[
[
"class MultiHeadedAttention(nn.Module):\n def __init__(self, h, d_model, dropout=0.1):\n \"Take in model size and number of heads.\"\n super(MultiHeadedAttention, self).__init__()\n assert d_model % h == 0\n # We assume d_v always equals d_k\n self.d_k = d_model // h\n self.h = h\n self.p = dropout\n self.linears = clones(nn.Linear(d_model, d_model), 4)\n self.attn = None\n \n def forward(self, query, key, value, mask=None):\n \"Implements Figure 2\"\n if mask is not None:\n # Same mask applied to all h heads.\n mask = mask.unsqueeze(1)\n nbatches = query.size(0)\n \n # 1) Do all the linear projections in batch from d_model => h x d_k \n query, key, value = [l(x).view(nbatches, -1, self.h, self.d_k).transpose(1, 2)\n for l, x in zip(self.linears, (query, key, value))]\n \n # 2) Apply attention on all the projected vectors in batch. \n x, self.attn = attention(query, key, value, mask=mask, dropout=self.p)\n \n # 3) \"Concat\" using a view and apply a final linear. \n x = x.transpose(1, 2).contiguous().view(nbatches, -1, self.h * self.d_k)\n return self.linears[-1](x)",
"_____no_output_____"
]
],
[
[
"### Applications of Attention in our Model \nThe Transformer uses multi-head attention in three different ways: \n1) In \"encoder-decoder attention\" layers, the queries come from the previous decoder layer, and the memory keys and values come from the output of the encoder. This allows every position in the decoder to attend over all positions in the input sequence. This mimics the typical encoder-decoder attention mechanisms in sequence-to-sequence models such as [(cite)](wu2016google, bahdanau2014neural,JonasFaceNet2017). \n\n\n2) The encoder contains self-attention layers. In a self-attention layer all of the keys, values and queries come from the same place, in this case, the output of the previous layer in the encoder. Each position in the encoder can attend to all positions in the previous layer of the encoder. \n\n\n3) Similarly, self-attention layers in the decoder allow each position in the decoder to attend to all positions in the decoder up to and including that position. We need to prevent leftward information flow in the decoder to preserve the auto-regressive property. We implement this inside of scaled dot-product attention by masking out (setting to $-\\infty$) all values in the input of the softmax which correspond to illegal connections. ",
"_____no_output_____"
],
[
"## Position-wise Feed-Forward Networks \n \nIn addition to attention sub-layers, each of the layers in our encoder and decoder contains a fully connected feed-forward network, which is applied to each position separately and identically. This consists of two linear transformations with a ReLU activation in between.\n\n$$\\mathrm{FFN}(x)=\\max(0, xW_1 + b_1) W_2 + b_2$$ \n \nWhile the linear transformations are the same across different positions, they use different parameters from layer to layer. Another way of describing this is as two convolutions with kernel size 1. The dimensionality of input and output is $d_{\\text{model}}=512$, and the inner-layer has dimensionality $d_{ff}=2048$. ",
"_____no_output_____"
]
],
[
[
"class PositionwiseFeedForward(nn.Module):\n \"Implements FFN equation.\"\n def __init__(self, d_model, d_ff, dropout=0.1):\n super(PositionwiseFeedForward, self).__init__()\n # Torch linears have a `b` by default. \n self.w_1 = nn.Linear(d_model, d_ff)\n self.w_2 = nn.Linear(d_ff, d_model)\n self.dropout = nn.Dropout(dropout)\n\n def forward(self, x):\n return self.w_2(self.dropout(F.relu(self.w_1(x))))",
"_____no_output_____"
]
],
[
[
"## Embeddings and Softmax \nSimilarly to other sequence transduction models, we use learned embeddings to convert the input tokens and output tokens to vectors of dimension $d_{\\text{model}}$. We also use the usual learned linear transformation and softmax function to convert the decoder output to predicted next-token probabilities. In our model, we share the same weight matrix between the two embedding layers and the pre-softmax linear transformation, similar to [(cite)](press2016using). In the embedding layers, we multiply those weights by $\\sqrt{d_{\\text{model}}}$. ",
"_____no_output_____"
]
],
[
[
"class Embeddings(nn.Module):\n def __init__(self, d_model, vocab):\n super(Embeddings, self).__init__()\n self.lut = nn.Embedding(vocab, d_model)\n self.d_model = d_model\n\n def forward(self, x):\n return self.lut(x) * math.sqrt(self.d_model)",
"_____no_output_____"
]
],
[
[
"## Positional Encoding \nSince our model contains no recurrence and no convolution, in order for the model to make use of the order of the sequence, we must inject some information about the relative or absolute position of the tokens in the sequence. To this end, we add \"positional encodings\" to the input embeddings at the bottoms of the encoder and decoder stacks. The positional encodings have the same dimension $d_{\\text{model}}$ as the embeddings, so that the two can be summed. There are many choices of positional encodings, learned and fixed [(cite)](JonasFaceNet2017). \n\nIn this work, we use sine and cosine functions of different frequencies: \n$$ \n PE_{(pos,2i)} = sin(pos / 10000^{2i/d_{\\text{model}}}) \\\\ \n PE_{(pos,2i+1)} = cos(pos / 10000^{2i/d_{\\text{model}}}) \n$$ \nwhere $pos$ is the position and $i$ is the dimension. That is, each dimension of the positional encoding corresponds to a sinusoid. The wavelengths form a geometric progression from $2\\pi$ to $10000 \\cdot 2\\pi$. We chose this function because we hypothesized it would allow the model to easily learn to attend by relative positions, since for any fixed offset $k$, $PE_{pos+k}$ can be represented as a linear function of $PE_{pos}$. \n\nIn addition, we apply dropout to the sums of the embeddings and the positional encodings in both the encoder and decoder stacks. For the base model, we use a rate of $P_{drop}=0.1$. \n \n",
"_____no_output_____"
]
],
[
[
"class PositionalEncoding(nn.Module):\n \"Implement the PE function.\"\n def __init__(self, d_model, dropout, max_len=5000):\n super(PositionalEncoding, self).__init__()\n self.dropout = nn.Dropout(p=dropout)\n \n # Compute the positional encodings once in log space.\n pe = torch.zeros(max_len, d_model)\n position = torch.arange(0., max_len).unsqueeze(1)\n div_term = torch.exp(torch.arange(0., d_model, 2) *\n -(math.log(10000.0) / d_model))\n pe[:, 0::2] = torch.sin(position * div_term)\n pe[:, 1::2] = torch.cos(position * div_term)\n pe = pe.unsqueeze(0)\n self.register_buffer('pe', pe)\n \n def forward(self, x):\n x = x + Variable(self.pe[:, :x.size(1)], requires_grad=False)\n return self.dropout(x)",
"_____no_output_____"
],
[
"# The positional encoding will add in a sine wave based on position.\n# The frequency and offset of the wave is different for each dimension.\nplt.figure(figsize=(15, 5))\npe = PositionalEncoding(20, 0)\ny = pe.forward(Variable(torch.zeros(1, 100, 20)))\nplt.plot(np.arange(100), y[0, :, 4:8].data.numpy())\nplt.legend([\"dim %d\"%p for p in [4,5,6,7]])\nNone",
"_____no_output_____"
]
],
[
[
"We also experimented with using learned positional embeddings [(cite)](JonasFaceNet2017) instead, and found that the two versions produced nearly identical results. We chose the sinusoidal version because it may allow the model to extrapolate to sequence lengths longer than the ones encountered during training. ",
"_____no_output_____"
],
[
"## Generation",
"_____no_output_____"
]
],
[
[
"class Generator(nn.Module):\n \"Standard generation step. (Not described in the paper.)\"\n def __init__(self, d_model, vocab):\n super(Generator, self).__init__()\n self.proj = nn.Linear(d_model, vocab)\n\n def forward(self, x):\n return F.log_softmax(self.proj(x), dim=-1)",
"_____no_output_____"
]
],
[
[
"## Full Model",
"_____no_output_____"
]
],
[
[
"def make_model(src_vocab, tgt_vocab, N=6, d_model=512, d_ff=2048, h=8, dropout=0.1):\n \"Construct a model object based on hyperparameters.\"\n c = copy.deepcopy\n attn = MultiHeadedAttention(h, d_model, dropout)\n ff = PositionwiseFeedForward(d_model, d_ff, dropout)\n position = PositionalEncoding(d_model, dropout)\n model = EncoderDecoder(\n Encoder(EncoderLayer(d_model, c(attn), c(ff), dropout), N),\n Decoder(DecoderLayer(d_model, c(attn), c(attn), c(ff), dropout), N),\n nn.Sequential(Embeddings(d_model, src_vocab), c(position)),\n nn.Sequential(Embeddings(d_model, tgt_vocab), c(position)),\n Generator(d_model, tgt_vocab))\n \n # This was important from their code. Initialize parameters with Glorot or fan_avg.\n for p in model.parameters():\n if p.dim() > 1:\n nn.init.xavier_uniform_(p)\n return model",
"_____no_output_____"
],
[
"# Small example model.\ntmp_model = make_model(10, 10, 2)\ntmp_model",
"_____no_output_____"
]
],
[
[
"# Training\n\nThis section describes the training regime for our models.\n",
"_____no_output_____"
],
[
"## Training Data and Batching\n\nWe trained on the standard WMT 2014 English-German dataset consisting of about 4.5 million sentence pairs. Sentences were encoded using byte-pair encoding \\citep{DBLP:journals/corr/BritzGLL17}, which has a shared source-target vocabulary of about 37000 tokens. For English-French, we used the significantly larger WMT 2014 English-French dataset consisting of 36M sentences and split tokens into a 32000 word-piece vocabulary [(cite)](wu2016google). \n\n\nSentence pairs were batched together by approximate sequence length. Each training batch contained a set of sentence pairs containing approximately 25000 source tokens and 25000 target tokens. ",
"_____no_output_____"
],
[
"## Hardware and Schedule \nWe trained our models on one machine with 8 NVIDIA P100 GPUs. For our base models using the hyperparameters described throughout the paper, each training step took about 0.4 seconds. We trained the base models for a total of 100,000 steps or 12 hours. For our big models, step time was 1.0 seconds. The big models were trained for 300,000 steps (3.5 days).",
"_____no_output_____"
],
[
"## Optimizer\n\nWe used the Adam optimizer [(cite)](kingma2014adam) with $\\beta_1=0.9$, $\\beta_2=0.98$ and $\\epsilon=10^{-9}$. We varied the learning rate over the course of training, according to the formula: \n$$ \nlrate = d_{\\text{model}}^{-0.5} \\cdot \n \\min({step\\_num}^{-0.5}, \n {step\\_num} \\cdot {warmup\\_steps}^{-1.5}) \n$$ \nThis corresponds to increasing the learning rate linearly for the first $warmup\\_steps$ training steps, and decreasing it thereafter proportionally to the inverse square root of the step number. We used $warmup\\_steps=4000$. ",
"_____no_output_____"
]
],
[
[
"# Note: This part is incredibly important. \n# Need to train with this setup of the model is very unstable.\nclass NoamOpt:\n \"Optim wrapper that implements rate.\"\n def __init__(self, model_size, factor, warmup, optimizer):\n self.optimizer = optimizer\n self._step = 0\n self.warmup = warmup\n self.factor = factor\n self.model_size = model_size\n self._rate = 0\n \n def step(self):\n \"Update parameters and rate\"\n self._step += 1\n rate = self.rate()\n for p in self.optimizer.param_groups:\n p['lr'] = rate\n self._rate = rate\n self.optimizer.step()\n \n def rate(self, step = None):\n \"Implement `lrate` above\"\n if step is None:\n step = self._step\n return self.factor * \\\n (self.model_size ** (-0.5) *\n min(step ** (-0.5), step * self.warmup**(-1.5)))\n \ndef get_std_opt(model):\n return NoamOpt(model.src_embed[0].d_model, 2, 4000,\n torch.optim.Adam(model.parameters(), lr=0, betas=(0.9, 0.98), eps=1e-9))",
"_____no_output_____"
],
[
"# Three settings of the lrate hyperparameters.\nopts = [NoamOpt(512, 1, 4000, None), \n NoamOpt(512, 1, 8000, None),\n NoamOpt(256, 1, 4000, None)]\nplt.plot(np.arange(1, 20000), [[opt.rate(i) for opt in opts] for i in range(1, 20000)])\nplt.legend([\"512:4000\", \"512:8000\", \"256:4000\"])\nNone",
"_____no_output_____"
]
],
[
[
"## Regularization \n \n### Label Smoothing\n\nDuring training, we employed label smoothing of value $\\epsilon_{ls}=0.1$ [(cite)](DBLP:journals/corr/SzegedyVISW15). This hurts perplexity, as the model learns to be more unsure, but improves accuracy and BLEU score. ",
"_____no_output_____"
]
],
[
[
"class LabelSmoothing(nn.Module):\n \"\"\"\n Implement label smoothing.\n \"\"\"\n def __init__(self, size, padding_idx, smoothing=0.0):\n super(LabelSmoothing, self).__init__()\n self.criterion = nn.KLDivLoss(size_average=False)\n self.padding_idx = padding_idx\n self.confidence = 1.0 - smoothing\n self.smoothing = smoothing\n self.size = size\n self.true_dist = None\n\n def forward(self, x, target):\n assert x.size(1) == self.size\n true_dist = x.data.clone()\n true_dist.fill_(self.smoothing / (self.size - 2))\n true_dist.scatter_(1, target.data.unsqueeze(1), self.confidence)\n true_dist[:, self.padding_idx] = 0\n mask = torch.nonzero(target.data == self.padding_idx)\n if mask.dim() > 0 and mask.size(-1) > 0:\n true_dist.index_fill_(0, mask.squeeze(), 0.0)\n self.true_dist = true_dist\n return self.criterion(x, Variable(true_dist, requires_grad=False))",
"_____no_output_____"
],
[
"#Example\ncrit = LabelSmoothing(5, 0, 0.5)\npredict = torch.FloatTensor([[0, 0.2, 0.7, 0.1, 0],\n [0, 0.2, 0.7, 0.1, 0], \n [0, 0.2, 0.7, 0.1, 0]])\nv = crit(Variable(predict.log()), \n Variable(torch.LongTensor([2, 1, 0])))\n\n# Show the target distributions expected by the system.\nplt.imshow(crit.true_dist)\nNone",
"/Users/dianang/.local/lib/python3.6/site-packages/torch/nn/functional.py:52: UserWarning: size_average and reduce args will be deprecated, please use reduction='sum' instead.\n warnings.warn(warning.format(ret))\n"
],
[
"# Label smoothing starts to penalize the model \n# if it gets very confident about a given choice\ncrit = LabelSmoothing(5, 0, 0.2)\ndef loss(x):\n d = x + 3 * 1\n predict = torch.FloatTensor([[0, x / d, 1 / d, 1 / d, 1 / d],\n ])\n #print(predict)\n return crit(Variable(predict.log()),\n Variable(torch.LongTensor([1]))).data[0]\nplt.plot(np.arange(1, 100), [loss(x) for x in range(1, 100)])",
"/Users/dianang/.local/lib/python3.6/site-packages/torch/nn/functional.py:52: UserWarning: size_average and reduce args will be deprecated, please use reduction='sum' instead.\n warnings.warn(warning.format(ret))\n/Users/dianang/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:10: UserWarning: invalid index of a 0-dim tensor. This will be an error in PyTorch 0.5. Use tensor.item() to convert a 0-dim tensor to a Python number\n # Remove the CWD from sys.path while we load stuff.\n"
]
],
[
[
"### Memory Optimization",
"_____no_output_____"
]
],
[
[
"def loss_backprop(generator, criterion, out, targets, normalize):\n \"\"\"\n Memory optmization. Compute each timestep separately and sum grads.\n \"\"\"\n assert out.size(1) == targets.size(1)\n total = 0.0\n out_grad = []\n for i in range(out.size(1)):\n out_column = Variable(out[:, i].data, requires_grad=True)\n gen = generator(out_column)\n loss = criterion(gen, targets[:, i].float()) / normalize\n total += loss.data[0]\n loss.backward()\n out_grad.append(out_column.grad.data.clone())\n out_grad = torch.stack(out_grad, dim=1)\n out.backward(gradient=out_grad)\n return total",
"_____no_output_____"
],
[
"def make_std_mask(src, tgt, pad):\n src_mask = (src != pad).unsqueeze(-2)\n tgt_mask = (tgt != pad).unsqueeze(-2)\n tgt_mask = tgt_mask & Variable(subsequent_mask(tgt.size(-1)).type_as(tgt_mask.data))\n return src_mask, tgt_mask",
"_____no_output_____"
],
[
"def train_epoch(train_iter, model, criterion, opt, transpose=False):\n model.train()\n for i, batch in enumerate(train_iter):\n src, trg, src_mask, trg_mask = \\\n batch.src, batch.trg, batch.src_mask, batch.trg_mask\n out = model.forward(src, trg[:, :-1], src_mask, trg_mask[:, :-1, :-1])\n loss = loss_backprop(model.generator, criterion, out, trg[:, 1:], batch.ntokens) \n \n model_opt.step()\n model_opt.optimizer.zero_grad()\n if i % 10 == 1:\n print(i, loss, model_opt._rate)",
"_____no_output_____"
],
[
"def valid_epoch(valid_iter, model, criterion, transpose=False):\n model.test()\n total = 0\n for batch in valid_iter:\n src, trg, src_mask, trg_mask = \\\n batch.src, batch.trg, batch.src_mask, batch.trg_mask\n out = model.forward(src, trg[:, :-1], src_mask, trg_mask[:, :-1, :-1])\n loss = loss_backprop(model.generator, criterion, out, trg[:, 1:], batch.ntokens) \n ",
"_____no_output_____"
],
[
"class Batch:\n def __init__(self, src, trg, src_mask, trg_mask, ntokens):\n self.src = src\n self.trg = trg\n self.src_mask = src_mask\n self.trg_mask = trg_mask\n self.ntokens = ntokens\n \ndef data_gen(V, batch, nbatches):\n for i in range(nbatches):\n data = torch.from_numpy(np.random.randint(1, V, size=(batch, 10)))\n src = Variable(data, requires_grad=False)\n tgt = Variable(data, requires_grad=False)\n src_mask, tgt_mask = make_std_mask(src, tgt, 0)\n yield Batch(src, tgt, src_mask, tgt_mask, (tgt[1:] != 0).data.sum())",
"_____no_output_____"
],
[
"V = 11\ncriterion = LabelSmoothing(size=V, padding_idx=0, smoothing=0.0)\nmodel = make_model(V, V, N=2)\nmodel_opt = get_std_opt(model)\nfor epoch in range(2):\n train_epoch(data_gen(V, 30, 20), model, criterion, model_opt)",
"/Users/dianang/.local/lib/python3.6/site-packages/torch/nn/functional.py:52: UserWarning: size_average and reduce args will be deprecated, please use reduction='sum' instead.\n warnings.warn(warning.format(ret))\n"
]
],
[
[
"# A Real World Example",
"_____no_output_____"
]
],
[
[
"# For data loading.\nfrom torchtext import data, datasets",
"_____no_output_____"
],
[
"!pip install torchtext spacy\n!python -m spacy download en\n!python -m spacy download de",
"Requirement already satisfied: torchtext in /usr/local/lib/python3.6/dist-packages\nRequirement already satisfied: spacy in /usr/local/lib/python3.6/dist-packages\nRequirement already satisfied: tqdm in /usr/local/lib/python3.6/dist-packages (from torchtext)\nRequirement already satisfied: requests in /usr/local/lib/python3.6/dist-packages (from torchtext)\nRequirement already satisfied: preshed<2.0.0,>=1.0.0 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: murmurhash<0.29,>=0.28 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: numpy>=1.7 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: plac<1.0.0,>=0.9.6 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: cymem<1.32,>=1.30 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: msgpack-python==0.5.4 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: html5lib==1.0b8 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: thinc<6.11.0,>=6.10.1 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: pathlib in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: ujson>=1.35 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: dill<0.3,>=0.2 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: msgpack-numpy==0.4.1 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: ftfy<5.0.0,>=4.4.2 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: regex==2017.4.5 in /usr/local/lib/python3.6/dist-packages (from spacy)\nRequirement already satisfied: chardet<3.1.0,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->torchtext)\nRequirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests->torchtext)\nRequirement already satisfied: urllib3<1.23,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests->torchtext)\nRequirement already satisfied: idna<2.7,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->torchtext)\nRequirement already satisfied: cytoolz<0.9,>=0.8 in /usr/local/lib/python3.6/dist-packages (from thinc<6.11.0,>=6.10.1->spacy)\nRequirement already satisfied: termcolor in /usr/local/lib/python3.6/dist-packages (from thinc<6.11.0,>=6.10.1->spacy)\nRequirement already satisfied: wrapt in /usr/local/lib/python3.6/dist-packages (from thinc<6.11.0,>=6.10.1->spacy)\nRequirement already satisfied: wcwidth in /usr/local/lib/python3.6/dist-packages (from ftfy<5.0.0,>=4.4.2->spacy)\nRequirement already satisfied: toolz>=0.8.0 in /usr/local/lib/python3.6/dist-packages (from cytoolz<0.9,>=0.8->thinc<6.11.0,>=6.10.1->spacy)\nCollecting https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-2.0.0/en_core_web_sm-2.0.0.tar.gz\n Downloading https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-2.0.0/en_core_web_sm-2.0.0.tar.gz (37.4MB)\n\u001b[K 6% |██ | 2.4MB 47.2MB/s eta 0:00:01\u001b[K 100% |████████████████████████████████| 37.4MB 50.2MB/s \n\u001b[?25hInstalling collected packages: en-core-web-sm\n Running setup.py install for en-core-web-sm ... \u001b[?25ldone\n\u001b[?25hSuccessfully installed en-core-web-sm-2.0.0\n\n\u001b[93m Linking successful\u001b[0m\n /usr/local/lib/python3.6/dist-packages/en_core_web_sm -->\n /usr/local/lib/python3.6/dist-packages/spacy/data/en\n\n You can now load the model via spacy.load('en')\n\nCollecting https://github.com/explosion/spacy-models/releases/download/de_core_news_sm-2.0.0/de_core_news_sm-2.0.0.tar.gz\n Downloading https://github.com/explosion/spacy-models/releases/download/de_core_news_sm-2.0.0/de_core_news_sm-2.0.0.tar.gz (38.2MB)\n\u001b[K 100% |████████████████████████████████| 38.2MB 50.0MB/s \n\u001b[?25hInstalling collected packages: de-core-news-sm\n Running setup.py install for de-core-news-sm ... \u001b[?25ldone\n\u001b[?25hSuccessfully installed de-core-news-sm-2.0.0\n\n\u001b[93m Linking successful\u001b[0m\n /usr/local/lib/python3.6/dist-packages/de_core_news_sm -->\n /usr/local/lib/python3.6/dist-packages/spacy/data/de\n\n You can now load the model via spacy.load('de')\n\n"
],
[
"# Load words from IWSLT\n\n#!pip install torchtext spacy\n#!python -m spacy download en\n#!python -m spacy download de\n\nimport spacy\nspacy_de = spacy.load('de')\nspacy_en = spacy.load('en')\n\ndef tokenize_de(text):\n return [tok.text for tok in spacy_de.tokenizer(text)]\n\ndef tokenize_en(text):\n return [tok.text for tok in spacy_en.tokenizer(text)]\n\nBOS_WORD = '<s>'\nEOS_WORD = '</s>'\nBLANK_WORD = \"<blank>\"\nSRC = data.Field(tokenize=tokenize_de, pad_token=BLANK_WORD)\nTGT = data.Field(tokenize=tokenize_en, init_token = BOS_WORD, \n eos_token = EOS_WORD, pad_token=BLANK_WORD)\n\nMAX_LEN = 100\ntrain, val, test = datasets.IWSLT.splits(exts=('.de', '.en'), fields=(SRC, TGT), \n filter_pred=lambda x: len(vars(x)['src']) <= MAX_LEN and \n len(vars(x)['trg']) <= MAX_LEN)\nMIN_FREQ = 1\nSRC.build_vocab(train.src, min_freq=MIN_FREQ)\nTGT.build_vocab(train.trg, min_freq=MIN_FREQ)",
"downloading de-en.tgz\n.data/iwslt/de-en/IWSLT16.TEDX.dev2012.de-en.de.xml\n.data/iwslt/de-en/IWSLT16.TED.tst2013.de-en.en.xml\n.data/iwslt/de-en/IWSLT16.TED.tst2014.de-en.de.xml\n.data/iwslt/de-en/IWSLT16.TED.tst2012.de-en.en.xml\n.data/iwslt/de-en/IWSLT16.TED.tst2010.de-en.en.xml\n.data/iwslt/de-en/IWSLT16.TED.dev2010.de-en.en.xml\n.data/iwslt/de-en/IWSLT16.TED.dev2010.de-en.de.xml\n.data/iwslt/de-en/IWSLT16.TEDX.tst2013.de-en.en.xml\n.data/iwslt/de-en/IWSLT16.TEDX.dev2012.de-en.en.xml\n.data/iwslt/de-en/IWSLT16.TED.tst2013.de-en.de.xml\n.data/iwslt/de-en/IWSLT16.TEDX.tst2013.de-en.de.xml\n.data/iwslt/de-en/IWSLT16.TEDX.tst2014.de-en.de.xml\n.data/iwslt/de-en/IWSLT16.TED.tst2014.de-en.en.xml\n.data/iwslt/de-en/IWSLT16.TED.tst2011.de-en.de.xml\n.data/iwslt/de-en/IWSLT16.TED.tst2011.de-en.en.xml\n.data/iwslt/de-en/IWSLT16.TED.tst2010.de-en.de.xml\n.data/iwslt/de-en/IWSLT16.TEDX.tst2014.de-en.en.xml\n.data/iwslt/de-en/IWSLT16.TED.tst2012.de-en.de.xml\n.data/iwslt/de-en/train.tags.de-en.en\n.data/iwslt/de-en/train.tags.de-en.de\n"
],
[
"# Detail. Batching seems to matter quite a bit. \n# This is temporary code for dynamic batching based on number of tokens.\n# This code should all go away once things get merged in this library.\n\nBATCH_SIZE = 4096\nglobal max_src_in_batch, max_tgt_in_batch\ndef batch_size_fn(new, count, sofar):\n \"Keep augmenting batch and calculate total number of tokens + padding.\"\n global max_src_in_batch, max_tgt_in_batch\n if count == 1:\n max_src_in_batch = 0\n max_tgt_in_batch = 0\n max_src_in_batch = max(max_src_in_batch, len(new.src))\n max_tgt_in_batch = max(max_tgt_in_batch, len(new.trg) + 2)\n src_elements = count * max_src_in_batch\n tgt_elements = count * max_tgt_in_batch\n return max(src_elements, tgt_elements)\n\nclass MyIterator(data.Iterator):\n def create_batches(self):\n if self.train:\n def pool(d, random_shuffler):\n for p in data.batch(d, self.batch_size * 100):\n p_batch = data.batch(\n sorted(p, key=self.sort_key),\n self.batch_size, self.batch_size_fn)\n for b in random_shuffler(list(p_batch)):\n yield b\n self.batches = pool(self.data(), self.random_shuffler)\n \n else:\n self.batches = []\n for b in data.batch(self.data(), self.batch_size,\n self.batch_size_fn):\n self.batches.append(sorted(b, key=self.sort_key))\n\ndef rebatch(pad_idx, batch):\n \"Fix order in torchtext to match ours\"\n src, trg = batch.src.transpose(0, 1), batch.trg.transpose(0, 1)\n src_mask, trg_mask = make_std_mask(src, trg, pad_idx)\n return Batch(src, trg, src_mask, trg_mask, (trg[1:] != pad_idx).data.sum())\n\ntrain_iter = MyIterator(train, batch_size=BATCH_SIZE, device=0,\n repeat=False, sort_key=lambda x: (len(x.src), len(x.trg)),\n batch_size_fn=batch_size_fn, train=True)\nvalid_iter = MyIterator(val, batch_size=BATCH_SIZE, device=0,\n repeat=False, sort_key=lambda x: (len(x.src), len(x.trg)),\n batch_size_fn=batch_size_fn, train=False)",
"_____no_output_____"
],
[
"# Create the model an load it onto our GPU.\npad_idx = TGT.vocab.stoi[\"<blank>\"]\nmodel = make_model(len(SRC.vocab), len(TGT.vocab), N=6)\nmodel_opt = get_std_opt(model)\nmodel.cuda()",
"_____no_output_____"
],
[
"\ncriterion = LabelSmoothing(size=len(TGT.vocab), padding_idx=pad_idx, smoothing=0.1)\ncriterion.cuda()\nfor epoch in range(15):\n train_epoch((rebatch(pad_idx, b) for b in train_iter), model, criterion, model_opt)\n valid_epoch((rebatch(pad_idx, b) for b in valid_iter), model, criterion)",
"1 9.299771845340729 6.987712429686844e-07\n11 9.415135336574167 4.192627457812107e-06\n21 8.813630282878876 7.686483672655528e-06\n31 9.112178653478622 1.118033988749895e-05\n41 8.607461810112 1.4674196102342371e-05\n51 8.913826749660075 1.8168052317185794e-05\n61 8.701497752219439 2.1661908532029216e-05\n71 8.373274087905884 2.515576474687264e-05\n81 8.454237446188927 2.8649620961716057e-05\n91 7.6996782422065735 3.214347717655948e-05\n101 8.037408232688904 3.56373333914029e-05\n111 7.704962134361267 3.913118960624633e-05\n121 7.699015600606799 4.262504582108975e-05\n131 7.367554426193237 4.611890203593317e-05\n141 7.2071177661418915 4.961275825077659e-05\n151 7.106400920893066 5.310661446562001e-05\n161 6.804656069725752 5.660047068046343e-05\n171 6.390337720513344 6.0094326895306855e-05\n181 5.687528342008591 6.358818311015028e-05\n191 6.122820109128952 6.70820393249937e-05\n201 5.829070374369621 7.057589553983712e-05\n"
]
],
[
[
"\nOTHER",
"_____no_output_____"
]
],
[
[
"BOS_WORD = '<s>'\nEOS_WORD = '</s>'\nBLANK_WORD = \"<blank>\"\nSRC = data.Field()\nTGT = data.Field(init_token = BOS_WORD, eos_token = EOS_WORD, pad_token=BLANK_WORD) # only target needs BOS/EOS\n\nMAX_LEN = 100\ntrain = datasets.TranslationDataset(path=\"/n/home00/srush/Data/baseline-1M_train.tok.shuf\", \n exts=('.en', '.fr'),\n fields=(SRC, TGT), \n filter_pred=lambda x: len(vars(x)['src']) <= MAX_LEN and \n len(vars(x)['trg']) <= MAX_LEN)\nSRC.build_vocab(train.src, max_size=50000)\nTGT.build_vocab(train.trg, max_size=50000)",
"_____no_output_____"
],
[
"pad_idx = TGT.vocab.stoi[\"<blank>\"]\nprint(pad_idx)\nmodel = make_model(len(SRC.vocab), len(TGT.vocab), pad_idx, N=6)\nmodel_opt = get_opt(model)\nmodel.cuda()",
"1\n"
],
[
"criterion = LabelSmoothing(size=len(TGT.vocab), padding_idx=pad_idx, label_smoothing=0.1)\ncriterion.cuda()\nfor epoch in range(15):\n train_epoch(train_iter, model, criterion, model_opt)\n valid_epoch()",
"_____no_output_____"
],
[
"print(pad_idx)\nprint(len(SRC.vocab))",
"1\n50002\n"
],
[
"torch.save(model, \"/n/rush_lab/trans_ipython.pt\")",
"/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type EncoderDecoder. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type Encoder. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type EncoderLayer. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type MultiHeadedAttention. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type PositionwiseFeedForward. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type SublayerConnection. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type LayerNorm. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type Decoder. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type DecoderLayer. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type Embeddings. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type PositionalEncoding. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n/n/home00/srush/.conda/envs/py3/lib/python3.6/site-packages/torch/serialization.py:158: UserWarning: Couldn't retrieve source code for container of type Generator. It won't be checked for correctness upon loading.\n \"type \" + obj.__name__ + \". It won't be checked \"\n"
],
[
"#weight = torch.ones(len(TGT.vocab))\n#weight[pad_idx] = 0\n#criterion = nn.NLLLoss(size_average=False, weight=weight.cuda())\ncriterion = LabelSmoothing(size=len(TGT.vocab), padding_idx=pad_idx, label_smoothing=0.1)\ncriterion.cuda()\nfor epoch in range(15):\n train_epoch(train_iter, model, criterion, model_opt)",
"1 3.269582842476666 0.0005377044714644026\n101 3.300532897672383 0.0005726430336128369\n201 3.3047672072425485 0.0006075815957612711\n301 2.7151080842595547 0.0006425201579097052\n401 2.6975380268413574 0.0006774587200581396\n501 3.051631387323141 0.0007123972822065737\n601 2.554425454698503 0.000747335844355008\n701 2.6254820519825444 0.0007822744065034422\n801 2.868743653933052 0.0008172129686518764\n901 2.5978208642918617 0.0008521515308003106\n1001 2.5955790174775757 0.0008870900929487448\n1101 2.6764775353949517 0.000922028655097179\n1201 2.464000296778977 0.0009569672172456132\n1301 2.0503073083236814 0.0009919057793940475\n1401 2.295472824771423 0.0010268443415424816\n1501 2.245281406212598 0.0010617829036909158\n1601 2.2577588511630893 0.00109672146583935\n1701 2.2232908592559397 0.0011316600279877844\n1801 2.357596361427568 0.0011665985901362186\n1901 2.121352154412307 0.0012015371522846527\n2001 2.5742998471250758 0.001236475714433087\n2101 2.2518509055953473 0.0012714142765815214\n2201 2.2251326659170445 0.0013063528387299553\n2301 2.078994876006618 0.0013412914008783896\n2401 2.068276036065072 0.001376229963026824\n2501 2.31435151558253 0.0013907788851585368\n2601 1.9106871648691595 0.0013738752565588634\n2701 2.183084836578928 0.0013575733592730722\n2801 2.4668076275847852 0.0013418383196400342\n2901 1.963176985620521 0.0013266380295186675\n3001 2.2140520309330896 0.0013119428705609764\n3101 2.6989458349489723 0.0012977254713568687\n3201 2.1293521663174033 0.0012839604929174666\n3301 2.1402786187827587 0.0012706244386700126\n3401 2.041781216394156 0.0012576954857216498\n3501 2.051893091876991 0.0012451533346344698\n3601 1.5498304846696556 0.001232979075358713\n3701 2.763939742697403 0.001221155067309524\n3801 2.7611468499198963 0.0012096648318570434\n3901 1.7321470333263278 0.0011984929557393293\n4001 2.139603299088776 0.0011876250041103701\n4101 2.1966493157087825 0.0011770474421074978\n4201 2.0962203710805625 0.0011667475639689723\n4301 1.9717675620922819 0.0011567134288575545\n4401 2.097687987901736 0.0011469338026529508\n4501 1.9319786678534001 0.001137398105067946\n4601 1.8846281475271098 0.0011280963615221983\n4701 1.9817245414596982 0.0011190191592759865\n4801 1.7659185670199804 0.0011101576073853326\n4901 2.188665813198895 0.0011015033000912066\n5001 2.1391192222399695 0.0010930482833001135\n5101 1.8125874139368534 0.0010847850238522342\n5201 1.6616800595074892 0.0010767063813072288\n5301 1.6544548005331308 0.0010688055820075176\n5401 1.9542939933016896 0.0010610761952049212\n5501 2.218412609123334 0.0010535121110594244\n5601 1.838119359650591 0.001046107520339004\n5701 1.892627771012485 0.0010388568956672375\n5801 2.2462481096954434 0.0010317549741811346\n5901 1.4471426841337234 0.0010247967414755423\n6001 1.9312338004237972 0.0010179774167228303\n6101 1.7303275546291843 0.001011292438867507\n6201 1.8833909621462226 0.0010047374538051973\n6301 1.8943474531406537 0.0009983083024640838\n"
],
[
"1 10.825187489390373 6.987712429686844e-07\n101 9.447168171405792 3.56373333914029e-05\n201 7.142856806516647 7.057589553983712e-05\n301 6.237934365868568 0.00010551445768827134\n401 5.762486848048866 0.00014045301983670557\n501 5.415792358107865 0.00017539158198513977\n601 5.081815680023283 0.000210330144133574\n701 4.788327748770826 0.00024526870628200823\n801 4.381739928154275 0.0002802072684304424\n901 4.55433791608084 0.00031514583057887664\n1001 4.911875109748507 0.0003500843927273108\n1101 4.0579032292589545 0.0003850229548757451\n1201 4.2276234351193125 0.0004199615170241793\n1301 3.932735869428143 0.00045490007917261356\n1401 3.8179439397063106 0.0004898386413210477\n1501 3.3608515430241823 0.000524777203469482\n1601 3.832796103321016 0.0005597157656179162\n1701 2.907085266895592 0.0005946543277663504\n1801 3.5280659823838505 0.0006295928899147847\n1901 2.895841649500653 0.0006645314520632189\n2001 3.273784235585481 0.000699470014211653\n2101 3.181488689899197 0.0007344085763600873\n2201 3.4151616653980454 0.0007693471385085215\n2301 3.4343731447588652 0.0008042857006569557\n2401 3.0505455391539726 0.0008392242628053899\n2501 2.8089329147478566 0.0008741628249538242\n2601 2.7827929875456903 0.0009091013871022583\n2701 2.4428516102489084 0.0009440399492506926\n2801 2.4015486147254705 0.0009789785113991267\n2901 2.3568112018401735 0.001013917073547561\n3001 2.6349758653668687 0.0010488556356959952\n3101 2.5981983028614195 0.0010837941978444295\n3201 2.666826274838968 0.0011187327599928637\n3301 3.0092043554177508 0.0011536713221412978\n3401 2.4580375660589198 0.0011886098842897321\n3501 2.586465588421561 0.0012235484464381662\n3601 2.5663993963389657 0.0012584870085866006\n3701 2.9430236657499336 0.0012934255707350347\n3801 2.464644919440616 0.001328364132883469\n3901 2.7124062888276512 0.0013633026950319032\n4001 2.646443709731102 0.0013971932312809247\n4101 2.7294750874862075 0.001380057517579748\n4201 2.1295202329056337 0.0013635372009002666\n4301 2.596563663915731 0.001347596306985731\n4401 2.1265982036820787 0.0013322017384983986\n4501 2.3880532500334084 0.0013173229858148\n4601 2.6129120760888327 0.0013029318725783852\n4701 2.2873719420749694 0.001289002331178292\n4801 2.4949760700110346 0.0012755102040816328\n4901 2.496607314562425 0.001262433067573089\n5001 2.1889712483389303 0.0012497500749750088\n5101 1.8677761815488338 0.0012374418168536253\n5201 2.2992054556962103 0.0012254901960784316\n5301 2.664361578106707 0.0012138783159049418\n5401 2.705850490485318 0.0012025903795063202\n5501 2.581445264921058 0.0011916115995949978\n5601 2.2480602325085783 0.0011809281169581616\n5701 1.9289666265249252 0.0011705269268863989\n5801 2.4863578918157145 0.0011603958126073107\n5901 2.632946971571073 0.0011505232849492607\n6001 2.496141305891797 0.0011408985275576757\n6101 2.6422974687084206 0.0011315113470699342\n6201 2.448802186456305 0.0011223521277270118",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb93f881d8c27aa94fa117fd28e745579943c4dc
| 95,929 |
ipynb
|
Jupyter Notebook
|
.ipynb_checkpoints/COVID and Ontario Licensed Child Care-checkpoint.ipynb
|
donsohn/COVID-for-Toronto-Licensed-Child-Care
|
139907e5623b4dff78a49239949412559c5c66e1
|
[
"CNRI-Python"
] | null | null | null |
.ipynb_checkpoints/COVID and Ontario Licensed Child Care-checkpoint.ipynb
|
donsohn/COVID-for-Toronto-Licensed-Child-Care
|
139907e5623b4dff78a49239949412559c5c66e1
|
[
"CNRI-Python"
] | null | null | null |
.ipynb_checkpoints/COVID and Ontario Licensed Child Care-checkpoint.ipynb
|
donsohn/COVID-for-Toronto-Licensed-Child-Care
|
139907e5623b4dff78a49239949412559c5c66e1
|
[
"CNRI-Python"
] | null | null | null | 65.614911 | 25,236 | 0.685163 |
[
[
[
"# COVID and Ontario Licensed Child Care\n",
"_____no_output_____"
]
],
[
[
"import pandas as pd\nimport datetime\nimport io\nfrom io import StringIO\nimport os\nimport requests\nimport urllib.request\nimport time\nfrom bs4 import BeautifulSoup\n\n%matplotlib inline\n# import naming conventions \nimport numpy as np\nimport matplotlib.pyplot as plt ",
"_____no_output_____"
],
[
"url = 'https://data.ontario.ca/dataset/5bf54477-6147-413f-bab0-312f06fcb388/resource/eee282d3-01e6-43ac-9159-4ba694757aea/download/lccactivecovid.csv'\nresponse = requests.get(url)\ns = requests.get(url).text\ncovid_df = pd.read_csv(StringIO(s))",
"_____no_output_____"
]
],
[
[
"## Toronto (new cases per date)",
"_____no_output_____"
]
],
[
[
"covid_tor_df = covid_df.loc[(covid_df['municipality'] == 'Toronto')]\ncovid_tor_df.set_index('collected_date')\ncovid_sum_tor = covid_tor_df.groupby('collected_date')['total_confirmed_cases'].sum().to_frame(name='sum')\ncovid_sum_tor\ncovid_sum_tor['sum'].plot(figsize=(10,5),kind='line', color='b')",
"_____no_output_____"
]
],
[
[
"## All municipalities in Ontario (inclusive of Toronto; new cases per date)",
"_____no_output_____"
]
],
[
[
"covid_all_df = covid_df\ncovid_all_df.set_index('collected_date')\ncovid_sum_all = covid_all_df.groupby('collected_date')['total_confirmed_cases'].sum().to_frame(name='sum')\ncovid_sum_all['sum'].plot(figsize=(10,5),kind='line', color='r')",
"_____no_output_____"
],
[
"covid_tor_df.head()\n",
"_____no_output_____"
],
[
"# looking at Ys for friends\n#covid_tor_df[covid_tor_df['lcc_name'].str.contains(\"YMCA\")]\n#ymca_df = covid_tor_df[covid_tor_df['lcc_name'].str.contains(\"YMCA\")]\n#ymca_df.groupby([\"lcc_name\",\"reported_date\"])[\"total_confirmed_cases\"].sum()",
"_____no_output_____"
],
[
"# looking at Kids and Co for friends\n#kids_df = covid_tor_df[(covid_tor_df['lcc_name']).str.contains(\"Kids &\")]\n#kids_df.groupby([\"lcc_name\",\"reported_date\"])[\"total_confirmed_cases\"].sum()",
"_____no_output_____"
],
[
"covid_tor_df.lcc_name",
"_____no_output_____"
],
[
"from arcgis.gis import GIS\nfrom arcgis.geocoding import geocode\nfrom IPython.display import display",
"_____no_output_____"
],
[
"gis = GIS()\nmap1 = gis.map()\nmap1",
"_____no_output_____"
],
[
"# set the map's extent by geocoding the location\ntoronto = geocode(\"Toronto\")[0]\nmap1.extent = toronto['extent']",
"_____no_output_____"
],
[
"map1.zoom = 14",
"_____no_output_____"
],
[
"url = 'http://opendata.toronto.ca/childrens.services/licensed-child-care-centres/child-care.csv'\nresponse = requests.get(url)\ns = requests.get(url).text\nlcc_df = pd.read_csv(StringIO(s))",
"_____no_output_____"
],
[
"gis = GIS()\nmap1 = gis.map()\nmap1",
"_____no_output_____"
],
[
"# set the map's extent by geocoding the location\ntoronto = geocode(\"Toronto\")[0]\nmap1.extent = toronto['extent']\n",
"_____no_output_____"
],
[
"map1.zoom = 14",
"_____no_output_____"
],
[
"lcc_df = lcc_df[['STR_NO', 'STREET','LOC_NAME','TOTSPACE','LONGITUDE','LATITUDE']]",
"_____no_output_____"
],
[
"lcc_df",
"_____no_output_____"
],
[
"lcc_df['ADDRESS'] = lcc_df['STR_NO']+ \" \" +lcc_df['STREET']",
"_____no_output_____"
],
[
"lcc_df = lcc_df[['LOC_NAME','ADDRESS','TOTSPACE','LONGITUDE','LATITUDE']]\nlcc_df.columns = ['Name','Address','Total Capacity','Longitude','Latitude']",
"_____no_output_____"
],
[
"lcc_df\n",
"_____no_output_____"
],
[
"lcc_df[lcc_df['LOC_NAME'].str.contains('Matthew-John Early Learning Centre')]",
"_____no_output_____"
],
[
"lcc_df.isna().sum()",
"_____no_output_____"
],
[
"lcc_df[lcc_df['Longitude'].isnull()]\n\n# NEED TO FIGURE OUT WHAT TO DO WITH THE 7 NAN",
"_____no_output_____"
],
[
"lcc_df = lcc_df[lcc_df['Longitude'].notna()]",
"_____no_output_____"
],
[
"lcc_df.info()",
"<class 'pandas.core.frame.DataFrame'>\nInt64Index: 1028 entries, 0 to 1034\nData columns (total 5 columns):\n # Column Non-Null Count Dtype \n--- ------ -------------- ----- \n 0 Name 1028 non-null object \n 1 Address 1028 non-null object \n 2 Total Capacity 1028 non-null int64 \n 3 Longitude 1028 non-null float64\n 4 Latitude 1028 non-null float64\ndtypes: float64(2), int64(1), object(2)\nmemory usage: 48.2+ KB\n"
],
[
"lcc_df = lcc_df.reset_index()\ndel lcc_df['index']",
"_____no_output_____"
]
],
[
[
"## what is wrong with row 1019 - why stops here\n## keep only first 1018 rows out of 1029 for now\n\nuse this later\nhttps://developers.arcgis.com/python/guide/accessing-and-creating-content/",
"_____no_output_____"
]
],
[
[
"lcc_df2 = lcc_df.head(1018)\n\n",
"_____no_output_____"
],
[
"lcc1 = gis.content.import_data(lcc_df2)",
"_____no_output_____"
],
[
"map1.add_layer(lcc1)",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
]
] |
cb93f89371a97090855166d67919bff7a1f5ea4c
| 19,361 |
ipynb
|
Jupyter Notebook
|
docs/_downloads/dd583ee0bd16160dc842469ec717b9d3/audio_preprocessing_tutorial.ipynb
|
pleiades-s/PyTorch-tutorials-kr
|
3d749ea2fe67363b5d46340b742308b744fa0419
|
[
"BSD-3-Clause"
] | 1 |
2021-12-24T23:26:33.000Z
|
2021-12-24T23:26:33.000Z
|
docs/_downloads/dd583ee0bd16160dc842469ec717b9d3/audio_preprocessing_tutorial.ipynb
|
pleiades-s/PyTorch-tutorials-kr
|
3d749ea2fe67363b5d46340b742308b744fa0419
|
[
"BSD-3-Clause"
] | null | null | null |
docs/_downloads/dd583ee0bd16160dc842469ec717b9d3/audio_preprocessing_tutorial.ipynb
|
pleiades-s/PyTorch-tutorials-kr
|
3d749ea2fe67363b5d46340b742308b744fa0419
|
[
"BSD-3-Clause"
] | null | null | null | 47.687192 | 1,579 | 0.617014 |
[
[
[
"%matplotlib inline",
"_____no_output_____"
]
],
[
[
"\ntorchaudio Tutorial\n===================\n\nPyTorch is an open source deep learning platform that provides a\nseamless path from research prototyping to production deployment with\nGPU support.\n\nSignificant effort in solving machine learning problems goes into data\npreparation. ``torchaudio`` leverages PyTorch’s GPU support, and provides\nmany tools to make data loading easy and more readable. In this\ntutorial, we will see how to load and preprocess data from a simple\ndataset.\n\nFor this tutorial, please make sure the ``matplotlib`` package is\ninstalled for easier visualization.\n\n\n",
"_____no_output_____"
]
],
[
[
"import torch\nimport torchaudio\nimport matplotlib.pyplot as plt",
"_____no_output_____"
]
],
[
[
"Opening a file\n-----------------\n\n``torchaudio`` also supports loading sound files in the wav and mp3 format. We\ncall waveform the resulting raw audio signal.\n\n\n",
"_____no_output_____"
]
],
[
[
"filename = \"../_static/img/steam-train-whistle-daniel_simon-converted-from-mp3.wav\"\nwaveform, sample_rate = torchaudio.load(filename)\n\nprint(\"Shape of waveform: {}\".format(waveform.size()))\nprint(\"Sample rate of waveform: {}\".format(sample_rate))\n\nplt.figure()\nplt.plot(waveform.t().numpy())",
"_____no_output_____"
]
],
[
[
"When you load a file in ``torchaudio``, you can optionally specify the backend to use either \n`SoX <https://pypi.org/project/sox/>`_ or `SoundFile <https://pypi.org/project/SoundFile/>`_ \nvia ``torchaudio.set_audio_backend``. These backends are loaded lazily when needed.\n\n``torchaudio`` also makes JIT compilation optional for functions, and uses ``nn.Module`` where possible.\n\n",
"_____no_output_____"
],
[
"Transformations\n---------------\n\n``torchaudio`` supports a growing list of\n`transformations <https://pytorch.org/audio/transforms.html>`_.\n\n- **Resample**: Resample waveform to a different sample rate.\n- **Spectrogram**: Create a spectrogram from a waveform.\n- **GriffinLim**: Compute waveform from a linear scale magnitude spectrogram using \n the Griffin-Lim transformation.\n- **ComputeDeltas**: Compute delta coefficients of a tensor, usually a spectrogram.\n- **ComplexNorm**: Compute the norm of a complex tensor.\n- **MelScale**: This turns a normal STFT into a Mel-frequency STFT,\n using a conversion matrix.\n- **AmplitudeToDB**: This turns a spectrogram from the\n power/amplitude scale to the decibel scale.\n- **MFCC**: Create the Mel-frequency cepstrum coefficients from a\n waveform.\n- **MelSpectrogram**: Create MEL Spectrograms from a waveform using the\n STFT function in PyTorch.\n- **MuLawEncoding**: Encode waveform based on mu-law companding.\n- **MuLawDecoding**: Decode mu-law encoded waveform.\n- **TimeStretch**: Stretch a spectrogram in time without modifying pitch for a given rate.\n- **FrequencyMasking**: Apply masking to a spectrogram in the frequency domain.\n- **TimeMasking**: Apply masking to a spectrogram in the time domain.\n\nEach transform supports batching: you can perform a transform on a single raw \naudio signal or spectrogram, or many of the same shape.\n\nSince all transforms are ``nn.Modules`` or ``jit.ScriptModules``, they can be\nused as part of a neural network at any point.\n\n\n",
"_____no_output_____"
],
[
"To start, we can look at the log of the spectrogram on a log scale.\n\n\n",
"_____no_output_____"
]
],
[
[
"specgram = torchaudio.transforms.Spectrogram()(waveform)\n\nprint(\"Shape of spectrogram: {}\".format(specgram.size()))\n\nplt.figure()\nplt.imshow(specgram.log2()[0,:,:].numpy(), cmap='gray')",
"_____no_output_____"
]
],
[
[
"Or we can look at the Mel Spectrogram on a log scale.\n\n\n",
"_____no_output_____"
]
],
[
[
"specgram = torchaudio.transforms.MelSpectrogram()(waveform)\n\nprint(\"Shape of spectrogram: {}\".format(specgram.size()))\n\nplt.figure()\np = plt.imshow(specgram.log2()[0,:,:].detach().numpy(), cmap='gray')",
"_____no_output_____"
]
],
[
[
"We can resample the waveform, one channel at a time.\n\n\n",
"_____no_output_____"
]
],
[
[
"new_sample_rate = sample_rate/10\n\n# Since Resample applies to a single channel, we resample first channel here\nchannel = 0\ntransformed = torchaudio.transforms.Resample(sample_rate, new_sample_rate)(waveform[channel,:].view(1,-1))\n\nprint(\"Shape of transformed waveform: {}\".format(transformed.size()))\n\nplt.figure()\nplt.plot(transformed[0,:].numpy())",
"_____no_output_____"
]
],
[
[
"As another example of transformations, we can encode the signal based on\nMu-Law enconding. But to do so, we need the signal to be between -1 and\n1. Since the tensor is just a regular PyTorch tensor, we can apply\nstandard operators on it.\n\n\n",
"_____no_output_____"
]
],
[
[
"# Let's check if the tensor is in the interval [-1,1]\nprint(\"Min of waveform: {}\\nMax of waveform: {}\\nMean of waveform: {}\".format(waveform.min(), waveform.max(), waveform.mean()))",
"_____no_output_____"
]
],
[
[
"Since the waveform is already between -1 and 1, we do not need to\nnormalize it.\n\n\n",
"_____no_output_____"
]
],
[
[
"def normalize(tensor):\n # Subtract the mean, and scale to the interval [-1,1]\n tensor_minusmean = tensor - tensor.mean()\n return tensor_minusmean/tensor_minusmean.abs().max()\n\n# Let's normalize to the full interval [-1,1]\n# waveform = normalize(waveform)",
"_____no_output_____"
]
],
[
[
"Let’s apply encode the waveform.\n\n\n",
"_____no_output_____"
]
],
[
[
"transformed = torchaudio.transforms.MuLawEncoding()(waveform)\n\nprint(\"Shape of transformed waveform: {}\".format(transformed.size()))\n\nplt.figure()\nplt.plot(transformed[0,:].numpy())",
"_____no_output_____"
]
],
[
[
"And now decode.\n\n\n",
"_____no_output_____"
]
],
[
[
"reconstructed = torchaudio.transforms.MuLawDecoding()(transformed)\n\nprint(\"Shape of recovered waveform: {}\".format(reconstructed.size()))\n\nplt.figure()\nplt.plot(reconstructed[0,:].numpy())",
"_____no_output_____"
]
],
[
[
"We can finally compare the original waveform with its reconstructed\nversion.\n\n\n",
"_____no_output_____"
]
],
[
[
"# Compute median relative difference\nerr = ((waveform-reconstructed).abs() / waveform.abs()).median()\n\nprint(\"Median relative difference between original and MuLaw reconstucted signals: {:.2%}\".format(err))",
"_____no_output_____"
]
],
[
[
"Functional\n---------------\n\nThe transformations seen above rely on lower level stateless functions for their computations. \nThese functions are available under ``torchaudio.functional``. The complete list is available \n`here <https://pytorch.org/audio/functional.html>`_ and includes:\n\n- **istft**: Inverse short time Fourier Transform.\n- **gain**: Applies amplification or attenuation to the whole waveform.\n- **dither**: Increases the perceived dynamic range of audio stored at a\n particular bit-depth.\n- **compute_deltas**: Compute delta coefficients of a tensor.\n- **equalizer_biquad**: Design biquad peaking equalizer filter and perform filtering.\n- **lowpass_biquad**: Design biquad lowpass filter and perform filtering.\n- **highpass_biquad**:Design biquad highpass filter and perform filtering.\n\nFor example, let's try the `mu_law_encoding` functional:\n\n",
"_____no_output_____"
]
],
[
[
"mu_law_encoding_waveform = torchaudio.functional.mu_law_encoding(waveform, quantization_channels=256)\n\nprint(\"Shape of transformed waveform: {}\".format(mu_law_encoding_waveform.size()))\n\nplt.figure()\nplt.plot(mu_law_encoding_waveform[0,:].numpy())",
"_____no_output_____"
]
],
[
[
"You can see how the output fron ``torchaudio.functional.mu_law_encoding`` is the same as \nthe output from ``torchaudio.transforms.MuLawEncoding``.\n\nNow let's experiment with a few of the other functionals and visualize their output. Taking our \nspectogram, we can compute it's deltas:\n\n",
"_____no_output_____"
]
],
[
[
"computed = torchaudio.functional.compute_deltas(specgram, win_length=3)\nprint(\"Shape of computed deltas: {}\".format(computed.shape))\n\nplt.figure()\nplt.imshow(computed.log2()[0,:,:].detach().numpy(), cmap='gray')",
"_____no_output_____"
]
],
[
[
"We can take the original waveform and apply different effects to it.\n\n\n",
"_____no_output_____"
]
],
[
[
"gain_waveform = torchaudio.functional.gain(waveform, gain_db=5.0)\nprint(\"Min of gain_waveform: {}\\nMax of gain_waveform: {}\\nMean of gain_waveform: {}\".format(gain_waveform.min(), gain_waveform.max(), gain_waveform.mean()))\n\ndither_waveform = torchaudio.functional.dither(waveform)\nprint(\"Min of dither_waveform: {}\\nMax of dither_waveform: {}\\nMean of dither_waveform: {}\".format(dither_waveform.min(), dither_waveform.max(), dither_waveform.mean()))",
"_____no_output_____"
]
],
[
[
"Another example of the capabilities in ``torchaudio.functional`` are applying filters to our\nwaveform. Applying the lowpass biquad filter to our waveform will output a new waveform with \nthe signal of the frequency modified.\n\n",
"_____no_output_____"
]
],
[
[
"lowpass_waveform = torchaudio.functional.lowpass_biquad(waveform, sample_rate, cutoff_freq=3000)\n\nprint(\"Min of lowpass_waveform: {}\\nMax of lowpass_waveform: {}\\nMean of lowpass_waveform: {}\".format(lowpass_waveform.min(), lowpass_waveform.max(), lowpass_waveform.mean()))\n\nplt.figure()\nplt.plot(lowpass_waveform.t().numpy())",
"_____no_output_____"
]
],
[
[
"We can also visualize a waveform with the highpass biquad filter.\n\n\n",
"_____no_output_____"
]
],
[
[
"highpass_waveform = torchaudio.functional.highpass_biquad(waveform, sample_rate, cutoff_freq=2000)\n\nprint(\"Min of highpass_waveform: {}\\nMax of highpass_waveform: {}\\nMean of highpass_waveform: {}\".format(highpass_waveform.min(), highpass_waveform.max(), highpass_waveform.mean()))\n\nplt.figure()\nplt.plot(highpass_waveform.t().numpy())",
"_____no_output_____"
]
],
[
[
"Migrating to torchaudio from Kaldi\n----------------------------------\n\nUsers may be familiar with\n`Kaldi <http://github.com/kaldi-asr/kaldi>`_, a toolkit for speech\nrecognition. ``torchaudio`` offers compatibility with it in\n``torchaudio.kaldi_io``. It can indeed read from kaldi scp, or ark file\nor streams with:\n\n- read_vec_int_ark\n- read_vec_flt_scp\n- read_vec_flt_arkfile/stream\n- read_mat_scp\n- read_mat_ark\n\n``torchaudio`` provides Kaldi-compatible transforms for ``spectrogram``,\n``fbank``, ``mfcc``, and ``resample_waveform with the benefit of GPU support, see\n`here <compliance.kaldi.html>`__ for more information.\n\n\n",
"_____no_output_____"
]
],
[
[
"n_fft = 400.0\nframe_length = n_fft / sample_rate * 1000.0\nframe_shift = frame_length / 2.0\n\nparams = {\n \"channel\": 0,\n \"dither\": 0.0,\n \"window_type\": \"hanning\",\n \"frame_length\": frame_length,\n \"frame_shift\": frame_shift,\n \"remove_dc_offset\": False,\n \"round_to_power_of_two\": False,\n \"sample_frequency\": sample_rate,\n}\n\nspecgram = torchaudio.compliance.kaldi.spectrogram(waveform, **params)\n\nprint(\"Shape of spectrogram: {}\".format(specgram.size()))\n\nplt.figure()\nplt.imshow(specgram.t().numpy(), cmap='gray')",
"_____no_output_____"
]
],
[
[
"We also support computing the filterbank features from waveforms,\nmatching Kaldi’s implementation.\n\n\n",
"_____no_output_____"
]
],
[
[
"fbank = torchaudio.compliance.kaldi.fbank(waveform, **params)\n\nprint(\"Shape of fbank: {}\".format(fbank.size()))\n\nplt.figure()\nplt.imshow(fbank.t().numpy(), cmap='gray')",
"_____no_output_____"
]
],
[
[
"You can create mel frequency cepstral coefficients from a raw audio signal\nThis matches the input/output of Kaldi’s compute-mfcc-feats.\n\n\n",
"_____no_output_____"
]
],
[
[
"mfcc = torchaudio.compliance.kaldi.mfcc(waveform, **params)\n\nprint(\"Shape of mfcc: {}\".format(mfcc.size()))\n\nplt.figure()\nplt.imshow(mfcc.t().numpy(), cmap='gray')",
"_____no_output_____"
]
],
[
[
"Available Datasets\n-----------------\n\nIf you do not want to create your own dataset to train your model, ``torchaudio`` offers a\nunified dataset interface. This interface supports lazy-loading of files to memory, download \nand extract functions, and datasets to build models.\n\nThe datasets ``torchaudio`` currently supports are:\n\n- **VCTK**: Speech data uttered by 109 native speakers of English with various accents\n (`Read more here <https://homepages.inf.ed.ac.uk/jyamagis/page3/page58/page58.html>`_).\n- **Yesno**: Sixty recordings of one individual saying yes or no in Hebrew; each\n recording is eight words long (`Read more here <https://www.openslr.org/1/>`_).\n- **Common Voice**: An open source, multi-language dataset of voices that anyone can use\n to train speech-enabled applications (`Read more here <https://voice.mozilla.org/en/datasets>`_).\n- **LibriSpeech**: Large-scale (1000 hours) corpus of read English speech (`Read more here <http://www.openslr.org/12>`_).\n\n\n",
"_____no_output_____"
]
],
[
[
"yesno_data = torchaudio.datasets.YESNO('./', download=True)\n\n# A data point in Yesno is a tuple (waveform, sample_rate, labels) where labels is a list of integers with 1 for yes and 0 for no.\n\n# Pick data point number 3 to see an example of the the yesno_data:\nn = 3\nwaveform, sample_rate, labels = yesno_data[n]\n\nprint(\"Waveform: {}\\nSample rate: {}\\nLabels: {}\".format(waveform, sample_rate, labels))\n\nplt.figure()\nplt.plot(waveform.t().numpy())",
"_____no_output_____"
]
],
[
[
"Now, whenever you ask for a sound file from the dataset, it is loaded in memory only when you ask for it.\nMeaning, the dataset only loads and keeps in memory the items that you want and use, saving on memory.\n\n\n",
"_____no_output_____"
],
[
"Conclusion\n----------\n\nWe used an example raw audio signal, or waveform, to illustrate how to\nopen an audio file using ``torchaudio``, and how to pre-process,\ntransform, and apply functions to such waveform. We also demonstrated how\nto use familiar Kaldi functions, as well as utilize built-in datasets to \nconstruct our models. Given that ``torchaudio`` is built on PyTorch,\nthese techniques can be used as building blocks for more advanced audio\napplications, such as speech recognition, while leveraging GPUs.\n\n\n",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
]
] |
cb93fbcb829923555cc9e44dd3b03aa981b92d0e
| 180,848 |
ipynb
|
Jupyter Notebook
|
dataStats.ipynb
|
fredrike/dataverse-social_network
|
f0ec845cdee4f2d47c2d2c65aa468632df6a33f4
|
[
"BSD-2-Clause"
] | null | null | null |
dataStats.ipynb
|
fredrike/dataverse-social_network
|
f0ec845cdee4f2d47c2d2c65aa468632df6a33f4
|
[
"BSD-2-Clause"
] | null | null | null |
dataStats.ipynb
|
fredrike/dataverse-social_network
|
f0ec845cdee4f2d47c2d2c65aa468632df6a33f4
|
[
"BSD-2-Clause"
] | null | null | null | 32.887434 | 131 | 0.491662 |
[
[
[
"import numpy as np\nimport pandas as pd\nfrom pandas.io.json import json_normalize\n\nfrom IPython.core.display import display, HTML\nimport locale\nlocale.setlocale(locale.LC_ALL, 'en_US')\n\nfrom concurrent.futures import ProcessPoolExecutor\nimport multiprocessing\nfrom tqdm import tqdm_notebook as tqdm\n\ntimeout = 3600 # Timeout for building bipartite graph.\nnum_processes = int(multiprocessing.cpu_count())\n\nDATA_DOI = 'doi:10.7910/DVN/DCBDEP'\nDATAFILE_URL = 'https://dataverse.harvard.edu/api/access/datafile/{id}?format=original'\nDATASETS_URL = 'https://dataverse.harvard.edu/api/datasets/:persistentId/versions/1/files?persistentId={}'.format(DATA_DOI)\n",
"_____no_output_____"
],
[
"import time\nimport graph_tool as gt\nfrom graph_tool.stats import vertex_average\n\ndef buildGraph(df, debug=False, timeout=900):\n if debug: tic = time.time()\n b = gt.Graph(directed=False)\n b.add_edge_list(df[['post', 'user']].values)\n if debug:\n print(\"Bipartite nodes: %d, vertices: %d\" % (b.num_vertices(), b.num_edges()))\n print(\"Average degree:\", vertex_average(b, \"total\"))\n print(\"Time\", time.time() - tic)\n tic = time.time()\n user_count = df.user.value_counts()\n p = projected_graph(b, user_count[user_count >= 2].index, timeout=timeout)\n if debug:\n print(\"Projection nodes: %d, vertices: %d\" % (p.num_vertices(), p.num_edges()))\n print(\"Average degree:\", vertex_average(p, \"total\"))\n print(\"Time\", time.time() - tic)\n\n # degree_sequence = sorted(p.get_out_degrees(p.get_vertices()), reverse=True)\n\n del b\n return p\n\n\ndef projected_graph(b, nodes, timeout=900):\n tic = time.time()\n p = gt.Graph(directed=False)\n try:\n for u in nodes:\n nbrs2 = [n for nbr in b.vertex(u).all_neighbours() for n in b.vertex(nbr).all_neighbours()]\n p.add_edge_list((u, n) for n in nbrs2)\n if time.time() - tic > timeout: raise\n except:\n p = gt.Graph(directed=False)\n return p\n\n\ndef extractUserInteractions(uri):\n df = pd.read_csv(uri, index_col=False, engine='c',\n dtype={'post': np.int64, 'user': np.float64, 'type': str})\n df = df[df['type'] == 'C']\n df.dropna(inplace=True)\n _map = df['user'].drop_duplicates().reset_index(drop=True)\n _map = pd.Series(_map.index.values, index=_map)\n df['user'] = df['user'].map(_map)\n _map = df['post'].drop_duplicates().reset_index(drop=True)\n _map = pd.Series(_map.index.values, index=_map)\n _map += len(df['user'].unique()) + 10\n df['post'] = df['post'].map(_map)\n \n df['user'] = df['user'].astype(np.uint32, casting='safe')\n df['post'] = df['post'].astype(np.uint32, casting='safe')\n # print(\"Read %d interactions for the page %d\" % (len(df), page))\n return df",
"_____no_output_____"
],
[
"def download_datafile(dataFile_id, overwrite=True):\n import shutil\n import requests\n import os.path\n\n try:\n r = requests.get(DATAFILE_URL.format(id=dataFile_id), stream=True, allow_redirects=True, timeout=1)\n filename = r.headers.get('content-disposition')\n filename = filename.split(\"'\")[2] if len(filename.split(\"'\")) > 2 else filename.split('\"')[1]\n try:\n if overwrite or not os.path.isfile('data/{}'.format(filename)):\n with open('data/{}'.format(filename), 'wb') as fd:\n shutil.copyfileobj(r.raw, fd)\n return (filename, dataFile_id)\n except:\n if os.path.isfile('data/{}'.format(filename)):\n os.remove('data/{}'.format(filename))\n except:\n pass\n\n return (None, dataFile_id)\n\n\ndef process_file(filename, timeout=900):\n try:\n if filename in ['00__combinedPageInteractions.csv']:\n raise\n df = pd.read_csv('data/' + filename, index_col=False,\n dtype={'post': np.int64, 'user': np.float64, 'type': str})\n types = df.type.value_counts()\n graph = buildGraph(df[df['type'] == 'C'].dropna(), debug=False, timeout=timeout)\n return {filename.split('.')[0]: [len(df.post.unique()), len(df.user.unique()), \n types['C'] if 'C' in types.index else 0, \n types['L'] if 'L' in types.index else 0,\n graph.num_edges() if graph.num_edges() else np.nan,\n graph.num_vertices() if graph.num_vertices() else np.nan]}\n except:\n pass\n return {filename.split('.')[0]: []}",
"_____no_output_____"
],
[
"files = json_normalize(data=pd.read_json(DATASETS_URL)['data']).set_index('label')\nfile_ids = files['dataFile.id']\nfilenames = set()\n\nfor _ in range(5): # Try five times to download all files from Dataverse.\n # Process the rows in chunks in parallel\n with ProcessPoolExecutor(num_processes) as pool:\n fn = list(tqdm(pool.map(download_datafile, file_ids, chunksize=1), total=len(file_ids)))\n filenames = filenames.union(set(x[0] for x in fn if x[0]))\n file_ids = [x[1] for x in fn if not x[0]]\n if len(file_ids) == 0:\n break\n",
"_____no_output_____"
],
[
"len(filenames)",
"_____no_output_____"
],
[
"with ProcessPoolExecutor(4) as pool:\n pages = list(tqdm(pool.map(process_file, filenames, chunksize=1), total=len(filenames)))\n \nstats = pd.DataFrame.from_dict(dict((key,d[key]) for d in pages for key in d), orient='index',\n columns=[\"Posts\", \"Users\", \"Comments\", \"Likes\", 'Edges', 'Nodes'])\n",
"_____no_output_____"
],
[
"stats.to_pickle('stats_01.pkl')",
"_____no_output_____"
],
[
"print(len(stats.drop(['00__combinedPageInteractions'], errors='ignore')))\ns = (stats.drop(['00__combinedPageInteractions'], errors='ignore').describe()\n .T[['mean','std','min','25%', '50%', '75%', 'max']])\n\ns.columns.name = 'Metric'\ns.index.name = None\ns['Sum'] = stats.sum()\ns.rename(columns={'mean': \"Mean\",'std': 'Std.','min': 'Min','25%': '$Q1$', \n '50%': 'Median', '75%': '$Q3$', 'max': 'Max'},\n inplace=True)\n\ns.to_latex('article/pageStats.tex', bold_rows=True, escape=False,\n float_format=lambda x: \"$%s$\" % locale.format(\"%d\", x, grouping=True))\n\ndisplay(HTML(s.to_html(float_format=lambda x: locale.format(\"%d\", x, grouping=True))))\n",
"160\n"
],
[
"names = pd.read_csv(\"data/names.csv\", dtype={0: str, 1: str}).set_index('Page')\n\ns = stats.copy().drop(['00__combinedPageInteractions'], errors='ignore')\ns['Id'] = s.index\ns['Edges'] = [pd.np.nan if x == 0.0 else x for x in s['Edges']]\ns['Nodes'] = [pd.np.nan if x == 0.0 else x for x in s['Nodes']]\n\nprint(len(s))\n\ndef name_mapping(x, latex_escape=False):\n if type(names.loc[x.Id].Name) is float: \n ret = str(x.Id)\n else:\n if latex_escape:\n ret = (str(names.loc[x.Id].Name).replace('\\\\', '\\\\textbackslash ')\n .replace('_', '\\\\_')\n .replace('%', '\\\\%').replace('$', '\\\\$')\n .replace('#', '\\\\#').replace('{', '\\\\{')\n .replace('}', '\\\\}').replace('~', '\\\\textasciitilde ')\n .replace('^', '\\\\textasciicircum ')\n .replace('&', '\\\\&'))\n else:\n ret = str(names.loc[x.Id].Name)\n \n if pd.np.isnan(x['Edges']):\n ret = ret + \"$^\\ddagger$\"\n\n return ret\n\n\"\"\"Save table as latex file.\"\"\"\ns['Name'] = s.apply(lambda x: name_mapping(x, True), axis=1)\n(s.fillna(np.inf).sort_values(['Edges','Nodes', 'Users'])\n .to_latex('article/allPages.tex', escape=False, longtable=True, index=False,\n columns=['Name', 'Id', 'Posts', 'Users', 'Comments', 'Likes','Edges','Nodes'],\n float_format=lambda x: \"\" if x == np.inf else locale.format(\"%d\", x, grouping=True)))\n\n\"\"\"Display table (after remapping the Name).\"\"\"\ns['Name'] = s.apply(name_mapping, axis=1)\ndisplay(HTML(s.fillna(pd.np.inf).sort_values(['Edges','Nodes', 'Users'])\n .to_html(index=False, \n columns=['Name', 'Id', 'Posts', 'Users', 'Comments', 'Likes','Edges','Nodes'],\n float_format=lambda x: \"\" if x == np.inf else locale.format(\"%d\", x, grouping=True))))\n\n",
"160\n"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb94108b1ef1afa1027ada257f28331b086c9d28
| 57,298 |
ipynb
|
Jupyter Notebook
|
HeroesOfPymoli/02-Homework_04-Pandas_Instructions_HeroesOfPymoli_HeroesOfPymoli_starter.ipynb
|
Sanskruti1982/pandas-challenge
|
7d187980ab4267a522335add26fa99330069a3be
|
[
"ADSL"
] | null | null | null |
HeroesOfPymoli/02-Homework_04-Pandas_Instructions_HeroesOfPymoli_HeroesOfPymoli_starter.ipynb
|
Sanskruti1982/pandas-challenge
|
7d187980ab4267a522335add26fa99330069a3be
|
[
"ADSL"
] | null | null | null |
HeroesOfPymoli/02-Homework_04-Pandas_Instructions_HeroesOfPymoli_HeroesOfPymoli_starter.ipynb
|
Sanskruti1982/pandas-challenge
|
7d187980ab4267a522335add26fa99330069a3be
|
[
"ADSL"
] | null | null | null | 60.250263 | 1,585 | 0.591295 |
[
[
[
"### Note\n* Instructions have been included for each segment. You do not have to follow them exactly, but they are included to help you think through the steps.",
"_____no_output_____"
]
],
[
[
"# Dependencies and Setup\nimport pandas as pd\n\n# File to Load (Remember to Change These)\nfile_to_load = \"purchase_data.csv\"\n\n# Read Purchasing File and store into Pandas data frame\npurchase_data_df = pd.read_csv(file_to_load)",
"_____no_output_____"
]
],
[
[
"## Player Count",
"_____no_output_____"
],
[
"* Display the total number of players\n",
"_____no_output_____"
]
],
[
[
"#counting length of the unique values in SN column of the data frame\ncount = len(purchase_data_df[\"SN\"].unique())\n#Putting the count into a dictonary of list\ncount1 = [{\"Total Players\" : count}]\n#Creating a data frame to display the total number of players\nnoofplayers_df = pd.DataFrame(count1)\n#printing the data frame\nnoofplayers_df",
"_____no_output_____"
]
],
[
[
"## Purchasing Analysis (Total)",
"_____no_output_____"
],
[
"* Run basic calculations to obtain number of unique items, average price, etc.\n\n\n* Create a summary data frame to hold the results\n\n\n* Optional: give the displayed data cleaner formatting\n\n\n* Display the summary data frame\n",
"_____no_output_____"
]
],
[
[
"#Calculating unique items, average price, number of purchases and total revenue\nuniitems = len(purchase_data_df[\"Item ID\"].unique())\navgprice = purchase_data_df[\"Price\"].mean(axis=0)\nnoofpurchases = purchase_data_df[\"Purchase ID\"].count()\ntotalrevenue = purchase_data_df[\"Price\"].sum()\n#Inserting all new calculations into a summary table\nsummary_df = [{\"Number of Unique Items\": uniitems, \"Average Price\": avgprice, \"Number of Purchases\" : noofpurchases, \"Total Revenue\" : totalrevenue}]\nsummary_df = pd.DataFrame(summary_df)\n#Formatting the Average Price and Total Revenue object to include $ sign and decimal points. \nsummary_df[\"Average Price\"] = summary_df[\"Average Price\"].map(\"${:.2f}\".format)\nsummary_df[\"Total Revenue\"] = summary_df[\"Total Revenue\"].map(\"${:,}\".format)\n#Aligning the data to align left\nsummary_df = summary_df.style.set_properties(**{'text-align':'left'})\n#Prining the data frame\nsummary_df",
"_____no_output_____"
]
],
[
[
"## Gender Demographics",
"_____no_output_____"
],
[
"* Percentage and Count of Male Players\n\n\n* Percentage and Count of Female Players\n\n\n* Percentage and Count of Other / Non-Disclosed\n\n\n",
"_____no_output_____"
]
],
[
[
"#Grouping the data frame by Gender and getting unique count of column SN\ngroupedpurchase = purchase_data_df.groupby('Gender').agg({ \"SN\": \"nunique\"})\n#Inserting new column called Total Count\ngroupedpurchase.columns=[\"Total Count\"]\n#Inserting new column called Percentage of Players and calculating the percentange of total players by Gender\ngroupedpurchase[\"Percentage of Players\"] = (groupedpurchase[\"Total Count\"]/purchase_data_df[\"SN\"].nunique())*100\n#Dropping the name of the index\ngroupedpurchase.index.name=None\n#Formatting the column Percentage of Players to have % sign and two decimal points\ngroupedpurchase[\"Percentage of Players\"] = groupedpurchase[\"Percentage of Players\"].map(\"{:.2f}%\".format)\n#Sorting the data frame by Total Count\ngroupedpurchase = groupedpurchase.sort_values(\"Total Count\", ascending=False)\n#Aligning all the text to be left indented\ngroupedpurchase = groupedpurchase.style.set_properties(**{'text-align':'left'})\n#Setting all table style to be left indented which causes the index to be left indented as well\ngroupedpurchase.set_table_styles([dict(selector='th', props=[('text-align', 'left')])])\n#Printing the table\ngroupedpurchase",
"_____no_output_____"
]
],
[
[
"\n## Purchasing Analysis (Gender)",
"_____no_output_____"
],
[
"* Run basic calculations to obtain purchase count, avg. purchase price, avg. purchase total per person etc. by gender\n\n\n\n\n* Create a summary data frame to hold the results\n\n\n* Optional: give the displayed data cleaner formatting\n\n\n* Display the summary data frame",
"_____no_output_____"
]
],
[
[
"#Grouping the data frame by Gender and getting unique count of column SN, count of Purchase Id and mean & sum of Price\npurchanalysis = purchase_data_df.groupby('Gender').agg({\"Purchase ID\": \"count\", \"Price\": [\"mean\", \"sum\"], \"SN\": \"nunique\"})\n#Inserting new columns\npurchanalysis.columns=[\"Purchase Count\", \"Average Purchase Price\", \"Total Purchase Value\", \"SN\"]\n#Calculating Avg Total Purchase per Person and inserting the column\npurchanalysis[\"Avg Total Purchase per Person\"] = purchanalysis[\"Total Purchase Value\"]/purchanalysis[\"SN\"]\n#Deleting column SN\ndel purchanalysis [\"SN\"]\n#Formatting Average Purchase Price, Total Purchase Value and Avg Total Purchase per Person to have $ sign and two decimal points\npurchanalysis[\"Average Purchase Price\"] = purchanalysis[\"Average Purchase Price\"].map(\"${:.2f}\".format)\npurchanalysis[\"Total Purchase Value\"] = purchanalysis[\"Total Purchase Value\"].map(\"${:.2f}\".format)\npurchanalysis[\"Avg Total Purchase per Person\"] = purchanalysis[\"Avg Total Purchase per Person\"].map(\"${:.2f}\".format)\n#Aligning all the text to be left indented\npurchanalysis = purchanalysis.style.set_properties(**{'text-align':'left'})\n#Setting all table style to be left indented which causes the index to be left indented as well\npurchanalysis.set_table_styles([dict(selector='th', props=[('text-align', 'left')])])\n\n",
"_____no_output_____"
]
],
[
[
"## Age Demographics",
"_____no_output_____"
],
[
"* Establish bins for ages\n\n\n* Categorize the existing players using the age bins. Hint: use pd.cut()\n\n\n* Calculate the numbers and percentages by age group\n\n\n* Create a summary data frame to hold the results\n\n\n* Optional: round the percentage column to two decimal points\n\n\n* Display Age Demographics Table\n",
"_____no_output_____"
]
],
[
[
"#Establishing bins for ages\nbins = [0, 9, 14, 19, 24, 29, 34, 39, 45]\n#Creating labels for the bin\ngroup_labels = [\"<10\",\"10-14\",\"15-19\",\"20-24\",\"25-29\", \"30-34\", \"35-39\", \"40+\"]\n#Using cut method to slice the data by Age, applying group labels and to include the lowest range \npd.cut(purchase_data_df[\"Age\"], bins, labels=group_labels).head()\npurchase_data_df[\"Age Ranges\"] = pd.cut(purchase_data_df[\"Age\"], bins, labels=group_labels, include_lowest=True)\n#Grouping the data frame by new column and include unique counts of column SN to get total count by age\nAgerange = purchase_data_df.groupby(\"Age Ranges\").agg({\"SN\": \"nunique\"})\n#Inser the new column\nAgerange.columns=[\"Total Count\"]\n#Calculate percentage of players in each age range and insert in new column of percentage of playes\nAgerange[\"Percentage of Players\"] = (Agerange[\"Total Count\"]/Agerange[\"Total Count\"].sum())*100\n#Formatting Percentage of Players to display % and two decimal points\nAgerange[\"Percentage of Players\"] = Agerange[\"Percentage of Players\"].map(\"{:.2f}%\".format)\n#Dropping the index name\nAgerange.index.name=None\n#Aligning all the text to be left indented\nAgerange = Agerange.style.set_properties(**{'text-align':'left'})\n#Setting all table style to be left indented which causes the index to be left indented as well\nAgerange.set_table_styles([dict(selector='th', props=[('text-align', 'left')])])",
"_____no_output_____"
]
],
[
[
"## Purchasing Analysis (Age)",
"_____no_output_____"
],
[
"* Bin the purchase_data data frame by age\n\n\n* Run basic calculations to obtain purchase count, avg. purchase price, avg. purchase total per person etc. in the table below\n\n\n* Create a summary data frame to hold the results\n\n\n* Optional: give the displayed data cleaner formatting\n\n\n* Display the summary data frame",
"_____no_output_____"
]
],
[
[
"#Grouping the data frame by alread established bin and group labels and getting unique count of column SN, count of Purchase Id and mean & sum of Price\nAgerange1 = purchase_data_df.groupby(\"Age Ranges\").agg({\"Purchase ID\": \"count\", \"Price\": [\"mean\", \"sum\"], \"SN\": \"nunique\"})\n#Inserting new columns\nAgerange1.columns=[\"Purchase Count\", \"Average Purchase Price\", \"Total Purchase Value\", \"SN\"]\n#Calculating Avg Total Purchase per Person\nAgerange1[\"Avg Total Purchase per Person\"] = Agerange1[\"Total Purchase Value\"]/Agerange1[\"SN\"]\n#Deleting column SN\ndel Agerange1 [\"SN\"]\n#Formatting Average Purchase Price, Total Purchase Value and Avg Total Purchase per Person to have $ sign and two decimal points\nAgerange1[\"Average Purchase Price\"] = Agerange1[\"Average Purchase Price\"].map(\"${:.2f}\".format)\nAgerange1[\"Total Purchase Value\"] = Agerange1[\"Total Purchase Value\"].map(\"${:.2f}\".format)\nAgerange1[\"Avg Total Purchase per Person\"] = Agerange1[\"Avg Total Purchase per Person\"].map(\"${:.2f}\".format)\n#Aligning all the text to be left indented\nAgerange1 = Agerange1.style.set_properties(**{'text-align':'left'})\n#Setting all table style to be left indented which causes the index to be left indented as well\nAgerange1.set_table_styles([dict(selector='th', props=[('text-align', 'left')])])\n",
"_____no_output_____"
]
],
[
[
"## Top Spenders",
"_____no_output_____"
],
[
"* Run basic calculations to obtain the results in the table below\n\n\n* Create a summary data frame to hold the results\n\n\n* Sort the total purchase value column in descending order\n\n\n* Optional: give the displayed data cleaner formatting\n\n\n* Display a preview of the summary data frame\n\n",
"_____no_output_____"
]
],
[
[
"#Grouping the data frame by SN and getting count of Purchase Id and mean & sum of Price\nSN1 = purchase_data_df.groupby(\"SN\").agg({\"Purchase ID\": \"count\", \"Price\": [\"mean\", \"sum\"]})\n#Inserting new columns\nSN1.columns=[\"Purchase Count\", \"Average Purchase Price\", \"Total Purchase Value\"]\n#Sorting by Total Purchase Value highest to lowest\nSN1 = SN1.sort_values(\"Total Purchase Value\", ascending=False)\n#Formatting Average Purchase Price and Total Purchase Value to have $ sign and two decimal points\nSN1[\"Average Purchase Price\"] = SN1[\"Average Purchase Price\"].map(\"${:.2f}\".format)\nSN1[\"Total Purchase Value\"] = SN1[\"Total Purchase Value\"].map(\"${:.2f}\".format)\n#Aligning all the text to be left indented\nSN1 = SN1.head(5).style.set_properties(**{'text-align':'left'})\n#Setting all table style to be left indented which causes the index to be left indented as well\nSN1.set_table_styles([dict(selector='th', props=[('text-align', 'left')])])\n",
"_____no_output_____"
]
],
[
[
"## Most Popular Items",
"_____no_output_____"
],
[
"* Retrieve the Item ID, Item Name, and Item Price columns\n\n\n* Group by Item ID and Item Name. Perform calculations to obtain purchase count, item price, and total purchase value\n\n\n* Create a summary data frame to hold the results\n\n\n* Sort the purchase count column in descending order\n\n\n* Optional: give the displayed data cleaner formatting\n\n\n* Display a preview of the summary data frame\n\n",
"_____no_output_____"
]
],
[
[
"#Grouping the data frame by Item Id and Item Name and getting count of Purchase Id and mean & sum of Price\nItem1 = purchase_data_df.groupby([\"Item ID\", \"Item Name\"]).agg({\"Purchase ID\": \"count\", \"Price\": [\"mean\", \"sum\"]})\n#Inserting new columns\nItem1.columns=[\"Purchase Count\", \"Item Price\", \"Total Purchase Value\"]\n#Sorting by Purchase Count highest to lowest\nItem1 = Item1.sort_values(\"Purchase Count\", ascending=False)\n#Formatting Item Price and Total Purchase Value to have $ sign and two decimal points\nItem1[\"Item Price\"] = Item1[\"Item Price\"].map(\"${:.2f}\".format)\nItem1[\"Total Purchase Value\"] = Item1[\"Total Purchase Value\"].map(\"${:.2f}\".format)\n#Aligning all the text to be left indented\nItem1 = Item1.head(5).style.set_properties(**{'text-align':'left'})\n#Setting all table style to be left indented which causes the index to be left indented as well\nItem1.set_table_styles([dict(selector='th', props=[('text-align', 'left')])])\n",
"_____no_output_____"
]
],
[
[
"## Most Profitable Items",
"_____no_output_____"
],
[
"* Sort the above table by total purchase value in descending order\n\n\n* Optional: give the displayed data cleaner formatting\n\n\n* Display a preview of the data frame\n\n",
"_____no_output_____"
]
],
[
[
"#Grouping the data frame by Item Id and Item Name and getting count of Purchase Id and mean & sum of Price\nItem2 = purchase_data_df.groupby([\"Item ID\", \"Item Name\"]).agg({\"Purchase ID\": \"count\", \"Price\": [\"mean\", \"sum\"]})\n#Inserting new columns\nItem2.columns=[\"Purchase Count\", \"Item Price\", \"Total Purchase Value\"]\n#Sorting by Total Purchase Value highest to lowest\nItem2 = Item2.sort_values(\"Total Purchase Value\", ascending=False)\n#Formatting Item Price and Total Purchase Value to have $ sign and two decimal points\nItem2[\"Item Price\"] = Item2[\"Item Price\"].map(\"${:.2f}\".format)\nItem2[\"Total Purchase Value\"] = Item2[\"Total Purchase Value\"].map(\"${:.2f}\".format)\n#Aligning all the text to be left indented\nItem2 = Item2.head(5).style.set_properties(**{'text-align':'left'})\n#Setting all table style to be left indented which causes the index to be left indented as well\nItem2.set_table_styles([dict(selector='th', props=[('text-align', 'left')])])\n",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
]
] |
cb9420baa73cfb632b456647fc1670e7861e5738
| 29,963 |
ipynb
|
Jupyter Notebook
|
Summer-19/Section/Section 6.ipynb
|
sberto2019/ENVECON-118
|
657628c4c29e1494bea871c7963e9452f7bf9c08
|
[
"BSD-3-Clause"
] | 2 |
2020-09-10T13:45:34.000Z
|
2021-11-01T21:41:59.000Z
|
Summer-19/Section/Section 6.ipynb
|
sberto2019/ENVECON-118
|
657628c4c29e1494bea871c7963e9452f7bf9c08
|
[
"BSD-3-Clause"
] | 1 |
2021-01-22T17:02:59.000Z
|
2021-06-20T14:03:54.000Z
|
Summer-19/Section/Section 6.ipynb
|
sberto2019/ENVECON-118
|
657628c4c29e1494bea871c7963e9452f7bf9c08
|
[
"BSD-3-Clause"
] | 7 |
2019-11-06T20:55:27.000Z
|
2021-06-20T08:14:41.000Z
| 47.560317 | 686 | 0.581951 |
[
[
[
"# EEP/IAS 118 - Section 6\n\n## Fixed Effects Regression\n\n### August 1, 2019\n\nToday we will practice with fixed effects regressions in __R__. We have two different ways to estimate the model, and we will see how to do both and the situations in which we might favor one versus the other.\n\nLet's give this a try using the dataset `wateruse.dta`, a panel of residential water use for residents in Alameda and Contra Costa Counties. The subset of households are high water users, people who used over 1,000 gallons per billing cycle. We have information on their water use, weather during the period, as well as information on the city and zipcode of where the home is located, and information on the size and value of the house.\n\nSuppose we are interested in running the following panel regression of residential water use:\n\n$$ GPD_{it} = \\beta_0 + \\beta_1 degree\\_days_{it} + \\beta_2 precip_{it} ~~~~~~~~~~~~~~~~~~~~~~~(1)$$\n\nWhere $GPD$ is the gallons used per day by household $i$ in billing cycle $t$, $degree\\_days$ the count of degree days experienced by the household in that billing cycle (degree days are a measure of cumulative time spent above a certain temperature threshold), and $precip$ the amount of precipitation in millimeters.",
"_____no_output_____"
]
],
[
[
"library(tidyverse)\nlibrary(haven)\nlibrary(lfe)\nwaterdata <- read_dta(\"wateruse.dta\") %>%\n mutate(gpd = (unit*748)/num_days)\nwaterdata <- mutate(waterdata,\n n = 1:nrow(waterdata))\nhead(waterdata)\ndim(waterdata)\nreg1 <- lm(gpd ~ degree_days + precip, data = waterdata)\nsummary(reg1)",
"_____no_output_____"
]
],
[
[
"Here we obtain an estimate of $\\hat\\beta_1 = 0.777$, telling us that an additional degree day per billing cycle is associated with an additional $0.7769$ gallon used per day. These billing cycles are roughly two months long, so this suggests an increase of roughly 47 gallons per billing cycle. Our estimate is statistically significant at all conventional levels, suggesting residential water use does respond to increased exposure to high heat.\n\nWe estimate a statistically insignificant coefficient on additional precipitation, which tells us that on average household water use in our sample doesn't adjust to how much it rains.\n\nWe might think that characteristics of the home impact how much water is used there, so we add in some home controls:\n\n\n$$ GPD_{it} = \\beta_0 + \\beta_1 degree\\_days_{it} + \\beta_2 precip_{it} + \\beta_3 lotsize_{i} + \\beta_4 homesize_i + \\beta_5 num\\_baths_i + \\beta_6 num\\_beds_i + \\beta_7 homeval_i~~~~~~~~~~~~~~~~~~~~~~~(2)$$",
"_____no_output_____"
]
],
[
[
"reg2 <- lm(gpd ~ degree_days + precip + lotsize + homesize + num_baths + num_beds + homeval, data = waterdata)\nsummary(reg2)",
"_____no_output_____"
]
],
[
[
"Our coefficient on $degree\\_days$ remains statistically significant and doesn't change much, so we find that $\\hat\\beta_1$ is robust to the addition of home characteristics. Of these characteristics, we obtain statistically significant coefficients on the size of the lot (in acres), the size of the home ($ft^2$), and the number of bedrooms in the home.\n\nWe get a curious result for $\\hat\\beta_6$: for each additional bedroom in the home we predict that water use will _fall_ by 48 gallons per day. \n\n### Discussion: what might be driving this effect? ",
"_____no_output_____"
]
],
[
[
"\nwaterdata %>%\n filter( city <= 9) %>%\nggplot(aes(x=num_beds, y=gpd)) +\ngeom_point() +\n facet_grid(. ~ city)\n\nwaterdata %>%\n filter(city> 9 & city <= 18) %>%\nggplot(aes(x=num_beds, y=gpd)) +\ngeom_point() +\n facet_grid(. ~ city)\n\nwaterdata %>%\n filter( city> 18) %>%\nggplot(aes(x=num_beds, y=gpd)) +\ngeom_point() +\n facet_grid(. ~ city)\n",
"_____no_output_____"
]
],
[
[
"Since there are likely a number of sources of omitted variable bias in the previous model, we think it might be worth including some fixed effects in our model. \n\n## Method 1: Fixed Effects with lm() \n\nUp to this point we have been running our regressions using the `lm()` function. We can still use `lm()` for our fixed effects models, but it takes some more work.\n\nRecall that we can write our general panel fixed effects model as \n\n$$ y_{it} = \\beta x_{it} + \\mathbf{a}_i + {d}_t + u_{it} $$\n\n* $y$ our outcome of interest, which varies in both the time and cross-sectional dimensions\n* $x_{it}$ our set of time-varying unit characteristics\n* $\\mathbf{a}_i$ our set of unit fixed effects\n* $d_t$ our time fixed effects\n\nWe can estimate this model in `lm()` provided we have variables in our dataframe that correspond to $a_i$ and $d_t$. This means we'll have to generate them before we can run any regression.\n\n### Generating Dummy Variables\n\nIn order to include fixed effects for our regression, we have to first generate the set of dummy variables that we want. For example, if we want to include a set of city fixed effects in our model, we need to generate them.\n\nWe can do this in a few ways.\n\n1. First, we can use `mutate()` and add a separate line for each individual city:",
"_____no_output_____"
]
],
[
[
"fe_1 <- waterdata %>%\n mutate(city_1 = as.numeric((city==1)),\n city_2 = as.numeric((city ==2)),\n city_3 = as.numeric((city ==3))) %>%\n select(n, city, city_1, city_2, city_3)\nhead(fe_1)",
"_____no_output_____"
]
],
[
[
"This can be super tedious though when we have a bunch of different levels of our variable that we want to make fixed effects for. In this case, we have 27 different cities.\n\n2. Alternatively, we can use the `spread()` function to help us out. Here we add in a constant variable `v` that is equal to one in all rows, and a copy of city that adds \"city_\" to the front of the city number. Then we pass the data to `spread`, telling it to split the variable `cty` into dummy variables for all its levels, with all the \"false\" cases filled with zeros.\n\n",
"_____no_output_____"
]
],
[
[
"fe_2 <- waterdata %>%\nselect(n, city)\n\nhead(fe_2)\n\nfe_2 %>%\n mutate(v = 1, cty = paste0(\"city_\", city)) %>% \n spread(cty, v, fill = 0)",
"_____no_output_____"
]
],
[
[
"That is much easier! Let's now do that so that they'll be in `waterdata`:",
"_____no_output_____"
]
],
[
[
"waterdata <- waterdata %>%\n mutate(v = 1, cty = paste0(\"city_\", city)) %>% \n spread(cty, v, fill = 0)\nhead(waterdata)\nnames(waterdata)",
"_____no_output_____"
]
],
[
[
"Note that both of the variables we used in `spread` are no longer in our dataset.\n\nWhile we're at it, let's also add in a set of billing cycle fixed effects.",
"_____no_output_____"
]
],
[
[
"waterdata <- waterdata %>%\n mutate(v = 1, cyc = paste0(\"cycle_\", billingcycle)) %>% \n spread(cyc, v, fill = 0)\nhead(waterdata)\n",
"_____no_output_____"
]
],
[
[
"Now we have all our variables to run the regression\n\n\n$$ GPD_{it} = \\beta_0 + \\beta_1 degree\\_days_{it} + \\beta_2 precip_{it} + \\mathbf{a}_i + \\mathbf{d}_t~~~~~~~~~~~~~~~~~~~~~~~(2)$$\n\nWhere $\\mathbf{a}_i$ are our city fixed effects, and $\\mathbf{d}_t$ our billing cycle fixed effects.\n\nNow we can run our model! Well, now we can _write out_ our model. The challenge here is that we need to specify all of the dummy variables in our formula. We could do this all by hand, but when we end up with a bunch of fixed effects it's easier to use the following trick: we can write `y ~ .` to tell __R__ we want it to put every variable in our dataset other than $y$ on the right hand side of our regression. That means we can create a version of our dataset with only $gpd$, $degree\\_days$, $precip$, and our fixed effects and won't have to write out all those fixed effects by hand!\n\nNote that we can use `select()` and `-` to remove variables from our dataframe. If there is a list of variables in order that we want to get rid of, we can do it easily by passing a vector through! For instance, we want to get rid of the first 12 variables in our data, so we can add `-unit:-hh` in select to get rid of them all at once. if we separate with a comma, we can drop other sections of our data too!",
"_____no_output_____"
]
],
[
[
"fe_data <- waterdata %>%\n select(-unit:-hh, -city, -n) \nhead(fe_data)\n\nfe_reg1 <- lm(gpd ~ ., data = fe_data)\nsummary(fe_reg1)",
"_____no_output_____"
]
],
[
[
"Since I specified it this way, __R__ chose the last dummy variable for each set of fixed effect to leave out as our omitted group. \n\nNow that we account for which billing cycle we're in (i.e. whether we're in the winter or whether we're in the summer), we find that the coefficient on $degree\\_days$ is now much smaller and statistically insignificant. This makes sense, as we were falsely attributing the extra water use that comes from seasonality to temperature on its own. Now that we control for the season we're in via billing cycle fixed effects, we find that deviations in temperature exposure during a billing cycle doesn't result in dramatically higher water use within the sample.\n\n### Discussion: Why did we drop the home characteristics from our model?",
"_____no_output_____"
],
[
"## Method 2: Fixed Effects with felm()\n\nAlternatively, we could do everything way faster using the `felm()` function from the package __lfe__. This package doesn't require us to produce all the dummy variables by hand. Further, it performs the background math way faster so will be much quicker to estimate models using large datasets and many variables.\n\nThe syntax we use is now \n\n`felm(y ~ x1 + x2 + ... + xk | FE_1 + FE_2, data = df)`\n\n* The first section $y \\sim x1 + x2 +... xk$ is our formula, written the same way as with `lm()`\n* We now add a `|` and in the second section we specify our fixed effects. Here we say $FE\\_1 + FE\\_2$ which tells __R__ to include fixed effects for each level of $FE\\_1$ and $FE\\_2$.\n* Note that our fixed effect variables must be of class \"factor\" - we can force our variables to take this class by adding them as `as.factor(FE_1) + as.factor(FE_2)`.\n* we add the data source after the comma, as before.\n\nLet's go ahead and try this now with our water data model:",
"_____no_output_____"
]
],
[
[
"fe_reg2 <- felm(gpd ~ degree_days + precip | as.factor(city) + as.factor(billingcycle), data = waterdata)\nsummary(fe_reg2)",
"_____no_output_____"
]
],
[
[
"And we estimate the exact same coefficients on $degree\\_days$ and $precip$ as in the case where we specified everything by hand! We didn't have to mutate our data or add any variables. The one potential downside is that this approach doesn't report the fixed effects themselves. However, the tradeoff is that `felm` runs a lot faster than `lm`. To see this, we can compare run times:",
"_____no_output_____"
]
],
[
[
"lm_start <- Sys.time()\nfe_data <- waterdata %>%\n mutate(v = 1, cyc = paste0(\"cycle_\", billingcycle)) %>% \n spread(cyc, v, fill = 0) %>%\n mutate(v = 1, cty = paste0(\"city_\", city)) %>% \n spread(cty, v, fill = 0) %>%\n select(-unit:-hh, -city, -n) \nlm(gpd ~ ., data = fe_data)\nlm_end <- Sys.time()\nlm_dur <- lm_end - lm_start\n\n\nfelm_start <- Sys.time()\nfelm(gpd ~ degree_days + precip | as.factor(city) + as.factor(billingcycle), data = waterdata)\nfelm_end <- Sys.time()\nfelm_dur <- felm_end - felm_start\n\nprint(paste0(\"lm() duration is \", lm_dur, \" seconds, while felm() duration is \", felm_dur, \" seconds.\"))\n\n",
"_____no_output_____"
]
],
[
[
"Okay, neither of these models took very long, but that's because we only have two covariates other than our fixed effects and only around 2300 observations. If we have hundreds of covariates and millions of observations, this time difference becomes massive.\n",
"_____no_output_____"
],
[
"# Regression Discontinuity\n\nLet's practice running a regression discontinuity model. Suppose we were interested in exploring the weird relationship we saw earlier between water use and number of bedrooms in a home. Let's take a look at that relationship a bit more closely.",
"_____no_output_____"
]
],
[
[
"\nwaterdata %>%\n ggplot(aes(x = num_beds, y = gpd)) +\n geom_point( alpha = 0.4, colour = \"royalblue\")",
"_____no_output_____"
]
],
[
[
"We see that average water use appears to rise as we add bedrooms to a house from a low number, peaks when households have five bedrooms, then begins to fall with larger and larger houses... though there are a few high outliers in the 6-9 bedroom cases as well, which might overshadow that trend.\n\nIs there something else that's correlated with the number of bedrooms in a home that may also be driving this?",
"_____no_output_____"
]
],
[
[
"\nwaterdata %>%\n ggplot(aes(x = num_beds, y = lotsize)) +\n geom_point( alpha = 0.4, colour = \"royalblue\")",
"_____no_output_____"
]
],
[
[
"It looks like lotsize and the number of bedrooms share a similar relationship - lot size increasing in \\# bedrooms up until 5, then declining from there.\n\nGiven that it looks like 5 bedrooms is where the relationship changes, let's use this as our running variable and allow the relationship for the number of bedrooms to differ around a threshold of five bedrooms. We can write an RD model as \n\n$$ GPD_{it} = \\beta_0 + \\beta_1 T_i + \\beta_2 (num\\_beds - 5) + \\beta_3\\left( T_i \\times (num\\_beds - 5) \\right) + u_{it} $$",
"_____no_output_____"
]
],
[
[
"rd_data <- waterdata %>%\n select(gpd, num_beds, lotsize) %>%\n mutate(treat = (num_beds > 5),\n beds_below = num_beds - 5,\n beds_above = treat * (num_beds - 5)) \n\nrd_reg <- lm(gpd ~ treat + beds_below + beds_above, data = rd_data)\nsummary(rd_reg)",
"_____no_output_____"
]
],
[
[
"What if we limit our comparison closer to around the threshold? Right now we're using data from the entire sample, but this might not be a valid comparison. Let's see what happens when we reduce our bandwidth to 3 and look at only homes with between 3 and 8 bedrooms.",
"_____no_output_____"
]
],
[
[
"rd_data_trim <- rd_data %>%\n filter(!(num_beds < 2)) %>%\n filter(!(num_beds > 8))\n\nrd_reg2 <- lm(gpd ~ treat + beds_below + beds_above, data = rd_data_trim)\nsummary(rd_reg2)",
"_____no_output_____"
]
],
[
[
"We now estimate a treatment effect at the discontinuity! Our model finds a discontinuity, estimating a jump down of 284 gallons per day right around the 5 bedroom threshold. \n\nHowever, we saw earlier that it appears that lotsize is correlated with the number of bedrooms in the home, and is definitely a factor correlated with residential water use. What happens to our LATE estimate when we control for lot size in our regression?",
"_____no_output_____"
]
],
[
[
"rd_reg3 <- lm(gpd ~ treat + beds_below + beds_above + lotsize, data = rd_data_trim)\nsummary(rd_reg3)",
"_____no_output_____"
]
],
[
[
"Once we control for lot size, our interpretation changes. Now we estimate a coefficient on treatment nearly half the magnitude as before and without statistical significance. \n\n### Discussion Q: What does this tell us about the covariance between lot size and the number of bedrooms?",
"_____no_output_____"
],
[
"# Fixed Effects Practice Question #1\n\n#### From a random sample of agricultural yields Y (1000 dollars per acre) for region $i$ in year $t$ for the US, we have estimated the following eqation:\n\n \\begin{align*} \\widehat{\\log(Y)}_{it} &=\t0.49\t+ .01 GE_{it} ~~~~ \tR^2 = .32\\\\\n\t&~~~~~(.11)\t ~~~~ (.01) ~~~~ n = 1526 \\end{align*}\n\n#### (a) Interpret the results on the Genetically engineered ($GE$) technology on yields. (follow SSS= Sign Size Significance)\n\n#### (b) Suppose $GE$ is used more on the West Coast, where crop yields are also higher. How would the estimated effect of GE change if we include a West Coast region dummy variable in the equation? Justify your answer.\n\n#### (c) If we include region fixed effects, would they control for the factors in (b)? Justify your answer.\n\n#### (d) If yields have been generally improving over time and GE adoption was only recently introduced in the USA, what would happen to the coefficient of GE if we included year fixed effects?\n\n\n\n# Fixed Effects Practice Question #2\n\n#### A recent paper investigates whether advertisement for Viagra causes increases in birth rates in the USA. Apparently, advertising for products, including Viagra, happens on TV and reaches households that have a TV within a Marketing region and does not happen in areas outside a designated marketing region. What the authors do is look at hospital birth rates in regions inside and near the advertising region border and collect data on dollars per 100 people (Ads) for a certain time, and compare those to the birth rates in hospitals located outside and near the advertising region designated border. They conduct a panel data analysis and estimate the following model:\n\n$$ Births_{it} = \\beta_0 + \\beta_1 Ads + \\beta_2 Ads^2 + Z_i + M_t + u_{it}$$\n\n#### Where $Z_i$ are zipcode fixed effects and $M_t$ monthly fixed effects.\n\n#### (a) Why do the authors include Zip Code Fixed Effects? In particular, what would be a variable that they are controlling for when adding Zip Code fixed effects that could cause a problem when interpreting the marginal effect of ad spending on birth rates? What would that (solved) problem be?\n\n#### (b) Why do they add month fixed effects? \n\n",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
]
] |
cb942c28e84c23d5a9af012e87580ad0d1bf03bc
| 134,478 |
ipynb
|
Jupyter Notebook
|
notebooks/SVM Classification - Individual Replicas, Only Regulator Genes.ipynb
|
uwescience/new_xstate
|
7df9f99b5e2caadc1db806a40ab30965c4c4212b
|
[
"MIT"
] | 3 |
2019-11-13T17:25:52.000Z
|
2021-12-21T02:25:03.000Z
|
notebooks/SVM Classification - Individual Replicas, Only Regulator Genes.ipynb
|
uwescience/xstate
|
f7563db624258f43c9547e974a0fd5929fc2fa5b
|
[
"MIT"
] | null | null | null |
notebooks/SVM Classification - Individual Replicas, Only Regulator Genes.ipynb
|
uwescience/xstate
|
f7563db624258f43c9547e974a0fd5929fc2fa5b
|
[
"MIT"
] | null | null | null | 271.125 | 22,780 | 0.924114 |
[
[
[
"# SVM Classification Using Individual Replicas\nThis notebook analyzes the quality of the classifiers resulting from training on individual replicas of read counts rather than averaged values. Data are adjusted for library size and gene length.\n\nTraining data\n1. Uses individual replicas (not averaged)\n1. Uses all genes\n1. Includes time T1 (normoxia is not combined with resuscitation)",
"_____no_output_____"
]
],
[
[
"import init\nfrom common import constants as cn\nfrom common.trinary_data import TrinaryData\nfrom common.data_provider import DataProvider\nfrom common_python.plots import util_plots\nfrom plots import util_plots as xutil_plots\nfrom common_python.classifier import classifier_ensemble\nfrom common_python.classifier import util_classifier\nfrom common_python.classifier import classifier_collection\nfrom common_python.classifier.classifier_ensemble_random_forest import ClassifierEnsembleRandomForest\nfrom common_python.plots import util_plots as common_plots\n\nimport collections\nimport os\nimport matplotlib.pyplot as plt\nimport numpy as np\nimport pandas as pd\nimport sklearn\nfrom sklearn.decomposition import PCA\nfrom sklearn import svm\nfrom sklearn.model_selection import cross_val_score\n\n%matplotlib inline",
"_____no_output_____"
]
],
[
[
"## Analyze Replica Data\nThe following shows the extent to which replicas agree with the tranary values that are assigned.",
"_____no_output_____"
]
],
[
[
"def compareDFValues(df1, df2, title):\n RANGE = [-8, 8]\n plt.figure()\n arr1 = df1.values.flatten()\n arr2 = df2.values.flatten()\n plt.scatter(arr1, arr2)\n # Define region of 0 values\n plt.plot([-1, -1], [-1, 1], color=\"b\")\n plt.plot([1, 1], [-1, 1], color=\"b\")\n plt.plot([-1, 1], [-1, -1], color=\"b\")\n plt.plot([-1, 1], [1, 1], color=\"b\")\n # Define region of 1 values\n plt.plot([1, 1], [1, RANGE[1]], color=\"b\")\n plt.plot([1, RANGE[1]], [1, 1], color=\"b\")\n # Define region of -1 values\n plt.plot([-1, -1], [-1, RANGE[0]], color=\"b\")\n plt.plot([-1, RANGE[0]], [-1, -1], color=\"b\")\n plt.plot(RANGE, RANGE, color=\"r\")\n plt.title(title)\n \nprovider = DataProvider()\nprovider.do()\ndfs = []\nfor idx in range(3):\n dfs.append(provider.dfs_adjusted_read_count_wrtT0_log2[idx])\ncompareDFValues(dfs[0], dfs[1], \"0 vs 1\")\ncompareDFValues(dfs[0], dfs[2], \"0 vs 2\")\ncompareDFValues(dfs[1], dfs[2], \"1 vs 2\")\n ",
"_____no_output_____"
],
[
"dfs[0].values.flatten()\n",
"_____no_output_____"
]
],
[
[
"## Accuracy With Replicas\nCompares accuracies with replicas (is_averaged=False) vs. with replicas as the data.",
"_____no_output_____"
]
],
[
[
"def clfEval(is_averaged, high_rank=15, ensemble_size=50, is_randomize=False, num_iterations=10):\n trinary = TrinaryData(is_averaged=is_averaged, is_dropT1=is_averaged)\n df_X = trinary.df_X.copy()\n df_X.columns = trinary.features\n ser_y = trinary.ser_y.copy()\n if is_randomize:\n # Randomize the relationship between features and state\n df_X = df_X.sample(frac=1)\n ser_y = ser_y.sample(frac=1)\n #\n svm_ensemble = classifier_ensemble.ClassifierEnsemble(\n classifier_ensemble.ClassifierDescriptorSVM(), size=ensemble_size,\n filter_high_rank=high_rank)\n return classifier_ensemble.ClassifierEnsemble.crossValidateByState(\n svm_ensemble, df_X, ser_y, num_iterations)",
"_____no_output_____"
],
[
"clfEval(True, ensemble_size=50, is_randomize=True)",
"_____no_output_____"
],
[
"clfEval(False, ensemble_size=50, is_randomize=True)",
"_____no_output_____"
],
[
"clfEval(True)",
"_____no_output_____"
],
[
"clfEval(False)",
"_____no_output_____"
]
],
[
[
"## Analysis of Classifier Accuracy by State",
"_____no_output_____"
]
],
[
[
"# Plot values by state\ndef plotValuesByState(states, values, stds=None, ylabel=\"percent\"):\n if stds is None:\n plt.bar(states, values)\n else:\n plt.bar(states, values, yerr=stds, alpha=0.5)\n plt.xticks(rotation=45)\n plt.xlabel(\"state\")\n plt.ylabel(ylabel)",
"_____no_output_____"
],
[
"# State statistics\ndef plotStateDistributions():\n PERCENT = \"percent\"\n VALUE = \"value\"\n NAME = \"name\"\n trinary = TrinaryData(is_averaged=False, is_dropT1=False)\n df = pd.DataFrame(trinary.ser_y)\n df[VALUE] = list(np.repeat(1, len(df)))\n df_group = pd.DataFrame(df.groupby(NAME).count())\n dct = {v: k for k, v in trinary.state_dict.items()}\n df_group.index = [dct[s] for s in df_group.index]\n df_group[PERCENT] = 100*df_group[VALUE] / len(df)\n plotValuesByState(df_group.index, df_group[PERCENT])\n \nplotStateDistributions()",
"_____no_output_____"
],
[
"# Classification accuracy by state\ndef stateClassificationAccuracy(state, num_iterations=10, is_averaged=False):\n NUM_HOLDOUTS = 1\n is_dropT1 = is_averaged\n trinary = TrinaryData(is_averaged=is_averaged, is_dropT1=is_dropT1)\n df_X = trinary.df_X.copy()\n df_X.columns = trinary.features\n ser_y = trinary.ser_y\n results = []\n for _ in range(num_iterations):\n test_indices = []\n ser_sample = ser_y[ser_y == state].sample(n=NUM_HOLDOUTS)\n test_indices.extend(list(ser_sample.index))\n train_indices = list(set(df_X.index).difference(test_indices))\n svm_ensemble = classifier_ensemble.ClassifierEnsemble(\n classifier_ensemble.ClassifierDescriptorSVM(), size=30,\n filter_high_rank=1500,\n classes=list(ser_y.values))\n svm_ensemble.fit(df_X.loc[train_indices, :], ser_y.loc[train_indices])\n results.append(svm_ensemble.score(df_X.loc[test_indices, :], ser_y[test_indices]))\n return results",
"_____no_output_____"
],
[
"def plotStateAccuracies(is_averaged=True):\n is_dropT1 = is_averaged\n trinary = TrinaryData(is_averaged=is_averaged, is_dropT1=is_dropT1)\n states = list(trinary.state_dict.values())\n avgs = []\n stds = []\n for state in states:\n values = stateClassificationAccuracy(state, is_averaged=is_averaged)\n avgs.append(np.mean(values))\n stds.append(np.std(values))\n plotValuesByState(list(trinary.state_dict.keys()), avgs, stds=stds, ylabel=\"accuracy\")",
"_____no_output_____"
],
[
"plotStateAccuracies(is_averaged=True)\nplt.figure()\nplotStateAccuracies(is_averaged=False)",
"_____no_output_____"
],
[
"if False:\n is_averaged = False\n df = df_confuse.applymap(lambda v: np.nan if v <= 0.2 else 1)\n df.columns = [c - 0.5 for c in df.columns]\n trinary = TrinaryData(is_averaged=is_averaged, is_dropT1=is_averaged)\n states = trinary.ser_y.values\n state_colors = [\"grey\", \"orange\", \"green\", \"pink\", \"peru\", \"greenyellow\"]\n heatmap = plt.pcolor(df.T, cmap='jet')\n #fig = heatmap.get_figure()\n #axes = fig.get_axes()[0]\n #yaxis = axes.get_yaxis()\n #xv = [x + 0.5 for x in range(len(df.T.columns))]\n #yv = [y + 0.5 for y in range(len(df.T))]\n #plt.xticks(xv)\n #plt.yticks(yv)\n positions = [p - 0.5 for p in range(-1, len(states))]\n labels = [str(int(c-.5)) if c >= 0 else \"\" for c in positions]\n plt.yticks(positions, labels)\n for idx, state in enumerate(states):\n color = state_colors[state]\n plt.scatter(idx, [-1], color=color)\n #plt.colorbar(heatmap)\n",
"_____no_output_____"
],
[
"is_averaged = False\ntrinary = TrinaryData(is_averaged=is_averaged, is_dropT1=is_averaged)\nsvm_ensemble = classifier_ensemble.ClassifierEnsemble(\n classifier_ensemble.ClassifierDescriptorSVM(), size=50,\n filter_high_rank=None)\nser_pred = svm_ensemble.makeInstancePredictionDF(trinary.df_X, trinary.ser_y)",
"_____no_output_____"
],
[
"util_classifier.plotInstancePredictions(trinary.ser_y, ser_pred, is_plot=True)\n",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb943faf537acadf9395bf4d2287618f33772e0e
| 14,831 |
ipynb
|
Jupyter Notebook
|
_notebooks/2020-10-09-Earth-Engine-Inference.ipynb
|
developmentseed/sat-ml-training
|
bb4fe7c42a593617136179d225f863d7a5e3883d
|
[
"Apache-2.0"
] | 16 |
2020-10-22T05:43:33.000Z
|
2022-03-03T12:38:57.000Z
|
_notebooks/2020-10-09-Earth-Engine-Inference.ipynb
|
PhuntshoPhuntsho/sat-ml-training
|
bb4fe7c42a593617136179d225f863d7a5e3883d
|
[
"Apache-2.0"
] | 17 |
2020-09-09T23:20:05.000Z
|
2022-02-26T09:49:34.000Z
|
_notebooks/2020-10-09-Earth-Engine-Inference.ipynb
|
PhuntshoPhuntsho/sat-ml-training
|
bb4fe7c42a593617136179d225f863d7a5e3883d
|
[
"Apache-2.0"
] | 7 |
2020-09-29T02:53:22.000Z
|
2022-01-12T16:52:37.000Z
| 29.841046 | 443 | 0.538602 |
[
[
[
"# Inference in Google Earth Engine + Colab\n> Scaling up machine learning with GEE and Google Colab.\n\n- toc: true \n- badges: true\n- author: Drew Bollinger\n- comments: false\n- hide: false\n- sticky_rank: 11",
"_____no_output_____"
],
[
"# Inference in Google Earth Engine + Colab\n\nHere we demonstrate how to take a trained model and apply to to imagery with Google Earth Engine + Colab + Tensorflow. This is adapted from an [Earth Engine <> TensorFlow demonstration notebook](https://developers.google.com/earth-engine/guides/tf_examples). We'll be taking the trained model from the [Deep Learning Crop Type Segmentation Model Example](https://developmentseed.org/sat-ml-training/DeepLearning_CropType_Segmentation).\n",
"_____no_output_____"
],
[
"# Setup software libraries\n\nAuthenticate and import as necessary.",
"_____no_output_____"
]
],
[
[
"# Import, authenticate and initialize the Earth Engine library.\nimport ee\nee.Authenticate()\nee.Initialize()",
"_____no_output_____"
],
[
"# Mount our Google Drive\nfrom google.colab import drive\ndrive.mount('/content/drive')",
"_____no_output_____"
],
[
"# Add necessary libraries.\n!pip install -q focal-loss\nimport os\nfrom os import path as op\nimport tensorflow as tf\nimport folium\nfrom focal_loss import SparseCategoricalFocalLoss",
"_____no_output_____"
]
],
[
[
"# Variables\n\nDeclare the variables that will be in use throughout the notebook.",
"_____no_output_____"
]
],
[
[
"# Specify names locations for outputs in Google Drive. \nFOLDER = 'servir-inference-demo'\nROOT_DIR = '/content/drive/My Drive/'\n\n# Specify inputs (Sentinel indexes) to the model.\nBANDS = ['NDVI', 'WDRVI', 'SAVI']\n\n# Specify the size and shape of patches expected by the model.\nKERNEL_SIZE = 224\nKERNEL_SHAPE = [KERNEL_SIZE, KERNEL_SIZE]",
"_____no_output_____"
]
],
[
[
"# Imagery\n\nGather and setup the imagery to use for inputs. It's important that we match the index inputs from the earlier analysis. This is a three-month Sentinel-2 composite. Display it in the notebook for a sanity check.",
"_____no_output_____"
]
],
[
[
"# Use Sentinel-2 data.\n\ndef add_indexes(img): \n ndvi = img.expression(\n '(nir - red) / (nir + red + a)', {\n 'a': 1e-5,\n 'nir': img.select('B8'),\n 'red': img.select('B4')\n }\n \n ).rename('NDVI')\n\n wdrvi = img.expression(\n '(a * nir - red) / (a * nir + red)', {\n 'a': 0.2,\n 'nir': img.select('B8'),\n 'red': img.select('B4')\n }\n ).rename('WDRVI')\n\n savi = img.expression(\n '1.5 * (nir - red) / (nir + red + 0.5)', {\n 'nir': img.select('B8'),\n 'red': img.select('B4')\n }\n ).rename('SAVI')\n\n return ee.Image.cat([ndvi, wdrvi, savi])\n\nimage = ee.ImageCollection('COPERNICUS/S2') \\\n .filterDate('2018-01-01', '2018-04-01') \\\n .filter(ee.Filter.lt('CLOUDY_PIXEL_PERCENTAGE', 20)) \\\n .map(add_indexes) \\\n .median()\n\n# Use folium to visualize the imagery.\nmapid = image.getMapId({'bands': BANDS, 'min': -1, 'max': 1})\nmap = folium.Map(location=[ \n -29.177943749121233,\n 30.55984497070313,\n])\nfolium.TileLayer(\n tiles=mapid['tile_fetcher'].url_format,\n attr='Map Data © <a href=\"https://earthengine.google.com/\">Google Earth Engine</a>',\n overlay=True,\n name='median composite',\n ).add_to(map)\n\nmap.add_child(folium.LayerControl())\nmap",
"_____no_output_____"
]
],
[
[
"# Load our saved model\n\n",
"_____no_output_____"
]
],
[
[
"# Load a trained model.\nMODEL_DIR = '/content/drive/Shared drives/servir-sat-ml/data/model_out/10062020/'\nmodel = tf.keras.models.load_model(MODEL_DIR)\nmodel.summary()",
"_____no_output_____"
]
],
[
[
"# Prediction\n\nThe prediction pipeline is:\n\n1. Export imagery on which to do predictions from Earth Engine in TFRecord format to Google Drive.\n2. Use the trained model to make the predictions.\n3. Write the predictions to a TFRecord file in Google Drive.\n4. Manually upload the predictions TFRecord file to Earth Engine.\n\nThe following functions handle this process. It's useful to separate the export from the predictions so that you can experiment with different models without running the export every time.",
"_____no_output_____"
]
],
[
[
"def doExport(out_image_base, shape, region):\n \"\"\"Run the image export task. Block until complete.\n \"\"\"\n task = ee.batch.Export.image.toDrive(\n image = image.select(BANDS),\n description = out_image_base,\n fileNamePrefix = out_image_base,\n folder = FOLDER,\n region = region.getInfo()['coordinates'],\n scale = 30,\n fileFormat = 'TFRecord',\n maxPixels = 1e10,\n formatOptions = {\n 'patchDimensions': shape,\n 'compressed': True,\n 'maxFileSize': 104857600\n }\n )\n task.start()\n\n # Block until the task completes.\n print('Running image export to Google Drive...')\n import time\n while task.active():\n time.sleep(30)\n\n # Error condition\n if task.status()['state'] != 'COMPLETED':\n print('Error with image export.')\n else:\n print('Image export completed.')",
"_____no_output_____"
],
[
"def doPrediction(out_image_base, kernel_shape, region):\n \"\"\"Perform inference on exported imagery.\n \"\"\"\n\n print('Looking for TFRecord files...')\n\n # Get a list of all the files in the output bucket.\n filesList = os.listdir(op.join(ROOT_DIR, FOLDER))\n\n # Get only the files generated by the image export.\n exportFilesList = [s for s in filesList if out_image_base in s]\n\n # Get the list of image files and the JSON mixer file.\n imageFilesList = []\n jsonFile = None\n for f in exportFilesList:\n if f.endswith('.tfrecord.gz'):\n imageFilesList.append(op.join(ROOT_DIR, FOLDER, f))\n elif f.endswith('.json'):\n jsonFile = f\n\n # Make sure the files are in the right order.\n imageFilesList.sort()\n\n from pprint import pprint\n pprint(imageFilesList)\n print(jsonFile)\n\n import json\n # Load the contents of the mixer file to a JSON object.\n with open(op.join(ROOT_DIR, FOLDER, jsonFile), 'r') as f:\n mixer = json.load(f)\n\n pprint(mixer)\n patches = mixer['totalPatches']\n\n # Get set up for prediction.\n\n imageColumns = [\n tf.io.FixedLenFeature(shape=kernel_shape, dtype=tf.float32) \n for k in BANDS\n ]\n\n imageFeaturesDict = dict(zip(BANDS, imageColumns))\n\n def parse_image(example_proto):\n return tf.io.parse_single_example(example_proto, imageFeaturesDict)\n\n def toTupleImage(inputs):\n inputsList = [inputs.get(key) for key in BANDS]\n stacked = tf.stack(inputsList, axis=0)\n stacked = tf.transpose(stacked, [1, 2, 0])\n return stacked\n\n # Create a dataset from the TFRecord file(s) in Cloud Storage.\n imageDataset = tf.data.TFRecordDataset(imageFilesList, compression_type='GZIP')\n imageDataset = imageDataset.map(parse_image, num_parallel_calls=5)\n imageDataset = imageDataset.map(toTupleImage).batch(1)\n\n # Perform inference.\n print('Running predictions...')\n predictions = model.predict(imageDataset, steps=patches, verbose=1)\n # print(predictions[0])\n\n print('Writing predictions...')\n out_image_file = op.join(ROOT_DIR, FOLDER, f'{out_image_base}pred.TFRecord')\n writer = tf.io.TFRecordWriter(out_image_file)\n patches = 0\n for predictionPatch in predictions:\n print('Writing patch ' + str(patches) + '...')\n predictionPatch = tf.argmax(predictionPatch, axis=2)\n\n # Create an example.\n example = tf.train.Example(\n features=tf.train.Features(\n feature={\n 'class': tf.train.Feature(\n float_list=tf.train.FloatList(\n value=predictionPatch.numpy().flatten()))\n }\n )\n )\n # Write the example.\n writer.write(example.SerializeToString())\n patches += 1\n\n writer.close()",
"_____no_output_____"
]
],
[
[
"Now there's all the code needed to run the prediction pipeline, all that remains is to specify the output region in which to do the prediction, the names of the output files, where to put them, and the shape of the outputs. ",
"_____no_output_____"
]
],
[
[
"# Base file name to use for TFRecord files and assets.\nimage_base = 'servir_inference_demo_'\n\n# South Africa (near training data)\nregion = ee.Geometry.Polygon(\n [[[\n 30.55984497070313,\n -29.177943749121233\n ],\n [\n 30.843429565429684,\n -29.177943749121233\n ],\n [\n 30.843429565429684,\n -28.994928377910732\n ],\n [\n 30.55984497070313,\n -28.994928377910732\n ]]], None, False)",
"_____no_output_____"
],
[
"# Run the export.\ndoExport(image_base, KERNEL_SHAPE, region)",
"_____no_output_____"
],
[
"# Run the prediction.\ndoPrediction(image_base, KERNEL_SHAPE, region)",
"_____no_output_____"
]
],
[
[
"# Display the output\n\nOne the data has been exported, the model has made predictions and the predictions have been written to a file, we need to [manually import the TFRecord to Earth Engine](https://developers.google.com/earth-engine/guides/tfrecord#uploading-tfrecords-to-earth-engine). Then we can display our crop type predictions as an image asset",
"_____no_output_____"
]
],
[
[
"out_image = ee.Image('users/drew/servir_inference_demo_-mixer')\nmapid = out_image.getMapId({'min': 0, 'max': 10, 'palette': ['00A600','63C600','E6E600','E9BD3A','ECB176','EFC2B3','F2F2F2']})\nmap = folium.Map(location=[ \n -29.177943749121233,\n 30.55984497070313,\n])\nfolium.TileLayer(\n tiles=mapid['tile_fetcher'].url_format,\n attr='Map Data © <a href=\"https://earthengine.google.com/\">Google Earth Engine</a>',\n overlay=True,\n name='predicted crop type',\n ).add_to(map)\nmap.add_child(folium.LayerControl())\nmap",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb9445a37d77546cac5c376afc4bceb5c2336dc7
| 30,530 |
ipynb
|
Jupyter Notebook
|
content/docs/data-science/labs/3-Understanding-to-Preparation-py.ipynb
|
nabeelvalley/docs
|
e953296c99ebbf495328bc47fb15757511d5766a
|
[
"Apache-2.0"
] | 3 |
2019-03-06T10:39:49.000Z
|
2020-05-26T14:12:01.000Z
|
content/docs/data-science/labs/3-Understanding-to-Preparation-py.ipynb
|
nabeelvalley/docs
|
e953296c99ebbf495328bc47fb15757511d5766a
|
[
"Apache-2.0"
] | 7 |
2020-05-15T06:20:07.000Z
|
2020-07-17T10:19:23.000Z
|
content/docs/data-science/labs/3-Understanding-to-Preparation-py.ipynb
|
nabeelvalley/docs
|
e953296c99ebbf495328bc47fb15757511d5766a
|
[
"Apache-2.0"
] | null | null | null | 24.601128 | 522 | 0.5585 |
[
[
[
"<a href=\"https://cognitiveclass.ai\"><img src = \"https://ibm.box.com/shared/static/9gegpsmnsoo25ikkbl4qzlvlyjbgxs5x.png\" width = 400> </a>\n\n<h1 align=center><font size = 5>From Understanding to Preparation</font></h1>",
"_____no_output_____"
],
[
"## Introduction\n\nIn this lab, we will continue learning about the data science methodology, and focus on the **Data Understanding** and the **Data Preparation** stages.",
"_____no_output_____"
],
[
"## Table of Contents\n\n<div class=\"alert alert-block alert-info\" style=\"margin-top: 20px\">\n1. [Recap](#0)<br>\n2. [Data Understanding](#2)<br>\n3. [Data Preparation](#4)<br>\n</div>\n<hr>",
"_____no_output_____"
],
[
"# Recap <a id=\"0\"></a>",
"_____no_output_____"
],
[
"In Lab **From Requirements to Collection**, we learned that the data we need to answer the question developed in the business understanding stage, namely *can we automate the process of determining the cuisine of a given recipe?*, is readily available. A researcher named Yong-Yeol Ahn scraped tens of thousands of food recipes (cuisines and ingredients) from three different websites, namely:",
"_____no_output_____"
],
[
"<img src = \"https://ibm.box.com/shared/static/4fruwan7wmjov3gywiz3swlojw0srv54.png\" width=500>\n\nwww.allrecipes.com\n\n<img src = \"https://ibm.box.com/shared/static/cebfdbr22fjxa47lltp0bs533r103g0z.png\" width=500>\n\nwww.epicurious.com\n\n<img src = \"https://ibm.box.com/shared/static/epk727njg7xrz49pbkpkzd05cm5ywqmu.png\" width=500>\n\nwww.menupan.com",
"_____no_output_____"
],
[
"For more information on Yong-Yeol Ahn and his research, you can read his paper on [Flavor Network and the Principles of Food Pairing](http://yongyeol.com/papers/ahn-flavornet-2011.pdf).",
"_____no_output_____"
],
[
"We also collected the data and placed it on an IBM server for your convenience.\n\n------------",
"_____no_output_____"
],
[
"# Data Understanding <a id=\"2\"></a>",
"_____no_output_____"
],
[
"<img src=\"https://ibm.box.com/shared/static/89geb3m0ge1z73s92hl8o8wdcpcrggtz.png\" width=500>",
"_____no_output_____"
],
[
"<strong> Important note:</strong> Please note that you are not expected to know how to program in python. The following code is meant to illustrate the stages of data understanding and data preparation, so it is totally fine if you do not understand the individual lines of code. We have a full course on programming in python, <a href=\"http://cocl.us/PY0101EN_DS0103EN_LAB3_PYTHON\">Python for Data Science</a>, so please feel free to complete the course if you are interested in learning how to program in python.",
"_____no_output_____"
],
[
"### Using this notebook:\n\nTo run any of the following cells of code, you can type **Shift + Enter** to excute the code in a cell.",
"_____no_output_____"
],
[
"Get the version of Python installed.",
"_____no_output_____"
]
],
[
[
"# check Python version\n!python -V",
"_____no_output_____"
]
],
[
[
"Download the library and dependencies that we will need to run this lab.",
"_____no_output_____"
]
],
[
[
"import pandas as pd # import library to read data into dataframe\npd.set_option('display.max_columns', None)\nimport numpy as np # import numpy library\nimport re # import library for regular expression",
"_____no_output_____"
]
],
[
[
"Download the data from the IBM server and read it into a *pandas* dataframe.",
"_____no_output_____"
]
],
[
[
"recipes = pd.read_csv(\"https://ibm.box.com/shared/static/5wah9atr5o1akuuavl2z9tkjzdinr1lv.csv\")\n\nprint(\"Data read into dataframe!\") # takes about 30 seconds",
"_____no_output_____"
]
],
[
[
"Show the first few rows.",
"_____no_output_____"
]
],
[
[
"recipes.head()",
"_____no_output_____"
]
],
[
[
"Get the dimensions of the dataframe.",
"_____no_output_____"
]
],
[
[
"recipes.shape",
"_____no_output_____"
]
],
[
[
"So our dataset consists of 57,691 recipes. Each row represents a recipe, and for each recipe, the corresponding cuisine is documented as well as whether 384 ingredients exist in the recipe or not, beginning with almond and ending with zucchini.",
"_____no_output_____"
],
[
"We know that a basic sushi recipe includes the ingredients:\n* rice\n* soy sauce\n* wasabi\n* some fish/vegetables",
"_____no_output_____"
],
[
"Let's check that these ingredients exist in our dataframe:",
"_____no_output_____"
]
],
[
[
"ingredients = list(recipes.columns.values)\n\nprint([match.group(0) for ingredient in ingredients for match in [(re.compile(\".*(rice).*\")).search(ingredient)] if match])\nprint([match.group(0) for ingredient in ingredients for match in [(re.compile(\".*(wasabi).*\")).search(ingredient)] if match])\nprint([match.group(0) for ingredient in ingredients for match in [(re.compile(\".*(soy).*\")).search(ingredient)] if match])",
"_____no_output_____"
]
],
[
[
"Yes, they do!\n\n* rice exists as rice.\n* wasabi exists as wasabi.\n* soy exists as soy_sauce.\n\nSo maybe if a recipe contains all three ingredients: rice, wasabi, and soy_sauce, then we can confidently say that the recipe is a **Japanese** cuisine! Let's keep this in mind!\n\n----------------",
"_____no_output_____"
],
[
"# Data Preparation <a id=\"4\"></a>",
"_____no_output_____"
],
[
"<img src=\"https://ibm.box.com/shared/static/lqc2j3r0ndhokh77mubohwjqybzf8dhk.png\" width=500>",
"_____no_output_____"
],
[
"In this section, we will prepare data for the next stage in the data science methodology, which is modeling. This stage involves exploring the data further and making sure that it is in the right format for the machine learning algorithm that we selected in the analytic approach stage, which is decision trees.",
"_____no_output_____"
],
[
"First, look at the data to see if it needs cleaning.",
"_____no_output_____"
]
],
[
[
"recipes[\"country\"].value_counts() # frequency table",
"_____no_output_____"
]
],
[
[
"By looking at the above table, we can make the following observations:\n\n1. Cuisine column is labeled as Country, which is inaccurate.\n2. Cuisine names are not consistent as not all of them start with an uppercase first letter.\n3. Some cuisines are duplicated as variation of the country name, such as Vietnam and Vietnamese.\n4. Some cuisines have very few recipes.",
"_____no_output_____"
],
[
"#### Let's fixes these problems.",
"_____no_output_____"
],
[
"Fix the name of the column showing the cuisine.",
"_____no_output_____"
]
],
[
[
"column_names = recipes.columns.values\ncolumn_names[0] = \"cuisine\"\nrecipes.columns = column_names\n\nrecipes",
"_____no_output_____"
]
],
[
[
"Make all the cuisine names lowercase.",
"_____no_output_____"
]
],
[
[
"recipes[\"cuisine\"] = recipes[\"cuisine\"].str.lower()",
"_____no_output_____"
]
],
[
[
"Make the cuisine names consistent.",
"_____no_output_____"
]
],
[
[
"recipes.loc[recipes[\"cuisine\"] == \"austria\", \"cuisine\"] = \"austrian\"\nrecipes.loc[recipes[\"cuisine\"] == \"belgium\", \"cuisine\"] = \"belgian\"\nrecipes.loc[recipes[\"cuisine\"] == \"china\", \"cuisine\"] = \"chinese\"\nrecipes.loc[recipes[\"cuisine\"] == \"canada\", \"cuisine\"] = \"canadian\"\nrecipes.loc[recipes[\"cuisine\"] == \"netherlands\", \"cuisine\"] = \"dutch\"\nrecipes.loc[recipes[\"cuisine\"] == \"france\", \"cuisine\"] = \"french\"\nrecipes.loc[recipes[\"cuisine\"] == \"germany\", \"cuisine\"] = \"german\"\nrecipes.loc[recipes[\"cuisine\"] == \"india\", \"cuisine\"] = \"indian\"\nrecipes.loc[recipes[\"cuisine\"] == \"indonesia\", \"cuisine\"] = \"indonesian\"\nrecipes.loc[recipes[\"cuisine\"] == \"iran\", \"cuisine\"] = \"iranian\"\nrecipes.loc[recipes[\"cuisine\"] == \"italy\", \"cuisine\"] = \"italian\"\nrecipes.loc[recipes[\"cuisine\"] == \"japan\", \"cuisine\"] = \"japanese\"\nrecipes.loc[recipes[\"cuisine\"] == \"israel\", \"cuisine\"] = \"jewish\"\nrecipes.loc[recipes[\"cuisine\"] == \"korea\", \"cuisine\"] = \"korean\"\nrecipes.loc[recipes[\"cuisine\"] == \"lebanon\", \"cuisine\"] = \"lebanese\"\nrecipes.loc[recipes[\"cuisine\"] == \"malaysia\", \"cuisine\"] = \"malaysian\"\nrecipes.loc[recipes[\"cuisine\"] == \"mexico\", \"cuisine\"] = \"mexican\"\nrecipes.loc[recipes[\"cuisine\"] == \"pakistan\", \"cuisine\"] = \"pakistani\"\nrecipes.loc[recipes[\"cuisine\"] == \"philippines\", \"cuisine\"] = \"philippine\"\nrecipes.loc[recipes[\"cuisine\"] == \"scandinavia\", \"cuisine\"] = \"scandinavian\"\nrecipes.loc[recipes[\"cuisine\"] == \"spain\", \"cuisine\"] = \"spanish_portuguese\"\nrecipes.loc[recipes[\"cuisine\"] == \"portugal\", \"cuisine\"] = \"spanish_portuguese\"\nrecipes.loc[recipes[\"cuisine\"] == \"switzerland\", \"cuisine\"] = \"swiss\"\nrecipes.loc[recipes[\"cuisine\"] == \"thailand\", \"cuisine\"] = \"thai\"\nrecipes.loc[recipes[\"cuisine\"] == \"turkey\", \"cuisine\"] = \"turkish\"\nrecipes.loc[recipes[\"cuisine\"] == \"vietnam\", \"cuisine\"] = \"vietnamese\"\nrecipes.loc[recipes[\"cuisine\"] == \"uk-and-ireland\", \"cuisine\"] = \"uk-and-irish\"\nrecipes.loc[recipes[\"cuisine\"] == \"irish\", \"cuisine\"] = \"uk-and-irish\"\n\nrecipes",
"_____no_output_____"
]
],
[
[
"Remove cuisines with < 50 recipes.",
"_____no_output_____"
]
],
[
[
"# get list of cuisines to keep\nrecipes_counts = recipes[\"cuisine\"].value_counts()\ncuisines_indices = recipes_counts > 50\n\ncuisines_to_keep = list(np.array(recipes_counts.index.values)[np.array(cuisines_indices)])",
"_____no_output_____"
],
[
"rows_before = recipes.shape[0] # number of rows of original dataframe\nprint(\"Number of rows of original dataframe is {}.\".format(rows_before))\n\nrecipes = recipes.loc[recipes['cuisine'].isin(cuisines_to_keep)]\n\nrows_after = recipes.shape[0] # number of rows of processed dataframe\nprint(\"Number of rows of processed dataframe is {}.\".format(rows_after))\n\nprint(\"{} rows removed!\".format(rows_before - rows_after))",
"_____no_output_____"
]
],
[
[
"Convert all Yes's to 1's and the No's to 0's",
"_____no_output_____"
]
],
[
[
"recipes = recipes.replace(to_replace=\"Yes\", value=1)\nrecipes = recipes.replace(to_replace=\"No\", value=0)",
"_____no_output_____"
]
],
[
[
"#### Let's analyze the data a little more in order to learn the data better and note any interesting preliminary observations.",
"_____no_output_____"
],
[
"Run the following cell to get the recipes that contain **rice** *and* **soy** *and* **wasabi** *and* **seaweed**.",
"_____no_output_____"
]
],
[
[
"recipes.head()",
"_____no_output_____"
],
[
"check_recipes = recipes.loc[\n (recipes[\"rice\"] == 1) &\n (recipes[\"soy_sauce\"] == 1) &\n (recipes[\"wasabi\"] == 1) &\n (recipes[\"seaweed\"] == 1)\n]\n\ncheck_recipes",
"_____no_output_____"
]
],
[
[
"Based on the results of the above code, can we classify all recipes that contain **rice** *and* **soy** *and* **wasabi** *and* **seaweed** as **Japanese** recipes? Why?",
"_____no_output_____"
]
],
[
[
"Your Answer:",
"_____no_output_____"
]
],
[
[
"Double-click __here__ for the solution.\n<!-- The correct answer is:\nNo, because other recipes such as Asian and East_Asian recipes also contain these ingredients.\n-->",
"_____no_output_____"
],
[
"Let's count the ingredients across all recipes.",
"_____no_output_____"
]
],
[
[
"# sum each column\ning = recipes.iloc[:, 1:].sum(axis=0)",
"_____no_output_____"
],
[
"# define each column as a pandas series\ningredient = pd.Series(ing.index.values, index = np.arange(len(ing)))\ncount = pd.Series(list(ing), index = np.arange(len(ing)))\n\n# create the dataframe\ning_df = pd.DataFrame(dict(ingredient = ingredient, count = count))\ning_df = ing_df[[\"ingredient\", \"count\"]]\nprint(ing_df.to_string())",
"_____no_output_____"
]
],
[
[
"Now we have a dataframe of ingredients and their total counts across all recipes. Let's sort this dataframe in descending order.",
"_____no_output_____"
]
],
[
[
"ing_df.sort_values([\"count\"], ascending=False, inplace=True)\ning_df.reset_index(inplace=True, drop=True)\n\nprint(ing_df)",
"_____no_output_____"
]
],
[
[
"#### What are the 3 most popular ingredients?",
"_____no_output_____"
]
],
[
[
"Your Answer:\n1.\n\n2.\n\n3.",
"_____no_output_____"
]
],
[
[
"Double-click __here__ for the solution.\n<!-- The correct answer is:\n// 1. Egg with <strong>21,025</strong> occurrences. \n// 2. Wheat with <strong>20,781</strong> occurrences. \n// 3. Butter with <strong>20,719</strong> occurrences.\n-->",
"_____no_output_____"
],
[
"However, note that there is a problem with the above table. There are ~40,000 American recipes in our dataset, which means that the data is biased towards American ingredients.",
"_____no_output_____"
],
[
"**Therefore**, let's compute a more objective summary of the ingredients by looking at the ingredients per cuisine.",
"_____no_output_____"
],
[
"#### Let's create a *profile* for each cuisine.\n\nIn other words, let's try to find out what ingredients Chinese people typically use, and what is **Canadian** food for example.",
"_____no_output_____"
]
],
[
[
"cuisines = recipes.groupby(\"cuisine\").mean()\ncuisines.head()",
"_____no_output_____"
]
],
[
[
"As shown above, we have just created a dataframe where each row is a cuisine and each column (except for the first column) is an ingredient, and the row values represent the percentage of each ingredient in the corresponding cuisine.\n\n**For example**:\n\n* *almond* is present across 15.65% of all of the **African** recipes.\n* *butter* is present across 38.11% of all of the **Canadian** recipes.",
"_____no_output_____"
],
[
"Let's print out the profile for each cuisine by displaying the top four ingredients in each cuisine.",
"_____no_output_____"
]
],
[
[
"num_ingredients = 4 # define number of top ingredients to print\n\n# define a function that prints the top ingredients for each cuisine\ndef print_top_ingredients(row):\n print(row.name.upper())\n row_sorted = row.sort_values(ascending=False)*100\n top_ingredients = list(row_sorted.index.values)[0:num_ingredients]\n row_sorted = list(row_sorted)[0:num_ingredients]\n\n for ind, ingredient in enumerate(top_ingredients):\n print(\"%s (%d%%)\" % (ingredient, row_sorted[ind]), end=' ')\n print(\"\\n\")\n\n# apply function to cuisines dataframe\ncreate_cuisines_profiles = cuisines.apply(print_top_ingredients, axis=1)",
"_____no_output_____"
]
],
[
[
"At this point, we feel that we have understood the data well and the data is ready and is in the right format for modeling!\n\n-----------",
"_____no_output_____"
],
[
"### Thank you for completing this lab!\n\nThis notebook was created by [Alex Aklson](https://www.linkedin.com/in/aklson/). We hope you found this lab session interesting. Feel free to contact us if you have any questions!",
"_____no_output_____"
],
[
"This notebook is part of the free course on **Cognitive Class** called *Data Science Methodology*. If you accessed this notebook outside the course, you can take this free self-paced course, online by clicking [here](https://cocl.us/DS0103EN_LAB3_PYTHON).",
"_____no_output_____"
],
[
"<hr>\nCopyright © 2018 [Cognitive Class](https://cognitiveclass.ai/?utm_source=bducopyrightlink&utm_medium=dswb&utm_campaign=bdu). This notebook and its source code are released under the terms of the [MIT License](https://bigdatauniversity.com/mit-license/).",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"raw",
"markdown",
"code",
"markdown",
"code",
"markdown",
"raw",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"raw"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"raw"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
]
] |
cb944e8161328436f4182d431d50c21d38129a39
| 41,410 |
ipynb
|
Jupyter Notebook
|
tinyNN.ipynb
|
Clark-Whitehead/tinyNN
|
3731f8827106c25f68051320f67efd9d7cc718a9
|
[
"MIT"
] | null | null | null |
tinyNN.ipynb
|
Clark-Whitehead/tinyNN
|
3731f8827106c25f68051320f67efd9d7cc718a9
|
[
"MIT"
] | null | null | null |
tinyNN.ipynb
|
Clark-Whitehead/tinyNN
|
3731f8827106c25f68051320f67efd9d7cc718a9
|
[
"MIT"
] | null | null | null | 131.044304 | 32,734 | 0.840546 |
[
[
[
"",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport keras\nfrom keras.models import Sequential\nfrom keras.layers import Dense",
"_____no_output_____"
],
[
"x_train = np.array([[150, 72],\n [125, 55],\n [115, 50],\n [165, 75],\n [180, 60],\n [135, 56]])\n \ny_train = np.array([1,0,0,1,1,0])\n",
"_____no_output_____"
]
],
[
[
"## Scale Data using z = (x - m) / s where m = mean, s = standard deviation",
"_____no_output_____"
]
],
[
[
"class scale:\n\n def mean(self, x):\n return x.sum(axis=0) / x_train.shape[0]\n\n def std(self, x):\n\n m = self.mean(x)\n\n x_minus_mean = x - m\n\n x_minus_mean2 = x_minus_mean ** 2\n\n summation = x_minus_mean2.sum(axis=0)\n\n divide_by_N = summation / (x_train.shape[0])\n\n standard_deviation = np.sqrt(divide_by_N)\n\n return standard_deviation\n\n def scale_data(self, x):\n return (x - self.mean(x)) / self.std(x)\n\nscaler = scale()",
"_____no_output_____"
],
[
"print(\"MEAN = {}\".format(scaler.mean(x_train)))\nprint(\"STD = {}\".format(scaler.std(x_train)))",
"MEAN = [145. 61.33333333]\nSTD = [22.54624876 9.12262146]\n"
],
[
"x_train = scaler.scale_data(x_train)\n\nprint(x_train)",
"[[ 0.22176638 1.16925455]\n [-0.88706553 -0.69424489]\n [-1.33059829 -1.24233296]\n [ 0.88706553 1.4981074 ]\n [ 1.55236467 -0.14615682]\n [-0.44353276 -0.58462728]]\n"
],
[
"#x_train shape = 11x4\n# x_train = np.array([[1.2, 4.3, 2.1, 1.9],\n# [6.2, 8.3, 5.1, 9.9],\n# [2.3, 4.3, 3.1, 0.9],\n# [4.1, 4.4, 1.1, 0.3],\n# [6.1, 7.1, 8.1, 9.1],\n# [1.0, 2.0, 1.0, 1.0],\n# [5.1, 5.1, 5.1, 5.1],\n# [1.8, 4.0, 3.9, 2.7],\n# [4.4, 0.8, 1.9, 2.7],\n# [6.9, 8.8, 5.7, 7.1]])\n\n# #y_train shape = 1x11\n# y_train = np.array([[1,0],\n# [0,1],\n# [1,0],\n# [1,0],\n# [0,1],\n# [1,0],\n# [0,1],\n# [1,0],\n# [1,0],\n# [0,1]])\n ",
"_____no_output_____"
],
[
"class tinyNN:\n def __init__(self):\n \n np.random.seed(1)\n\n self.w1 = (2 * np.random.rand(2,50)) - 1\n self.w2 = (2 * np.random.rand(50,1)) - 1\n\n self.lr = .001\n\n def tanh(self, x):\n return (np.exp(x)-np.exp(-x)) / (np.exp(x)+np.exp(-x))\n\n def derivTanh(self, x):\n return (4*np.exp(2*x)) / ((np.exp(2*x) + 1)**2)\n\n def sigmoid(self, x):\n return 1 / (1+np.exp(-x))\n\n def derivSigmoid(self, x):\n return np.exp(x) / (np.exp(x) + 1)**2\n\n def loss(self, y_pred, y_true):\n return (y_pred - y_true)**2\n\n def derivLoss(self, error):\n return error\n\n def forward(self, inputs):\n \n self.hidden = np.dot(inputs, self.w1)\n self.h1 = self.tanh(self.hidden)\n\n outputs = np.dot(self.h1, self.w2)\n \n return outputs\n\n def calcError(self, y_pred, y):\n\n return y_pred.T - y\n\n def backProp(self, error):\n\n dypred_dw2 = self.h1\n dse_dypred = error\n\n delta_w2 = np.dot(dypred_dw2.T, dse_dypred.T)\n\n dse_dh = np.dot(dse_dypred.T, self.w2.T)\n\n dh_dp = self.derivTanh(self.hidden)\n\n dse_dp = dse_dh * dh_dp\n\n delta_w1 = np.dot(x_train.T, dse_dp)\n\n self.w2 -= self.lr * delta_w2\n self.w1 -= self.lr * delta_w1\n\n\n \n\ngarvis = tinyNN()",
"_____no_output_____"
],
[
"for i in range(1000):\n\n outputs = garvis.forward(x_train)\n error = garvis.calcError(outputs, y_train)\n garvis.backProp(error)\n\noutputs = garvis.forward(x_train)\n\nprint(outputs)\n\noutputs = np.where(outputs >= .5, 1, 0)\n\noutputs = np.where(outputs == 1, \"Male\", \"Female\")\n\nprint(outputs)",
"[[ 0.93975543]\n [-0.02139624]\n [-0.12630504]\n [ 0.84544591]\n [ 0.9055826 ]\n [-0.19441367]]\n[['Male']\n ['Female']\n ['Female']\n ['Male']\n ['Male']\n ['Female']]\n"
]
]
] |
[
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb945106bd6b7ddd79683ad8194f8ff9839b9188
| 8,925 |
ipynb
|
Jupyter Notebook
|
Data Science Academy/PythonFundamentos/Cap04/Notebooks/DSA-Python-Cap04-Exercicios-Solucao.ipynb
|
ivanbergon/Github_privado
|
2468e4914dfdf41dbcdc7b964f863cb2c3f375f5
|
[
"MIT"
] | null | null | null |
Data Science Academy/PythonFundamentos/Cap04/Notebooks/DSA-Python-Cap04-Exercicios-Solucao.ipynb
|
ivanbergon/Github_privado
|
2468e4914dfdf41dbcdc7b964f863cb2c3f375f5
|
[
"MIT"
] | null | null | null |
Data Science Academy/PythonFundamentos/Cap04/Notebooks/DSA-Python-Cap04-Exercicios-Solucao.ipynb
|
ivanbergon/Github_privado
|
2468e4914dfdf41dbcdc7b964f863cb2c3f375f5
|
[
"MIT"
] | null | null | null | 23.8 | 129 | 0.487507 |
[
[
[
"# <font color='blue'>Data Science Academy - Python Fundamentos - Capítulo 4</font>\n\n## Download: http://github.com/dsacademybr",
"_____no_output_____"
]
],
[
[
"# Versão da Linguagem Python\nfrom platform import python_version\nprint('Versão da Linguagem Python Usada Neste Jupyter Notebook:', python_version())",
"Versão da Linguagem Python Usada Neste Jupyter Notebook: 3.8.8\n"
]
],
[
[
"# Versão da Linguagem Python\nfrom platform import python_version\nprint('Versão da Linguagem Python Usada Neste Jupyter Notebook:', python_version())",
"_____no_output_____"
],
[
"## Exercícios ",
"_____no_output_____"
]
],
[
[
"# Exercício 1 - Crie uma lista de 3 elementos e calcule a terceira potência de cada elemento.\nlist1 = [3,4,5]\nquadrado = [item**3 for item in list1] \nprint(quadrado)",
"[27, 64, 125]\n"
],
[
"# Exercício 2 - Reescreva o código abaixo, usando a função map(). O resultado final deve ser o mesmo!\npalavras = 'A Data Science Academy oferce os melhores cursos de análise de dados do Brasil'.split()\nresultado = [[w.upper(), w.lower(), len(w)] for w in palavras]\nfor i in resultado:\n print (i)",
"['A', 'a', 1]\n['DATA', 'data', 4]\n['SCIENCE', 'science', 7]\n['ACADEMY', 'academy', 7]\n['OFERCE', 'oferce', 6]\n['OS', 'os', 2]\n['MELHORES', 'melhores', 8]\n['CURSOS', 'cursos', 6]\n['DE', 'de', 2]\n['ANÁLISE', 'análise', 7]\n['DE', 'de', 2]\n['DADOS', 'dados', 5]\n['DO', 'do', 2]\n['BRASIL', 'brasil', 6]\n"
],
[
"resultado = map(lambda w: [w.upper(), w.lower(), len(w)], palavras)\nfor i in resultado:\n print (i)",
"['A', 'a', 1]\n['DATA', 'data', 4]\n['SCIENCE', 'science', 7]\n['ACADEMY', 'academy', 7]\n['OFERCE', 'oferce', 6]\n['OS', 'os', 2]\n['MELHORES', 'melhores', 8]\n['CURSOS', 'cursos', 6]\n['DE', 'de', 2]\n['ANÁLISE', 'análise', 7]\n['DE', 'de', 2]\n['DADOS', 'dados', 5]\n['DO', 'do', 2]\n['BRASIL', 'brasil', 6]\n"
],
[
"# Exercício 3 - Calcule a matriz transposta da matriz abaixo.\n# Caso não saiba o que é matriz transposta, visite este link: https://pt.wikipedia.org/wiki/Matriz_transposta\n# Matriz transposta é um conceito fundamental na construção de redes neurais artificiais, base de sistemas de IA.\nmatrix = [[1, 2],[3,4],[5,6],[7,8]]\ntranspose = [[row[i] for row in matrix] for i in range(2)]\nprint(transpose)",
"[[1, 3, 5, 7], [2, 4, 6, 8]]\n"
],
[
"# Exercício 4 - Crie duas funções, uma para elevar um número ao quadrado e outra para elevar ao cubo. \n# Aplique as duas funções aos elementos da lista abaixo. \n# Obs: as duas funções devem ser aplicadas simultaneamente.\nlista = [0, 1, 2, 3, 4]\n\ndef square(x):\n return (x**2)\n \ndef cube(x):\n return (x**3)\n\nfuncs = [square, cube]\n\nfor i in lista:\n valor = map(lambda x: x(i), funcs)\n print(list((valor)))",
"[0, 0]\n[1, 1]\n[4, 8]\n[9, 27]\n[16, 64]\n"
],
[
"# Exercício 5 - Abaixo você encontra duas listas. Faça com que cada elemento da listaA seja elevado \n# ao elemento correspondente na listaB.\nlistaA = [2, 3, 4]\nlistaB = [10, 11, 12]\nlist(map(pow, listaA, listaB))",
"_____no_output_____"
],
[
"# Exercício 6 - Considerando o range de valores abaixo, use a função filter() para retornar apenas os valores negativos.\nrange(-5, 5)\nlist(filter((lambda x: x < 0), range(-5,5)))",
"_____no_output_____"
],
[
"# Exercício 7 - Usando a função filter(), encontre os valores que são comuns às duas listas abaixo.\na = [1,2,3,5,7,9]\nb = [2,3,5,6,7,8]\nprint (list(filter(lambda x: x in a, b)))",
"[2, 3, 5, 7]\n"
],
[
"# Exercício 8 - Considere o código abaixo. Obtenha o mesmo resultado usando o pacote time. \n# Não conhece o pacote time? Pesquise!\nimport datetime\nprint (datetime.datetime.now().strftime(\"%d/%m/%Y %H:%M\"))\n\nimport time\nprint (time.strftime(\"%d/%m/%Y %H:%M\"))",
"05/07/2021 12:12\n05/07/2021 12:12\n"
],
[
"# Exercício 9 - Considere os dois dicionários abaixo. \n# Crie um terceiro dicionário com as chaves do dicionário 1 e os valores do dicionário 2.\ndict1 = {'a':1,'b':2}\ndict2 = {'c':4,'d':5}\n\ndef trocaValores(d1, d2):\n dicTemp = {}\n \n for d1key, d2val in zip(d1,d2.values()):\n dicTemp[d1key] = d2val\n \n return dicTemp\n\ndict3 = trocaValores(dict1, dict2)\nprint(dict3)",
"{'a': 4, 'b': 5}\n"
],
[
"# Exercício 10 - Considere a lista abaixo e retorne apenas os elementos cujo índice for maior que 5.\nlista = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h']\nfor indice, valor in enumerate(lista):\n if indice <= 5:\n continue\n else:\n print (valor)",
"g\nh\n"
]
],
[
[
"# Fim",
"_____no_output_____"
],
[
"### Obrigado\n\n### Visite o Blog da Data Science Academy - <a href=\"http://blog.dsacademy.com.br\">Blog DSA</a>",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown",
"markdown"
]
] |
cb94625e6480f76a5a918a701c563e1fd43943f7
| 30,357 |
ipynb
|
Jupyter Notebook
|
main-Copy1.ipynb
|
GreggRoll/GEO-CSV-EXPLOITER
|
090d33d66b8c2925ab90638933181ad7cebee2a1
|
[
"MIT"
] | null | null | null |
main-Copy1.ipynb
|
GreggRoll/GEO-CSV-EXPLOITER
|
090d33d66b8c2925ab90638933181ad7cebee2a1
|
[
"MIT"
] | null | null | null |
main-Copy1.ipynb
|
GreggRoll/GEO-CSV-EXPLOITER
|
090d33d66b8c2925ab90638933181ad7cebee2a1
|
[
"MIT"
] | null | null | null | 42.636236 | 1,663 | 0.58596 |
[
[
[
"import pandas as pd\nimport bs4\nimport itertools\nfrom scipy.stats import chi2_contingency",
"_____no_output_____"
]
],
[
[
"## Report Writing Defs (To be moved)",
"_____no_output_____"
]
],
[
[
"from scipy import stats\nfrom scipy.stats import linregress\nfrom scipy.stats import chi2_contingency\nimport statsmodels.api as sm\nfrom statsmodels.formula.api import ols\n\n\n# Creating a function that provide an overview of the data\ndef overview(df, numerical_variable, report):\n data_head = df.head()\n data_shape = df.shape\n data_type = df.dtypes\n df = (df.drop(numerical_variable, axis=1).join(df[numerical_variable].apply(pd.to_numeric, errors='coerce'))) # Converts any non-numeric values in a numerical column into NaN\n null_values = df.isnull().sum()\n zero_prop = ((df[df == 0].count(axis=0)/len(df.index)).round(2)* 100)\n data_summary = df.describe()\n report.write(f\"\"\"<h1>Exploratory data analysis summary of {file_name}</h1><br>\n <h3>The first 5 rows of content comprise of:</h3>\n <table>{data_head.to_html()}</table><br>\n <h3>There are a total of {data_shape[0]} rows and {data_shape[1]} columns.</h3><br>\n <table>\n <tr>\n <th>The data type for each column is:</th>\n <th>Number of NaN values for each column:</th>\n <th>% of zeros in each column:</th>\n </tr><tr>\n <td>{data_type.to_frame().to_html()}</td>\n <td>{null_values.to_frame().to_html()}</td>\n <td>{zero_prop.to_frame().to_html()}</td>\n </tr>\n </table><br>\n <h3>The summary of data:</h3>\n <table>{data_summary.to_html()}</table>\"\"\")\n return df\n\n\n# Creating report for correlation\ndef run(num_var_combination, catnum_combination, cat_var_combination, report,data):\n## For numeric variables\n# Pearson correlation (Numerical)\n report.write(\"<h1>Pearson Correlation Summary (numerical vs. numerical)</h1>\")\n for i in num_var_combination:\n var1 = i[0]\n var2 = i[1]\n pearson_data = linregress(data[var1], data[var2])\n pearson_r2, pearson_pvalue = ((pearson_data[2]**2), pearson_data[3])\n report.write(f\"<p>The Pearson R_Square and Pearson P-values between {var1} and {var2} are {pearson_r2} and {pearson_pvalue} respectively.</p>\")\n\n# Spearsman correlation (Ordinal)\n report.write(\"<h1>Spearsman Correlation Summary (Ordinal numerical vs. numerical)</h1>\")\n for q in num_var_combination:\n var1 = q[0]\n var2 = q[1]\n spearsman_data = stats.spearmanr(data[var1], data[var2])\n spearsman_r2, spearsman_pvalue = ((spearsman_data[0]**2), spearsman_data[1])\n report.write(f\"<p>The Spearsman R_Square and Spearsman P-values between {var1} and {var2} are {spearsman_r2} and {spearsman_pvalue} respectively.</p>\")\n \n ## For categorical-categorical variables\n # Chi-Sq test\n report.write(\"<h1>Chi Square Test Correlation Summary (categorical - categorical)</h1>\")\n for k in cat_var_combination:\n cat1 = k[0]\n cat2 = k[1]\n chi_sq = pd.crosstab(data[cat1], data[cat2])\n chi_sq_result = chi2_contingency(chi_sq)\n report.write(f\"<p>The Chi-Square P-value between {cat1} and {cat2} is {chi_sq_result[1]}.\")\n \n# ## For numeric-categorical variables\n# # ONE WAY ANOVA (Cat-num variables)\n# report.write(\"\\n\\n\\n\\n__________Correlation Summary (One Way ANOVA)__________\")\n# for j in catnum_combination:\n# var1 = j[0]\n# var2 = j[1]\n# lm = ols('{} ~ {}'.format(var1,var2), data = data).fit()\n# table = sm.stats.anova_lm(lm)\n# one_way_anova_pvalue = table.loc[var2,'PR(>F)']\n# report.write(\"\\n\\nThe One Way ANOVA P-value between {} and {} is {}.\"\n# .format(var1, var2, one_way_anova_pvalue))\n report.write(\"<h1>Summary Complete</h1>\")\n\n report.close() ",
"_____no_output_____"
]
],
[
[
"## Plot Creation Defs (to be moved)",
"_____no_output_____"
]
],
[
[
"#TODO: Create HTML docustructure to contain all plots\n#TODO: Change to accept path\n#TODO: Change plots to be scrollable\nimport seaborn as sns\nimport itertools\nfrom sklearn.preprocessing import LabelEncoder\nimport matplotlib.pyplot as plt\n\n# Create a function for plots\n ## Look into making count plots for categorical data\n ## Look into making scatterplots and barplot for numerical data\n ## Export plots into plot path\ndef plot_run(data,categorical_variable,numerical_variable):\n \n ## Set Unique categorical values that are < 5 as hue\n hue_lst = []\n for x in categorical_variable:\n if len(set(data[x])) <= 5: # if we have less than 5 unique values, we will use it for hue attributes\n hue_lst.append(x)\n ## Creating possible combinations among a list of numerical variables\n num_var_combination = list(itertools.combinations(numerical_variable, 2))\n ## Creating possible combinations among a list of categorical variables\n cat_var_combination = list(itertools.combinations(categorical_variable, 2))\n ## Creating possible combinations among a list of numerical and categorical variuable\n catnum_combination = list(itertools.product(numerical_variable, categorical_variable))\n\n ## Using scatterplot for numerical-numerical variables\n if len(categorical_variable) > 1: \n num_var_hue_combination = list(itertools.product(num_var_combination, hue_lst))\n for i in num_var_hue_combination:\n var1 = i[0][0]\n var2 = i[0][1]\n hue1 = i[1]\n plot1 = sns.scatterplot(data = data, x = var1, y = var2, hue = hue1, figsize = (len(), len()))\n plot1.set_xticklabels(plot1.get_xticklabels(), rotation=45, horizontalalignment='right')\n fig1 = plot1.get_figure()\n fig1.savefig(\"plots/{} vs {} by {} scatterplot.png\".format(var1,var2, hue1))\n fig1.clf()\n else:\n for l in num_var_combination:\n var1 = l[0]\n var2 = l[1]\n plot1 = sns.scatterplot(data = data, x = var1, y = var2)\n plot1.set_xticklabels(plot1.get_xticklabels(), rotation=45, horizontalalignment='right')\n fig1 = plot1.get_figure()\n fig1.savefig(\"plots/{} vs {} scatterplot.png\".format(var1,var2))\n fig1.clf()\n\n\n ## Using countplot for categorical data\n for j in categorical_variable:\n plot2 = sns.countplot(data = data, x = j)\n plot2.set_xticklabels(plot2.get_xticklabels(), rotation=45, horizontalalignment='right')\n fig2 = plot2.get_figure()\n fig2.savefig(\"plots/{}_countplot.png\".format(j))\n fig2.clf()\n\n ## Using boxplot for numerical + Categorical data\n for k in catnum_combination:\n num1 = k[0]\n cat1 = k[1]\n plt.figure(figsize=(len(cat1)/3, 6))\n plot3 = sns.boxplot(data = data, x = cat1, y = num1)\n plot3.set_xticklabels(plot3.get_xticklabels(), rotation=45, horizontalalignment='right')\n fig3 = plot3.get_figure()\n fig3.savefig(\"plots/{}_{}_barplot.png\".format(num1,cat1))\n fig3.clf()\n\n ## Creating heatmap to show correlation\n le = LabelEncoder()\n for cat in data[categorical_variable]:\n data[cat] = le.fit_transform(data[cat])\n plt.figure(figsize=(15,10))\n corrMatrix = data.corr()\n plot4 = sns.heatmap(corrMatrix, annot=True)\n plot4.set_xticklabels(plot4.get_xticklabels(), rotation=45, horizontalalignment='right')\n fig4 = plot4.get_figure()\n fig4.savefig(\"plots/heatplot.png\")\n fig4.clf()",
"_____no_output_____"
]
],
[
[
"# Data Intake",
"_____no_output_____"
]
],
[
[
"file_name = 'Video Game Sales'\ndata = pd.read_csv(r\"C:\\Users\\602387\\Desktop\\Data sets\\vgsales.csv\")",
"_____no_output_____"
],
[
"#Defining columns as categorical or numerical\n#TODO: def geo_variable\ndef cat_variable(df):\n return list(df.select_dtypes(include = ['category', 'object']))\n\ndef num_variable(df):\n return list(df.select_dtypes(exclude = ['category', 'object']))\n\ncategorical_variable = cat_variable(data)\nnumerical_variable = num_variable(data)",
"_____no_output_____"
],
[
"print(categorical_variable)\nprint(numerical_variable)",
"['Name', 'Platform', 'Genre', 'Publisher']\n['Rank', 'Year', 'NA_Sales', 'EU_Sales', 'JP_Sales', 'Other_Sales', 'Global_Sales']\n"
],
[
"#TODO create tab structure for Summary, Correlations, Plots\n#creating blank report\nreport = open(\"report.html\", \"w\")\n#template for report\nreport.write(\"\"\"\n<!DOCTYPE html>\n<html>\n<head>\n <style>\n table {\n font-family: \"Trebuchet MS\", Arial, Helvetica, sans-serif;\n border-collapse: collapse;\n \n }\n\n td, th {\n border: 1px solid #ddd;\n padding: 8px;\n }\n\n tr:nth-child(even){background-color: #f2f2f2;}\n\n tr:hover {background-color: #ddd;}\n\n th {\n padding-top: 12px;\n padding-bottom: 12px;\n text-align: left;\n background-color: #269bd1;\n color: white;\n }\n </style>\n <title>Correlation Reprot</title>\n</head>\n<body>\n</body>\n</html>\"\"\")",
"_____no_output_____"
],
[
"# Execute overview function in model module\ndata = overview(data, numerical_variable, report)",
"_____no_output_____"
],
[
"## Creating possible combinations among a list of numerical variables\nnum_var_combination = list(itertools.combinations(numerical_variable, 2))\n\n## Creating possible combinations among a list of categorical variables\ncat_var_combination = list(itertools.combinations(categorical_variable, 2))\n\n## Creating possible combinations among a list of numerical and categorical variuable\ncatnum_combination = list(itertools.product(numerical_variable, categorical_variable))",
"_____no_output_____"
],
[
"## Running the report now\nrun(num_var_combination,catnum_combination,cat_var_combination,report,data)",
"C:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\scipy\\stats\\_distn_infrastructure.py:903: RuntimeWarning: invalid value encountered in greater\n return (a < x) & (x < b)\nC:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\scipy\\stats\\_distn_infrastructure.py:903: RuntimeWarning: invalid value encountered in less\n return (a < x) & (x < b)\nC:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\scipy\\stats\\_distn_infrastructure.py:1912: RuntimeWarning: invalid value encountered in less_equal\n cond2 = cond0 & (x <= _a)\n"
],
[
"## Running Plots\nplot_run(data, categorical_variable,numerical_variable)",
"C:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:62: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\nC:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:62: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\nC:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:62: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\nC:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:62: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\nC:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:62: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\nC:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:62: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\nC:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:62: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\nC:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:62: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\nC:\\Users\\602387\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:62: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\n"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb9478f3b6949e93b8b07eddea9e6ccfe894e9d3
| 3,968 |
ipynb
|
Jupyter Notebook
|
Python basics practice/Python 3 (2)/Arithmetic Operators - Exercise_Py3.ipynb
|
rachithh/data-science
|
1f7c5678094fc3acfda30cb00f9de93a2974f505
|
[
"MIT"
] | null | null | null |
Python basics practice/Python 3 (2)/Arithmetic Operators - Exercise_Py3.ipynb
|
rachithh/data-science
|
1f7c5678094fc3acfda30cb00f9de93a2974f505
|
[
"MIT"
] | null | null | null |
Python basics practice/Python 3 (2)/Arithmetic Operators - Exercise_Py3.ipynb
|
rachithh/data-science
|
1f7c5678094fc3acfda30cb00f9de93a2974f505
|
[
"MIT"
] | null | null | null | 15.087452 | 54 | 0.435736 |
[
[
[
"## Arithmetic operators",
"_____no_output_____"
],
[
"Combine 15 and 23.",
"_____no_output_____"
]
],
[
[
"print(1523)",
"31\n"
]
],
[
[
"Subtract 50 from 26.",
"_____no_output_____"
]
],
[
[
"26-50",
"_____no_output_____"
]
],
[
[
"Divide 20 by 4.",
"_____no_output_____"
]
],
[
[
"20/4",
"_____no_output_____"
]
],
[
[
"Divide 22 by 4.",
"_____no_output_____"
]
],
[
[
"22/4",
"_____no_output_____"
]
],
[
[
"Obtain the remainder of the division of 22 by 4.",
"_____no_output_____"
]
],
[
[
"22%4",
"_____no_output_____"
]
],
[
[
"Divide the float 22 by 4.",
"_____no_output_____"
]
],
[
[
"float(22)/4",
"_____no_output_____"
],
[
"6*8",
"_____no_output_____"
]
],
[
[
"Multiply 6 by 8.",
"_____no_output_____"
],
[
"Raise 15 to the power of 2.",
"_____no_output_____"
]
],
[
[
"15**2",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
]
] |
cb94859ceed029f6ee580c9d81442c8ee66f1721
| 27,377 |
ipynb
|
Jupyter Notebook
|
09. Cloud/bnp/BNP_XGB_No tuning.ipynb
|
vm1729/Openthink
|
63b3ae65c85b88bdffb71f5789a7799ae389fc82
|
[
"MIT"
] | 1 |
2021-01-14T14:32:20.000Z
|
2021-01-14T14:32:20.000Z
|
09. Cloud/bnp/BNP_XGB_No tuning.ipynb
|
vm1729/Openthink
|
63b3ae65c85b88bdffb71f5789a7799ae389fc82
|
[
"MIT"
] | 2 |
2020-09-12T10:45:18.000Z
|
2020-09-12T10:45:18.000Z
|
09. Cloud/bnp/BNP_XGB_No tuning.ipynb
|
vm1729/Openthink
|
63b3ae65c85b88bdffb71f5789a7799ae389fc82
|
[
"MIT"
] | 2 |
2020-04-21T15:09:00.000Z
|
2020-12-01T07:27:10.000Z
| 59.775109 | 6,296 | 0.689082 |
[
[
[
"import pandas as pd\n\nfrom sklearn import model_selection\nfrom sklearn.metrics import accuracy_score\nfrom sklearn.preprocessing import LabelEncoder\n\nimport xgboost as xgb\nfrom xgboost.sklearn import XGBClassifier\nfrom sklearn import metrics #Additional scklearn functions\n#from sklearn import cross_validation\nfrom sklearn.model_selection import GridSearchCV",
"_____no_output_____"
],
[
"target='target'\nIDcol='ID'\ntrain=pd.read_csv('train_modified2.csv')",
"_____no_output_____"
],
[
"train.head()",
"_____no_output_____"
],
[
"test=pd.read_csv('test_modified2.csv')",
"_____no_output_____"
],
[
"predictors=[x for x in train.columns if x not in [target,IDcol]]",
"_____no_output_____"
],
[
"import seaborn as sns\nsns.countplot(train['target'])",
"_____no_output_____"
],
[
"model=xgb.XGBClassifier()\nmodel.fit(train[predictors],train[target],eval_metric='auc')\nprint(model)",
"XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n colsample_bytree=1, gamma=0, learning_rate=0.1, max_delta_step=0,\n max_depth=3, min_child_weight=1, missing=None, n_estimators=100,\n n_jobs=1, nthread=None, objective='binary:logistic', random_state=0,\n reg_alpha=0, reg_lambda=1, scale_pos_weight=1, seed=None,\n silent=True, subsample=1)\n"
],
[
"y_pred = model.predict(train[predictors])\ny_predprob = model.predict_proba(train[predictors])[:,1]\n",
"_____no_output_____"
],
[
"print (\"Accuracy : %.4g\" % metrics.accuracy_score(train['target'].values, y_pred))\nprint (\"AUC Score (Train): %f\" % metrics.roc_auc_score(train['target'], y_predprob))\n",
"Accuracy : 0.7874\nAUC Score (Train): 0.764417\n"
],
[
"import numpy as np\nimport matplotlib.pyplot as plt\n#fscore = model.get_booster().get_score(importance_type='weight')\nfscore = pd.Series(model.get_booster().get_score(importance_type='weight'))\nfscore.plot(kind='bar', title='Feature Importances')\nplt.ylabel('Feature Importance Score')",
"_____no_output_____"
],
[
"import matplotlib.pyplot as plt\nplt.hist(model.predict(test[predictors]))",
"_____no_output_____"
],
[
"y_xg = model.predict(test[predictors])\npredictions = [value for value in y_xg]\naccuracy = accuracy_score(y_test, predictions)\nprecision = precision_score(y_test, predictions)\nrecall = recall_score(y_test, predictions)\nf1 = f1_score(y_test, predictions)\nprint(\"Accuracy_score: %.2f%% on test dataset\" % (accuracy * 100.0))\nprint(\"precision_score: %.2f%% on test dataset\" % (precision * 100.0))\nprint(\"recall_score: %.2f%% on test dataset\" % (recall * 100.0))\nprint(\"f1_score: %.2f%% on test dataset\" % (f1 * 100.0))\nprint(\"roc_auc test set\", roc_auc_score(y_test, modelXg.predict_proba(X_test)[:,1]))\nprint(\"roc_auc training set\", roc_auc_score(y_train, modelXg.predict_proba(X_train)[:,1]))",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb9487092923a9bb8adb47e658f8a8cab93895b9
| 58,313 |
ipynb
|
Jupyter Notebook
|
docs/tutorials/part2_arrays_dicts.ipynb
|
mattias-ek/isopy
|
96d5530034655c7f9559568ab9b0879b978ef566
|
[
"MIT"
] | null | null | null |
docs/tutorials/part2_arrays_dicts.ipynb
|
mattias-ek/isopy
|
96d5530034655c7f9559568ab9b0879b978ef566
|
[
"MIT"
] | 1 |
2021-08-23T08:48:04.000Z
|
2021-08-23T08:48:04.000Z
|
docs/tutorials/part2_arrays_dicts.ipynb
|
mattias-ek/isopy
|
96d5530034655c7f9559568ab9b0879b978ef566
|
[
"MIT"
] | null | null | null | 25.632088 | 436 | 0.493046 |
[
[
[
"# Tutorial Part 2: Arrays, Dictionaries and Reference Values\n\n**Table of Content**\n\n* [Isopy Arrays](#Isopy-Arrays)\n * [Creating Arrays](#Creating-Arrays)\n * [Array Attributes](#Array-Attributes)\n * [Array Methods](#Array-Methods)\n * [Array Functions](#Array-Functions)\n* [Isopy Dictionaries](#Isopy-Dictionaries)\n* [Reference Values](#Reference-Values)",
"_____no_output_____"
]
],
[
[
"import isopy\nimport numpy as np\nimport pandas as pd\nimport pyperclip # library for interacting with the clipboard",
"_____no_output_____"
]
],
[
[
"## Isopy Arrays\nAn isopy array can be seen as a table of data with a certain number of rows and columns. Each column has a key in the form of an isopy key string. These arrays allow you to easily manipulate data while keeping track of what the values represents. Technically, isopy arrays are a custom view of a structured numpy array. This means that they inherit much of the functionality of a numpy array. ",
"_____no_output_____"
],
[
"### Creating Arrays\nYou can create arrays directly using the ``array`` and ``asarray`` functions or by directly using the ``IsopyArray`` class. Isopy arrays can be created from a range of different data, described below.",
"_____no_output_____"
],
[
"---\nWhen the input is a list/tuple or a numpy array we have to pass along the keys for each column in the input",
"_____no_output_____"
]
],
[
[
"data = [10, 20, 30] # Produces a 0-dimensional array\nisopy.array(data, ['ru', 'pd', 'cd']) ",
"_____no_output_____"
],
[
"data = np.array([[10, 20, 30], [11, 21, 31]]) #Produces a 1-dimensional array\nisopy.array(data, ['ru', 'pd', 'cd']) ",
"_____no_output_____"
]
],
[
[
"**Note** The data type of each column defaults to ``numpy.float64`` for values that do not have a numpy dtype. If a value cannot be represented as a float the default data type inferred by numpy will be used. In the examples above the first array created will use the default data type while the second array will inherit the data type of the input, ``np.int32`` in this case.",
"_____no_output_____"
],
[
"---\nUsing the ``dtype`` keyword you can specify the data type for columns in the array. Any data type accepted by numpy is valid. You can either pass a single data type or a tuple of data types. The latter will use the first the data type which is valid for all the data in the column.",
"_____no_output_____"
]
],
[
[
"isopy.array([10, 20, 30], ['ru', 'pd', 'cd']).dtype # f8 stands for np.float64",
"_____no_output_____"
],
[
"isopy.array([10, 20, 30], ['ru', 'pd', 'cd'], dtype=np.int32).dtype # i4 stands for np.int32",
"_____no_output_____"
],
[
"isopy.array(['ten', 20, 30], ['ru', 'pd', 'cd'], dtype=(np.int32, str)).dtype # U stands for unicode string and the next number is the maximum length",
"_____no_output_____"
]
],
[
[
"To specify different data types for different columns pass a list of equal length to the number of columns.",
"_____no_output_____"
]
],
[
[
"isopy.array([10, 20, 30], ['ru', 'pd', 'cd'], dtype=[str, np.int32, np.float64]).dtype",
"_____no_output_____"
]
],
[
[
"---\nUsing the ``ndim`` keyword you can specify the number of dimensions of the array. To return a 0-dimensional array if possible, otherwise return a 1-dimensional array specify ``ndim`` as ``-1``. The row number of 0-dimensional arrays will appear as ``\"None\"`` in the ``repr()`` output of an array. You can also check the dimensionality of an array using the ``ndim`` attribute.",
"_____no_output_____"
]
],
[
[
"isopy.array([10, 20, 30], ['ru', 'pd', 'cd'], ndim=0) # Make 0-dimensional",
"_____no_output_____"
],
[
"isopy.array([10, 20, 30], ['ru', 'pd', 'cd'], ndim=1) # Make 1-dimensional",
"_____no_output_____"
],
[
"isopy.array([10, 20, 30], ['ru', 'pd', 'cd'], ndim=-1) # Make 0-dimesional if possible otherwise 1-dimensional.",
"_____no_output_____"
],
[
"isopy.array([[10, 20, 30], [11, 21, 31]], ['ru', 'pd', 'cd'], ndim = -1) ",
"_____no_output_____"
]
],
[
[
"---\nIf the input is a dictionary or a structured numpy array the name of each column will be automatically inferred from the first argument.",
"_____no_output_____"
]
],
[
[
"data = dict(ru = [10, 11], pd= [20, 21], cd = [30, 31])\nisopy.array(data)",
"_____no_output_____"
],
[
"data = np.array([(10, 20, 30), (11, 21, 31)], dtype = [('ru', float), ('pd', float), ('cd', float)])\nisopy.array(data)",
"_____no_output_____"
]
],
[
[
"You can overwrite the inferred column keys by passing keys during creation",
"_____no_output_____"
]
],
[
[
"data = dict(ru = [10, 11], pd= [20, 21], cd = [30, 31])\nisopy.array(data, ['101ru', '105pd', '111cd'])",
"_____no_output_____"
]
],
[
[
"You can convert an isopy array back into a numpy array or a dictionary using the ``to_ndarray()`` and ``to_dict()`` methods.",
"_____no_output_____"
],
[
"---\nThere are a number of methods for converting isopy arrays into other python objects",
"_____no_output_____"
]
],
[
[
"a = isopy.array(dict(ru = [10, 11], pd= [20, 21], cd = [30, 31]))\na.to_ndarray() # Converts array into a structured numpy array",
"_____no_output_____"
],
[
"a.to_dict() # Converts array into a dictionary",
"_____no_output_____"
]
],
[
[
"---\nSimilarly you can create isopy arrays from a pandas dataframe",
"_____no_output_____"
]
],
[
[
"df = pd.DataFrame(dict(ru = [10, 11], pd= [20, 21], cd = [30, 31]))\ndf",
"_____no_output_____"
],
[
"isopy.array(df)",
"_____no_output_____"
]
],
[
[
"You can convert a isopy array back into a dataframe using the ``to_dataframe()`` method or by passing it directly to ``DataFrame()``",
"_____no_output_____"
]
],
[
[
"array = isopy.array(df)\narray.to_dataframe()",
"_____no_output_____"
],
[
"pd.DataFrame(array)",
"_____no_output_____"
]
],
[
[
"---\nYou can create arrays from existing isopy arrays",
"_____no_output_____"
]
],
[
[
"a = isopy.array([10, 20, 30], ['ru', 'pd', 'cd']) \nisopy.array(a)",
"_____no_output_____"
]
],
[
[
"There key difference between ``array`` and ``asarray`` is that if the first argument is an isopy array then ``asarray`` will return a reference that array, rather than a copy, if no other arguments are given while ``array`` will return a copy.",
"_____no_output_____"
]
],
[
[
"a = isopy.array([10, 20, 30], ['ru', 'pd', 'cd']) \nisopy.array(a) is a, isopy.asarray(a) is a",
"_____no_output_____"
]
],
[
[
"#### Filled Arrays\nYou can create an array of uninitiated values, zeros or ones using the ``empty``, ``zeros`` and ``one`` functions.",
"_____no_output_____"
]
],
[
[
"isopy.empty(None, ['ru', 'pd', 'cd']) #None, or -1, creates a 0-dimensional array",
"_____no_output_____"
],
[
"isopy.zeros(1, ['ru', 'pd', 'cd'])",
"_____no_output_____"
],
[
"isopy.ones(2, ['ru', 'pd', 'cd'])",
"_____no_output_____"
]
],
[
[
"To create an array filled with a specific value use the ``full`` function. The second arguement is the fill value. This can either be a single value used for all rows in the column or a sequence of values of of the same length as the number of rows.",
"_____no_output_____"
]
],
[
[
"isopy.full(2, np.nan, ['ru', 'pd', 'cd'])",
"_____no_output_____"
],
[
"isopy.full(2, [1,2], ['ru', 'pd', 'cd'])",
"_____no_output_____"
]
],
[
[
"---\nIf no keys are given, or can be inferred, a normal numpy array is returned.",
"_____no_output_____"
]
],
[
[
"isopy.ones(5) # same as np.ones(5)",
"_____no_output_____"
]
],
[
[
"---\n#### Random Arrays\nTo create an array of random values use the ``random`` function. The second argument is either a single argument or a tuple of arguments that will be passed to the random generator. By default this function draws values from a normal distribution. The following example draws values from a normal distribution with a center of 1 and standard deviation of 0.1",
"_____no_output_____"
]
],
[
[
"isopy.random(10, (1, 0.1), ['ru', 'pd', 'cd'])",
"_____no_output_____"
]
],
[
[
"You can specify different distributions for different columns by passing a list as the second argument",
"_____no_output_____"
]
],
[
[
"isopy.random(10, [(1, 0.1), (0,1), (10, 1)], ['ru', 'pd', 'cd'])",
"_____no_output_____"
]
],
[
[
"---\nIf no keys are given, or can be inferred, a normal numpy array is returned.",
"_____no_output_____"
]
],
[
[
"isopy.random(10)",
"_____no_output_____"
]
],
[
[
"### Array Attributes\nSince isopy arrays are custom implementation of a numpy arrays they have all the attributes you would find in numpy arrays, e.g. ``size``, ``.ndim``, ``.shape`` and ``.dtype``.",
"_____no_output_____"
]
],
[
[
"a = isopy.array(dict(ru = [10, 11], pd= [20, 21], cd = [30, 31]))\na.size, a.ndim, a.shape, a.dtype",
"_____no_output_____"
]
],
[
[
"In addition to the numpy attributes ``.nrows`` and ``.ncols`` are also available for isopy arrays. There return the number of rows and number of columns in the array respectively.",
"_____no_output_____"
]
],
[
[
"a.nrows, a.ncols",
"_____no_output_____"
]
],
[
[
"**Note** That ``.size`` will return ``1`` for both 0-dimensional arrays and 1-dimensional arrays with 1 row. ``.nrows`` on the other hand will return ``-1`` 0-dimensional arrays.",
"_____no_output_____"
]
],
[
[
"a = isopy.array(dict(ru = 10, pd= 20, cd = 30))\na.size, a.nrows",
"_____no_output_____"
]
],
[
[
"The column keys are available through the ``.keys`` attribute",
"_____no_output_____"
]
],
[
[
"a.keys # a.keys() also works fine",
"_____no_output_____"
]
],
[
[
"### Array Methods\nWhile isopy arrays also contain all the methods found in numpy arrays many of these are not relevant to isopy arrays and may therefore not work as expected, if at all. See the reference documentation for a list of all methods that have been implemented for isopy arrays. Any methods not listed there should be used with **caution** as the behavior is undefined.\n\n---\nIsopy arrays have a number of methods that mimic those found in dictionaries. In addition to the ``.keys`` attribute, that can be used as a method, arrays also have ``values()``, ``items()`` and ``get()`` methods.",
"_____no_output_____"
]
],
[
[
"a = isopy.array(dict(ru = [10, 11], pd= [20, 21], cd = [30, 31]))\na.values()",
"_____no_output_____"
],
[
"a.items()",
"_____no_output_____"
]
],
[
[
"**Note** both``values()`` and ``items()`` both return a tuple (Unlike dictionaries where they return iterators). ",
"_____no_output_____"
]
],
[
[
"a.get('ru')",
"_____no_output_____"
]
],
[
[
"If a column with the specified key is not present in the array a default value is return with the same shape as a column in the array.",
"_____no_output_____"
]
],
[
[
"a.get('ag') # If not specified the default value is np.nan",
"_____no_output_____"
],
[
"a.get('ag', 40) # Second argument is the default value",
"_____no_output_____"
],
[
"a.get('ag', [40, 41]) # A sequence the same shape as a valid column is also accepted",
"_____no_output_____"
]
],
[
[
"---\nThe ``filter()`` method return a view containing only the columns that match the supplied key filters. See the ``filter()`` method for the different key lists for available filter keywords.",
"_____no_output_____"
]
],
[
[
"a = isopy.array(dict(ru101 = [10, 11], pd105= [20, 21], cd111 = [30, 31]))\na.filter(mass_number_gt=104) # Return only the columns with that have a mass number greater than 104",
"_____no_output_____"
]
],
[
[
"---\nThe ``copy()`` method can be used to return a copy of the array",
"_____no_output_____"
]
],
[
[
"a = isopy.array(dict(ru101 = [10, 11], pd105= [20, 21], cd111 = [30, 31]))\nb = a.copy()\nb is a, a == b",
"_____no_output_____"
]
],
[
[
"You can copy only those columns that meet a certain criteria by passing filter keywords as in the ``filter()`` method described above.",
"_____no_output_____"
]
],
[
[
"a = isopy.array(dict(ru101 = [10, 11], pd105= [20, 21], cd111 = [30, 31]))\na.copy(mass_number_gt = 104) # Return only the columns with that have a mass number greater than 104",
"_____no_output_____"
]
],
[
[
"---\nYou can create a ratio from data within an array using the ``ratio()`` method",
"_____no_output_____"
]
],
[
[
"c = a.ratio('105Pd'); c",
"_____no_output_____"
]
],
[
[
"Ratio arrays have a ``deratio()`` method for flattening a ratio array. This requires that all column keys in the array have a common denominator",
"_____no_output_____"
]
],
[
[
"c.deratio()",
"_____no_output_____"
],
[
"c.deratio([20, 21]) #You can specify the value(s) for the denominator",
"_____no_output_____"
]
],
[
[
"---\nThe ``normalise`` function allows you to normalise the data in the array to a certian value. Calling the function without any arguments will normalise all the values to that the sum of each row is ``1``",
"_____no_output_____"
]
],
[
[
"a = isopy.array(dict(ru = [10, 11], pd= [20, 21], cd = [30, 31]))\na.normalise()",
"_____no_output_____"
]
],
[
[
"The optional arguments are 1) the value you with to normalise to and 2) the key(s) of the columns that the normalisation should be based on.",
"_____no_output_____"
]
],
[
[
"a.normalise(100, 'pd')",
"_____no_output_____"
],
[
"a.normalise([100, 1000], ['ru', 'cd']) # The sum of the specified keys will be equal to the values given",
"_____no_output_____"
],
[
"a.normalise([100, 1000], ['ru', 'cd']).normalise([20, 21], 'pd')",
"_____no_output_____"
]
],
[
[
"---\nWith the ``to_text()`` method of arrays you can convert the contents of an array to a text string. The ``str()`` ``repr()`` functions both call this function to produce two slightly different strings. The *str* function will return the raw data in each column of the array.",
"_____no_output_____"
]
],
[
[
"array = isopy.random(20, keys='ru pd cd'.split())",
"_____no_output_____"
],
[
"print(array) # Same as print( str(array) ) and print( array.to_text() )",
"Ru , Pd , Cd \n0.7423098454624386 , 0.2630859993817623 , -2.052605165325265 \n-0.036241987546634755 , 2.765418917186138 , 0.02123660227846849 \n0.10526077208384534 , -0.7252290622088389 , 0.28237653594763196 \n1.4642978852086517 , -0.36530535185645874 , -1.419658403506246 \n-0.4071008728918706 , -0.10614162076449779 , 1.0301482740113062 \n1.8112995186818222 , -0.47476146077698905 , 1.570544281316929 \n1.8299243646994559 , 0.290408799843238 , -0.8940795968570514 \n-0.01980863219333887 , -0.6340970761629324 , 1.2604906968615817 \n0.2996284335088453 , 1.136140795492163 , 1.1535760309650112 \n0.9237767535294121 , -0.019482550940763656 , -0.7597607491314028 \n0.9539298028846708 , -0.6675101173727683 , -0.7978596359795319 \n0.7208150546289938 , 0.8172728209094148 , 0.9830731210309498 \n-1.4473959961523544 , 0.2666358957982169 , 0.051893166569739406 \n-1.2705739976837838 , 1.3596889515584312 , -0.3753310459831876 \n0.7547719852115573 , 0.4327369158617446 , -0.6455168936091344 \n1.4389314187427644 , 0.9223732916625811 , 0.8515130659123246 \n-0.38593409510733895 , 1.793076738891864 , -0.7300233613272916 \n0.9781076555830062 , 1.6513424987462482 , -0.2693850346623433 \n-1.386425412616868 , 1.5576442193412907 , 1.5902268158777644 \n-0.41603641339752806 , -0.24021959750039967 , 0.7038661921414116 \n"
]
],
[
[
"The *repr* function will format *float* values to 5 significant digits, include an additional column with the row number, and it will only show a maximum of 10 rows.",
"_____no_output_____"
]
],
[
[
"array # Same as print( repr(array) ) and print( array.to_text(nrows = 10, include_row=True, f='{:.5g}') )",
"_____no_output_____"
]
],
[
[
"There are a number of optional arguments you can specify to change things like number formats and the delimiter for the string",
"_____no_output_____"
]
],
[
[
"a.to_text(delimiter = '\\t', include_row=True) #includes the row number and uses a tab delimiter",
"_____no_output_____"
]
],
[
[
"The method ``to_clipboard()`` takes the same arguments as ``to_text()`` but copies the string to the clipboard.",
"_____no_output_____"
]
],
[
[
"a.to_clipboard() # It also returns the copied string",
"_____no_output_____"
],
[
"pyperclip.paste() # Paste whatever is currently is in the clipboard",
"_____no_output_____"
]
],
[
[
"If you are using jupyter you can use the ``display_table()`` method to render a nice table in a cell. Except from *delimiter* it takes the same arguments as ``to_text()``",
"_____no_output_____"
]
],
[
[
"array.display_table(include_row=True)",
"_____no_output_____"
]
],
[
[
"### Array Functions\nThe isopy package comes with several custom made array functions and isopy arrays support a large number of the numpy array functions. An array function is a function that perform an action on one or more arrays, e.g. adding arrays together or finding the mean of values in an array. See [Introduction Part 3: Working with arrays]() **LINK MISSING** for a comprehensive explanation of array functions with lots examples.\n\nA few quick examples are",
"_____no_output_____"
]
],
[
[
"a1 = isopy.array(dict(ru = 1, pd= 2, cd = 3))\na2 = isopy.array(dict(ru = 10, pd= 20, ag = 25))\na1 + a2 # Columns not present in all arrays are assinged a value of np.nan",
"_____no_output_____"
],
[
"a = isopy.random(100, [(1, 0.1), (0, 1), (10, 2)], ['ru', 'pd', 'cd'])\nnp.mean(a) #Calculate the mean of each column",
"_____no_output_____"
],
[
"np.std(a) # Calculate the standard deviation of each column",
"_____no_output_____"
]
],
[
[
"---\nOne useful feature of the array function implementation for isopy arrays is that they can be used in conjunction with dictionaries. Only keys present in the array are included in the output meanign the dictionary can contain as many keys as you want. Thus dictionaries are useful for storing reference values.",
"_____no_output_____"
]
],
[
[
"a = isopy.array(dict(ru = 10, pd= 20, cd = 30))\nd = dict(ru = 1, rh = 2, pd=3, ag=4, cd = 5)\na + d # Only column keys in the array are present in the output",
"_____no_output_____"
]
],
[
[
"**Note** The dictionary keys do not need to formatted to match the proper key string format.",
"_____no_output_____"
],
[
"## Isopy Dictionaries",
"_____no_output_____"
],
[
"Isopy has two special dictionaries, ``IsopyDict`` and ``ScalarDict``. These function much like normal dictionaries with a few enhancements. First, all values are stored as isopy key strings. Second, they can be readonly and have predefined default values. Third, create a subsection of the dictionary using ``copy()`` by passing filter keywords.",
"_____no_output_____"
]
],
[
[
"d = isopy.IsopyDict(ru101 = 1, rh103 = 2, pd105=3, ag107=4, cd111 = 5, default_value=0); d # The input can also be a another dictionary",
"_____no_output_____"
],
[
"d.get('76ge'), d.get('80se', 100) # If not specified the default value of the dictionary is used",
"_____no_output_____"
],
[
"d = isopy.IsopyDict(ru101 = 1, rh103 = 2, pd105=3, ag107=4, cd111 = 5, default_value=0)\nd.copy(mass_number_gt = 104) # Returns a new dict containing only isotopes with a mass number greater than 104",
"_____no_output_____"
]
],
[
[
"---\n``ScalarDict`` works just like an ``IsopyDict`` with three exceptions. First it can only store scalars, that is a single numerical value. Second, the ``get()`` method can calculate the ratio of two values in the dictionary if a ratio key string is not present in the dictionary. Finally, you can create an array directly from the dictionary with the ``to_array()`` method.",
"_____no_output_____"
]
],
[
[
"d = isopy.ScalarDict(ru101 = 1, rh103 = 2, pd105=3, ag107=4, cd111 = 5) # Default value is by default np.nan\nd.get('pd105/cd111') # Automatically calculated from the numerator and denominator values",
"_____no_output_____"
],
[
"d.to_array()",
"_____no_output_____"
],
[
"d.to_array(mass_number_gt = 104) # You can specify key filters too",
"_____no_output_____"
]
],
[
[
"## Reference Values\nThere are a number of reference values included with isopy under the ``refval`` namespace. You can find the available reference values listed [here](https://isopy.readthedocs.io/en/latest/refpages/reference_values.html) toghether with a short description. There are currently three categories of reference values, ``mass``, ``element`` and ``isotope`` referring to the flavour of the key string of the values in the dictionaries. ",
"_____no_output_____"
]
],
[
[
"isopy.refval.element.atomic_number.get('pd') # The atomic number of palladium",
"_____no_output_____"
],
[
"isopy.refval.isotope.fraction.get('105pd') # The natural isotope fraction of 105Pd",
"_____no_output_____"
],
[
"isopy.refval.element.isotopes.get('pd') # Returns a list of all naturally occuring isotopes of palladium",
"_____no_output_____"
]
],
[
[
"---\nMany reference values are isopy dictionaries so the ``copy()`` method accepts filter keywords",
"_____no_output_____"
]
],
[
[
"isopy.refval.isotope.fraction.copy(element_symbol='pd')",
"_____no_output_____"
]
],
[
[
"Similarly many reference values has the ``to_array()`` method.",
"_____no_output_____"
]
],
[
[
"isopy.refval.isotope.fraction.to_array(element_symbol='pd')",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb948fdd75609991536c50a3bbf16c88e0b64c6e
| 6,408 |
ipynb
|
Jupyter Notebook
|
src/visualization/draft-ppic.ipynb
|
chekos/voila-ppic-mockup
|
d7c676e42ec2e7b8bf4b56dd7d3c33317d61204b
|
[
"RSA-MD"
] | null | null | null |
src/visualization/draft-ppic.ipynb
|
chekos/voila-ppic-mockup
|
d7c676e42ec2e7b8bf4b56dd7d3c33317d61204b
|
[
"RSA-MD"
] | null | null | null |
src/visualization/draft-ppic.ipynb
|
chekos/voila-ppic-mockup
|
d7c676e42ec2e7b8bf4b56dd7d3c33317d61204b
|
[
"RSA-MD"
] | null | null | null | 26.589212 | 395 | 0.539014 |
[
[
[
"from ipyleaflet import Map, GeoData, basemaps, LayersControl\nfrom ipywidgets import Dropdown, interactive\nimport geopandas as gpd\nimport pandas as pd\nimport json\nfrom pathlib import Path\nfrom IPython.display import display",
"_____no_output_____"
],
[
"PROCESSED_DATA = Path(\"../../data/processed/\")",
"_____no_output_____"
],
[
"datos = gpd.read_file(PROCESSED_DATA / 'CA_data.shp')",
"_____no_output_____"
],
[
"datos.dropna(inplace = True)",
"_____no_output_____"
],
[
"datos.crs = {'proj': 'aea',\n 'lat_1': \"29.5\",\n 'lat_2': \"45.5\",\n 'lat_0': \"37.5\",\n 'lon_0': \"-96\",\n 'x_0': \"0\",\n 'y_0': \"0\",\n 'datum': 'NAD83',\n 'units': 'm',\n 'no_defs': \"true\"}",
"_____no_output_____"
],
[
"datos = datos.to_crs(epsg=4326)",
"_____no_output_____"
],
[
"def geo_center(df):\n _bounds = df.bounds\n xmin, xmax, ymin, ymax = _bounds.minx.min(), _bounds.maxx.max(), _bounds.miny.min(), _bounds.maxy.max()\n y_center = (ymax + ymin) / 2\n x_center = (xmax + xmin) / 2\n center = (y_center, x_center)\n \n return center",
"_____no_output_____"
],
[
"datos_female = datos[datos['sex'] == 'female'].copy()",
"_____no_output_____"
],
[
"datos_female['bins'] = pd.cut(datos_female['count'], bins = 5)",
"_____no_output_____"
],
[
"#datos_female['bins'].unique()",
"_____no_output_____"
],
[
"datos_female['bins'] = datos_female['bins'].astype(str)",
"_____no_output_____"
],
[
"_orden = [\"(153686.0, 305094.0]\", \"(1520.96, 153686.0]\", \"(607910.0, 759318.0]\"]\n_colores = \"#81d1e6\",\"#dc0d12\",[\"#fbe1e1\",]",
"_____no_output_____"
],
[
"_escala = dict(zip(_orden, _colores))",
"_____no_output_____"
],
[
"def crea_visualizacion(datos):\n _centro = geo_center(datos)\n \n m = Map(center = _centro, zoom = 5, basemap = basemaps.OpenStreetMap.BlackAndWhite, scroll_wheel_zoom = True, )\n \n for option in datos[\"bins\"].unique():\n \n option_data = datos[datos[\"bins\"] == option]\n \n geo_data = GeoData(\n geo_dataframe = option_data,\n style={'color': _escala[option], 'fillColor': _escala[option], 'opacity':0.35, 'weight':1.9, 'dashArray':'2', 'fillOpacity':0.7}, \n hover_style={'fillColor': _escala[option] , 'fillOpacity': 0.9},\n name = f'{option}'\n )\n \n # add layers\n m.add_layer(geo_data)\n \n m.add_control(LayersControl())\n \n return m",
"_____no_output_____"
]
],
[
[
"<div class='ppic-blog-titlearea'><h1>Where are the nerds?</h1><div class='ppic-author-line'><span class='ppic-blog-authors'><a href='https://www.ppic.org/person/sergio-sanchez/'>Sergio Sánchez</a>, <a href='https://www.ppic.org/person/joseph-hayes/'>Joseph Hayes</a></span> <span class='ppic-date'>June 26, 2019</span></div></div>",
"_____no_output_____"
],
[
"Lorem ipsum dolor amet dreamcatcher pabst DIY four loko disrupt. Disrupt literally deep v keytar banh mi live-edge kickstarter locavore af. Trust fund cray vexillologist drinking vinegar blog. Hoodie small batch chartreuse crucifix kickstarter health goth squid raclette 8-bit activated charcoal stumptown dreamcatcher glossier. Gochujang four loko church-key, brunch farm-to-table mlkshk.",
"_____no_output_____"
],
[
"Gastropub hashtag jianbing, man bun trust fund street art shaman kombucha retro master cleanse selvage 90's godard. Selfies echo park lumbersexual, migas sartorial four loko master fingerstache shoreditch try-hard. Shaman four loko stumptown viral, cold-pressed fixie pabst. Pickled master cleanse 90's williamsburg kombucha humblebrag wolf.",
"_____no_output_____"
]
],
[
[
"crea_visualizacion(datos_female)",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code"
]
] |
cb94903f02c68a4bd3eb87671c81a6166918903b
| 16,072 |
ipynb
|
Jupyter Notebook
|
Part 2 - Regression/Section 8 - Decision Tree Regression/DTREGRESSION.ipynb
|
divyanshchoubisa/Machine-Learning-Super-Data-Science-
|
c7c0e02ea8b4cd01ba8c06c4eb6ecdba62477bdc
|
[
"MIT"
] | null | null | null |
Part 2 - Regression/Section 8 - Decision Tree Regression/DTREGRESSION.ipynb
|
divyanshchoubisa/Machine-Learning-Super-Data-Science-
|
c7c0e02ea8b4cd01ba8c06c4eb6ecdba62477bdc
|
[
"MIT"
] | null | null | null |
Part 2 - Regression/Section 8 - Decision Tree Regression/DTREGRESSION.ipynb
|
divyanshchoubisa/Machine-Learning-Super-Data-Science-
|
c7c0e02ea8b4cd01ba8c06c4eb6ecdba62477bdc
|
[
"MIT"
] | null | null | null | 119.051852 | 13,160 | 0.887008 |
[
[
[
"import pandas as pd\nimport numpy as np\nimport matplotlib.pyplot as plt",
"_____no_output_____"
],
[
"df = pd.read_csv('Position_Salaries.csv')\nX = df.iloc[:, 1:2].values\ny = df.iloc[:, 2].values",
"_____no_output_____"
],
[
"from sklearn.tree import DecisionTreeRegressor\nregressor = DecisionTreeRegressor(random_state = 0)\nregressor.fit(X, y)",
"_____no_output_____"
],
[
"y_pred = regressor.predict([[6.5]])\nprint(y_pred)",
"[150000.]\n"
],
[
"X_grid = np.arange(min(X), max(X), 0.01)\nX_grid = X_grid.reshape((len(X_grid), 1))\nplt.scatter(X, y, color = 'red')\nplt.plot(X_grid, regressor.predict(X_grid), color = 'blue')\nplt.title('Truth or Bluff (Decision Tree Regression)')\nplt.xlabel('Position level')\nplt.ylabel('Salary')\nplt.show()",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code"
]
] |
cb94938839ec3b172245aaa03b4a94a291d6becc
| 315,961 |
ipynb
|
Jupyter Notebook
|
notebooks/6_Autoencoder.ipynb
|
davidggphy/modelling_week_2019
|
73b12328e1a2bbb0062c7d4f3f09774a43cd296d
|
[
"MIT"
] | 2 |
2019-06-29T17:10:18.000Z
|
2019-07-01T06:38:18.000Z
|
notebooks/6_Autoencoder.ipynb
|
davidggphy/modelling_week_2019
|
73b12328e1a2bbb0062c7d4f3f09774a43cd296d
|
[
"MIT"
] | 5 |
2020-03-24T17:08:11.000Z
|
2021-12-13T20:00:45.000Z
|
notebooks/6_Autoencoder.ipynb
|
davidggphy/modelling_week_2019
|
73b12328e1a2bbb0062c7d4f3f09774a43cd296d
|
[
"MIT"
] | null | null | null | 286.196558 | 58,028 | 0.913841 |
[
[
[
"# Autoencoder (Semi-supervised)",
"_____no_output_____"
]
],
[
[
"%load_ext autoreload\n%autoreload 2",
"_____no_output_____"
],
[
"# Seed value\n# Apparently you may use different seed values at each stage\nseed_value= 0\n\n# 1. Set the `PYTHONHASHSEED` environment variable at a fixed value\nimport os\nos.environ['PYTHONHASHSEED']=str(seed_value)\n\n# 2. Set the `python` built-in pseudo-random generator at a fixed value\nimport random\nrandom.seed(seed_value)\n\n# 3. Set the `numpy` pseudo-random generator at a fixed value\nimport numpy as np\nnp.random.seed(seed_value)\n\n# 4. Set the `tensorflow` pseudo-random generator at a fixed value\nimport tensorflow as tf\ntf.set_random_seed(seed_value)\n\n# 5. Configure a new global `tensorflow` session\nfrom keras import backend as K\nsession_conf = tf.ConfigProto(intra_op_parallelism_threads=1, inter_op_parallelism_threads=1)\nsess = tf.Session(graph=tf.get_default_graph(), config=session_conf)\nK.set_session(sess)",
"Using TensorFlow backend.\n"
],
[
"import keras\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport seaborn as sns\nimport sklearn\n# plt.style.use('fivethirtyeight')\nsns.set_style(\"whitegrid\")\nsns.set_context(\"notebook\")\n\n\nDATA_PATH = '../data/'\nVAL_SPLITS = 4",
"_____no_output_____"
],
[
"from plot_utils import plot_confusion_matrix\nfrom cv_utils import run_cv_f1\nfrom cv_utils import plot_cv_roc\nfrom cv_utils import plot_cv_roc_prc\nfrom sklearn.model_selection import StratifiedShuffleSplit\nfrom sklearn.model_selection import cross_val_score\nfrom sklearn.model_selection import cross_validate",
"_____no_output_____"
]
],
[
[
"For this part of the project, we will only work with the training set, that we will split again into train and validation to perform the hyperparameter tuning.\n\nWe will save the test set for the final part, when we have already tuned our hyperparameters.",
"_____no_output_____"
]
],
[
[
"df = pd.read_csv(os.path.join(DATA_PATH,'df_train.csv'))\ndf.drop(columns= df.columns[0:2],inplace=True)\ndf.head()",
"_____no_output_____"
]
],
[
[
"## Preprocessing the data",
"_____no_output_____"
],
[
"Although we are always using cross validation with `VAL_SPLITS` folds, (in general, 4), here we are gonna set only one split in order to explore how the Autoencoder works and get intuition.",
"_____no_output_____"
]
],
[
[
"cv = StratifiedShuffleSplit(n_splits=1,test_size=0.15,random_state=0)",
"_____no_output_____"
],
[
"# In case we want to select a subset of features\n# df_ = df[['Class','V9','V14','V16','V2','V3','V17']]\ndf_ = df[['Class','V4','V14','V16','V12','V3','V17']]\nX = df_.drop(columns='Class').to_numpy()\ny = df_['Class'].to_numpy()\n\nfor idx_t, idx_v in cv.split(X,y):\n X_train = X[idx_t]\n y_train = y[idx_t]\n X_val = X[idx_v]\n y_val = y[idx_v]\n \n # Now we need to erase the FRAUD cases on the TRAINING set \n X_train_normal = X_train[y_train==0]",
"_____no_output_____"
]
],
[
[
"## Defining the model",
"_____no_output_____"
]
],
[
[
"# this is the size of our encoded representations\nENCODED_DIM = 2\nINPUT_DIM = X.shape[1]",
"_____no_output_____"
],
[
"from keras.layers import Input, Dense\nfrom keras.models import Model\nfrom keras.models import Sequential\nfrom keras.layers import Dense, Activation\nfrom keras.layers import LeakyReLU\n\ndef create_encoder(input_dim, encoded_dim):\n encoder = Sequential([\n Dense(32, input_shape=(input_dim,)),\n LeakyReLU(),\n Dense(16),\n LeakyReLU(),\n Dense(8),\n LeakyReLU(),\n Dense(encoded_dim)\n ], name='encoder')\n return encoder\n\ndef create_decoder(input_dim, encoded_dim):\n decoder = Sequential([\n Dense(8, input_shape=(encoded_dim,) ),\n LeakyReLU(),\n Dense(16),\n LeakyReLU(),\n Dense(8),\n LeakyReLU(),\n Dense(input_dim)\n],name='decoder')\n return decoder\n\n\n\ndef create_autoencoder(input_dim, encoded_dim, return_encoder = True):\n encoder = create_encoder(input_dim,encoded_dim)\n decoder = create_decoder(input_dim,encoded_dim)\n inp = Input(shape=(INPUT_DIM,),name='Input_Layer')\n\n # a layer instance is callable on a tensor, and returns a tensor\n x_enc = encoder(inp)\n x_out = decoder(x_enc)\n\n\n # This creates a model that includes\n # the Input layer and three Dense layers\n autoencoder = Model(inputs=inp, outputs=x_out)\n if return_encoder:\n return autoencoder, encoder\n else:\n return autoencoder",
"_____no_output_____"
],
[
"autoencoder, encoder = create_autoencoder(INPUT_DIM,ENCODED_DIM)\nprint('ENCODER SUMMARY\\n')\nprint(encoder.summary())\nprint('AUTOENCODER SUMMARY\\n')\nprint(autoencoder.summary())",
"WARNING:tensorflow:From /Users/danky/anaconda3/envs/fraud_credit/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\nInstructions for updating:\nColocations handled automatically by placer.\nENCODER SUMMARY\n\n_________________________________________________________________\nLayer (type) Output Shape Param # \n=================================================================\ndense_1 (Dense) (None, 32) 224 \n_________________________________________________________________\nleaky_re_lu_1 (LeakyReLU) (None, 32) 0 \n_________________________________________________________________\ndense_2 (Dense) (None, 16) 528 \n_________________________________________________________________\nleaky_re_lu_2 (LeakyReLU) (None, 16) 0 \n_________________________________________________________________\ndense_3 (Dense) (None, 8) 136 \n_________________________________________________________________\nleaky_re_lu_3 (LeakyReLU) (None, 8) 0 \n_________________________________________________________________\ndense_4 (Dense) (None, 2) 18 \n=================================================================\nTotal params: 906\nTrainable params: 906\nNon-trainable params: 0\n_________________________________________________________________\nNone\nAUTOENCODER SUMMARY\n\n_________________________________________________________________\nLayer (type) Output Shape Param # \n=================================================================\nInput_Layer (InputLayer) (None, 6) 0 \n_________________________________________________________________\nencoder (Sequential) (None, 2) 906 \n_________________________________________________________________\ndecoder (Sequential) (None, 6) 358 \n=================================================================\nTotal params: 1,264\nTrainable params: 1,264\nNon-trainable params: 0\n_________________________________________________________________\nNone\n"
],
[
"autoencoder.compile(optimizer='adam',\n loss='mean_squared_error')",
"_____no_output_____"
]
],
[
[
"## Training the model",
"_____no_output_____"
]
],
[
[
"autoencoder.fit(x=X_train_normal, y= X_train_normal,\n batch_size=512,epochs=40, validation_split=0.1) # starts training",
"WARNING:tensorflow:From /Users/danky/anaconda3/envs/fraud_credit/lib/python3.6/site-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\nInstructions for updating:\nUse tf.cast instead.\nTrain on 184184 samples, validate on 20465 samples\nEpoch 1/40\n184184/184184 [==============================] - 2s 9us/step - loss: 0.6179 - val_loss: 0.4537\nEpoch 2/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.4595 - val_loss: 0.4329\nEpoch 3/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.4411 - val_loss: 0.4208\nEpoch 4/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.4235 - val_loss: 0.4015\nEpoch 5/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.4033 - val_loss: 0.3827\nEpoch 6/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3900 - val_loss: 0.3725\nEpoch 7/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3808 - val_loss: 0.3635\nEpoch 8/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3734 - val_loss: 0.3575\nEpoch 9/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3677 - val_loss: 0.3512\nEpoch 10/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3628 - val_loss: 0.3484\nEpoch 11/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3584 - val_loss: 0.3455\nEpoch 12/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3549 - val_loss: 0.3429\nEpoch 13/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3514 - val_loss: 0.3375\nEpoch 14/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3481 - val_loss: 0.3386\nEpoch 15/40\n184184/184184 [==============================] - 1s 6us/step - loss: 0.3452 - val_loss: 0.3345\nEpoch 16/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3424 - val_loss: 0.3301\nEpoch 17/40\n184184/184184 [==============================] - 1s 5us/step - loss: 0.3407 - val_loss: 0.3395\nEpoch 18/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3391 - val_loss: 0.3255\nEpoch 19/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3363 - val_loss: 0.3249\nEpoch 20/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3346 - val_loss: 0.3217\nEpoch 21/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3337 - val_loss: 0.3193\nEpoch 22/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3316 - val_loss: 0.3217\nEpoch 23/40\n184184/184184 [==============================] - 1s 5us/step - loss: 0.3311 - val_loss: 0.3230\nEpoch 24/40\n184184/184184 [==============================] - 1s 6us/step - loss: 0.3294 - val_loss: 0.3163\nEpoch 25/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3279 - val_loss: 0.3165\nEpoch 26/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3267 - val_loss: 0.3153\nEpoch 27/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3256 - val_loss: 0.3157\nEpoch 28/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3252 - val_loss: 0.3110\nEpoch 29/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3234 - val_loss: 0.3098\nEpoch 30/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3222 - val_loss: 0.3092\nEpoch 31/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3220 - val_loss: 0.3118\nEpoch 32/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3201 - val_loss: 0.3080\nEpoch 33/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3189 - val_loss: 0.3061\nEpoch 34/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3182 - val_loss: 0.3057\nEpoch 35/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3185 - val_loss: 0.3057\nEpoch 36/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3173 - val_loss: 0.3058\nEpoch 37/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3163 - val_loss: 0.3057\nEpoch 38/40\n184184/184184 [==============================] - 1s 4us/step - loss: 0.3150 - val_loss: 0.3034\nEpoch 39/40\n184184/184184 [==============================] - 1s 5us/step - loss: 0.3144 - val_loss: 0.3017\nEpoch 40/40\n184184/184184 [==============================] - 1s 5us/step - loss: 0.3142 - val_loss: 0.3031\n"
]
],
[
[
"## Testing",
"_____no_output_____"
]
],
[
[
"X_enc = encoder.predict(X_val)\nX_enc_normal = X_enc[y_val==0]\nX_enc_fraud = X_enc[y_val==1]\nsns.scatterplot(x = X_enc_normal[:,0], y = X_enc_normal[:,1] ,label='Normal', alpha=0.5)\nsns.scatterplot(x = X_enc_fraud[:,0], y = X_enc_fraud[:,1] ,label='Fraud')",
"_____no_output_____"
],
[
"X_out = autoencoder.predict(X_val)\nprint(X_out.shape)",
"(36176, 6)\n"
],
[
"X_val.shape",
"_____no_output_____"
],
[
"distances = np.sum((X_out-X_val)**2,axis=1)",
"_____no_output_____"
],
[
"bins = np.linspace(0,np.max(distances),40)\nsns.distplot(distances[y_val==0],label='Normal',kde=False, \n bins=bins, norm_hist=True, axlabel='Distance')\nsns.distplot(distances[y_val==1],label='Fraud',kde=False, bins=bins, norm_hist=True)",
"_____no_output_____"
],
[
"bins = np.linspace(0,100,40)\nsns.distplot(distances[y_val==0],label='Normal',kde=False, \n bins=bins, norm_hist=True, axlabel='Distance')\nsns.distplot(distances[y_val==1],label='Fraud',kde=False, bins=bins, norm_hist=True)\nplt.xlim((0,100))",
"_____no_output_____"
]
],
[
[
"## Validating the model",
"_____no_output_____"
]
],
[
[
"from sklearn.model_selection import cross_val_score\nfrom sklearn.model_selection import cross_validate\nfrom sklearn.metrics import f1_score",
"_____no_output_____"
],
[
"def clf_autoencoder(X,autoencoder,threshold):\n \"\"\"\n Classifier based in the autoencoder.\n A datapoint is a nomaly if the distance of the original points\n and the result of the autoencoder is greater than the threshold.\n \"\"\"\n X_out = autoencoder.predict(X)\n distances = np.sum((X_out-X)**2,axis=1).reshape((-1,1))\n # y_pred = 1 if it is anomaly\n y_pred = 1.*(distances > threshold )\n return y_pred",
"_____no_output_____"
],
[
"cv = StratifiedShuffleSplit(n_splits=VAL_SPLITS,test_size=0.15,random_state=0)\n# Thresholds to validate\nthresholds = np.linspace(0,100,100)\n# List with the f1 of all the thresholds at each validation fold\nf1_all = [] \nfor i,(idx_t, idx_v) in enumerate(cv.split(X,y)):\n X_train = X[idx_t]\n y_train = y[idx_t]\n X_val = X[idx_v]\n y_val = y[idx_v]\n # Now we need to erase the FRAUD cases on the TRAINING set \n X_train_normal = X_train[y_train==0]\n \n # Train the autoencoder\n autoencoder, encoder = create_autoencoder(INPUT_DIM,ENCODED_DIM)\n autoencoder.compile(optimizer='adam',\n loss='mean_squared_error')\n autoencoder.fit(x=X_train_normal, y= X_train_normal,\n batch_size=512,epochs=30, shuffle=True,\n verbose=0) # starts training\n \n # Plot of the validation set in the embedding space\n X_enc = encoder.predict(X_val)\n X_enc_normal = X_enc[y_val==0]\n X_enc_fraud = X_enc[y_val==1]\n sns.scatterplot(x = X_enc_normal[:,0], y = X_enc_normal[:,1] ,label='Normal', alpha=0.5)\n sns.scatterplot(x = X_enc_fraud[:,0], y = X_enc_fraud[:,1] ,label='Fraud')\n plt.show()\n \n # Transformation of the points through the autoencoder \n # and calculate the predictions\n y_preds=clf_autoencoder(X_val,autoencoder,thresholds)\n \n metrics_f1 = np.array([ f1_score(y_val,y_pred) for y_pred in y_preds.T ])\n f1_all.append(metrics_f1)\n \n # Save the models into files for future use\n autoencoder.save('models_autoencoder/autoencoder_fold_'+str(i+1)+'.h5')\n encoder.save('models_autoencoder/encoder_fold_'+str(i+1)+'.h5')\n del(autoencoder,encoder)\n\nf1_mean = np.mean(f1_all,axis=0) \n# Plot of F1-Threshold curves\nfor i,f1_fold in enumerate(f1_all):\n sns.lineplot(thresholds,f1_fold, label='Fold '+str(i+1))\nsns.scatterplot(thresholds,f1_mean,label='Mean')\nplt.show()\nf1_opt = f1_mean.max()\nthreshold_opt = thresholds[np.argmax(f1_mean)]\nprint('F1 max = {:.3f} at threshold = {:.3f}'.format(f1_opt,threshold_opt))",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
]
] |
cb949588eb3da18629ccbd2060966662dcd5036c
| 1,031,457 |
ipynb
|
Jupyter Notebook
|
1. Load and Visualize Data.ipynb
|
SyedaZainabAkhtar/Facial-Keypoint-Detection
|
361b40fca78e85f9edcaabfcc3de4cb4de9eeab8
|
[
"MIT"
] | 1 |
2020-05-07T23:29:49.000Z
|
2020-05-07T23:29:49.000Z
|
1. Load and Visualize Data.ipynb
|
SyedaZainabAkhtar/Facial-Keypoint-Detection
|
361b40fca78e85f9edcaabfcc3de4cb4de9eeab8
|
[
"MIT"
] | null | null | null |
1. Load and Visualize Data.ipynb
|
SyedaZainabAkhtar/Facial-Keypoint-Detection
|
361b40fca78e85f9edcaabfcc3de4cb4de9eeab8
|
[
"MIT"
] | null | null | null | 159.0037 | 193,300 | 0.850087 |
[
[
[
"# Facial Keypoint Detection\n \nThis project will be all about defining and training a convolutional neural network to perform facial keypoint detection, and using computer vision techniques to transform images of faces. The first step in any challenge like this will be to load and visualize the data you'll be working with. \n\nLet's take a look at some examples of images and corresponding facial keypoints.\n\n<img src='images/key_pts_example.png' width=50% height=50%/>\n\nFacial keypoints (also called facial landmarks) are the small magenta dots shown on each of the faces in the image above. In each training and test image, there is a single face and **68 keypoints, with coordinates (x, y), for that face**. These keypoints mark important areas of the face: the eyes, corners of the mouth, the nose, etc. These keypoints are relevant for a variety of tasks, such as face filters, emotion recognition, pose recognition, and so on. Here they are, numbered, and you can see that specific ranges of points match different portions of the face.\n\n<img src='images/landmarks_numbered.jpg' width=30% height=30%/>\n\n---",
"_____no_output_____"
],
[
"## Load and Visualize Data\n\nThe first step in working with any dataset is to become familiar with your data; you'll need to load in the images of faces and their keypoints and visualize them! This set of image data has been extracted from the [YouTube Faces Dataset](https://www.cs.tau.ac.il/~wolf/ytfaces/), which includes videos of people in YouTube videos. These videos have been fed through some processing steps and turned into sets of image frames containing one face and the associated keypoints.\n\n#### Training and Testing Data\n\nThis facial keypoints dataset consists of 5770 color images. All of these images are separated into either a training or a test set of data.\n\n* 3462 of these images are training images, for you to use as you create a model to predict keypoints.\n* 2308 are test images, which will be used to test the accuracy of your model.\n\nThe information about the images and keypoints in this dataset are summarized in CSV files, which we can read in using `pandas`. Let's read the training CSV and get the annotations in an (N, 2) array where N is the number of keypoints and 2 is the dimension of the keypoint coordinates (x, y).\n\n---",
"_____no_output_____"
],
[
"First, before we do anything, we have to load in our image data. This data is stored in a zip file and in the below cell, we access it by it's URL and unzip the data in a `/data/` directory that is separate from the workspace home directory.",
"_____no_output_____"
]
],
[
[
"# -- DO NOT CHANGE THIS CELL -- #\n!mkdir /data\n!wget -P /data/ https://s3.amazonaws.com/video.udacity-data.com/topher/2018/May/5aea1b91_train-test-data/train-test-data.zip\n!unzip -n /data/train-test-data.zip -d /data",
"--2019-02-24 09:43:07-- https://s3.amazonaws.com/video.udacity-data.com/topher/2018/May/5aea1b91_train-test-data/train-test-data.zip\nResolving s3.amazonaws.com (s3.amazonaws.com)... 52.216.179.69\nConnecting to s3.amazonaws.com (s3.amazonaws.com)|52.216.179.69|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 338613624 (323M) [application/zip]\nSaving to: ‘/data/train-test-data.zip’\n\ntrain-test-data.zip 100%[===================>] 322.93M 78.0MB/s in 4.4s \n\n2019-02-24 09:43:12 (73.5 MB/s) - ‘/data/train-test-data.zip’ saved [338613624/338613624]\n\nArchive: /data/train-test-data.zip\n creating: /data/test/\n inflating: /data/test/Abdel_Aziz_Al-Hakim_00.jpg \n inflating: /data/test/Abdel_Aziz_Al-Hakim_01.jpg \n inflating: /data/test/Abdel_Aziz_Al-Hakim_10.jpg \n inflating: /data/test/Abdel_Aziz_Al-Hakim_11.jpg \n inflating: /data/test/Abdel_Aziz_Al-Hakim_40.jpg \n inflating: /data/test/Abdel_Aziz_Al-Hakim_41.jpg \n inflating: /data/test/Abdullah_Gul_10.jpg \n inflating: /data/test/Abdullah_Gul_11.jpg \n inflating: /data/test/Abdullah_Gul_30.jpg \n inflating: /data/test/Abdullah_Gul_31.jpg \n inflating: /data/test/Abdullah_Gul_50.jpg \n inflating: /data/test/Abdullah_Gul_51.jpg \n inflating: /data/test/Adam_Sandler_00.jpg \n inflating: /data/test/Adam_Sandler_01.jpg \n inflating: /data/test/Adam_Sandler_10.jpg \n inflating: /data/test/Adam_Sandler_11.jpg \n inflating: /data/test/Adam_Sandler_40.jpg \n inflating: /data/test/Adam_Sandler_41.jpg \n inflating: /data/test/Adrian_Nastase_10.jpg \n inflating: /data/test/Adrian_Nastase_11.jpg \n inflating: /data/test/Adrian_Nastase_40.jpg \n inflating: /data/test/Adrian_Nastase_41.jpg \n inflating: /data/test/Adrian_Nastase_50.jpg \n inflating: /data/test/Adrian_Nastase_51.jpg \n inflating: /data/test/Agbani_Darego_00.jpg \n inflating: /data/test/Agbani_Darego_01.jpg \n inflating: /data/test/Agbani_Darego_20.jpg \n inflating: /data/test/Agbani_Darego_21.jpg \n inflating: /data/test/Agbani_Darego_40.jpg \n inflating: /data/test/Agbani_Darego_41.jpg \n inflating: /data/test/Agbani_Darego_50.jpg \n inflating: /data/test/Agbani_Darego_51.jpg \n inflating: /data/test/Agnes_Bruckner_00.jpg \n inflating: /data/test/Agnes_Bruckner_01.jpg \n inflating: /data/test/Agnes_Bruckner_10.jpg \n inflating: /data/test/Agnes_Bruckner_11.jpg \n inflating: /data/test/Agnes_Bruckner_20.jpg \n inflating: /data/test/Agnes_Bruckner_21.jpg \n inflating: /data/test/Agnes_Bruckner_40.jpg \n inflating: /data/test/Agnes_Bruckner_41.jpg \n inflating: /data/test/Ahmad_Masood_00.jpg \n inflating: /data/test/Ahmad_Masood_01.jpg \n inflating: /data/test/Ahmad_Masood_30.jpg \n inflating: /data/test/Ahmad_Masood_31.jpg \n inflating: /data/test/Ahmad_Masood_40.jpg \n inflating: /data/test/Ahmad_Masood_41.jpg \n inflating: /data/test/Ahmed_Ahmed_00.jpg \n inflating: /data/test/Ahmed_Ahmed_01.jpg \n inflating: /data/test/Ahmed_Ahmed_10.jpg \n inflating: /data/test/Ahmed_Ahmed_11.jpg \n inflating: /data/test/Ahmed_Ahmed_40.jpg \n inflating: /data/test/Ahmed_Ahmed_41.jpg \n inflating: /data/test/Ahmed_Ahmed_50.jpg \n inflating: /data/test/Ahmed_Ahmed_51.jpg \n inflating: /data/test/Aidan_Quinn_00.jpg \n inflating: /data/test/Aidan_Quinn_01.jpg \n inflating: /data/test/Aidan_Quinn_10.jpg \n inflating: /data/test/Aidan_Quinn_11.jpg \n inflating: /data/test/Aidan_Quinn_20.jpg \n inflating: /data/test/Aidan_Quinn_21.jpg \n inflating: /data/test/Aidan_Quinn_30.jpg \n inflating: /data/test/Aidan_Quinn_31.jpg \n inflating: /data/test/Aishwarya_Rai_00.jpg \n inflating: /data/test/Aishwarya_Rai_01.jpg \n inflating: /data/test/Aishwarya_Rai_10.jpg \n inflating: /data/test/Aishwarya_Rai_11.jpg \n inflating: /data/test/Aishwarya_Rai_40.jpg \n inflating: /data/test/Aishwarya_Rai_41.jpg \n inflating: /data/test/Aishwarya_Rai_50.jpg \n inflating: /data/test/Aishwarya_Rai_51.jpg \n inflating: /data/test/Albert_Brooks_00.jpg \n inflating: /data/test/Albert_Brooks_01.jpg \n inflating: /data/test/Albert_Brooks_10.jpg \n inflating: /data/test/Albert_Brooks_11.jpg \n inflating: /data/test/Albert_Brooks_30.jpg \n inflating: /data/test/Albert_Brooks_31.jpg \n inflating: /data/test/Alejandro_Toledo_10.jpg \n inflating: /data/test/Alejandro_Toledo_11.jpg \n inflating: /data/test/Alejandro_Toledo_30.jpg \n inflating: /data/test/Alejandro_Toledo_31.jpg \n inflating: /data/test/Alejandro_Toledo_50.jpg \n inflating: /data/test/Alejandro_Toledo_51.jpg \n inflating: /data/test/Aleksander_Kwasniewski_00.jpg \n inflating: /data/test/Aleksander_Kwasniewski_01.jpg \n inflating: /data/test/Aleksander_Kwasniewski_10.jpg \n inflating: /data/test/Aleksander_Kwasniewski_11.jpg \n inflating: /data/test/Aleksander_Kwasniewski_20.jpg \n inflating: /data/test/Aleksander_Kwasniewski_21.jpg \n inflating: /data/test/Aleksander_Kwasniewski_30.jpg \n inflating: /data/test/Aleksander_Kwasniewski_31.jpg \n inflating: /data/test/Alex_Ferguson_00.jpg \n inflating: /data/test/Alex_Ferguson_01.jpg \n inflating: /data/test/Alex_Ferguson_10.jpg \n inflating: /data/test/Alex_Ferguson_11.jpg \n inflating: /data/test/Alex_Ferguson_50.jpg \n inflating: /data/test/Alex_Ferguson_51.jpg \n inflating: /data/test/Alexandra_Pelosi_00.jpg \n inflating: /data/test/Alexandra_Pelosi_01.jpg \n inflating: /data/test/Alexandra_Pelosi_10.jpg \n inflating: /data/test/Alexandra_Pelosi_11.jpg \n inflating: /data/test/Alexandra_Pelosi_30.jpg \n inflating: /data/test/Alexandra_Pelosi_31.jpg \n inflating: /data/test/Alfredo_di_Stefano_00.jpg \n inflating: /data/test/Alfredo_di_Stefano_01.jpg \n inflating: /data/test/Alfredo_di_Stefano_20.jpg \n inflating: /data/test/Alfredo_di_Stefano_21.jpg \n inflating: /data/test/Alfredo_di_Stefano_50.jpg \n inflating: /data/test/Alfredo_di_Stefano_51.jpg \n inflating: /data/test/Ali_Abbas_20.jpg \n inflating: /data/test/Ali_Abbas_21.jpg \n inflating: /data/test/Ali_Abbas_30.jpg \n inflating: /data/test/Ali_Abbas_31.jpg \n inflating: /data/test/Ali_Abbas_40.jpg \n inflating: /data/test/Ali_Abbas_41.jpg \n inflating: /data/test/Ali_Abbas_50.jpg \n inflating: /data/test/Ali_Abbas_51.jpg \n inflating: /data/test/Alicia_Silverstone_00.jpg \n inflating: /data/test/Alicia_Silverstone_01.jpg \n inflating: /data/test/Alicia_Silverstone_10.jpg \n inflating: /data/test/Alicia_Silverstone_11.jpg \n inflating: /data/test/Alicia_Silverstone_20.jpg \n inflating: /data/test/Alicia_Silverstone_21.jpg \n inflating: /data/test/Alicia_Silverstone_50.jpg \n inflating: /data/test/Alicia_Silverstone_51.jpg \n inflating: /data/test/Alma_Powell_00.jpg \n inflating: /data/test/Alma_Powell_01.jpg \n inflating: /data/test/Alma_Powell_10.jpg \n inflating: /data/test/Alma_Powell_11.jpg \n inflating: /data/test/Alma_Powell_40.jpg \n inflating: /data/test/Alma_Powell_41.jpg \n inflating: /data/test/Alma_Powell_50.jpg \n inflating: /data/test/Alma_Powell_51.jpg \n inflating: /data/test/Alvaro_Silva_Calderon_00.jpg \n inflating: /data/test/Alvaro_Silva_Calderon_01.jpg \n inflating: /data/test/Alvaro_Silva_Calderon_10.jpg \n inflating: /data/test/Alvaro_Silva_Calderon_11.jpg \n inflating: /data/test/Alvaro_Silva_Calderon_20.jpg \n inflating: /data/test/Alvaro_Silva_Calderon_21.jpg \n inflating: /data/test/Alvaro_Silva_Calderon_30.jpg \n inflating: /data/test/Alvaro_Silva_Calderon_31.jpg \n inflating: /data/test/Amelia_Vega_10.jpg \n inflating: /data/test/Amelia_Vega_11.jpg \n inflating: /data/test/Amelia_Vega_20.jpg \n inflating: /data/test/Amelia_Vega_21.jpg \n inflating: /data/test/Amelia_Vega_30.jpg \n inflating: /data/test/Amelia_Vega_31.jpg \n inflating: /data/test/Amelia_Vega_40.jpg \n inflating: /data/test/Amelia_Vega_41.jpg \n inflating: /data/test/Amy_Brenneman_10.jpg \n inflating: /data/test/Amy_Brenneman_11.jpg \n inflating: /data/test/Amy_Brenneman_30.jpg \n inflating: /data/test/Amy_Brenneman_31.jpg \n inflating: /data/test/Amy_Brenneman_50.jpg \n inflating: /data/test/Amy_Brenneman_51.jpg \n inflating: /data/test/Andrea_Bocelli_10.jpg \n inflating: /data/test/Andrea_Bocelli_11.jpg \n inflating: /data/test/Andrea_Bocelli_20.jpg \n inflating: /data/test/Andrea_Bocelli_21.jpg \n inflating: /data/test/Andrea_Bocelli_30.jpg \n inflating: /data/test/Andrea_Bocelli_31.jpg \n inflating: /data/test/Andy_Roddick_20.jpg \n inflating: /data/test/Andy_Roddick_21.jpg \n inflating: /data/test/Andy_Roddick_40.jpg \n inflating: /data/test/Andy_Roddick_41.jpg \n inflating: /data/test/Andy_Roddick_50.jpg \n inflating: /data/test/Andy_Roddick_51.jpg \n inflating: /data/test/Andy_Rooney_10.jpg \n inflating: /data/test/Andy_Rooney_11.jpg \n inflating: /data/test/Andy_Rooney_20.jpg \n inflating: /data/test/Andy_Rooney_21.jpg \n inflating: /data/test/Andy_Rooney_50.jpg \n inflating: /data/test/Andy_Rooney_51.jpg \n inflating: /data/test/Angel_Lockward_30.jpg \n inflating: /data/test/Angel_Lockward_31.jpg \n inflating: /data/test/Angel_Lockward_40.jpg \n inflating: /data/test/Angel_Lockward_41.jpg \n inflating: /data/test/Angel_Lockward_50.jpg \n inflating: /data/test/Angel_Lockward_51.jpg \n inflating: /data/test/Angela_Bassett_20.jpg \n inflating: /data/test/Angela_Bassett_21.jpg \n inflating: /data/test/Angela_Bassett_30.jpg \n inflating: /data/test/Angela_Bassett_31.jpg \n inflating: /data/test/Angela_Bassett_40.jpg \n inflating: /data/test/Angela_Bassett_41.jpg \n inflating: /data/test/Angelo_Reyes_20.jpg \n inflating: /data/test/Angelo_Reyes_21.jpg \n inflating: /data/test/Angelo_Reyes_30.jpg \n inflating: /data/test/Angelo_Reyes_31.jpg \n inflating: /data/test/Angelo_Reyes_50.jpg \n inflating: /data/test/Angelo_Reyes_51.jpg \n inflating: /data/test/Baburam_Bhattari_00.jpg \n inflating: /data/test/Baburam_Bhattari_01.jpg \n inflating: /data/test/Baburam_Bhattari_20.jpg \n inflating: /data/test/Baburam_Bhattari_21.jpg \n inflating: /data/test/Baburam_Bhattari_30.jpg \n inflating: /data/test/Baburam_Bhattari_31.jpg \n inflating: /data/test/Barbara_Bodine_00.jpg \n inflating: /data/test/Barbara_Bodine_01.jpg \n inflating: /data/test/Barbara_Bodine_20.jpg \n inflating: /data/test/Barbara_Bodine_21.jpg \n inflating: /data/test/Barbara_Bodine_40.jpg \n inflating: /data/test/Barbara_Bodine_41.jpg \n inflating: /data/test/Barbara_Bodine_50.jpg \n inflating: /data/test/Barbara_Bodine_51.jpg \n inflating: /data/test/Barbara_Boxer_10.jpg \n inflating: /data/test/Barbara_Boxer_11.jpg \n inflating: /data/test/Barbara_Boxer_40.jpg \n inflating: /data/test/Barbara_Boxer_41.jpg \n inflating: /data/test/Barbara_Boxer_50.jpg \n inflating: /data/test/Barbara_Boxer_51.jpg \n inflating: /data/test/Barbara_Walters_00.jpg \n inflating: /data/test/Barbara_Walters_01.jpg \n inflating: /data/test/Barbara_Walters_20.jpg \n inflating: /data/test/Barbara_Walters_21.jpg \n inflating: /data/test/Barbara_Walters_40.jpg \n inflating: /data/test/Barbara_Walters_41.jpg \n inflating: /data/test/Barbara_Walters_50.jpg \n inflating: /data/test/Barbara_Walters_51.jpg \n inflating: /data/test/Barry_Alvarez_00.jpg \n inflating: /data/test/Barry_Alvarez_01.jpg \n inflating: /data/test/Barry_Alvarez_10.jpg \n inflating: /data/test/Barry_Alvarez_11.jpg \n inflating: /data/test/Barry_Alvarez_20.jpg \n inflating: /data/test/Barry_Alvarez_21.jpg \n inflating: /data/test/Barry_Alvarez_30.jpg \n inflating: /data/test/Barry_Alvarez_31.jpg \n inflating: /data/test/Ben_Kingsley_10.jpg \n inflating: /data/test/Ben_Kingsley_11.jpg \n inflating: /data/test/Ben_Kingsley_20.jpg \n inflating: /data/test/Ben_Kingsley_21.jpg \n inflating: /data/test/Ben_Kingsley_50.jpg \n inflating: /data/test/Ben_Kingsley_51.jpg \n inflating: /data/test/Ben_Stein_10.jpg \n inflating: /data/test/Ben_Stein_11.jpg \n inflating: /data/test/Ben_Stein_30.jpg \n inflating: /data/test/Ben_Stein_31.jpg \n inflating: /data/test/Ben_Stein_40.jpg \n inflating: /data/test/Ben_Stein_41.jpg \n inflating: /data/test/Ben_Stein_50.jpg \n inflating: /data/test/Ben_Stein_51.jpg \n inflating: /data/test/Benedita_da_Silva_10.jpg \n inflating: /data/test/Benedita_da_Silva_11.jpg \n inflating: /data/test/Benedita_da_Silva_20.jpg \n inflating: /data/test/Benedita_da_Silva_21.jpg \n inflating: /data/test/Benedita_da_Silva_50.jpg \n inflating: /data/test/Benedita_da_Silva_51.jpg \n inflating: /data/test/Benjamin_McKenzie_30.jpg \n inflating: /data/test/Benjamin_McKenzie_31.jpg \n inflating: /data/test/Benjamin_McKenzie_40.jpg \n inflating: /data/test/Benjamin_McKenzie_41.jpg \n inflating: /data/test/Benjamin_McKenzie_50.jpg \n inflating: /data/test/Benjamin_McKenzie_51.jpg \n inflating: /data/test/Benjamin_Netanyahu_00.jpg \n inflating: /data/test/Benjamin_Netanyahu_01.jpg \n inflating: /data/test/Benjamin_Netanyahu_10.jpg \n inflating: /data/test/Benjamin_Netanyahu_11.jpg \n inflating: /data/test/Benjamin_Netanyahu_30.jpg \n inflating: /data/test/Benjamin_Netanyahu_31.jpg \n inflating: /data/test/Benjamin_Netanyahu_40.jpg \n inflating: /data/test/Benjamin_Netanyahu_41.jpg \n inflating: /data/test/Beyonce_Knowles_00.jpg \n inflating: /data/test/Beyonce_Knowles_01.jpg \n inflating: /data/test/Beyonce_Knowles_30.jpg \n inflating: /data/test/Beyonce_Knowles_31.jpg \n inflating: /data/test/Beyonce_Knowles_50.jpg \n inflating: /data/test/Beyonce_Knowles_51.jpg \n inflating: /data/test/Bianca_Jagger_20.jpg \n inflating: /data/test/Bianca_Jagger_21.jpg \n inflating: /data/test/Bianca_Jagger_30.jpg \n inflating: /data/test/Bianca_Jagger_31.jpg \n inflating: /data/test/Bianca_Jagger_40.jpg \n inflating: /data/test/Bianca_Jagger_41.jpg \n inflating: /data/test/Biljana_Plavsic_00.jpg \n inflating: /data/test/Biljana_Plavsic_01.jpg \n inflating: /data/test/Biljana_Plavsic_10.jpg \n inflating: /data/test/Biljana_Plavsic_11.jpg \n inflating: /data/test/Biljana_Plavsic_30.jpg \n inflating: /data/test/Biljana_Plavsic_31.jpg \n inflating: /data/test/Bill_Bradley_10.jpg \n inflating: /data/test/Bill_Bradley_11.jpg \n inflating: /data/test/Bill_Bradley_30.jpg \n inflating: /data/test/Bill_Bradley_31.jpg \n inflating: /data/test/Bill_Bradley_50.jpg \n inflating: /data/test/Bill_Bradley_51.jpg \n inflating: /data/test/Bill_Clinton_00.jpg \n inflating: /data/test/Bill_Clinton_01.jpg \n inflating: /data/test/Bill_Clinton_10.jpg \n inflating: /data/test/Bill_Clinton_11.jpg \n inflating: /data/test/Bill_Clinton_50.jpg \n inflating: /data/test/Bill_Clinton_51.jpg \n inflating: /data/test/Bill_Frist_00.jpg \n inflating: /data/test/Bill_Frist_01.jpg \n inflating: /data/test/Bill_Frist_10.jpg \n inflating: /data/test/Bill_Frist_11.jpg \n inflating: /data/test/Bill_Frist_20.jpg \n inflating: /data/test/Bill_Frist_21.jpg \n inflating: /data/test/Cameron_Diaz_20.jpg \n inflating: /data/test/Cameron_Diaz_21.jpg \n inflating: /data/test/Cameron_Diaz_40.jpg \n inflating: /data/test/Cameron_Diaz_41.jpg \n inflating: /data/test/Cameron_Diaz_50.jpg \n inflating: /data/test/Cameron_Diaz_51.jpg \n inflating: /data/test/Carey_Lowell_00.jpg \n inflating: /data/test/Carey_Lowell_01.jpg \n inflating: /data/test/Carey_Lowell_20.jpg \n inflating: /data/test/Carey_Lowell_21.jpg \n inflating: /data/test/Carey_Lowell_30.jpg \n inflating: /data/test/Carey_Lowell_31.jpg \n inflating: /data/test/Carla_Gugino_10.jpg \n inflating: /data/test/Carla_Gugino_11.jpg \n inflating: /data/test/Carla_Gugino_30.jpg \n inflating: /data/test/Carla_Gugino_31.jpg \n inflating: /data/test/Carla_Gugino_50.jpg \n inflating: /data/test/Carla_Gugino_51.jpg \n inflating: /data/test/Carlo_Azeglio_Ciampi_00.jpg \n inflating: /data/test/Carlo_Azeglio_Ciampi_01.jpg \n inflating: /data/test/Carlo_Azeglio_Ciampi_10.jpg \n inflating: /data/test/Carlo_Azeglio_Ciampi_11.jpg \n inflating: /data/test/Carlo_Azeglio_Ciampi_20.jpg \n inflating: /data/test/Carlo_Azeglio_Ciampi_21.jpg \n inflating: /data/test/Carlo_Azeglio_Ciampi_50.jpg \n inflating: /data/test/Carlo_Azeglio_Ciampi_51.jpg \n inflating: /data/test/Carlos_Ghosn_10.jpg \n inflating: /data/test/Carlos_Ghosn_11.jpg \n inflating: /data/test/Carlos_Ghosn_20.jpg \n inflating: /data/test/Carlos_Ghosn_21.jpg \n inflating: /data/test/Carlos_Ghosn_40.jpg \n inflating: /data/test/Carlos_Ghosn_41.jpg \n inflating: /data/test/Carlos_Iturgaitz_10.jpg \n inflating: /data/test/Carlos_Iturgaitz_11.jpg \n inflating: /data/test/Carlos_Iturgaitz_20.jpg \n inflating: /data/test/Carlos_Iturgaitz_21.jpg \n inflating: /data/test/Carlos_Iturgaitz_30.jpg \n inflating: /data/test/Carlos_Iturgaitz_31.jpg \n inflating: /data/test/Carlos_Iturgaitz_50.jpg \n inflating: /data/test/Carlos_Iturgaitz_51.jpg \n inflating: /data/test/Carlos_Menem_00.jpg \n inflating: /data/test/Carlos_Menem_01.jpg \n inflating: /data/test/Carlos_Menem_20.jpg \n inflating: /data/test/Carlos_Menem_21.jpg \n inflating: /data/test/Carlos_Menem_30.jpg \n inflating: /data/test/Carlos_Menem_31.jpg \n inflating: /data/test/Carlos_Queiroz_00.jpg \n inflating: /data/test/Carlos_Queiroz_01.jpg \n inflating: /data/test/Carlos_Queiroz_10.jpg \n inflating: /data/test/Carlos_Queiroz_11.jpg \n inflating: /data/test/Carlos_Queiroz_50.jpg \n inflating: /data/test/Carlos_Queiroz_51.jpg \n inflating: /data/test/Carrie-Anne_Moss_00.jpg \n inflating: /data/test/Carrie-Anne_Moss_01.jpg \n inflating: /data/test/Carrie-Anne_Moss_10.jpg \n inflating: /data/test/Carrie-Anne_Moss_11.jpg \n inflating: /data/test/Carrie-Anne_Moss_50.jpg \n inflating: /data/test/Carrie-Anne_Moss_51.jpg \n inflating: /data/test/Catriona_Le_May_Doan_10.jpg \n inflating: /data/test/Catriona_Le_May_Doan_11.jpg \n inflating: /data/test/Catriona_Le_May_Doan_30.jpg \n inflating: /data/test/Catriona_Le_May_Doan_31.jpg \n inflating: /data/test/Catriona_Le_May_Doan_40.jpg \n inflating: /data/test/Catriona_Le_May_Doan_41.jpg \n inflating: /data/test/Cecilia_Cheung_00.jpg \n inflating: /data/test/Cecilia_Cheung_01.jpg \n inflating: /data/test/Cecilia_Cheung_10.jpg \n inflating: /data/test/Cecilia_Cheung_11.jpg \n inflating: /data/test/Cecilia_Cheung_20.jpg \n inflating: /data/test/Cecilia_Cheung_21.jpg \n inflating: /data/test/Cecilia_Cheung_50.jpg \n inflating: /data/test/Cecilia_Cheung_51.jpg \n inflating: /data/test/Celso_Amorim_10.jpg \n inflating: /data/test/Celso_Amorim_11.jpg \n inflating: /data/test/Celso_Amorim_40.jpg \n inflating: /data/test/Celso_Amorim_41.jpg \n inflating: /data/test/Celso_Amorim_50.jpg \n inflating: /data/test/Celso_Amorim_51.jpg \n inflating: /data/test/Celso_Lafer_00.jpg \n inflating: /data/test/Celso_Lafer_01.jpg \n inflating: /data/test/Celso_Lafer_10.jpg \n inflating: /data/test/Celso_Lafer_11.jpg \n inflating: /data/test/Celso_Lafer_20.jpg \n inflating: /data/test/Celso_Lafer_21.jpg \n inflating: /data/test/Chadha_Gurinder_10.jpg \n inflating: /data/test/Chadha_Gurinder_11.jpg \n inflating: /data/test/Chadha_Gurinder_20.jpg \n inflating: /data/test/Chadha_Gurinder_21.jpg \n inflating: /data/test/Chadha_Gurinder_50.jpg \n inflating: /data/test/Chadha_Gurinder_51.jpg \n inflating: /data/test/Charles_Bronson_00.jpg \n inflating: /data/test/Charles_Bronson_01.jpg \n inflating: /data/test/Charles_Bronson_10.jpg \n inflating: /data/test/Charles_Bronson_11.jpg \n inflating: /data/test/Charles_Bronson_50.jpg \n inflating: /data/test/Charles_Bronson_51.jpg \n inflating: /data/test/Charlie_Coles_00.jpg \n inflating: /data/test/Charlie_Coles_01.jpg \n inflating: /data/test/Charlie_Coles_10.jpg \n inflating: /data/test/Charlie_Coles_11.jpg \n inflating: /data/test/Charlie_Coles_20.jpg \n inflating: /data/test/Charlie_Coles_21.jpg \n inflating: /data/test/Charlize_Theron_10.jpg \n inflating: /data/test/Charlize_Theron_11.jpg \n inflating: /data/test/Charlize_Theron_30.jpg \n inflating: /data/test/Charlize_Theron_31.jpg \n inflating: /data/test/Charlize_Theron_50.jpg \n inflating: /data/test/Charlize_Theron_51.jpg \n inflating: /data/test/Charlotte_Casiraghi_00.jpg \n inflating: /data/test/Charlotte_Casiraghi_01.jpg \n inflating: /data/test/Charlotte_Casiraghi_10.jpg \n inflating: /data/test/Charlotte_Casiraghi_11.jpg \n inflating: /data/test/Charlotte_Casiraghi_20.jpg \n inflating: /data/test/Charlotte_Casiraghi_21.jpg \n inflating: /data/test/Charlotte_Rampling_00.jpg \n inflating: /data/test/Charlotte_Rampling_01.jpg \n inflating: /data/test/Charlotte_Rampling_30.jpg \n inflating: /data/test/Charlotte_Rampling_31.jpg \n inflating: /data/test/Charlotte_Rampling_40.jpg \n inflating: /data/test/Charlotte_Rampling_41.jpg \n inflating: /data/test/Charlotte_Rampling_50.jpg \n inflating: /data/test/Charlotte_Rampling_51.jpg \n inflating: /data/test/Cherie_Blair_00.jpg \n inflating: /data/test/Cherie_Blair_01.jpg \n inflating: /data/test/Cherie_Blair_20.jpg \n inflating: /data/test/Cherie_Blair_21.jpg \n inflating: /data/test/Cherie_Blair_30.jpg \n inflating: /data/test/Cherie_Blair_31.jpg \n inflating: /data/test/Cherie_Blair_40.jpg \n inflating: /data/test/Cherie_Blair_41.jpg \n inflating: /data/test/Chita_Rivera_00.jpg \n inflating: /data/test/Chita_Rivera_01.jpg \n inflating: /data/test/Chita_Rivera_10.jpg \n inflating: /data/test/Chita_Rivera_11.jpg \n inflating: /data/test/Chita_Rivera_30.jpg \n inflating: /data/test/Chita_Rivera_31.jpg \n inflating: /data/test/Chris_Cirino_20.jpg \n inflating: /data/test/Chris_Cirino_21.jpg \n inflating: /data/test/Chris_Cirino_30.jpg \n inflating: /data/test/Chris_Cirino_31.jpg \n inflating: /data/test/Chris_Cirino_50.jpg \n inflating: /data/test/Chris_Cirino_51.jpg \n inflating: /data/test/Chris_Cooper_00.jpg \n inflating: /data/test/Chris_Cooper_01.jpg \n inflating: /data/test/Chris_Cooper_30.jpg \n inflating: /data/test/Chris_Cooper_31.jpg \n inflating: /data/test/Chris_Cooper_40.jpg \n inflating: /data/test/Chris_Cooper_41.jpg \n inflating: /data/test/Chris_Matthews_00.jpg \n inflating: /data/test/Chris_Matthews_01.jpg \n inflating: /data/test/Chris_Matthews_10.jpg \n inflating: /data/test/Chris_Matthews_11.jpg \n inflating: /data/test/Chris_Matthews_30.jpg \n inflating: /data/test/Chris_Matthews_31.jpg \n inflating: /data/test/Chris_Matthews_50.jpg \n inflating: /data/test/Chris_Matthews_51.jpg \n inflating: /data/test/Chris_Noth_00.jpg \n inflating: /data/test/Chris_Noth_01.jpg \n inflating: /data/test/Chris_Noth_10.jpg \n inflating: /data/test/Chris_Noth_11.jpg \n inflating: /data/test/Chris_Noth_30.jpg \n inflating: /data/test/Chris_Noth_31.jpg \n inflating: /data/test/Chris_Rock_00.jpg \n inflating: /data/test/Chris_Rock_01.jpg \n inflating: /data/test/Chris_Rock_10.jpg \n inflating: /data/test/Chris_Rock_11.jpg \n inflating: /data/test/Chris_Rock_20.jpg \n inflating: /data/test/Chris_Rock_21.jpg \n inflating: /data/test/Christine_Ebersole_00.jpg \n inflating: /data/test/Christine_Ebersole_01.jpg \n inflating: /data/test/Christine_Ebersole_20.jpg \n inflating: /data/test/Christine_Ebersole_21.jpg \n inflating: /data/test/Christine_Ebersole_50.jpg \n inflating: /data/test/Christine_Ebersole_51.jpg \n inflating: /data/test/Christopher_Amolsch_20.jpg \n inflating: /data/test/Christopher_Amolsch_21.jpg \n inflating: /data/test/Christopher_Amolsch_40.jpg \n inflating: /data/test/Christopher_Amolsch_41.jpg \n inflating: /data/test/Christopher_Amolsch_50.jpg \n inflating: /data/test/Christopher_Amolsch_51.jpg \n inflating: /data/test/Christopher_Reeve_10.jpg \n inflating: /data/test/Christopher_Reeve_11.jpg \n inflating: /data/test/Christopher_Reeve_20.jpg \n inflating: /data/test/Christopher_Reeve_21.jpg \n inflating: /data/test/Christopher_Reeve_40.jpg \n inflating: /data/test/Christopher_Reeve_41.jpg \n inflating: /data/test/Christopher_Walken_00.jpg \n inflating: /data/test/Christopher_Walken_01.jpg \n inflating: /data/test/Christopher_Walken_20.jpg \n inflating: /data/test/Christopher_Walken_21.jpg \n inflating: /data/test/Christopher_Walken_40.jpg \n inflating: /data/test/Christopher_Walken_41.jpg \n inflating: /data/test/Christopher_Walken_50.jpg \n inflating: /data/test/Christopher_Walken_51.jpg \n inflating: /data/test/Chuck_Hagel_20.jpg \n inflating: /data/test/Chuck_Hagel_21.jpg \n inflating: /data/test/Chuck_Hagel_30.jpg \n inflating: /data/test/Chuck_Hagel_31.jpg \n inflating: /data/test/Chuck_Hagel_40.jpg \n inflating: /data/test/Chuck_Hagel_41.jpg \n inflating: /data/test/Chuck_Hagel_50.jpg \n inflating: /data/test/Chuck_Hagel_51.jpg \n inflating: /data/test/Chuck_Woolery_10.jpg \n inflating: /data/test/Chuck_Woolery_11.jpg \n inflating: /data/test/Chuck_Woolery_20.jpg \n inflating: /data/test/Chuck_Woolery_21.jpg \n inflating: /data/test/Chuck_Woolery_40.jpg \n inflating: /data/test/Chuck_Woolery_41.jpg \n inflating: /data/test/Cindy_Crawford_00.jpg \n inflating: /data/test/Cindy_Crawford_01.jpg \n inflating: /data/test/Cindy_Crawford_30.jpg \n inflating: /data/test/Cindy_Crawford_31.jpg \n inflating: /data/test/Cindy_Crawford_50.jpg \n inflating: /data/test/Cindy_Crawford_51.jpg \n inflating: /data/test/Cindy_Klassen_00.jpg \n inflating: /data/test/Cindy_Klassen_01.jpg \n inflating: /data/test/Cindy_Klassen_30.jpg \n inflating: /data/test/Cindy_Klassen_31.jpg \n inflating: /data/test/Cindy_Klassen_40.jpg \n inflating: /data/test/Cindy_Klassen_41.jpg \n inflating: /data/test/Claire_Danes_30.jpg \n inflating: /data/test/Claire_Danes_31.jpg \n inflating: /data/test/Claire_Danes_40.jpg \n inflating: /data/test/Claire_Danes_41.jpg \n inflating: /data/test/Claire_Danes_50.jpg \n inflating: /data/test/Claire_Danes_51.jpg \n inflating: /data/test/Clark_Randt_10.jpg \n inflating: /data/test/Clark_Randt_11.jpg \n inflating: /data/test/Clark_Randt_20.jpg \n inflating: /data/test/Clark_Randt_21.jpg \n inflating: /data/test/Clark_Randt_40.jpg \n inflating: /data/test/Clark_Randt_41.jpg \n inflating: /data/test/Clark_Randt_50.jpg \n inflating: /data/test/Clark_Randt_51.jpg \n inflating: /data/test/Clay_Aiken_00.jpg \n inflating: /data/test/Clay_Aiken_01.jpg \n inflating: /data/test/Clay_Aiken_30.jpg \n inflating: /data/test/Clay_Aiken_31.jpg \n inflating: /data/test/Clay_Aiken_40.jpg \n inflating: /data/test/Clay_Aiken_41.jpg \n inflating: /data/test/Clay_Aiken_50.jpg \n inflating: /data/test/Clay_Aiken_51.jpg \n inflating: /data/test/Clint_Howard_00.jpg \n inflating: /data/test/Clint_Howard_01.jpg \n inflating: /data/test/Clint_Howard_10.jpg \n inflating: /data/test/Clint_Howard_11.jpg \n inflating: /data/test/Clint_Howard_20.jpg \n inflating: /data/test/Clint_Howard_21.jpg \n inflating: /data/test/Clint_Howard_30.jpg \n inflating: /data/test/Clint_Howard_31.jpg \n inflating: /data/test/Clive_Lloyd_30.jpg \n inflating: /data/test/Clive_Lloyd_31.jpg \n inflating: /data/test/Clive_Lloyd_40.jpg \n inflating: /data/test/Clive_Lloyd_41.jpg \n inflating: /data/test/Clive_Lloyd_50.jpg \n inflating: /data/test/Clive_Lloyd_51.jpg \n inflating: /data/test/Colin_Powell_10.jpg \n inflating: /data/test/Colin_Powell_11.jpg \n inflating: /data/test/Colin_Powell_40.jpg \n inflating: /data/test/Colin_Powell_41.jpg \n inflating: /data/test/Colin_Powell_50.jpg \n inflating: /data/test/Colin_Powell_51.jpg \n inflating: /data/test/Conan_OBrien_00.jpg \n inflating: /data/test/Conan_OBrien_01.jpg \n inflating: /data/test/Conan_OBrien_10.jpg \n inflating: /data/test/Conan_OBrien_11.jpg \n inflating: /data/test/Conan_OBrien_50.jpg \n inflating: /data/test/Conan_OBrien_51.jpg \n inflating: /data/test/Condoleezza_Rice_10.jpg \n inflating: /data/test/Condoleezza_Rice_11.jpg \n inflating: /data/test/Condoleezza_Rice_20.jpg \n inflating: /data/test/Condoleezza_Rice_21.jpg \n inflating: /data/test/Condoleezza_Rice_30.jpg \n inflating: /data/test/Condoleezza_Rice_31.jpg \n inflating: /data/test/Condoleezza_Rice_40.jpg \n inflating: /data/test/Condoleezza_Rice_41.jpg \n inflating: /data/test/Connie_Chung_20.jpg \n inflating: /data/test/Connie_Chung_21.jpg \n inflating: /data/test/Connie_Chung_30.jpg \n inflating: /data/test/Connie_Chung_31.jpg \n inflating: /data/test/Connie_Chung_40.jpg \n inflating: /data/test/Connie_Chung_41.jpg \n inflating: /data/test/Connie_Chung_50.jpg \n inflating: /data/test/Connie_Chung_51.jpg \n inflating: /data/test/Craig_David_10.jpg \n inflating: /data/test/Craig_David_11.jpg \n inflating: /data/test/Craig_David_20.jpg \n inflating: /data/test/Craig_David_21.jpg \n inflating: /data/test/Craig_David_30.jpg \n inflating: /data/test/Craig_David_31.jpg \n inflating: /data/test/Craig_David_50.jpg \n inflating: /data/test/Craig_David_51.jpg \n inflating: /data/test/Cristina_Fernandez_30.jpg \n inflating: /data/test/Cristina_Fernandez_31.jpg \n inflating: /data/test/Cristina_Fernandez_40.jpg \n inflating: /data/test/Cristina_Fernandez_41.jpg \n inflating: /data/test/Cristina_Fernandez_50.jpg \n inflating: /data/test/Cristina_Fernandez_51.jpg \n inflating: /data/test/Cristina_Saralegui_00.jpg \n inflating: /data/test/Cristina_Saralegui_01.jpg \n inflating: /data/test/Cristina_Saralegui_30.jpg \n inflating: /data/test/Cristina_Saralegui_31.jpg \n inflating: /data/test/Cristina_Saralegui_50.jpg \n inflating: /data/test/Cristina_Saralegui_51.jpg \n inflating: /data/test/Dai_Bachtiar_00.jpg \n inflating: /data/test/Dai_Bachtiar_01.jpg \n inflating: /data/test/Dai_Bachtiar_10.jpg \n inflating: /data/test/Dai_Bachtiar_11.jpg \n inflating: /data/test/Dai_Bachtiar_20.jpg \n inflating: /data/test/Dai_Bachtiar_21.jpg \n inflating: /data/test/Dai_Bachtiar_50.jpg \n inflating: /data/test/Dai_Bachtiar_51.jpg \n inflating: /data/test/Daisy_Fuentes_20.jpg \n inflating: /data/test/Daisy_Fuentes_21.jpg \n inflating: /data/test/Daisy_Fuentes_30.jpg \n inflating: /data/test/Daisy_Fuentes_31.jpg \n inflating: /data/test/Daisy_Fuentes_40.jpg \n inflating: /data/test/Daisy_Fuentes_41.jpg \n inflating: /data/test/Dan_Ackroyd_00.jpg \n inflating: /data/test/Dan_Ackroyd_01.jpg \n inflating: /data/test/Dan_Ackroyd_20.jpg \n inflating: /data/test/Dan_Ackroyd_21.jpg \n inflating: /data/test/Dan_Ackroyd_30.jpg \n inflating: /data/test/Dan_Ackroyd_31.jpg \n inflating: /data/test/Daniel_Radcliffe_00.jpg \n inflating: /data/test/Daniel_Radcliffe_01.jpg \n inflating: /data/test/Daniel_Radcliffe_20.jpg \n inflating: /data/test/Daniel_Radcliffe_21.jpg \n inflating: /data/test/Daniel_Radcliffe_50.jpg \n inflating: /data/test/Daniel_Radcliffe_51.jpg \n inflating: /data/test/Daniel_Rouse_00.jpg \n inflating: /data/test/Daniel_Rouse_01.jpg \n inflating: /data/test/Daniel_Rouse_10.jpg \n inflating: /data/test/Daniel_Rouse_11.jpg \n inflating: /data/test/Daniel_Rouse_20.jpg \n inflating: /data/test/Daniel_Rouse_21.jpg \n inflating: /data/test/Daniel_Rouse_30.jpg \n inflating: /data/test/Daniel_Rouse_31.jpg \n inflating: /data/test/Daniell_Sunjata_10.jpg \n inflating: /data/test/Daniell_Sunjata_11.jpg \n inflating: /data/test/Daniell_Sunjata_20.jpg \n inflating: /data/test/Daniell_Sunjata_21.jpg \n inflating: /data/test/Daniell_Sunjata_40.jpg \n inflating: /data/test/Daniell_Sunjata_41.jpg \n inflating: /data/test/Danny_Glover_10.jpg \n inflating: /data/test/Danny_Glover_11.jpg \n inflating: /data/test/Danny_Glover_30.jpg \n inflating: /data/test/Danny_Glover_31.jpg \n inflating: /data/test/Danny_Glover_50.jpg \n inflating: /data/test/Danny_Glover_51.jpg \n inflating: /data/test/Darrell_Issa_00.jpg \n inflating: /data/test/Darrell_Issa_01.jpg \n inflating: /data/test/Darrell_Issa_20.jpg \n inflating: /data/test/Darrell_Issa_21.jpg \n inflating: /data/test/Darrell_Issa_30.jpg \n inflating: /data/test/Darrell_Issa_31.jpg \n inflating: /data/test/Darrell_Issa_40.jpg \n inflating: /data/test/Darrell_Issa_41.jpg \n inflating: /data/test/Dave_Campo_10.jpg \n inflating: /data/test/Dave_Campo_11.jpg \n inflating: /data/test/Dave_Campo_20.jpg \n inflating: /data/test/Dave_Campo_21.jpg \n inflating: /data/test/Dave_Campo_30.jpg \n inflating: /data/test/Dave_Campo_31.jpg \n inflating: /data/test/David_Brent_00.jpg \n inflating: /data/test/David_Brent_01.jpg \n inflating: /data/test/David_Brent_10.jpg \n inflating: /data/test/David_Brent_11.jpg \n inflating: /data/test/David_Brent_20.jpg \n inflating: /data/test/David_Brent_21.jpg \n inflating: /data/test/David_Brent_30.jpg \n inflating: /data/test/David_Brent_31.jpg \n inflating: /data/test/David_Caruso_00.jpg \n inflating: /data/test/David_Caruso_01.jpg \n inflating: /data/test/David_Caruso_10.jpg \n inflating: /data/test/David_Caruso_11.jpg \n inflating: /data/test/David_Caruso_30.jpg \n inflating: /data/test/David_Caruso_31.jpg \n inflating: /data/test/David_Caruso_40.jpg \n inflating: /data/test/David_Caruso_41.jpg \n inflating: /data/test/Ed_Rendell_00.jpg \n inflating: /data/test/Ed_Rendell_01.jpg \n inflating: /data/test/Ed_Rendell_20.jpg \n inflating: /data/test/Ed_Rendell_21.jpg \n inflating: /data/test/Ed_Rendell_50.jpg \n inflating: /data/test/Ed_Rendell_51.jpg \n inflating: /data/test/Ed_Smart_10.jpg \n inflating: /data/test/Ed_Smart_11.jpg \n inflating: /data/test/Ed_Smart_30.jpg \n inflating: /data/test/Ed_Smart_31.jpg \n inflating: /data/test/Ed_Smart_50.jpg \n inflating: /data/test/Ed_Smart_51.jpg \n inflating: /data/test/Edie_Falco_20.jpg \n inflating: /data/test/Edie_Falco_21.jpg \n inflating: /data/test/Edie_Falco_30.jpg \n inflating: /data/test/Edie_Falco_31.jpg \n inflating: /data/test/Edie_Falco_40.jpg \n inflating: /data/test/Edie_Falco_41.jpg \n inflating: /data/test/Edie_Falco_50.jpg \n inflating: /data/test/Edie_Falco_51.jpg \n inflating: /data/test/Eduardo_Duhalde_00.jpg \n inflating: /data/test/Eduardo_Duhalde_01.jpg \n inflating: /data/test/Eduardo_Duhalde_10.jpg \n inflating: /data/test/Eduardo_Duhalde_11.jpg \n inflating: /data/test/Eduardo_Duhalde_30.jpg \n inflating: /data/test/Eduardo_Duhalde_31.jpg \n inflating: /data/test/Edward_Burns_10.jpg \n inflating: /data/test/Edward_Burns_11.jpg \n inflating: /data/test/Edward_Burns_20.jpg \n inflating: /data/test/Edward_Burns_21.jpg \n inflating: /data/test/Edward_Burns_30.jpg \n inflating: /data/test/Edward_Burns_31.jpg \n inflating: /data/test/Edward_Burns_50.jpg \n inflating: /data/test/Edward_Burns_51.jpg \n inflating: /data/test/Edward_Norton_10.jpg \n inflating: /data/test/Edward_Norton_11.jpg \n inflating: /data/test/Edward_Norton_30.jpg \n inflating: /data/test/Edward_Norton_31.jpg \n inflating: /data/test/Edward_Norton_40.jpg \n inflating: /data/test/Edward_Norton_41.jpg \n inflating: /data/test/Edward_Norton_50.jpg \n inflating: /data/test/Edward_Norton_51.jpg \n inflating: /data/test/Elaine_Chao_00.jpg \n inflating: /data/test/Elaine_Chao_01.jpg \n inflating: /data/test/Elaine_Chao_20.jpg \n inflating: /data/test/Elaine_Chao_21.jpg \n inflating: /data/test/Elaine_Chao_50.jpg \n inflating: /data/test/Elaine_Chao_51.jpg \n inflating: /data/test/Elaine_Stritch_10.jpg \n inflating: /data/test/Elaine_Stritch_11.jpg \n inflating: /data/test/Elaine_Stritch_40.jpg \n inflating: /data/test/Elaine_Stritch_41.jpg \n inflating: /data/test/Elaine_Stritch_50.jpg \n inflating: /data/test/Elaine_Stritch_51.jpg \n inflating: /data/test/Eliane_Karp_00.jpg \n inflating: /data/test/Eliane_Karp_01.jpg \n inflating: /data/test/Eliane_Karp_10.jpg \n inflating: /data/test/Eliane_Karp_11.jpg \n inflating: /data/test/Eliane_Karp_30.jpg \n inflating: /data/test/Eliane_Karp_31.jpg \n inflating: /data/test/Eliane_Karp_40.jpg \n inflating: /data/test/Eliane_Karp_41.jpg \n inflating: /data/test/Elijah_Wood_00.jpg \n inflating: /data/test/Elijah_Wood_01.jpg \n inflating: /data/test/Elijah_Wood_10.jpg \n inflating: /data/test/Elijah_Wood_11.jpg \n inflating: /data/test/Elijah_Wood_30.jpg \n inflating: /data/test/Elijah_Wood_31.jpg \n inflating: /data/test/Eliza_Dushku_00.jpg \n inflating: /data/test/Eliza_Dushku_01.jpg \n inflating: /data/test/Eliza_Dushku_10.jpg \n inflating: /data/test/Eliza_Dushku_11.jpg \n inflating: /data/test/Eliza_Dushku_20.jpg \n inflating: /data/test/Eliza_Dushku_21.jpg \n inflating: /data/test/Eliza_Dushku_30.jpg \n inflating: /data/test/Eliza_Dushku_31.jpg \n inflating: /data/test/Elizabeth_Dole_00.jpg \n inflating: /data/test/Elizabeth_Dole_01.jpg \n inflating: /data/test/Elizabeth_Dole_10.jpg \n inflating: /data/test/Elizabeth_Dole_11.jpg \n inflating: /data/test/Elizabeth_Dole_30.jpg \n inflating: /data/test/Elizabeth_Dole_31.jpg \n inflating: /data/test/Elizabeth_Shue_00.jpg \n inflating: /data/test/Elizabeth_Shue_01.jpg \n inflating: /data/test/Elizabeth_Shue_20.jpg \n inflating: /data/test/Elizabeth_Shue_21.jpg \n inflating: /data/test/Elizabeth_Shue_40.jpg \n inflating: /data/test/Elizabeth_Shue_41.jpg \n inflating: /data/test/Ellen_DeGeneres_10.jpg \n inflating: /data/test/Ellen_DeGeneres_11.jpg \n inflating: /data/test/Ellen_DeGeneres_40.jpg \n inflating: /data/test/Ellen_DeGeneres_41.jpg \n inflating: /data/test/Ellen_DeGeneres_50.jpg \n inflating: /data/test/Ellen_DeGeneres_51.jpg \n inflating: /data/test/Elmar_Brok_00.jpg \n inflating: /data/test/Elmar_Brok_01.jpg \n inflating: /data/test/Elmar_Brok_20.jpg \n inflating: /data/test/Elmar_Brok_21.jpg \n inflating: /data/test/Elmar_Brok_30.jpg \n inflating: /data/test/Elmar_Brok_31.jpg \n inflating: /data/test/Elsa_Zylberstein_00.jpg \n inflating: /data/test/Elsa_Zylberstein_01.jpg \n inflating: /data/test/Elsa_Zylberstein_10.jpg \n inflating: /data/test/Elsa_Zylberstein_11.jpg \n inflating: /data/test/Elsa_Zylberstein_40.jpg \n inflating: /data/test/Elsa_Zylberstein_41.jpg \n inflating: /data/test/Elton_John_10.jpg \n inflating: /data/test/Elton_John_11.jpg \n inflating: /data/test/Elton_John_20.jpg \n inflating: /data/test/Elton_John_21.jpg \n inflating: /data/test/Elton_John_30.jpg \n inflating: /data/test/Elton_John_31.jpg \n inflating: /data/test/Elton_John_40.jpg \n inflating: /data/test/Elton_John_41.jpg \n inflating: /data/test/Emile_Lahoud_00.jpg \n inflating: /data/test/Emile_Lahoud_01.jpg \n inflating: /data/test/Emile_Lahoud_30.jpg \n inflating: /data/test/Emile_Lahoud_31.jpg \n inflating: /data/test/Emile_Lahoud_40.jpg \n inflating: /data/test/Emile_Lahoud_41.jpg \n inflating: /data/test/Emilio_Botin_00.jpg \n inflating: /data/test/Emilio_Botin_01.jpg \n inflating: /data/test/Emilio_Botin_10.jpg \n inflating: /data/test/Emilio_Botin_11.jpg \n inflating: /data/test/Emilio_Botin_20.jpg \n inflating: /data/test/Emilio_Botin_21.jpg \n inflating: /data/test/Emilio_Botin_40.jpg \n inflating: /data/test/Emilio_Botin_41.jpg \n inflating: /data/test/Emma_Nicholson_10.jpg \n inflating: /data/test/Emma_Nicholson_11.jpg \n inflating: /data/test/Emma_Nicholson_20.jpg \n inflating: /data/test/Emma_Nicholson_21.jpg \n inflating: /data/test/Emma_Nicholson_30.jpg \n inflating: /data/test/Emma_Nicholson_31.jpg \n inflating: /data/test/Emma_Thompson_20.jpg \n inflating: /data/test/Emma_Thompson_21.jpg \n inflating: /data/test/Emma_Thompson_30.jpg \n inflating: /data/test/Emma_Thompson_31.jpg \n inflating: /data/test/Emma_Thompson_40.jpg \n inflating: /data/test/Emma_Thompson_41.jpg \n inflating: /data/test/Emma_Thompson_50.jpg \n inflating: /data/test/Emma_Thompson_51.jpg \n inflating: /data/test/Emmy_Rossum_20.jpg \n inflating: /data/test/Emmy_Rossum_21.jpg \n inflating: /data/test/Emmy_Rossum_30.jpg \n inflating: /data/test/Emmy_Rossum_31.jpg \n inflating: /data/test/Emmy_Rossum_40.jpg \n inflating: /data/test/Emmy_Rossum_41.jpg \n inflating: /data/test/Emmy_Rossum_50.jpg \n inflating: /data/test/Emmy_Rossum_51.jpg \n inflating: /data/test/Eric_Benet_00.jpg \n inflating: /data/test/Eric_Benet_01.jpg \n inflating: /data/test/Eric_Benet_10.jpg \n inflating: /data/test/Eric_Benet_11.jpg \n inflating: /data/test/Eric_Benet_30.jpg \n inflating: /data/test/Eric_Benet_31.jpg \n inflating: /data/test/Erin_Hershey_Presley_10.jpg \n inflating: /data/test/Erin_Hershey_Presley_11.jpg \n inflating: /data/test/Erin_Hershey_Presley_30.jpg \n inflating: /data/test/Erin_Hershey_Presley_31.jpg \n inflating: /data/test/Erin_Hershey_Presley_40.jpg \n inflating: /data/test/Erin_Hershey_Presley_41.jpg \n inflating: /data/test/Ernest_Hollings_00.jpg \n inflating: /data/test/Ernest_Hollings_01.jpg \n inflating: /data/test/Ernest_Hollings_10.jpg \n inflating: /data/test/Ernest_Hollings_11.jpg \n inflating: /data/test/Ernest_Hollings_20.jpg \n inflating: /data/test/Ernest_Hollings_21.jpg \n inflating: /data/test/Ernesto_Zedillo_10.jpg \n inflating: /data/test/Ernesto_Zedillo_11.jpg \n inflating: /data/test/Ernesto_Zedillo_20.jpg \n inflating: /data/test/Ernesto_Zedillo_21.jpg \n inflating: /data/test/Ernesto_Zedillo_30.jpg \n inflating: /data/test/Ernesto_Zedillo_31.jpg \n inflating: /data/test/Ernesto_Zedillo_40.jpg \n inflating: /data/test/Ernesto_Zedillo_41.jpg \n inflating: /data/test/Ernie_Grunfeld_20.jpg \n inflating: /data/test/Ernie_Grunfeld_21.jpg \n inflating: /data/test/Ernie_Grunfeld_30.jpg \n inflating: /data/test/Ernie_Grunfeld_31.jpg \n inflating: /data/test/Ernie_Grunfeld_40.jpg \n inflating: /data/test/Ernie_Grunfeld_41.jpg \n inflating: /data/test/Ernie_Grunfeld_50.jpg \n inflating: /data/test/Ernie_Grunfeld_51.jpg \n inflating: /data/test/Estelle_Morris_10.jpg \n inflating: /data/test/Estelle_Morris_11.jpg \n inflating: /data/test/Estelle_Morris_20.jpg \n inflating: /data/test/Estelle_Morris_21.jpg \n inflating: /data/test/Estelle_Morris_30.jpg \n inflating: /data/test/Estelle_Morris_31.jpg \n inflating: /data/test/Ethan_Hawke_00.jpg \n inflating: /data/test/Ethan_Hawke_01.jpg \n inflating: /data/test/Ethan_Hawke_10.jpg \n inflating: /data/test/Ethan_Hawke_11.jpg \n inflating: /data/test/Ethan_Hawke_30.jpg \n inflating: /data/test/Ethan_Hawke_31.jpg \n inflating: /data/test/Ethan_Hawke_40.jpg \n inflating: /data/test/Ethan_Hawke_41.jpg \n inflating: /data/test/Eunice_Barber_00.jpg \n inflating: /data/test/Eunice_Barber_01.jpg \n inflating: /data/test/Eunice_Barber_10.jpg \n inflating: /data/test/Eunice_Barber_11.jpg \n inflating: /data/test/Eunice_Barber_50.jpg \n inflating: /data/test/Eunice_Barber_51.jpg \n inflating: /data/test/Fernando_Henrique_Cardoso_00.jpg \n inflating: /data/test/Fernando_Henrique_Cardoso_01.jpg \n inflating: /data/test/Fernando_Henrique_Cardoso_20.jpg \n inflating: /data/test/Fernando_Henrique_Cardoso_21.jpg \n inflating: /data/test/Fernando_Henrique_Cardoso_30.jpg \n inflating: /data/test/Fernando_Henrique_Cardoso_31.jpg \n inflating: /data/test/Fernando_Sanz_30.jpg \n inflating: /data/test/Fernando_Sanz_31.jpg \n inflating: /data/test/Fernando_Sanz_40.jpg \n inflating: /data/test/Fernando_Sanz_41.jpg \n inflating: /data/test/Fernando_Sanz_50.jpg \n inflating: /data/test/Fernando_Sanz_51.jpg \n inflating: /data/test/Fidel_Castro_Daiz-Balart_10.jpg \n inflating: /data/test/Fidel_Castro_Daiz-Balart_11.jpg \n inflating: /data/test/Fidel_Castro_Daiz-Balart_30.jpg \n inflating: /data/test/Fidel_Castro_Daiz-Balart_31.jpg \n inflating: /data/test/Fidel_Castro_Daiz-Balart_40.jpg \n inflating: /data/test/Fidel_Castro_Daiz-Balart_41.jpg \n inflating: /data/test/Flavia_Pennetta_30.jpg \n inflating: /data/test/Flavia_Pennetta_31.jpg \n inflating: /data/test/Flavia_Pennetta_40.jpg \n inflating: /data/test/Flavia_Pennetta_41.jpg \n inflating: /data/test/Flavia_Pennetta_50.jpg \n inflating: /data/test/Flavia_Pennetta_51.jpg \n inflating: /data/test/Florecita_Cobian_00.jpg \n inflating: /data/test/Florecita_Cobian_01.jpg \n inflating: /data/test/Florecita_Cobian_10.jpg \n inflating: /data/test/Florecita_Cobian_11.jpg \n inflating: /data/test/Florecita_Cobian_20.jpg \n inflating: /data/test/Florecita_Cobian_21.jpg \n inflating: /data/test/Frances_Fisher_20.jpg \n inflating: /data/test/Frances_Fisher_21.jpg \n inflating: /data/test/Frances_Fisher_30.jpg \n inflating: /data/test/Frances_Fisher_31.jpg \n inflating: /data/test/Frances_Fisher_40.jpg \n inflating: /data/test/Frances_Fisher_41.jpg \n inflating: /data/test/Francis_Collins_00.jpg \n inflating: /data/test/Francis_Collins_01.jpg \n inflating: /data/test/Francis_Collins_10.jpg \n inflating: /data/test/Francis_Collins_11.jpg \n inflating: /data/test/Francis_Collins_20.jpg \n inflating: /data/test/Francis_Collins_21.jpg \n inflating: /data/test/Francis_Collins_40.jpg \n inflating: /data/test/Francis_Collins_41.jpg \n inflating: /data/test/Frank_Beamer_00.jpg \n inflating: /data/test/Frank_Beamer_01.jpg \n inflating: /data/test/Frank_Beamer_20.jpg \n inflating: /data/test/Frank_Beamer_21.jpg \n inflating: /data/test/Frank_Beamer_30.jpg \n inflating: /data/test/Frank_Beamer_31.jpg \n inflating: /data/test/Frank_Caliendo_10.jpg \n inflating: /data/test/Frank_Caliendo_11.jpg \n inflating: /data/test/Frank_Caliendo_30.jpg \n inflating: /data/test/Frank_Caliendo_31.jpg \n inflating: /data/test/Frank_Caliendo_40.jpg \n inflating: /data/test/Frank_Caliendo_41.jpg \n inflating: /data/test/Frank_Caliendo_50.jpg \n inflating: /data/test/Frank_Caliendo_51.jpg \n inflating: /data/test/Frank_Keating_30.jpg \n inflating: /data/test/Frank_Keating_31.jpg \n inflating: /data/test/Frank_Keating_40.jpg \n inflating: /data/test/Frank_Keating_41.jpg \n inflating: /data/test/Frank_Keating_50.jpg \n inflating: /data/test/Frank_Keating_51.jpg \n inflating: /data/test/Frank_Solich_10.jpg \n inflating: /data/test/Frank_Solich_11.jpg \n inflating: /data/test/Frank_Solich_20.jpg \n inflating: /data/test/Frank_Solich_21.jpg \n inflating: /data/test/Frank_Solich_30.jpg \n inflating: /data/test/Frank_Solich_31.jpg \n inflating: /data/test/Franz_Fischler_00.jpg \n inflating: /data/test/Franz_Fischler_01.jpg \n inflating: /data/test/Franz_Fischler_30.jpg \n inflating: /data/test/Franz_Fischler_31.jpg \n inflating: /data/test/Franz_Fischler_40.jpg \n inflating: /data/test/Franz_Fischler_41.jpg \n inflating: /data/test/Franz_Fischler_50.jpg \n inflating: /data/test/Franz_Fischler_51.jpg \n inflating: /data/test/Gabi_Zimmer_00.jpg \n inflating: /data/test/Gabi_Zimmer_01.jpg \n inflating: /data/test/Gabi_Zimmer_10.jpg \n inflating: /data/test/Gabi_Zimmer_11.jpg \n inflating: /data/test/Gabi_Zimmer_20.jpg \n inflating: /data/test/Gabi_Zimmer_21.jpg \n inflating: /data/test/Gabi_Zimmer_50.jpg \n inflating: /data/test/Gabi_Zimmer_51.jpg \n inflating: /data/test/Gary_Bettman_10.jpg \n inflating: /data/test/Gary_Bettman_11.jpg \n inflating: /data/test/Gary_Bettman_30.jpg \n inflating: /data/test/Gary_Bettman_31.jpg \n inflating: /data/test/Gary_Bettman_40.jpg \n inflating: /data/test/Gary_Bettman_41.jpg \n inflating: /data/test/Gary_Coleman_30.jpg \n inflating: /data/test/Gary_Coleman_31.jpg \n inflating: /data/test/Gary_Coleman_40.jpg \n inflating: /data/test/Gary_Coleman_41.jpg \n inflating: /data/test/Gary_Coleman_50.jpg \n inflating: /data/test/Gary_Coleman_51.jpg \n inflating: /data/test/Gary_Condit_00.jpg \n inflating: /data/test/Gary_Condit_01.jpg \n inflating: /data/test/Gary_Condit_10.jpg \n inflating: /data/test/Gary_Condit_11.jpg \n inflating: /data/test/Gary_Condit_30.jpg \n inflating: /data/test/Gary_Condit_31.jpg \n inflating: /data/test/Gene_Hackman_20.jpg \n inflating: /data/test/Gene_Hackman_21.jpg \n inflating: /data/test/Gene_Hackman_30.jpg \n inflating: /data/test/Gene_Hackman_31.jpg \n inflating: /data/test/Gene_Hackman_40.jpg \n inflating: /data/test/Gene_Hackman_41.jpg \n inflating: /data/test/Geoffrey_Rush_00.jpg \n inflating: /data/test/Geoffrey_Rush_01.jpg \n inflating: /data/test/Geoffrey_Rush_10.jpg \n inflating: /data/test/Geoffrey_Rush_11.jpg \n inflating: /data/test/Geoffrey_Rush_20.jpg \n inflating: /data/test/Geoffrey_Rush_21.jpg \n inflating: /data/test/George_Galloway_00.jpg \n inflating: /data/test/George_Galloway_01.jpg \n inflating: /data/test/George_Galloway_20.jpg \n inflating: /data/test/George_Galloway_21.jpg \n inflating: /data/test/George_Galloway_40.jpg \n inflating: /data/test/George_Galloway_41.jpg \n inflating: /data/test/George_Galloway_50.jpg \n inflating: /data/test/George_Galloway_51.jpg \n inflating: /data/test/George_Karl_10.jpg \n inflating: /data/test/George_Karl_11.jpg \n inflating: /data/test/George_Karl_20.jpg \n inflating: /data/test/George_Karl_21.jpg \n inflating: /data/test/George_Karl_50.jpg \n inflating: /data/test/George_Karl_51.jpg \n inflating: /data/test/GL_Peiris_00.jpg \n inflating: /data/test/GL_Peiris_01.jpg \n inflating: /data/test/GL_Peiris_10.jpg \n inflating: /data/test/GL_Peiris_11.jpg \n inflating: /data/test/GL_Peiris_30.jpg \n inflating: /data/test/GL_Peiris_31.jpg \n inflating: /data/test/Hanan_Ashrawi_10.jpg \n inflating: /data/test/Hanan_Ashrawi_11.jpg \n inflating: /data/test/Hanan_Ashrawi_20.jpg \n inflating: /data/test/Hanan_Ashrawi_21.jpg \n inflating: /data/test/Hanan_Ashrawi_40.jpg \n inflating: /data/test/Hanan_Ashrawi_41.jpg \n inflating: /data/test/Harrison_Ford_10.jpg \n inflating: /data/test/Harrison_Ford_11.jpg \n inflating: /data/test/Harrison_Ford_20.jpg \n inflating: /data/test/Harrison_Ford_21.jpg \n inflating: /data/test/Harrison_Ford_50.jpg \n inflating: /data/test/Harrison_Ford_51.jpg \n inflating: /data/test/Hassan_Nasrallah_30.jpg \n inflating: /data/test/Hassan_Nasrallah_31.jpg \n inflating: /data/test/Hassan_Nasrallah_40.jpg \n inflating: /data/test/Hassan_Nasrallah_41.jpg \n inflating: /data/test/Hassan_Nasrallah_50.jpg \n inflating: /data/test/Hassan_Nasrallah_51.jpg \n inflating: /data/test/Irene_Kahn_00.jpg \n inflating: /data/test/Irene_Kahn_01.jpg \n inflating: /data/test/Irene_Kahn_30.jpg \n inflating: /data/test/Irene_Kahn_31.jpg \n inflating: /data/test/Irene_Kahn_40.jpg \n inflating: /data/test/Irene_Kahn_41.jpg \n inflating: /data/test/Isabella_Rossellini_00.jpg \n inflating: /data/test/Isabella_Rossellini_01.jpg \n inflating: /data/test/Isabella_Rossellini_10.jpg \n inflating: /data/test/Isabella_Rossellini_11.jpg \n inflating: /data/test/Isabella_Rossellini_20.jpg \n inflating: /data/test/Isabella_Rossellini_21.jpg \n inflating: /data/test/Isabelle_Huppert_20.jpg \n inflating: /data/test/Isabelle_Huppert_21.jpg \n inflating: /data/test/Isabelle_Huppert_30.jpg \n inflating: /data/test/Isabelle_Huppert_31.jpg \n inflating: /data/test/Isabelle_Huppert_40.jpg \n inflating: /data/test/Isabelle_Huppert_41.jpg \n inflating: /data/test/Itzhak_Perlman_10.jpg \n inflating: /data/test/Itzhak_Perlman_11.jpg \n inflating: /data/test/Itzhak_Perlman_30.jpg \n inflating: /data/test/Itzhak_Perlman_31.jpg \n inflating: /data/test/Itzhak_Perlman_40.jpg \n inflating: /data/test/Itzhak_Perlman_41.jpg \n inflating: /data/test/Jack_Welch_10.jpg \n inflating: /data/test/Jack_Welch_11.jpg \n inflating: /data/test/Jack_Welch_30.jpg \n inflating: /data/test/Jack_Welch_31.jpg \n inflating: /data/test/Jack_Welch_40.jpg \n inflating: /data/test/Jack_Welch_41.jpg \n inflating: /data/test/Jack_Welch_50.jpg \n inflating: /data/test/Jack_Welch_51.jpg \n inflating: /data/test/Jackie_Sherrill_20.jpg \n inflating: /data/test/Jackie_Sherrill_21.jpg \n inflating: /data/test/Jackie_Sherrill_40.jpg \n inflating: /data/test/Jackie_Sherrill_41.jpg \n inflating: /data/test/Jackie_Sherrill_50.jpg \n inflating: /data/test/Jackie_Sherrill_51.jpg \n inflating: /data/test/Jacqueline_Gold_00.jpg \n inflating: /data/test/Jacqueline_Gold_01.jpg \n inflating: /data/test/Jacqueline_Gold_20.jpg \n inflating: /data/test/Jacqueline_Gold_21.jpg \n inflating: /data/test/Jacqueline_Gold_30.jpg \n inflating: /data/test/Jacqueline_Gold_31.jpg \n inflating: /data/test/Jafar_Umar_Thalib_00.jpg \n inflating: /data/test/Jafar_Umar_Thalib_01.jpg \n inflating: /data/test/Jafar_Umar_Thalib_20.jpg \n inflating: /data/test/Jafar_Umar_Thalib_21.jpg \n inflating: /data/test/Jafar_Umar_Thalib_30.jpg \n inflating: /data/test/Jafar_Umar_Thalib_31.jpg \n inflating: /data/test/Jafar_Umar_Thalib_50.jpg \n inflating: /data/test/Jafar_Umar_Thalib_51.jpg \n inflating: /data/test/Jaime_Pressly_00.jpg \n inflating: /data/test/Jaime_Pressly_01.jpg \n inflating: /data/test/Jaime_Pressly_10.jpg \n inflating: /data/test/Jaime_Pressly_11.jpg \n inflating: /data/test/Jaime_Pressly_40.jpg \n inflating: /data/test/Jaime_Pressly_41.jpg \n inflating: /data/test/Jake_Gyllenhaal_00.jpg \n inflating: /data/test/Jake_Gyllenhaal_01.jpg \n inflating: /data/test/Jake_Gyllenhaal_40.jpg \n inflating: /data/test/Jake_Gyllenhaal_41.jpg \n inflating: /data/test/Jake_Gyllenhaal_50.jpg \n inflating: /data/test/Jake_Gyllenhaal_51.jpg \n inflating: /data/test/Jake_Plummer_20.jpg \n inflating: /data/test/Jake_Plummer_21.jpg \n inflating: /data/test/Jake_Plummer_40.jpg \n inflating: /data/test/Jake_Plummer_41.jpg \n inflating: /data/test/Jake_Plummer_50.jpg \n inflating: /data/test/Jake_Plummer_51.jpg \n inflating: /data/test/James_Carville_00.jpg \n inflating: /data/test/James_Carville_01.jpg \n inflating: /data/test/James_Carville_10.jpg \n inflating: /data/test/James_Carville_11.jpg \n inflating: /data/test/James_Carville_30.jpg \n inflating: /data/test/James_Carville_31.jpg \n inflating: /data/test/James_Carville_50.jpg \n inflating: /data/test/James_Carville_51.jpg \n inflating: /data/test/James_Cunningham_00.jpg \n inflating: /data/test/James_Cunningham_01.jpg \n inflating: /data/test/James_Cunningham_20.jpg \n inflating: /data/test/James_Cunningham_21.jpg \n inflating: /data/test/James_Cunningham_30.jpg \n inflating: /data/test/James_Cunningham_31.jpg \n inflating: /data/test/James_Cunningham_40.jpg \n inflating: /data/test/James_Cunningham_41.jpg \n inflating: /data/test/James_Hoffa_10.jpg \n inflating: /data/test/James_Hoffa_11.jpg \n inflating: /data/test/James_Hoffa_20.jpg \n inflating: /data/test/James_Hoffa_21.jpg \n inflating: /data/test/James_Hoffa_40.jpg \n inflating: /data/test/James_Hoffa_41.jpg \n inflating: /data/test/James_Hoffa_50.jpg \n inflating: /data/test/James_Hoffa_51.jpg \n inflating: /data/test/James_Lockhart_00.jpg \n inflating: /data/test/James_Lockhart_01.jpg \n inflating: /data/test/James_Lockhart_10.jpg \n inflating: /data/test/James_Lockhart_11.jpg \n inflating: /data/test/James_Lockhart_50.jpg \n inflating: /data/test/James_Lockhart_51.jpg \n inflating: /data/test/James_McPherson_00.jpg \n inflating: /data/test/James_McPherson_01.jpg \n inflating: /data/test/James_McPherson_10.jpg \n inflating: /data/test/James_McPherson_11.jpg \n inflating: /data/test/James_McPherson_20.jpg \n inflating: /data/test/James_McPherson_21.jpg \n inflating: /data/test/James_Wolfensohn_00.jpg \n inflating: /data/test/James_Wolfensohn_01.jpg \n inflating: /data/test/James_Wolfensohn_20.jpg \n inflating: /data/test/James_Wolfensohn_21.jpg \n inflating: /data/test/James_Wolfensohn_30.jpg \n inflating: /data/test/James_Wolfensohn_31.jpg \n inflating: /data/test/James_Wolfensohn_50.jpg \n inflating: /data/test/James_Wolfensohn_51.jpg \n inflating: /data/test/Jan_Peter_Balkenende_00.jpg \n inflating: /data/test/Jan_Peter_Balkenende_01.jpg \n inflating: /data/test/Jan_Peter_Balkenende_10.jpg \n inflating: /data/test/Jan_Peter_Balkenende_11.jpg \n inflating: /data/test/Jan_Peter_Balkenende_30.jpg \n inflating: /data/test/Jan_Peter_Balkenende_31.jpg \n inflating: /data/test/Jan_Peter_Balkenende_50.jpg \n inflating: /data/test/Jan_Peter_Balkenende_51.jpg \n inflating: /data/test/Jane_Krakowski_00.jpg \n inflating: /data/test/Jane_Krakowski_01.jpg \n inflating: /data/test/Jane_Krakowski_10.jpg \n inflating: /data/test/Jane_Krakowski_11.jpg \n inflating: /data/test/Jane_Krakowski_40.jpg \n inflating: /data/test/Jane_Krakowski_41.jpg \n inflating: /data/test/Jane_Krakowski_50.jpg \n inflating: /data/test/Jane_Krakowski_51.jpg \n inflating: /data/test/Jane_Pauley_10.jpg \n inflating: /data/test/Jane_Pauley_11.jpg \n inflating: /data/test/Jane_Pauley_30.jpg \n inflating: /data/test/Jane_Pauley_31.jpg \n inflating: /data/test/Jane_Pauley_40.jpg \n inflating: /data/test/Jane_Pauley_41.jpg \n inflating: /data/test/Jane_Rooney_00.jpg \n inflating: /data/test/Jane_Rooney_01.jpg \n inflating: /data/test/Jane_Rooney_10.jpg \n inflating: /data/test/Jane_Rooney_11.jpg \n inflating: /data/test/Jane_Rooney_20.jpg \n inflating: /data/test/Jane_Rooney_21.jpg \n inflating: /data/test/Janis_Ruth_Coulter_00.jpg \n inflating: /data/test/Janis_Ruth_Coulter_01.jpg \n inflating: /data/test/Janis_Ruth_Coulter_20.jpg \n inflating: /data/test/Janis_Ruth_Coulter_21.jpg \n inflating: /data/test/Janis_Ruth_Coulter_40.jpg \n inflating: /data/test/Janis_Ruth_Coulter_41.jpg \n inflating: /data/test/Janis_Ruth_Coulter_50.jpg \n inflating: /data/test/Janis_Ruth_Coulter_51.jpg \n inflating: /data/test/JK_Rowling_20.jpg \n inflating: /data/test/JK_Rowling_21.jpg \n inflating: /data/test/JK_Rowling_30.jpg \n inflating: /data/test/JK_Rowling_31.jpg \n inflating: /data/test/JK_Rowling_40.jpg \n inflating: /data/test/JK_Rowling_41.jpg \n inflating: /data/test/JK_Rowling_50.jpg \n inflating: /data/test/JK_Rowling_51.jpg \n inflating: /data/test/Kate_Capshaw_10.jpg \n inflating: /data/test/Kate_Capshaw_11.jpg \n inflating: /data/test/Kate_Capshaw_20.jpg \n inflating: /data/test/Kate_Capshaw_21.jpg \n inflating: /data/test/Kate_Capshaw_40.jpg \n inflating: /data/test/Kate_Capshaw_41.jpg \n inflating: /data/test/Kate_Winslet_00.jpg \n inflating: /data/test/Kate_Winslet_01.jpg \n inflating: /data/test/Kate_Winslet_10.jpg \n inflating: /data/test/Kate_Winslet_11.jpg \n inflating: /data/test/Kate_Winslet_50.jpg \n inflating: /data/test/Kate_Winslet_51.jpg \n inflating: /data/test/Katharine_Hepburn_10.jpg \n inflating: /data/test/Katharine_Hepburn_11.jpg \n inflating: /data/test/Katharine_Hepburn_30.jpg \n inflating: /data/test/Katharine_Hepburn_31.jpg \n inflating: /data/test/Katharine_Hepburn_40.jpg \n inflating: /data/test/Katharine_Hepburn_41.jpg \n inflating: /data/test/Kathryn_Morris_10.jpg \n inflating: /data/test/Kathryn_Morris_11.jpg \n inflating: /data/test/Kathryn_Morris_20.jpg \n inflating: /data/test/Kathryn_Morris_21.jpg \n inflating: /data/test/Kathryn_Morris_40.jpg \n inflating: /data/test/Kathryn_Morris_41.jpg \n inflating: /data/test/Kathryn_Morris_50.jpg \n inflating: /data/test/Kathryn_Morris_51.jpg \n inflating: /data/test/Katja_Riemann_00.jpg \n inflating: /data/test/Katja_Riemann_01.jpg \n inflating: /data/test/Katja_Riemann_10.jpg \n inflating: /data/test/Katja_Riemann_11.jpg \n inflating: /data/test/Katja_Riemann_20.jpg \n inflating: /data/test/Katja_Riemann_21.jpg \n inflating: /data/test/Keith_Olbermann_00.jpg \n inflating: /data/test/Keith_Olbermann_01.jpg \n inflating: /data/test/Keith_Olbermann_10.jpg \n inflating: /data/test/Keith_Olbermann_11.jpg \n inflating: /data/test/Keith_Olbermann_20.jpg \n inflating: /data/test/Keith_Olbermann_21.jpg \n inflating: /data/test/Keith_Olbermann_50.jpg \n inflating: /data/test/Keith_Olbermann_51.jpg \n inflating: /data/test/Keith_Tyson_00.jpg \n inflating: /data/test/Keith_Tyson_01.jpg \n inflating: /data/test/Keith_Tyson_10.jpg \n inflating: /data/test/Keith_Tyson_11.jpg \n inflating: /data/test/Keith_Tyson_20.jpg \n inflating: /data/test/Keith_Tyson_21.jpg \n inflating: /data/test/Kemal_Dervis_00.jpg \n inflating: /data/test/Kemal_Dervis_01.jpg \n inflating: /data/test/Kemal_Dervis_10.jpg \n inflating: /data/test/Kemal_Dervis_11.jpg \n inflating: /data/test/Kemal_Dervis_30.jpg \n inflating: /data/test/Kemal_Dervis_31.jpg \n inflating: /data/test/Kevin_Satterfield_00.jpg \n inflating: /data/test/Kevin_Satterfield_01.jpg \n inflating: /data/test/Kevin_Satterfield_10.jpg \n inflating: /data/test/Kevin_Satterfield_11.jpg \n inflating: /data/test/Kevin_Satterfield_20.jpg \n inflating: /data/test/Kevin_Satterfield_21.jpg \n inflating: /data/test/Kieran_Culkin_00.jpg \n inflating: /data/test/Kieran_Culkin_01.jpg \n inflating: /data/test/Kieran_Culkin_10.jpg \n inflating: /data/test/Kieran_Culkin_11.jpg \n inflating: /data/test/Kieran_Culkin_20.jpg \n inflating: /data/test/Kieran_Culkin_21.jpg \n inflating: /data/test/Kirk_Ferentz_00.jpg \n inflating: /data/test/Kirk_Ferentz_01.jpg \n inflating: /data/test/Kirk_Ferentz_20.jpg \n inflating: /data/test/Kirk_Ferentz_21.jpg \n inflating: /data/test/Kirk_Ferentz_40.jpg \n inflating: /data/test/Kirk_Ferentz_41.jpg \n inflating: /data/test/Kirk_Ferentz_50.jpg \n inflating: /data/test/Kirk_Ferentz_51.jpg \n inflating: /data/test/Kirsten_Dunst_00.jpg \n inflating: /data/test/Kirsten_Dunst_01.jpg \n inflating: /data/test/Kirsten_Dunst_20.jpg \n inflating: /data/test/Kirsten_Dunst_21.jpg \n inflating: /data/test/Kirsten_Dunst_30.jpg \n inflating: /data/test/Kirsten_Dunst_31.jpg \n inflating: /data/test/Kit_Bond_10.jpg \n inflating: /data/test/Kit_Bond_11.jpg \n inflating: /data/test/Kit_Bond_20.jpg \n inflating: /data/test/Kit_Bond_21.jpg \n inflating: /data/test/Kit_Bond_30.jpg \n inflating: /data/test/Kit_Bond_31.jpg \n inflating: /data/test/Kit_Bond_50.jpg \n inflating: /data/test/Kit_Bond_51.jpg \n inflating: /data/test/Kristen_Breitweiser_00.jpg \n inflating: /data/test/Kristen_Breitweiser_01.jpg \n inflating: /data/test/Kristen_Breitweiser_10.jpg \n inflating: /data/test/Kristen_Breitweiser_11.jpg \n inflating: /data/test/Kristen_Breitweiser_20.jpg \n inflating: /data/test/Kristen_Breitweiser_21.jpg \n inflating: /data/test/Kristen_Breitweiser_50.jpg \n inflating: /data/test/Kristen_Breitweiser_51.jpg \n inflating: /data/test/Kristin_Chenoweth_10.jpg \n inflating: /data/test/Kristin_Chenoweth_11.jpg \n inflating: /data/test/Kristin_Chenoweth_40.jpg \n inflating: /data/test/Kristin_Chenoweth_41.jpg \n inflating: /data/test/Kristin_Chenoweth_50.jpg \n inflating: /data/test/Kristin_Chenoweth_51.jpg \n inflating: /data/test/Kristin_Scott_10.jpg \n inflating: /data/test/Kristin_Scott_11.jpg \n inflating: /data/test/Kristin_Scott_40.jpg \n inflating: /data/test/Kristin_Scott_41.jpg \n inflating: /data/test/Kristin_Scott_50.jpg \n inflating: /data/test/Kristin_Scott_51.jpg \n inflating: /data/test/Kristy_Curry_00.jpg \n inflating: /data/test/Kristy_Curry_01.jpg \n inflating: /data/test/Kristy_Curry_20.jpg \n inflating: /data/test/Kristy_Curry_21.jpg \n inflating: /data/test/Kristy_Curry_30.jpg \n inflating: /data/test/Kristy_Curry_31.jpg \n inflating: /data/test/Kurt_Warner_00.jpg \n inflating: /data/test/Kurt_Warner_01.jpg \n inflating: /data/test/Kurt_Warner_10.jpg \n inflating: /data/test/Kurt_Warner_11.jpg \n inflating: /data/test/Kurt_Warner_40.jpg \n inflating: /data/test/Kurt_Warner_41.jpg \n inflating: /data/test/Kweisi_Mfume_00.jpg \n inflating: /data/test/Kweisi_Mfume_01.jpg \n inflating: /data/test/Kweisi_Mfume_10.jpg \n inflating: /data/test/Kweisi_Mfume_11.jpg \n inflating: /data/test/Kweisi_Mfume_40.jpg \n inflating: /data/test/Kweisi_Mfume_41.jpg \n inflating: /data/test/Kweisi_Mfume_50.jpg \n inflating: /data/test/Kweisi_Mfume_51.jpg \n inflating: /data/test/Kyle_Shewfelt_00.jpg \n inflating: /data/test/Kyle_Shewfelt_01.jpg \n inflating: /data/test/Kyle_Shewfelt_10.jpg \n inflating: /data/test/Kyle_Shewfelt_11.jpg \n inflating: /data/test/Kyle_Shewfelt_20.jpg \n inflating: /data/test/Kyle_Shewfelt_21.jpg \n inflating: /data/test/Kyle_Shewfelt_40.jpg \n inflating: /data/test/Kyle_Shewfelt_41.jpg \n inflating: /data/test/Larry_Flynt_00.jpg \n inflating: /data/test/Larry_Flynt_01.jpg \n inflating: /data/test/Larry_Flynt_10.jpg \n inflating: /data/test/Larry_Flynt_11.jpg \n inflating: /data/test/Larry_Flynt_20.jpg \n inflating: /data/test/Larry_Flynt_21.jpg \n inflating: /data/test/Laura_Bozzo_00.jpg \n inflating: /data/test/Laura_Bozzo_01.jpg \n inflating: /data/test/Laura_Bozzo_10.jpg \n inflating: /data/test/Laura_Bozzo_11.jpg \n inflating: /data/test/Laura_Bozzo_40.jpg \n inflating: /data/test/Laura_Bozzo_41.jpg \n inflating: /data/test/Laura_Bush_10.jpg \n inflating: /data/test/Laura_Bush_11.jpg \n inflating: /data/test/Laura_Bush_20.jpg \n inflating: /data/test/Laura_Bush_21.jpg \n inflating: /data/test/Laura_Bush_40.jpg \n inflating: /data/test/Laura_Bush_41.jpg \n inflating: /data/test/Laura_Bush_50.jpg \n inflating: /data/test/Laura_Bush_51.jpg \n inflating: /data/test/Laura_Elena_Harring_00.jpg \n inflating: /data/test/Laura_Elena_Harring_01.jpg \n inflating: /data/test/Laura_Elena_Harring_20.jpg \n inflating: /data/test/Laura_Elena_Harring_21.jpg \n inflating: /data/test/Laura_Elena_Harring_40.jpg \n inflating: /data/test/Laura_Elena_Harring_41.jpg \n inflating: /data/test/Laura_Elena_Harring_50.jpg \n inflating: /data/test/Laura_Elena_Harring_51.jpg \n inflating: /data/test/Laurence_Fishburne_20.jpg \n inflating: /data/test/Laurence_Fishburne_21.jpg \n inflating: /data/test/Laurence_Fishburne_40.jpg \n inflating: /data/test/Laurence_Fishburne_41.jpg \n inflating: /data/test/Laurence_Fishburne_50.jpg \n inflating: /data/test/Laurence_Fishburne_51.jpg \n inflating: /data/test/Lee_Baca_00.jpg \n inflating: /data/test/Lee_Baca_01.jpg \n inflating: /data/test/Lee_Baca_10.jpg \n inflating: /data/test/Lee_Baca_11.jpg \n inflating: /data/test/Lee_Baca_40.jpg \n inflating: /data/test/Lee_Baca_41.jpg \n inflating: /data/test/Lene_Espersen_10.jpg \n inflating: /data/test/Lene_Espersen_11.jpg \n inflating: /data/test/Lene_Espersen_20.jpg \n inflating: /data/test/Lene_Espersen_21.jpg \n inflating: /data/test/Lene_Espersen_40.jpg \n inflating: /data/test/Lene_Espersen_41.jpg \n inflating: /data/test/Lesia_Burlak_00.jpg \n inflating: /data/test/Lesia_Burlak_01.jpg \n inflating: /data/test/Lesia_Burlak_20.jpg \n inflating: /data/test/Lesia_Burlak_21.jpg \n inflating: /data/test/Lesia_Burlak_30.jpg \n inflating: /data/test/Lesia_Burlak_31.jpg \n inflating: /data/test/Lester_Holt_00.jpg \n inflating: /data/test/Lester_Holt_01.jpg \n inflating: /data/test/Lester_Holt_30.jpg \n inflating: /data/test/Lester_Holt_31.jpg \n inflating: /data/test/Lester_Holt_40.jpg \n inflating: /data/test/Lester_Holt_41.jpg \n inflating: /data/test/Leszek_Miller_00.jpg \n inflating: /data/test/Leszek_Miller_01.jpg \n inflating: /data/test/Leszek_Miller_10.jpg \n inflating: /data/test/Leszek_Miller_11.jpg \n inflating: /data/test/Leszek_Miller_30.jpg \n inflating: /data/test/Leszek_Miller_31.jpg \n inflating: /data/test/Leticia_Van_de_Putte_00.jpg \n inflating: /data/test/Leticia_Van_de_Putte_01.jpg \n inflating: /data/test/Leticia_Van_de_Putte_10.jpg \n inflating: /data/test/Leticia_Van_de_Putte_11.jpg \n inflating: /data/test/Leticia_Van_de_Putte_40.jpg \n inflating: /data/test/Leticia_Van_de_Putte_41.jpg \n inflating: /data/test/Leuris_Pupo_00.jpg \n inflating: /data/test/Leuris_Pupo_01.jpg \n inflating: /data/test/Leuris_Pupo_20.jpg \n inflating: /data/test/Leuris_Pupo_21.jpg \n inflating: /data/test/Leuris_Pupo_30.jpg \n inflating: /data/test/Leuris_Pupo_31.jpg \n inflating: /data/test/Leuris_Pupo_40.jpg \n inflating: /data/test/Leuris_Pupo_41.jpg \n inflating: /data/test/Li_Zhaoxing_00.jpg \n inflating: /data/test/Li_Zhaoxing_01.jpg \n inflating: /data/test/Li_Zhaoxing_30.jpg \n inflating: /data/test/Li_Zhaoxing_31.jpg \n inflating: /data/test/Li_Zhaoxing_40.jpg \n inflating: /data/test/Li_Zhaoxing_41.jpg \n inflating: /data/test/Lincoln_Chafee_20.jpg \n inflating: /data/test/Lincoln_Chafee_21.jpg \n inflating: /data/test/Lincoln_Chafee_30.jpg \n inflating: /data/test/Lincoln_Chafee_31.jpg \n inflating: /data/test/Lincoln_Chafee_50.jpg \n inflating: /data/test/Lincoln_Chafee_51.jpg \n inflating: /data/test/Linda_Dano_00.jpg \n inflating: /data/test/Linda_Dano_01.jpg \n inflating: /data/test/Linda_Dano_20.jpg \n inflating: /data/test/Linda_Dano_21.jpg \n inflating: /data/test/Linda_Dano_30.jpg \n inflating: /data/test/Linda_Dano_31.jpg \n inflating: /data/test/Linda_Dano_50.jpg \n inflating: /data/test/Linda_Dano_51.jpg \n inflating: /data/test/Linda_Franklin_00.jpg \n inflating: /data/test/Linda_Franklin_01.jpg \n inflating: /data/test/Linda_Franklin_10.jpg \n inflating: /data/test/Linda_Franklin_11.jpg \n inflating: /data/test/Linda_Franklin_20.jpg \n inflating: /data/test/Linda_Franklin_21.jpg \n inflating: /data/test/Linda_Franklin_40.jpg \n inflating: /data/test/Linda_Franklin_41.jpg \n inflating: /data/test/Linda_Sanchez_00.jpg \n inflating: /data/test/Linda_Sanchez_01.jpg \n inflating: /data/test/Linda_Sanchez_10.jpg \n inflating: /data/test/Linda_Sanchez_11.jpg \n inflating: /data/test/Linda_Sanchez_20.jpg \n inflating: /data/test/Linda_Sanchez_21.jpg \n inflating: /data/test/Linda_Sanchez_40.jpg \n inflating: /data/test/Linda_Sanchez_41.jpg \n inflating: /data/test/Lindsey_Graham_00.jpg \n inflating: /data/test/Lindsey_Graham_01.jpg \n inflating: /data/test/Lindsey_Graham_10.jpg \n inflating: /data/test/Lindsey_Graham_11.jpg \n inflating: /data/test/Lindsey_Graham_20.jpg \n inflating: /data/test/Lindsey_Graham_21.jpg \n inflating: /data/test/Lindsey_Graham_30.jpg \n inflating: /data/test/Lindsey_Graham_31.jpg \n inflating: /data/test/Lino_Oviedo_00.jpg \n inflating: /data/test/Lino_Oviedo_01.jpg \n inflating: /data/test/Lino_Oviedo_30.jpg \n inflating: /data/test/Lino_Oviedo_31.jpg \n inflating: /data/test/Lino_Oviedo_50.jpg \n inflating: /data/test/Lino_Oviedo_51.jpg \n inflating: /data/test/Lisa_Ling_00.jpg \n inflating: /data/test/Lisa_Ling_01.jpg \n inflating: /data/test/Lisa_Ling_10.jpg \n inflating: /data/test/Lisa_Ling_11.jpg \n inflating: /data/test/Lisa_Ling_20.jpg \n inflating: /data/test/Lisa_Ling_21.jpg \n inflating: /data/test/Liu_Ye_00.jpg \n inflating: /data/test/Liu_Ye_01.jpg \n inflating: /data/test/Liu_Ye_10.jpg \n inflating: /data/test/Liu_Ye_11.jpg \n inflating: /data/test/Liu_Ye_20.jpg \n inflating: /data/test/Liu_Ye_21.jpg \n inflating: /data/test/Liu_Ye_50.jpg \n inflating: /data/test/Liu_Ye_51.jpg \n inflating: /data/test/Loretta_Lynn_Harper_00.jpg \n inflating: /data/test/Loretta_Lynn_Harper_01.jpg \n inflating: /data/test/Loretta_Lynn_Harper_30.jpg \n inflating: /data/test/Loretta_Lynn_Harper_31.jpg \n inflating: /data/test/Loretta_Lynn_Harper_40.jpg \n inflating: /data/test/Loretta_Lynn_Harper_41.jpg \n inflating: /data/test/Loretta_Lynn_Harper_50.jpg \n inflating: /data/test/Loretta_Lynn_Harper_51.jpg \n inflating: /data/test/Louis_Van_Gaal_00.jpg \n inflating: /data/test/Louis_Van_Gaal_01.jpg \n inflating: /data/test/Louis_Van_Gaal_10.jpg \n inflating: /data/test/Louis_Van_Gaal_11.jpg \n inflating: /data/test/Louis_Van_Gaal_40.jpg \n inflating: /data/test/Louis_Van_Gaal_41.jpg \n inflating: /data/test/Louisa_Baileche_00.jpg \n inflating: /data/test/Louisa_Baileche_01.jpg \n inflating: /data/test/Louisa_Baileche_10.jpg \n inflating: /data/test/Louisa_Baileche_11.jpg \n inflating: /data/test/Louisa_Baileche_20.jpg \n inflating: /data/test/Louisa_Baileche_21.jpg \n inflating: /data/test/Luc_Montagnier_20.jpg \n inflating: /data/test/Luc_Montagnier_21.jpg \n inflating: /data/test/Luc_Montagnier_40.jpg \n inflating: /data/test/Luc_Montagnier_41.jpg \n inflating: /data/test/Luc_Montagnier_50.jpg \n inflating: /data/test/Luc_Montagnier_51.jpg \n inflating: /data/test/Lucia_Kenny_Anthony_00.jpg \n inflating: /data/test/Lucia_Kenny_Anthony_01.jpg \n inflating: /data/test/Lucia_Kenny_Anthony_10.jpg \n inflating: /data/test/Lucia_Kenny_Anthony_11.jpg \n inflating: /data/test/Lucia_Kenny_Anthony_40.jpg \n inflating: /data/test/Lucia_Kenny_Anthony_41.jpg \n inflating: /data/test/Lucia_Kenny_Anthony_50.jpg \n inflating: /data/test/Lucia_Kenny_Anthony_51.jpg \n inflating: /data/test/Lucio_Stanca_00.jpg \n inflating: /data/test/Lucio_Stanca_01.jpg \n inflating: /data/test/Lucio_Stanca_20.jpg \n inflating: /data/test/Lucio_Stanca_21.jpg \n inflating: /data/test/Lucio_Stanca_30.jpg \n inflating: /data/test/Lucio_Stanca_31.jpg \n inflating: /data/test/Lucio_Stanca_40.jpg \n inflating: /data/test/Lucio_Stanca_41.jpg \n inflating: /data/test/Luis_Ernesto_Derbez_Bautista_00.jpg \n inflating: /data/test/Luis_Ernesto_Derbez_Bautista_01.jpg \n inflating: /data/test/Luis_Ernesto_Derbez_Bautista_10.jpg \n inflating: /data/test/Luis_Ernesto_Derbez_Bautista_11.jpg \n inflating: /data/test/Luis_Ernesto_Derbez_Bautista_50.jpg \n inflating: /data/test/Luis_Ernesto_Derbez_Bautista_51.jpg \n inflating: /data/test/Luis_Fonsi_20.jpg \n inflating: /data/test/Luis_Fonsi_21.jpg \n inflating: /data/test/Luis_Fonsi_40.jpg \n inflating: /data/test/Luis_Fonsi_41.jpg \n inflating: /data/test/Luis_Fonsi_50.jpg \n inflating: /data/test/Luis_Fonsi_51.jpg \n inflating: /data/test/Lyle_Lovett_20.jpg \n inflating: /data/test/Lyle_Lovett_21.jpg \n inflating: /data/test/Lyle_Lovett_40.jpg \n inflating: /data/test/Lyle_Lovett_41.jpg \n inflating: /data/test/Lyle_Lovett_50.jpg \n inflating: /data/test/Lyle_Lovett_51.jpg \n inflating: /data/test/Mack_Brown_00.jpg \n inflating: /data/test/Mack_Brown_01.jpg \n inflating: /data/test/Mack_Brown_40.jpg \n inflating: /data/test/Mack_Brown_41.jpg \n inflating: /data/test/Mack_Brown_50.jpg \n inflating: /data/test/Mack_Brown_51.jpg \n inflating: /data/test/Maggie_Cheung_00.jpg \n inflating: /data/test/Maggie_Cheung_01.jpg \n inflating: /data/test/Maggie_Cheung_30.jpg \n inflating: /data/test/Maggie_Cheung_31.jpg \n inflating: /data/test/Maggie_Cheung_50.jpg \n inflating: /data/test/Maggie_Cheung_51.jpg \n inflating: /data/test/Maggie_Smith_00.jpg \n inflating: /data/test/Maggie_Smith_01.jpg \n inflating: /data/test/Maggie_Smith_30.jpg \n inflating: /data/test/Maggie_Smith_31.jpg \n inflating: /data/test/Maggie_Smith_40.jpg \n inflating: /data/test/Maggie_Smith_41.jpg \n inflating: /data/test/Mahathir_Mohamad_00.jpg \n inflating: /data/test/Mahathir_Mohamad_01.jpg \n inflating: /data/test/Mahathir_Mohamad_10.jpg \n inflating: /data/test/Mahathir_Mohamad_11.jpg \n inflating: /data/test/Mahathir_Mohamad_20.jpg \n inflating: /data/test/Mahathir_Mohamad_21.jpg \n inflating: /data/test/Mahathir_Mohamad_30.jpg \n inflating: /data/test/Mahathir_Mohamad_31.jpg \n inflating: /data/test/Malcolm_Jamal_Warner_00.jpg \n inflating: /data/test/Malcolm_Jamal_Warner_01.jpg \n inflating: /data/test/Malcolm_Jamal_Warner_10.jpg \n inflating: /data/test/Malcolm_Jamal_Warner_11.jpg \n inflating: /data/test/Malcolm_Jamal_Warner_20.jpg \n inflating: /data/test/Malcolm_Jamal_Warner_21.jpg \n inflating: /data/test/Manuel_Pellegrini_10.jpg \n inflating: /data/test/Manuel_Pellegrini_11.jpg \n inflating: /data/test/Manuel_Pellegrini_20.jpg \n inflating: /data/test/Manuel_Pellegrini_21.jpg \n inflating: /data/test/Manuel_Pellegrini_30.jpg \n inflating: /data/test/Manuel_Pellegrini_31.jpg \n inflating: /data/test/Marc_Anthony_10.jpg \n inflating: /data/test/Marc_Anthony_11.jpg \n inflating: /data/test/Marc_Anthony_20.jpg \n inflating: /data/test/Marc_Anthony_21.jpg \n inflating: /data/test/Marc_Anthony_50.jpg \n inflating: /data/test/Marc_Anthony_51.jpg \n inflating: /data/test/Marc_Racicot_00.jpg \n inflating: /data/test/Marc_Racicot_01.jpg \n inflating: /data/test/Marc_Racicot_20.jpg \n inflating: /data/test/Marc_Racicot_21.jpg \n inflating: /data/test/Marc_Racicot_40.jpg \n inflating: /data/test/Marc_Racicot_41.jpg \n inflating: /data/test/Marc_Racicot_50.jpg \n inflating: /data/test/Marc_Racicot_51.jpg \n inflating: /data/test/Marc_Shaiman_10.jpg \n inflating: /data/test/Marc_Shaiman_11.jpg \n inflating: /data/test/Marc_Shaiman_20.jpg \n inflating: /data/test/Marc_Shaiman_21.jpg \n inflating: /data/test/Marc_Shaiman_30.jpg \n inflating: /data/test/Marc_Shaiman_31.jpg \n inflating: /data/test/Margaret_Thatcher_10.jpg \n inflating: /data/test/Margaret_Thatcher_11.jpg \n inflating: /data/test/Margaret_Thatcher_30.jpg \n inflating: /data/test/Margaret_Thatcher_31.jpg \n inflating: /data/test/Margaret_Thatcher_40.jpg \n inflating: /data/test/Margaret_Thatcher_41.jpg \n inflating: /data/test/Margaret_Thatcher_50.jpg \n inflating: /data/test/Margaret_Thatcher_51.jpg \n inflating: /data/test/Maria_Soledad_Alvear_Valenzuela_10.jpg \n inflating: /data/test/Maria_Soledad_Alvear_Valenzuela_11.jpg \n inflating: /data/test/Maria_Soledad_Alvear_Valenzuela_30.jpg \n inflating: /data/test/Maria_Soledad_Alvear_Valenzuela_31.jpg \n inflating: /data/test/Maria_Soledad_Alvear_Valenzuela_40.jpg \n inflating: /data/test/Maria_Soledad_Alvear_Valenzuela_41.jpg \n inflating: /data/test/Mariana_Ohata_00.jpg \n inflating: /data/test/Mariana_Ohata_01.jpg \n inflating: /data/test/Mariana_Ohata_20.jpg \n inflating: /data/test/Mariana_Ohata_21.jpg \n inflating: /data/test/Mariana_Ohata_30.jpg \n inflating: /data/test/Mariana_Ohata_31.jpg \n inflating: /data/test/Marieta_Chrousala_00.jpg \n inflating: /data/test/Marieta_Chrousala_01.jpg \n inflating: /data/test/Marieta_Chrousala_10.jpg \n inflating: /data/test/Marieta_Chrousala_11.jpg \n inflating: /data/test/Marieta_Chrousala_40.jpg \n inflating: /data/test/Marieta_Chrousala_41.jpg \n inflating: /data/test/Marina_Silva_10.jpg \n inflating: /data/test/Marina_Silva_11.jpg \n inflating: /data/test/Marina_Silva_20.jpg \n inflating: /data/test/Marina_Silva_21.jpg \n inflating: /data/test/Marina_Silva_40.jpg \n inflating: /data/test/Marina_Silva_41.jpg \n inflating: /data/test/Marina_Silva_50.jpg \n inflating: /data/test/Marina_Silva_51.jpg \n inflating: /data/test/Mario_Kreutzberger_20.jpg \n inflating: /data/test/Mario_Kreutzberger_21.jpg \n inflating: /data/test/Mario_Kreutzberger_30.jpg \n inflating: /data/test/Mario_Kreutzberger_31.jpg \n inflating: /data/test/Mario_Kreutzberger_40.jpg \n inflating: /data/test/Mario_Kreutzberger_41.jpg \n inflating: /data/test/Marisa_Tomei_10.jpg \n inflating: /data/test/Marisa_Tomei_11.jpg \n inflating: /data/test/Marisa_Tomei_20.jpg \n inflating: /data/test/Marisa_Tomei_21.jpg \n inflating: /data/test/Marisa_Tomei_40.jpg \n inflating: /data/test/Marisa_Tomei_41.jpg \n inflating: /data/test/Marissa_Jaret_Winokur_00.jpg \n inflating: /data/test/Marissa_Jaret_Winokur_01.jpg \n inflating: /data/test/Marissa_Jaret_Winokur_30.jpg \n inflating: /data/test/Marissa_Jaret_Winokur_31.jpg \n inflating: /data/test/Marissa_Jaret_Winokur_40.jpg \n inflating: /data/test/Marissa_Jaret_Winokur_41.jpg \n inflating: /data/test/Mark_Foley_10.jpg \n inflating: /data/test/Mark_Foley_11.jpg \n inflating: /data/test/Mark_Foley_40.jpg \n inflating: /data/test/Mark_Foley_41.jpg \n inflating: /data/test/Mark_Foley_50.jpg \n inflating: /data/test/Mark_Foley_51.jpg \n inflating: /data/test/Mark_Leno_10.jpg \n inflating: /data/test/Mark_Leno_11.jpg \n inflating: /data/test/Mark_Leno_20.jpg \n inflating: /data/test/Mark_Leno_21.jpg \n inflating: /data/test/Mark_Leno_30.jpg \n inflating: /data/test/Mark_Leno_31.jpg \n inflating: /data/test/Martin_Luther_King_III_00.jpg \n inflating: /data/test/Martin_Luther_King_III_01.jpg \n inflating: /data/test/Martin_Luther_King_III_30.jpg \n inflating: /data/test/Martin_Luther_King_III_31.jpg \n inflating: /data/test/Martin_Luther_King_III_50.jpg \n inflating: /data/test/Martin_Luther_King_III_51.jpg \n inflating: /data/test/Martin_Sheen_00.jpg \n inflating: /data/test/Martin_Sheen_01.jpg \n inflating: /data/test/Martin_Sheen_30.jpg \n inflating: /data/test/Martin_Sheen_31.jpg \n inflating: /data/test/Martin_Sheen_40.jpg \n inflating: /data/test/Martin_Sheen_41.jpg \n inflating: /data/test/Martin_Sheen_50.jpg \n inflating: /data/test/Martin_Sheen_51.jpg \n inflating: /data/test/Mary_Landrieu_00.jpg \n inflating: /data/test/Mary_Landrieu_01.jpg \n inflating: /data/test/Mary_Landrieu_20.jpg \n inflating: /data/test/Mary_Landrieu_21.jpg \n inflating: /data/test/Mary_Landrieu_30.jpg \n inflating: /data/test/Mary_Landrieu_31.jpg \n inflating: /data/test/Mary_Robinson_10.jpg \n inflating: /data/test/Mary_Robinson_11.jpg \n inflating: /data/test/Mary_Robinson_20.jpg \n inflating: /data/test/Mary_Robinson_21.jpg \n inflating: /data/test/Mary_Robinson_40.jpg \n inflating: /data/test/Mary_Robinson_41.jpg \n inflating: /data/test/Mary_Robinson_50.jpg \n inflating: /data/test/Mary_Robinson_51.jpg \n inflating: /data/test/Massoud_Barzani_00.jpg \n inflating: /data/test/Massoud_Barzani_01.jpg \n inflating: /data/test/Massoud_Barzani_10.jpg \n inflating: /data/test/Massoud_Barzani_11.jpg \n inflating: /data/test/Massoud_Barzani_20.jpg \n inflating: /data/test/Massoud_Barzani_21.jpg \n inflating: /data/test/Massoud_Barzani_40.jpg \n inflating: /data/test/Massoud_Barzani_41.jpg \n inflating: /data/test/Matt_LeBlanc_00.jpg \n inflating: /data/test/Matt_LeBlanc_01.jpg \n inflating: /data/test/Matt_LeBlanc_20.jpg \n inflating: /data/test/Matt_LeBlanc_21.jpg \n inflating: /data/test/Matt_LeBlanc_30.jpg \n inflating: /data/test/Matt_LeBlanc_31.jpg \n inflating: /data/test/Nancy_Kerrigan_00.jpg \n inflating: /data/test/Nancy_Kerrigan_01.jpg \n inflating: /data/test/Nancy_Kerrigan_20.jpg \n inflating: /data/test/Nancy_Kerrigan_21.jpg \n inflating: /data/test/Nancy_Kerrigan_30.jpg \n inflating: /data/test/Nancy_Kerrigan_31.jpg \n inflating: /data/test/Nancy_Kerrigan_40.jpg \n inflating: /data/test/Nancy_Kerrigan_41.jpg \n inflating: /data/test/Nancy_Reagan_00.jpg \n inflating: /data/test/Nancy_Reagan_01.jpg \n inflating: /data/test/Nancy_Reagan_10.jpg \n inflating: /data/test/Nancy_Reagan_11.jpg \n inflating: /data/test/Nancy_Reagan_30.jpg \n inflating: /data/test/Nancy_Reagan_31.jpg \n inflating: /data/test/Nancy_Reagan_40.jpg \n inflating: /data/test/Nancy_Reagan_41.jpg \n inflating: /data/test/Nanni_Moretti_10.jpg \n inflating: /data/test/Nanni_Moretti_11.jpg \n inflating: /data/test/Nanni_Moretti_20.jpg \n inflating: /data/test/Nanni_Moretti_21.jpg \n inflating: /data/test/Nanni_Moretti_40.jpg \n inflating: /data/test/Nanni_Moretti_41.jpg \n inflating: /data/test/Natalia_Vodonova_00.jpg \n inflating: /data/test/Natalia_Vodonova_01.jpg \n inflating: /data/test/Natalia_Vodonova_10.jpg \n inflating: /data/test/Natalia_Vodonova_11.jpg \n inflating: /data/test/Natalia_Vodonova_20.jpg \n inflating: /data/test/Natalia_Vodonova_21.jpg \n inflating: /data/test/Natasha_Lyonne_00.jpg \n inflating: /data/test/Natasha_Lyonne_01.jpg \n inflating: /data/test/Natasha_Lyonne_10.jpg \n inflating: /data/test/Natasha_Lyonne_11.jpg \n inflating: /data/test/Natasha_Lyonne_40.jpg \n inflating: /data/test/Natasha_Lyonne_41.jpg \n inflating: /data/test/Nick_Reilly_10.jpg \n inflating: /data/test/Nick_Reilly_11.jpg \n inflating: /data/test/Nick_Reilly_40.jpg \n inflating: /data/test/Nick_Reilly_41.jpg \n inflating: /data/test/Nick_Reilly_50.jpg \n inflating: /data/test/Nick_Reilly_51.jpg \n inflating: /data/test/Nicolas_Eyzaguirre_00.jpg \n inflating: /data/test/Nicolas_Eyzaguirre_01.jpg \n inflating: /data/test/Nicolas_Eyzaguirre_10.jpg \n inflating: /data/test/Nicolas_Eyzaguirre_11.jpg \n inflating: /data/test/Nicolas_Eyzaguirre_20.jpg \n inflating: /data/test/Nicolas_Eyzaguirre_21.jpg \n inflating: /data/test/Nicolas_Sarkozy_00.jpg \n inflating: /data/test/Nicolas_Sarkozy_01.jpg \n inflating: /data/test/Nicolas_Sarkozy_10.jpg \n inflating: /data/test/Nicolas_Sarkozy_11.jpg \n inflating: /data/test/Nicolas_Sarkozy_20.jpg \n inflating: /data/test/Nicolas_Sarkozy_21.jpg \n inflating: /data/test/Nicolas_Sarkozy_50.jpg \n inflating: /data/test/Nicolas_Sarkozy_51.jpg \n inflating: /data/test/Nina_Jacobson_00.jpg \n inflating: /data/test/Nina_Jacobson_01.jpg \n inflating: /data/test/Nina_Jacobson_10.jpg \n inflating: /data/test/Nina_Jacobson_11.jpg \n inflating: /data/test/Nina_Jacobson_30.jpg \n inflating: /data/test/Nina_Jacobson_31.jpg \n inflating: /data/test/Norah_Jones_10.jpg \n inflating: /data/test/Norah_Jones_11.jpg \n inflating: /data/test/Norah_Jones_20.jpg \n inflating: /data/test/Norah_Jones_21.jpg \n inflating: /data/test/Norah_Jones_40.jpg \n inflating: /data/test/Norah_Jones_41.jpg \n inflating: /data/test/Norah_Jones_50.jpg \n inflating: /data/test/Norah_Jones_51.jpg \n inflating: /data/test/Norman_Mineta_00.jpg \n inflating: /data/test/Norman_Mineta_01.jpg \n inflating: /data/test/Norman_Mineta_30.jpg \n inflating: /data/test/Norman_Mineta_31.jpg \n inflating: /data/test/Norman_Mineta_50.jpg \n inflating: /data/test/Norman_Mineta_51.jpg \n inflating: /data/test/Olene_Walker_00.jpg \n inflating: /data/test/Olene_Walker_01.jpg \n inflating: /data/test/Olene_Walker_10.jpg \n inflating: /data/test/Olene_Walker_11.jpg \n inflating: /data/test/Olene_Walker_30.jpg \n inflating: /data/test/Olene_Walker_31.jpg \n inflating: /data/test/Olene_Walker_40.jpg \n inflating: /data/test/Olene_Walker_41.jpg \n inflating: /data/test/Olivia_Newton-John_00.jpg \n inflating: /data/test/Olivia_Newton-John_01.jpg \n inflating: /data/test/Olivia_Newton-John_10.jpg \n inflating: /data/test/Olivia_Newton-John_11.jpg \n inflating: /data/test/Olivia_Newton-John_40.jpg \n inflating: /data/test/Olivia_Newton-John_41.jpg \n inflating: /data/test/Orlando_Bloom_00.jpg \n inflating: /data/test/Orlando_Bloom_01.jpg \n inflating: /data/test/Orlando_Bloom_30.jpg \n inflating: /data/test/Orlando_Bloom_31.jpg \n inflating: /data/test/Orlando_Bloom_40.jpg \n inflating: /data/test/Orlando_Bloom_41.jpg \n inflating: /data/test/Orlando_Bloom_50.jpg \n inflating: /data/test/Orlando_Bloom_51.jpg \n inflating: /data/test/Otto_Reich_00.jpg \n inflating: /data/test/Otto_Reich_01.jpg \n inflating: /data/test/Otto_Reich_10.jpg \n inflating: /data/test/Otto_Reich_11.jpg \n inflating: /data/test/Otto_Reich_30.jpg \n inflating: /data/test/Otto_Reich_31.jpg \n inflating: /data/test/Otto_Reich_40.jpg \n inflating: /data/test/Otto_Reich_41.jpg \n inflating: /data/test/Pat_Riley_00.jpg \n inflating: /data/test/Pat_Riley_01.jpg \n inflating: /data/test/Pat_Riley_20.jpg \n inflating: /data/test/Pat_Riley_21.jpg \n inflating: /data/test/Pat_Riley_50.jpg \n inflating: /data/test/Pat_Riley_51.jpg \n inflating: /data/test/Patrick_Leahy_10.jpg \n inflating: /data/test/Patrick_Leahy_11.jpg \n inflating: /data/test/Patrick_Leahy_20.jpg \n inflating: /data/test/Patrick_Leahy_21.jpg \n inflating: /data/test/Patrick_Leahy_30.jpg \n inflating: /data/test/Patrick_Leahy_31.jpg \n inflating: /data/test/Paul_Otellini_00.jpg \n inflating: /data/test/Paul_Otellini_01.jpg \n inflating: /data/test/Paul_Otellini_10.jpg \n inflating: /data/test/Paul_Otellini_11.jpg \n inflating: /data/test/Paul_Otellini_20.jpg \n inflating: /data/test/Paul_Otellini_21.jpg \n inflating: /data/test/Paul_Reiser_00.jpg \n inflating: /data/test/Paul_Reiser_01.jpg \n inflating: /data/test/Paul_Reiser_20.jpg \n inflating: /data/test/Paul_Reiser_21.jpg \n inflating: /data/test/Paul_Reiser_30.jpg \n inflating: /data/test/Paul_Reiser_31.jpg \n inflating: /data/test/Pedro_Solbes_00.jpg \n inflating: /data/test/Pedro_Solbes_01.jpg \n inflating: /data/test/Pedro_Solbes_20.jpg \n inflating: /data/test/Pedro_Solbes_21.jpg \n inflating: /data/test/Pedro_Solbes_30.jpg \n inflating: /data/test/Pedro_Solbes_31.jpg \n inflating: /data/test/Penelope_Ann_Miller_00.jpg \n inflating: /data/test/Penelope_Ann_Miller_01.jpg \n inflating: /data/test/Penelope_Ann_Miller_20.jpg \n inflating: /data/test/Penelope_Ann_Miller_21.jpg \n inflating: /data/test/Penelope_Ann_Miller_50.jpg \n inflating: /data/test/Penelope_Ann_Miller_51.jpg \n inflating: /data/test/Peter_Goldmark_10.jpg \n inflating: /data/test/Peter_Goldmark_11.jpg \n inflating: /data/test/Peter_Goldmark_40.jpg \n inflating: /data/test/Peter_Goldmark_41.jpg \n inflating: /data/test/Peter_Goldmark_50.jpg \n inflating: /data/test/Peter_Goldmark_51.jpg \n inflating: /data/test/Peter_Medgyessy_10.jpg \n inflating: /data/test/Peter_Medgyessy_11.jpg \n inflating: /data/test/Peter_Medgyessy_30.jpg \n inflating: /data/test/Peter_Medgyessy_31.jpg \n inflating: /data/test/Peter_Medgyessy_40.jpg \n inflating: /data/test/Peter_Medgyessy_41.jpg \n inflating: /data/test/Peter_Medgyessy_50.jpg \n inflating: /data/test/Peter_Medgyessy_51.jpg \n inflating: /data/test/Philippe_Gagnon_00.jpg \n inflating: /data/test/Philippe_Gagnon_01.jpg \n inflating: /data/test/Philippe_Gagnon_10.jpg \n inflating: /data/test/Philippe_Gagnon_11.jpg \n inflating: /data/test/Philippe_Gagnon_20.jpg \n inflating: /data/test/Philippe_Gagnon_21.jpg \n inflating: /data/test/Philippe_Gagnon_30.jpg \n inflating: /data/test/Philippe_Gagnon_31.jpg \n inflating: /data/test/Philippe_Noiret_10.jpg \n inflating: /data/test/Philippe_Noiret_11.jpg \n inflating: /data/test/Philippe_Noiret_30.jpg \n inflating: /data/test/Philippe_Noiret_31.jpg \n inflating: /data/test/Philippe_Noiret_50.jpg \n inflating: /data/test/Philippe_Noiret_51.jpg \n inflating: /data/test/Picabo_Street_00.jpg \n inflating: /data/test/Picabo_Street_01.jpg \n inflating: /data/test/Picabo_Street_20.jpg \n inflating: /data/test/Picabo_Street_21.jpg \n inflating: /data/test/Picabo_Street_40.jpg \n inflating: /data/test/Picabo_Street_41.jpg \n inflating: /data/test/Pilar_Montenegro_10.jpg \n inflating: /data/test/Pilar_Montenegro_11.jpg \n inflating: /data/test/Pilar_Montenegro_20.jpg \n inflating: /data/test/Pilar_Montenegro_21.jpg \n inflating: /data/test/Pilar_Montenegro_50.jpg \n inflating: /data/test/Pilar_Montenegro_51.jpg \n inflating: /data/test/Piotr_Anderszewski_20.jpg \n inflating: /data/test/Piotr_Anderszewski_21.jpg \n inflating: /data/test/Piotr_Anderszewski_30.jpg \n inflating: /data/test/Piotr_Anderszewski_31.jpg \n inflating: /data/test/Piotr_Anderszewski_50.jpg \n inflating: /data/test/Piotr_Anderszewski_51.jpg \n inflating: /data/test/Poala_Suarez_30.jpg \n inflating: /data/test/Poala_Suarez_31.jpg \n inflating: /data/test/Poala_Suarez_40.jpg \n inflating: /data/test/Poala_Suarez_41.jpg \n inflating: /data/test/Poala_Suarez_50.jpg \n inflating: /data/test/Poala_Suarez_51.jpg \n inflating: /data/test/Prince_Harry_10.jpg \n inflating: /data/test/Prince_Harry_11.jpg \n inflating: /data/test/Prince_Harry_20.jpg \n inflating: /data/test/Prince_Harry_21.jpg \n inflating: /data/test/Prince_Harry_40.jpg \n inflating: /data/test/Prince_Harry_41.jpg \n inflating: /data/test/Princess_Stephanie_00.jpg \n inflating: /data/test/Princess_Stephanie_01.jpg \n inflating: /data/test/Princess_Stephanie_20.jpg \n inflating: /data/test/Princess_Stephanie_21.jpg \n inflating: /data/test/Princess_Stephanie_40.jpg \n inflating: /data/test/Princess_Stephanie_41.jpg \n inflating: /data/test/Princess_Stephanie_50.jpg \n inflating: /data/test/Princess_Stephanie_51.jpg \n inflating: /data/test/Priyanka_Chopra_10.jpg \n inflating: /data/test/Priyanka_Chopra_11.jpg \n inflating: /data/test/Priyanka_Chopra_40.jpg \n inflating: /data/test/Priyanka_Chopra_41.jpg \n inflating: /data/test/Priyanka_Chopra_50.jpg \n inflating: /data/test/Priyanka_Chopra_51.jpg \n inflating: /data/test/Queen_Noor_10.jpg \n inflating: /data/test/Queen_Noor_11.jpg \n inflating: /data/test/Queen_Noor_30.jpg \n inflating: /data/test/Queen_Noor_31.jpg \n inflating: /data/test/Queen_Noor_50.jpg \n inflating: /data/test/Queen_Noor_51.jpg \n inflating: /data/test/Queen_Rania_10.jpg \n inflating: /data/test/Queen_Rania_11.jpg \n inflating: /data/test/Queen_Rania_30.jpg \n inflating: /data/test/Queen_Rania_31.jpg \n inflating: /data/test/Queen_Rania_50.jpg \n inflating: /data/test/Queen_Rania_51.jpg \n inflating: /data/test/Rachel_Hunter_30.jpg \n inflating: /data/test/Rachel_Hunter_31.jpg \n inflating: /data/test/Rachel_Hunter_40.jpg \n inflating: /data/test/Rachel_Hunter_41.jpg \n inflating: /data/test/Rachel_Hunter_50.jpg \n inflating: /data/test/Rachel_Hunter_51.jpg \n inflating: /data/test/Raja_Zafar-ul-Haq_00.jpg \n inflating: /data/test/Raja_Zafar-ul-Haq_01.jpg \n inflating: /data/test/Raja_Zafar-ul-Haq_10.jpg \n inflating: /data/test/Raja_Zafar-ul-Haq_11.jpg \n inflating: /data/test/Raja_Zafar-ul-Haq_20.jpg \n inflating: /data/test/Raja_Zafar-ul-Haq_21.jpg \n inflating: /data/test/Raja_Zafar-ul-Haq_30.jpg \n inflating: /data/test/Raja_Zafar-ul-Haq_31.jpg \n inflating: /data/test/Ralph_Klein_00.jpg \n inflating: /data/test/Ralph_Klein_01.jpg \n inflating: /data/test/Ralph_Klein_10.jpg \n inflating: /data/test/Ralph_Klein_11.jpg \n inflating: /data/test/Ralph_Klein_30.jpg \n inflating: /data/test/Ralph_Klein_31.jpg \n inflating: /data/test/Raza_Rabbani_20.jpg \n inflating: /data/test/Raza_Rabbani_21.jpg \n inflating: /data/test/Raza_Rabbani_30.jpg \n inflating: /data/test/Raza_Rabbani_31.jpg \n inflating: /data/test/Raza_Rabbani_50.jpg \n inflating: /data/test/Raza_Rabbani_51.jpg \n inflating: /data/test/Recep_Tayyip_Erdogan_00.jpg \n inflating: /data/test/Recep_Tayyip_Erdogan_01.jpg \n inflating: /data/test/Recep_Tayyip_Erdogan_20.jpg \n inflating: /data/test/Recep_Tayyip_Erdogan_21.jpg \n inflating: /data/test/Recep_Tayyip_Erdogan_40.jpg \n inflating: /data/test/Recep_Tayyip_Erdogan_41.jpg \n inflating: /data/test/Reese_Witherspoon_00.jpg \n inflating: /data/test/Reese_Witherspoon_01.jpg \n inflating: /data/test/Reese_Witherspoon_10.jpg \n inflating: /data/test/Reese_Witherspoon_11.jpg \n inflating: /data/test/Reese_Witherspoon_40.jpg \n inflating: /data/test/Reese_Witherspoon_41.jpg \n inflating: /data/test/Ricardo_Lopez_Murphy_10.jpg \n inflating: /data/test/Ricardo_Lopez_Murphy_11.jpg \n inflating: /data/test/Ricardo_Lopez_Murphy_30.jpg \n inflating: /data/test/Ricardo_Lopez_Murphy_31.jpg \n inflating: /data/test/Ricardo_Lopez_Murphy_40.jpg \n inflating: /data/test/Ricardo_Lopez_Murphy_41.jpg \n inflating: /data/test/Ricardo_Sanchez_20.jpg \n inflating: /data/test/Ricardo_Sanchez_21.jpg \n inflating: /data/test/Ricardo_Sanchez_30.jpg \n inflating: /data/test/Ricardo_Sanchez_31.jpg \n inflating: /data/test/Ricardo_Sanchez_40.jpg \n inflating: /data/test/Ricardo_Sanchez_41.jpg \n inflating: /data/test/Richard_Branson_00.jpg \n inflating: /data/test/Richard_Branson_01.jpg \n inflating: /data/test/Richard_Branson_10.jpg \n inflating: /data/test/Richard_Branson_11.jpg \n inflating: /data/test/Richard_Branson_50.jpg \n inflating: /data/test/Richard_Branson_51.jpg \n inflating: /data/test/Richard_Lennon_00.jpg \n inflating: /data/test/Richard_Lennon_01.jpg \n inflating: /data/test/Richard_Lennon_30.jpg \n inflating: /data/test/Richard_Lennon_31.jpg \n inflating: /data/test/Richard_Lennon_40.jpg \n inflating: /data/test/Richard_Lennon_41.jpg \n inflating: /data/test/Richard_Lugar_00.jpg \n inflating: /data/test/Richard_Lugar_01.jpg \n inflating: /data/test/Richard_Lugar_10.jpg \n inflating: /data/test/Richard_Lugar_11.jpg \n inflating: /data/test/Richard_Lugar_20.jpg \n inflating: /data/test/Richard_Lugar_21.jpg \n inflating: /data/test/Richard_Lugar_50.jpg \n inflating: /data/test/Richard_Lugar_51.jpg \n inflating: /data/test/Richard_Paul_Evans_00.jpg \n inflating: /data/test/Richard_Paul_Evans_01.jpg \n inflating: /data/test/Richard_Paul_Evans_20.jpg \n inflating: /data/test/Richard_Paul_Evans_21.jpg \n inflating: /data/test/Richard_Paul_Evans_40.jpg \n inflating: /data/test/Richard_Paul_Evans_41.jpg \n inflating: /data/test/Richard_Paul_Evans_50.jpg \n inflating: /data/test/Richard_Paul_Evans_51.jpg \n inflating: /data/test/Rick_Bragg_20.jpg \n inflating: /data/test/Rick_Bragg_21.jpg \n inflating: /data/test/Rick_Bragg_30.jpg \n inflating: /data/test/Rick_Bragg_31.jpg \n inflating: /data/test/Rick_Bragg_50.jpg \n inflating: /data/test/Rick_Bragg_51.jpg \n inflating: /data/test/Ridley_Scott_10.jpg \n inflating: /data/test/Ridley_Scott_11.jpg \n inflating: /data/test/Ridley_Scott_20.jpg \n inflating: /data/test/Ridley_Scott_21.jpg \n inflating: /data/test/Ridley_Scott_30.jpg \n inflating: /data/test/Ridley_Scott_31.jpg \n inflating: /data/test/Robbie_Coltrane_00.jpg \n inflating: /data/test/Robbie_Coltrane_01.jpg \n inflating: /data/test/Robbie_Coltrane_10.jpg \n inflating: /data/test/Robbie_Coltrane_11.jpg \n inflating: /data/test/Robbie_Coltrane_20.jpg \n inflating: /data/test/Robbie_Coltrane_21.jpg \n inflating: /data/test/Robbie_Coltrane_50.jpg \n inflating: /data/test/Robbie_Coltrane_51.jpg \n inflating: /data/test/Robert_Altman_10.jpg \n inflating: /data/test/Robert_Altman_11.jpg \n inflating: /data/test/Robert_Altman_20.jpg \n inflating: /data/test/Robert_Altman_21.jpg \n inflating: /data/test/Robert_Altman_50.jpg \n inflating: /data/test/Robert_Altman_51.jpg \n inflating: /data/test/Roberto_Benigni_00.jpg \n inflating: /data/test/Roberto_Benigni_01.jpg \n inflating: /data/test/Roberto_Benigni_10.jpg \n inflating: /data/test/Roberto_Benigni_11.jpg \n inflating: /data/test/Roberto_Benigni_50.jpg \n inflating: /data/test/Roberto_Benigni_51.jpg \n inflating: /data/test/Rocco_Buttiglione_00.jpg \n inflating: /data/test/Rocco_Buttiglione_01.jpg \n inflating: /data/test/Rocco_Buttiglione_10.jpg \n inflating: /data/test/Rocco_Buttiglione_11.jpg \n inflating: /data/test/Rocco_Buttiglione_30.jpg \n inflating: /data/test/Rocco_Buttiglione_31.jpg \n inflating: /data/test/Rocco_Buttiglione_40.jpg \n inflating: /data/test/Rocco_Buttiglione_41.jpg \n inflating: /data/test/Rodrigo_Borja_00.jpg \n inflating: /data/test/Rodrigo_Borja_01.jpg \n inflating: /data/test/Rodrigo_Borja_40.jpg \n inflating: /data/test/Rodrigo_Borja_41.jpg \n inflating: /data/test/Rodrigo_Borja_50.jpg \n inflating: /data/test/Rodrigo_Borja_51.jpg \n inflating: /data/test/Saeed_Mortazavi_10.jpg \n inflating: /data/test/Saeed_Mortazavi_11.jpg \n inflating: /data/test/Saeed_Mortazavi_20.jpg \n inflating: /data/test/Saeed_Mortazavi_21.jpg \n inflating: /data/test/Saeed_Mortazavi_50.jpg \n inflating: /data/test/Saeed_Mortazavi_51.jpg \n inflating: /data/test/Sally_Ride_00.jpg \n inflating: /data/test/Sally_Ride_01.jpg \n inflating: /data/test/Sally_Ride_40.jpg \n inflating: /data/test/Sally_Ride_41.jpg \n inflating: /data/test/Sally_Ride_50.jpg \n inflating: /data/test/Sally_Ride_51.jpg \n inflating: /data/test/Sanjay_Gupta_10.jpg \n inflating: /data/test/Sanjay_Gupta_11.jpg \n inflating: /data/test/Sanjay_Gupta_20.jpg \n inflating: /data/test/Sanjay_Gupta_21.jpg \n inflating: /data/test/Sanjay_Gupta_40.jpg \n inflating: /data/test/Sanjay_Gupta_41.jpg \n inflating: /data/test/Sara_Silverman_10.jpg \n inflating: /data/test/Sara_Silverman_11.jpg \n inflating: /data/test/Sara_Silverman_20.jpg \n inflating: /data/test/Sara_Silverman_21.jpg \n inflating: /data/test/Sara_Silverman_40.jpg \n inflating: /data/test/Sara_Silverman_41.jpg \n inflating: /data/test/Sara_Silverman_50.jpg \n inflating: /data/test/Sara_Silverman_51.jpg \n inflating: /data/test/Sarah_Wynter_00.jpg \n inflating: /data/test/Sarah_Wynter_01.jpg \n inflating: /data/test/Sarah_Wynter_40.jpg \n inflating: /data/test/Sarah_Wynter_41.jpg \n inflating: /data/test/Sarah_Wynter_50.jpg \n inflating: /data/test/Sarah_Wynter_51.jpg \n inflating: /data/test/Sasha_Cohen_20.jpg \n inflating: /data/test/Sasha_Cohen_21.jpg \n inflating: /data/test/Sasha_Cohen_40.jpg \n inflating: /data/test/Sasha_Cohen_41.jpg \n inflating: /data/test/Sasha_Cohen_50.jpg \n inflating: /data/test/Sasha_Cohen_51.jpg \n inflating: /data/test/T_Boone_Pickens_10.jpg \n inflating: /data/test/T_Boone_Pickens_11.jpg \n inflating: /data/test/T_Boone_Pickens_20.jpg \n inflating: /data/test/T_Boone_Pickens_21.jpg \n inflating: /data/test/T_Boone_Pickens_30.jpg \n inflating: /data/test/T_Boone_Pickens_31.jpg \n inflating: /data/test/T_Boone_Pickens_50.jpg \n inflating: /data/test/T_Boone_Pickens_51.jpg \n inflating: /data/test/Takeo_Hiranuma_00.jpg \n inflating: /data/test/Takeo_Hiranuma_01.jpg \n inflating: /data/test/Takeo_Hiranuma_10.jpg \n inflating: /data/test/Takeo_Hiranuma_11.jpg \n inflating: /data/test/Takeo_Hiranuma_30.jpg \n inflating: /data/test/Takeo_Hiranuma_31.jpg \n inflating: /data/test/Ted_Turner_20.jpg \n inflating: /data/test/Ted_Turner_21.jpg \n inflating: /data/test/Ted_Turner_30.jpg \n inflating: /data/test/Ted_Turner_31.jpg \n inflating: /data/test/Ted_Turner_50.jpg \n inflating: /data/test/Ted_Turner_51.jpg \n inflating: /data/test/Teresa_Heinz_Kerry_00.jpg \n inflating: /data/test/Teresa_Heinz_Kerry_01.jpg \n inflating: /data/test/Teresa_Heinz_Kerry_10.jpg \n inflating: /data/test/Teresa_Heinz_Kerry_11.jpg \n inflating: /data/test/Teresa_Heinz_Kerry_20.jpg \n inflating: /data/test/Teresa_Heinz_Kerry_21.jpg \n inflating: /data/test/Terje_Roed-Larsen_00.jpg \n inflating: /data/test/Terje_Roed-Larsen_01.jpg \n inflating: /data/test/Terje_Roed-Larsen_20.jpg \n inflating: /data/test/Terje_Roed-Larsen_21.jpg \n inflating: /data/test/Terje_Roed-Larsen_30.jpg \n inflating: /data/test/Terje_Roed-Larsen_31.jpg \n inflating: /data/test/Tessa_Jowell_00.jpg \n inflating: /data/test/Tessa_Jowell_01.jpg \n inflating: /data/test/Tessa_Jowell_20.jpg \n inflating: /data/test/Tessa_Jowell_21.jpg \n inflating: /data/test/Tessa_Jowell_30.jpg \n inflating: /data/test/Tessa_Jowell_31.jpg \n inflating: /data/test/Tessa_Jowell_50.jpg \n inflating: /data/test/Tessa_Jowell_51.jpg \n inflating: /data/test/Thomas_Ferguson_00.jpg \n inflating: /data/test/Thomas_Ferguson_01.jpg \n inflating: /data/test/Thomas_Ferguson_10.jpg \n inflating: /data/test/Thomas_Ferguson_11.jpg \n inflating: /data/test/Thomas_Ferguson_50.jpg \n inflating: /data/test/Thomas_Ferguson_51.jpg \n inflating: /data/test/Tim_Howard_00.jpg \n inflating: /data/test/Tim_Howard_01.jpg \n inflating: /data/test/Tim_Howard_10.jpg \n inflating: /data/test/Tim_Howard_11.jpg \n inflating: /data/test/Tim_Howard_30.jpg \n inflating: /data/test/Tim_Howard_31.jpg \n inflating: /data/test/Tim_Pawlenty_00.jpg \n inflating: /data/test/Tim_Pawlenty_01.jpg \n inflating: /data/test/Tim_Pawlenty_30.jpg \n inflating: /data/test/Tim_Pawlenty_31.jpg \n inflating: /data/test/Tim_Pawlenty_40.jpg \n inflating: /data/test/Tim_Pawlenty_41.jpg \n inflating: /data/test/Tim_Pawlenty_50.jpg \n inflating: /data/test/Tim_Pawlenty_51.jpg \n inflating: /data/test/Timothy_Goebel_00.jpg \n inflating: /data/test/Timothy_Goebel_01.jpg \n inflating: /data/test/Timothy_Goebel_30.jpg \n inflating: /data/test/Timothy_Goebel_31.jpg \n inflating: /data/test/Timothy_Goebel_40.jpg \n inflating: /data/test/Timothy_Goebel_41.jpg \n inflating: /data/test/Timothy_Goebel_50.jpg \n inflating: /data/test/Timothy_Goebel_51.jpg \n inflating: /data/test/Tina_Brown_20.jpg \n inflating: /data/test/Tina_Brown_21.jpg \n inflating: /data/test/Tina_Brown_40.jpg \n inflating: /data/test/Tina_Brown_41.jpg \n inflating: /data/test/Tina_Brown_50.jpg \n inflating: /data/test/Tina_Brown_51.jpg \n inflating: /data/test/Tom_Coughlin_20.jpg \n inflating: /data/test/Tom_Coughlin_21.jpg \n inflating: /data/test/Tom_Coughlin_30.jpg \n inflating: /data/test/Tom_Coughlin_31.jpg \n inflating: /data/test/Tom_Coughlin_50.jpg \n inflating: /data/test/Tom_Coughlin_51.jpg \n inflating: /data/test/Tom_Hanks_30.jpg \n inflating: /data/test/Tom_Hanks_31.jpg \n inflating: /data/test/Tom_Hanks_40.jpg \n inflating: /data/test/Tom_Hanks_41.jpg \n inflating: /data/test/Tom_Hanks_50.jpg \n inflating: /data/test/Tom_Hanks_51.jpg \n inflating: /data/test/Tom_Harkin_00.jpg \n inflating: /data/test/Tom_Harkin_01.jpg \n inflating: /data/test/Tom_Harkin_30.jpg \n inflating: /data/test/Tom_Harkin_31.jpg \n inflating: /data/test/Tom_Harkin_40.jpg \n inflating: /data/test/Tom_Harkin_41.jpg \n inflating: /data/test/Tom_Osborne_20.jpg \n inflating: /data/test/Tom_Osborne_21.jpg \n inflating: /data/test/Tom_Osborne_30.jpg \n inflating: /data/test/Tom_Osborne_31.jpg \n inflating: /data/test/Tom_Osborne_50.jpg \n inflating: /data/test/Tom_Osborne_51.jpg \n inflating: /data/test/Tom_Ridge_20.jpg \n inflating: /data/test/Tom_Ridge_21.jpg \n inflating: /data/test/Tom_Ridge_30.jpg \n inflating: /data/test/Tom_Ridge_31.jpg \n inflating: /data/test/Tom_Ridge_50.jpg \n inflating: /data/test/Tom_Ridge_51.jpg \n inflating: /data/test/Tom_Sizemore_00.jpg \n inflating: /data/test/Tom_Sizemore_01.jpg \n inflating: /data/test/Tom_Sizemore_10.jpg \n inflating: /data/test/Tom_Sizemore_11.jpg \n inflating: /data/test/Tom_Sizemore_20.jpg \n inflating: /data/test/Tom_Sizemore_21.jpg \n inflating: /data/test/Valerie_Harper_00.jpg \n inflating: /data/test/Valerie_Harper_01.jpg \n inflating: /data/test/Valerie_Harper_30.jpg \n inflating: /data/test/Valerie_Harper_31.jpg \n inflating: /data/test/Valerie_Harper_40.jpg \n inflating: /data/test/Valerie_Harper_41.jpg \n inflating: /data/test/Valerie_Harper_50.jpg \n inflating: /data/test/Valerie_Harper_51.jpg \n inflating: /data/test/Vicente_Fox_10.jpg \n inflating: /data/test/Vicente_Fox_11.jpg \n inflating: /data/test/Vicente_Fox_20.jpg \n inflating: /data/test/Vicente_Fox_21.jpg \n inflating: /data/test/Vicente_Fox_30.jpg \n inflating: /data/test/Vicente_Fox_31.jpg \n inflating: /data/test/Vojislav_Seselj_00.jpg \n inflating: /data/test/Vojislav_Seselj_01.jpg \n inflating: /data/test/Vojislav_Seselj_20.jpg \n inflating: /data/test/Vojislav_Seselj_21.jpg \n inflating: /data/test/Vojislav_Seselj_40.jpg \n inflating: /data/test/Vojislav_Seselj_41.jpg \n inflating: /data/test/Vojislav_Seselj_50.jpg \n inflating: /data/test/Vojislav_Seselj_51.jpg \n inflating: /data/test/Warren_Beatty_10.jpg \n inflating: /data/test/Warren_Beatty_11.jpg \n inflating: /data/test/Warren_Beatty_30.jpg \n inflating: /data/test/Warren_Beatty_31.jpg \n inflating: /data/test/Warren_Beatty_50.jpg \n inflating: /data/test/Warren_Beatty_51.jpg \n inflating: /data/test/Warren_Buffett_00.jpg \n inflating: /data/test/Warren_Buffett_01.jpg \n inflating: /data/test/Warren_Buffett_30.jpg \n inflating: /data/test/Warren_Buffett_31.jpg \n inflating: /data/test/Warren_Buffett_40.jpg \n inflating: /data/test/Warren_Buffett_41.jpg \n inflating: /data/test/Warren_Buffett_50.jpg \n inflating: /data/test/Warren_Buffett_51.jpg \n inflating: /data/test/Wayne_Allard_00.jpg \n inflating: /data/test/Wayne_Allard_01.jpg \n inflating: /data/test/Wayne_Allard_20.jpg \n inflating: /data/test/Wayne_Allard_21.jpg \n inflating: /data/test/Wayne_Allard_50.jpg \n inflating: /data/test/Wayne_Allard_51.jpg \n inflating: /data/test/Wayne_Gretzky_20.jpg \n inflating: /data/test/Wayne_Gretzky_21.jpg \n inflating: /data/test/Wayne_Gretzky_30.jpg \n inflating: /data/test/Wayne_Gretzky_31.jpg \n inflating: /data/test/Wayne_Gretzky_40.jpg \n inflating: /data/test/Wayne_Gretzky_41.jpg \n inflating: /data/test/Wayne_Newton_10.jpg \n inflating: /data/test/Wayne_Newton_11.jpg \n inflating: /data/test/Wayne_Newton_20.jpg \n inflating: /data/test/Wayne_Newton_21.jpg \n inflating: /data/test/Wayne_Newton_40.jpg \n inflating: /data/test/Wayne_Newton_41.jpg \n inflating: /data/test/Wes_Craven_00.jpg \n inflating: /data/test/Wes_Craven_01.jpg \n inflating: /data/test/Wes_Craven_20.jpg \n inflating: /data/test/Wes_Craven_21.jpg \n inflating: /data/test/Wes_Craven_30.jpg \n inflating: /data/test/Wes_Craven_31.jpg \n inflating: /data/test/Wes_Craven_50.jpg \n inflating: /data/test/Wes_Craven_51.jpg \n inflating: /data/test/Wesley_Clark_00.jpg \n inflating: /data/test/Wesley_Clark_01.jpg \n inflating: /data/test/Wesley_Clark_20.jpg \n inflating: /data/test/Wesley_Clark_21.jpg \n inflating: /data/test/Wesley_Clark_30.jpg \n inflating: /data/test/Wesley_Clark_31.jpg \n inflating: /data/test/Wesley_Clark_40.jpg \n inflating: /data/test/Wesley_Clark_41.jpg \n inflating: /data/test/Whoopi_Goldberg_00.jpg \n inflating: /data/test/Whoopi_Goldberg_01.jpg \n inflating: /data/test/Whoopi_Goldberg_40.jpg \n inflating: /data/test/Whoopi_Goldberg_41.jpg \n inflating: /data/test/Whoopi_Goldberg_50.jpg \n inflating: /data/test/Whoopi_Goldberg_51.jpg \n inflating: /data/test/William_Delahunt_00.jpg \n inflating: /data/test/William_Delahunt_01.jpg \n inflating: /data/test/William_Delahunt_10.jpg \n inflating: /data/test/William_Delahunt_11.jpg \n inflating: /data/test/William_Delahunt_20.jpg \n inflating: /data/test/William_Delahunt_21.jpg \n inflating: /data/test/William_Donaldson_00.jpg \n inflating: /data/test/William_Donaldson_01.jpg \n inflating: /data/test/William_Donaldson_10.jpg \n inflating: /data/test/William_Donaldson_11.jpg \n inflating: /data/test/William_Donaldson_50.jpg \n inflating: /data/test/William_Donaldson_51.jpg \n inflating: /data/test/William_McDonough_00.jpg \n inflating: /data/test/William_McDonough_01.jpg \n inflating: /data/test/William_McDonough_20.jpg \n inflating: /data/test/William_McDonough_21.jpg \n inflating: /data/test/William_McDonough_30.jpg \n inflating: /data/test/William_McDonough_31.jpg \n inflating: /data/test/William_McDonough_40.jpg \n inflating: /data/test/William_McDonough_41.jpg \n inflating: /data/test/Yang_Jianli_00.jpg \n inflating: /data/test/Yang_Jianli_01.jpg \n inflating: /data/test/Yang_Jianli_10.jpg \n inflating: /data/test/Yang_Jianli_11.jpg \n inflating: /data/test/Yang_Jianli_30.jpg \n inflating: /data/test/Yang_Jianli_31.jpg \n inflating: /data/test/Yuri_Fedotov_20.jpg \n inflating: /data/test/Yuri_Fedotov_21.jpg \n inflating: /data/test/Yuri_Fedotov_30.jpg \n inflating: /data/test/Yuri_Fedotov_31.jpg \n inflating: /data/test/Yuri_Fedotov_40.jpg \n inflating: /data/test/Yuri_Fedotov_41.jpg \n inflating: /data/test/Zhang_Ziyi_10.jpg \n inflating: /data/test/Zhang_Ziyi_11.jpg \n inflating: /data/test/Zhang_Ziyi_20.jpg \n inflating: /data/test/Zhang_Ziyi_21.jpg \n inflating: /data/test/Zhang_Ziyi_40.jpg \n inflating: /data/test/Zhang_Ziyi_41.jpg \n inflating: /data/test/Zhong_Nanshan_00.jpg \n inflating: /data/test/Zhong_Nanshan_01.jpg \n inflating: /data/test/Zhong_Nanshan_10.jpg \n inflating: /data/test/Zhong_Nanshan_11.jpg \n inflating: /data/test/Zhong_Nanshan_50.jpg \n inflating: /data/test/Zhong_Nanshan_51.jpg \n inflating: /data/test_frames_keypoints.csv \n creating: /data/training/\n inflating: /data/training/Abdel_Aziz_Al-Hakim_00.jpg \n inflating: /data/training/Abdel_Aziz_Al-Hakim_01.jpg \n inflating: /data/training/Abdel_Aziz_Al-Hakim_02.jpg \n inflating: /data/training/Abdel_Aziz_Al-Hakim_10.jpg \n inflating: /data/training/Abdel_Aziz_Al-Hakim_11.jpg \n inflating: /data/training/Abdel_Aziz_Al-Hakim_12.jpg \n inflating: /data/training/Abdel_Aziz_Al-Hakim_40.jpg \n inflating: /data/training/Abdel_Aziz_Al-Hakim_41.jpg \n inflating: /data/training/Abdel_Aziz_Al-Hakim_42.jpg \n inflating: /data/training/Abdullah_Gul_10.jpg \n inflating: /data/training/Abdullah_Gul_11.jpg \n inflating: /data/training/Abdullah_Gul_12.jpg \n inflating: /data/training/Abdullah_Gul_30.jpg \n inflating: /data/training/Abdullah_Gul_31.jpg \n inflating: /data/training/Abdullah_Gul_32.jpg \n inflating: /data/training/Abdullah_Gul_50.jpg \n inflating: /data/training/Abdullah_Gul_51.jpg \n inflating: /data/training/Abdullah_Gul_52.jpg \n inflating: /data/training/Adam_Sandler_00.jpg \n inflating: /data/training/Adam_Sandler_01.jpg \n inflating: /data/training/Adam_Sandler_02.jpg \n inflating: /data/training/Adam_Sandler_10.jpg \n inflating: /data/training/Adam_Sandler_11.jpg \n inflating: /data/training/Adam_Sandler_12.jpg \n inflating: /data/training/Adam_Sandler_40.jpg \n inflating: /data/training/Adam_Sandler_41.jpg \n inflating: /data/training/Adam_Sandler_42.jpg \n inflating: /data/training/Adrian_Nastase_10.jpg \n inflating: /data/training/Adrian_Nastase_11.jpg \n inflating: /data/training/Adrian_Nastase_12.jpg \n inflating: /data/training/Adrian_Nastase_40.jpg \n inflating: /data/training/Adrian_Nastase_41.jpg \n inflating: /data/training/Adrian_Nastase_42.jpg \n inflating: /data/training/Adrian_Nastase_50.jpg \n inflating: /data/training/Adrian_Nastase_51.jpg \n inflating: /data/training/Adrian_Nastase_52.jpg \n inflating: /data/training/Agbani_Darego_00.jpg \n inflating: /data/training/Agbani_Darego_01.jpg \n inflating: /data/training/Agbani_Darego_02.jpg \n inflating: /data/training/Agbani_Darego_20.jpg \n inflating: /data/training/Agbani_Darego_21.jpg \n inflating: /data/training/Agbani_Darego_22.jpg \n inflating: /data/training/Agbani_Darego_40.jpg \n inflating: /data/training/Agbani_Darego_41.jpg \n inflating: /data/training/Agbani_Darego_42.jpg \n inflating: /data/training/Agbani_Darego_50.jpg \n inflating: /data/training/Agbani_Darego_51.jpg \n inflating: /data/training/Agbani_Darego_52.jpg \n inflating: /data/training/Agnes_Bruckner_00.jpg \n inflating: /data/training/Agnes_Bruckner_01.jpg \n inflating: /data/training/Agnes_Bruckner_02.jpg \n inflating: /data/training/Agnes_Bruckner_10.jpg \n inflating: /data/training/Agnes_Bruckner_11.jpg \n inflating: /data/training/Agnes_Bruckner_12.jpg \n inflating: /data/training/Agnes_Bruckner_20.jpg \n inflating: /data/training/Agnes_Bruckner_21.jpg \n inflating: /data/training/Agnes_Bruckner_22.jpg \n inflating: /data/training/Agnes_Bruckner_40.jpg \n inflating: /data/training/Agnes_Bruckner_41.jpg \n inflating: /data/training/Agnes_Bruckner_42.jpg \n inflating: /data/training/Ahmad_Masood_00.jpg \n inflating: /data/training/Ahmad_Masood_01.jpg \n inflating: /data/training/Ahmad_Masood_02.jpg \n inflating: /data/training/Ahmad_Masood_30.jpg \n inflating: /data/training/Ahmad_Masood_31.jpg \n inflating: /data/training/Ahmad_Masood_32.jpg \n inflating: /data/training/Ahmad_Masood_40.jpg \n inflating: /data/training/Ahmad_Masood_41.jpg \n inflating: /data/training/Ahmad_Masood_42.jpg \n inflating: /data/training/Ahmed_Ahmed_00.jpg \n inflating: /data/training/Ahmed_Ahmed_01.jpg \n inflating: /data/training/Ahmed_Ahmed_02.jpg \n inflating: /data/training/Ahmed_Ahmed_10.jpg \n inflating: /data/training/Ahmed_Ahmed_11.jpg \n inflating: /data/training/Ahmed_Ahmed_12.jpg \n inflating: /data/training/Ahmed_Ahmed_40.jpg \n inflating: /data/training/Ahmed_Ahmed_41.jpg \n inflating: /data/training/Ahmed_Ahmed_42.jpg \n inflating: /data/training/Ahmed_Ahmed_50.jpg \n inflating: /data/training/Ahmed_Ahmed_51.jpg \n inflating: /data/training/Ahmed_Ahmed_52.jpg \n inflating: /data/training/Aidan_Quinn_00.jpg \n inflating: /data/training/Aidan_Quinn_01.jpg \n inflating: /data/training/Aidan_Quinn_02.jpg \n inflating: /data/training/Aidan_Quinn_10.jpg \n inflating: /data/training/Aidan_Quinn_11.jpg \n inflating: /data/training/Aidan_Quinn_12.jpg \n inflating: /data/training/Aidan_Quinn_20.jpg \n inflating: /data/training/Aidan_Quinn_21.jpg \n inflating: /data/training/Aidan_Quinn_22.jpg \n inflating: /data/training/Aidan_Quinn_30.jpg \n inflating: /data/training/Aidan_Quinn_31.jpg \n inflating: /data/training/Aidan_Quinn_32.jpg \n inflating: /data/training/Aishwarya_Rai_00.jpg \n inflating: /data/training/Aishwarya_Rai_01.jpg \n inflating: /data/training/Aishwarya_Rai_02.jpg \n inflating: /data/training/Aishwarya_Rai_10.jpg \n inflating: /data/training/Aishwarya_Rai_11.jpg \n inflating: /data/training/Aishwarya_Rai_12.jpg \n inflating: /data/training/Aishwarya_Rai_40.jpg \n inflating: /data/training/Aishwarya_Rai_41.jpg \n inflating: /data/training/Aishwarya_Rai_42.jpg \n inflating: /data/training/Aishwarya_Rai_50.jpg \n inflating: /data/training/Aishwarya_Rai_51.jpg \n inflating: /data/training/Aishwarya_Rai_52.jpg \n inflating: /data/training/Albert_Brooks_00.jpg \n inflating: /data/training/Albert_Brooks_01.jpg \n inflating: /data/training/Albert_Brooks_02.jpg \n inflating: /data/training/Albert_Brooks_10.jpg \n inflating: /data/training/Albert_Brooks_11.jpg \n inflating: /data/training/Albert_Brooks_12.jpg \n inflating: /data/training/Albert_Brooks_30.jpg \n inflating: /data/training/Albert_Brooks_31.jpg \n inflating: /data/training/Albert_Brooks_32.jpg \n inflating: /data/training/Alejandro_Toledo_10.jpg \n inflating: /data/training/Alejandro_Toledo_11.jpg \n inflating: /data/training/Alejandro_Toledo_12.jpg \n inflating: /data/training/Alejandro_Toledo_30.jpg \n inflating: /data/training/Alejandro_Toledo_31.jpg \n inflating: /data/training/Alejandro_Toledo_32.jpg \n inflating: /data/training/Alejandro_Toledo_50.jpg \n inflating: /data/training/Alejandro_Toledo_51.jpg \n inflating: /data/training/Alejandro_Toledo_52.jpg \n inflating: /data/training/Aleksander_Kwasniewski_00.jpg \n inflating: /data/training/Aleksander_Kwasniewski_01.jpg \n inflating: /data/training/Aleksander_Kwasniewski_02.jpg \n inflating: /data/training/Aleksander_Kwasniewski_10.jpg \n inflating: /data/training/Aleksander_Kwasniewski_11.jpg \n inflating: /data/training/Aleksander_Kwasniewski_12.jpg \n inflating: /data/training/Aleksander_Kwasniewski_20.jpg \n inflating: /data/training/Aleksander_Kwasniewski_21.jpg \n inflating: /data/training/Aleksander_Kwasniewski_22.jpg \n inflating: /data/training/Aleksander_Kwasniewski_30.jpg \n inflating: /data/training/Aleksander_Kwasniewski_31.jpg \n inflating: /data/training/Aleksander_Kwasniewski_32.jpg \n inflating: /data/training/Alex_Ferguson_00.jpg \n inflating: /data/training/Alex_Ferguson_01.jpg \n inflating: /data/training/Alex_Ferguson_02.jpg \n inflating: /data/training/Alex_Ferguson_10.jpg \n inflating: /data/training/Alex_Ferguson_11.jpg \n inflating: /data/training/Alex_Ferguson_12.jpg \n inflating: /data/training/Alex_Ferguson_50.jpg \n inflating: /data/training/Alex_Ferguson_51.jpg \n inflating: /data/training/Alex_Ferguson_52.jpg \n inflating: /data/training/Alexandra_Pelosi_00.jpg \n inflating: /data/training/Alexandra_Pelosi_01.jpg \n inflating: /data/training/Alexandra_Pelosi_02.jpg \n inflating: /data/training/Alexandra_Pelosi_10.jpg \n inflating: /data/training/Alexandra_Pelosi_11.jpg \n inflating: /data/training/Alexandra_Pelosi_12.jpg \n inflating: /data/training/Alexandra_Pelosi_30.jpg \n inflating: /data/training/Alexandra_Pelosi_31.jpg \n inflating: /data/training/Alexandra_Pelosi_32.jpg \n inflating: /data/training/Alfredo_di_Stefano_00.jpg \n inflating: /data/training/Alfredo_di_Stefano_01.jpg \n inflating: /data/training/Alfredo_di_Stefano_02.jpg \n inflating: /data/training/Alfredo_di_Stefano_20.jpg \n inflating: /data/training/Alfredo_di_Stefano_21.jpg \n inflating: /data/training/Alfredo_di_Stefano_22.jpg \n inflating: /data/training/Alfredo_di_Stefano_50.jpg \n inflating: /data/training/Alfredo_di_Stefano_51.jpg \n inflating: /data/training/Alfredo_di_Stefano_52.jpg \n inflating: /data/training/Ali_Abbas_20.jpg \n inflating: /data/training/Ali_Abbas_21.jpg \n inflating: /data/training/Ali_Abbas_22.jpg \n inflating: /data/training/Ali_Abbas_30.jpg \n inflating: /data/training/Ali_Abbas_31.jpg \n inflating: /data/training/Ali_Abbas_32.jpg \n inflating: /data/training/Ali_Abbas_40.jpg \n inflating: /data/training/Ali_Abbas_41.jpg \n inflating: /data/training/Ali_Abbas_42.jpg \n inflating: /data/training/Ali_Abbas_50.jpg \n inflating: /data/training/Ali_Abbas_51.jpg \n inflating: /data/training/Ali_Abbas_52.jpg \n inflating: /data/training/Alicia_Silverstone_00.jpg \n inflating: /data/training/Alicia_Silverstone_01.jpg \n inflating: /data/training/Alicia_Silverstone_02.jpg \n inflating: /data/training/Alicia_Silverstone_10.jpg \n inflating: /data/training/Alicia_Silverstone_11.jpg \n inflating: /data/training/Alicia_Silverstone_12.jpg \n inflating: /data/training/Alicia_Silverstone_20.jpg \n inflating: /data/training/Alicia_Silverstone_21.jpg \n inflating: /data/training/Alicia_Silverstone_22.jpg \n inflating: /data/training/Alicia_Silverstone_50.jpg \n inflating: /data/training/Alicia_Silverstone_51.jpg \n inflating: /data/training/Alicia_Silverstone_52.jpg \n inflating: /data/training/Alma_Powell_00.jpg \n inflating: /data/training/Alma_Powell_01.jpg \n inflating: /data/training/Alma_Powell_02.jpg \n inflating: /data/training/Alma_Powell_10.jpg \n inflating: /data/training/Alma_Powell_11.jpg \n inflating: /data/training/Alma_Powell_12.jpg \n inflating: /data/training/Alma_Powell_40.jpg \n inflating: /data/training/Alma_Powell_41.jpg \n inflating: /data/training/Alma_Powell_42.jpg \n inflating: /data/training/Alma_Powell_50.jpg \n inflating: /data/training/Alma_Powell_51.jpg \n inflating: /data/training/Alma_Powell_52.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_00.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_01.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_02.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_10.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_11.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_12.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_20.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_21.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_22.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_30.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_31.jpg \n inflating: /data/training/Alvaro_Silva_Calderon_32.jpg \n inflating: /data/training/Amelia_Vega_10.jpg \n inflating: /data/training/Amelia_Vega_11.jpg \n inflating: /data/training/Amelia_Vega_12.jpg \n inflating: /data/training/Amelia_Vega_20.jpg \n inflating: /data/training/Amelia_Vega_21.jpg \n inflating: /data/training/Amelia_Vega_22.jpg \n inflating: /data/training/Amelia_Vega_30.jpg \n inflating: /data/training/Amelia_Vega_31.jpg \n inflating: /data/training/Amelia_Vega_32.jpg \n inflating: /data/training/Amelia_Vega_40.jpg \n inflating: /data/training/Amelia_Vega_41.jpg \n inflating: /data/training/Amelia_Vega_42.jpg \n inflating: /data/training/Amy_Brenneman_10.jpg \n inflating: /data/training/Amy_Brenneman_11.jpg \n inflating: /data/training/Amy_Brenneman_12.jpg \n inflating: /data/training/Amy_Brenneman_30.jpg \n inflating: /data/training/Amy_Brenneman_31.jpg \n inflating: /data/training/Amy_Brenneman_32.jpg \n inflating: /data/training/Amy_Brenneman_50.jpg \n inflating: /data/training/Amy_Brenneman_51.jpg \n inflating: /data/training/Amy_Brenneman_52.jpg \n inflating: /data/training/Andrea_Bocelli_10.jpg \n inflating: /data/training/Andrea_Bocelli_11.jpg \n inflating: /data/training/Andrea_Bocelli_12.jpg \n inflating: /data/training/Andrea_Bocelli_20.jpg \n inflating: /data/training/Andrea_Bocelli_21.jpg \n inflating: /data/training/Andrea_Bocelli_22.jpg \n inflating: /data/training/Andrea_Bocelli_30.jpg \n inflating: /data/training/Andrea_Bocelli_31.jpg \n inflating: /data/training/Andrea_Bocelli_32.jpg \n inflating: /data/training/Andy_Roddick_20.jpg \n inflating: /data/training/Andy_Roddick_21.jpg \n inflating: /data/training/Andy_Roddick_22.jpg \n inflating: /data/training/Andy_Roddick_40.jpg \n inflating: /data/training/Andy_Roddick_41.jpg \n inflating: /data/training/Andy_Roddick_42.jpg \n inflating: /data/training/Andy_Roddick_50.jpg \n inflating: /data/training/Andy_Roddick_51.jpg \n inflating: /data/training/Andy_Roddick_52.jpg \n inflating: /data/training/Andy_Rooney_10.jpg \n inflating: /data/training/Andy_Rooney_11.jpg \n inflating: /data/training/Andy_Rooney_12.jpg \n inflating: /data/training/Andy_Rooney_20.jpg \n inflating: /data/training/Andy_Rooney_21.jpg \n inflating: /data/training/Andy_Rooney_22.jpg \n inflating: /data/training/Andy_Rooney_50.jpg \n inflating: /data/training/Andy_Rooney_51.jpg \n inflating: /data/training/Andy_Rooney_52.jpg \n inflating: /data/training/Angel_Lockward_30.jpg \n inflating: /data/training/Angel_Lockward_31.jpg \n inflating: /data/training/Angel_Lockward_32.jpg \n inflating: /data/training/Angel_Lockward_40.jpg \n inflating: /data/training/Angel_Lockward_41.jpg \n inflating: /data/training/Angel_Lockward_42.jpg \n inflating: /data/training/Angel_Lockward_50.jpg \n inflating: /data/training/Angel_Lockward_51.jpg \n inflating: /data/training/Angel_Lockward_52.jpg \n inflating: /data/training/Angela_Bassett_20.jpg \n inflating: /data/training/Angela_Bassett_21.jpg \n inflating: /data/training/Angela_Bassett_22.jpg \n inflating: /data/training/Angela_Bassett_30.jpg \n inflating: /data/training/Angela_Bassett_31.jpg \n inflating: /data/training/Angela_Bassett_32.jpg \n inflating: /data/training/Angela_Bassett_40.jpg \n inflating: /data/training/Angela_Bassett_41.jpg \n inflating: /data/training/Angela_Bassett_42.jpg \n inflating: /data/training/Angelo_Reyes_20.jpg \n inflating: /data/training/Angelo_Reyes_21.jpg \n inflating: /data/training/Angelo_Reyes_22.jpg \n inflating: /data/training/Angelo_Reyes_30.jpg \n inflating: /data/training/Angelo_Reyes_31.jpg \n inflating: /data/training/Angelo_Reyes_32.jpg \n inflating: /data/training/Angelo_Reyes_50.jpg \n inflating: /data/training/Angelo_Reyes_51.jpg \n inflating: /data/training/Angelo_Reyes_52.jpg \n inflating: /data/training/Baburam_Bhattari_00.jpg \n inflating: /data/training/Baburam_Bhattari_01.jpg \n inflating: /data/training/Baburam_Bhattari_02.jpg \n inflating: /data/training/Baburam_Bhattari_20.jpg \n inflating: /data/training/Baburam_Bhattari_21.jpg \n inflating: /data/training/Baburam_Bhattari_22.jpg \n inflating: /data/training/Baburam_Bhattari_30.jpg \n inflating: /data/training/Baburam_Bhattari_31.jpg \n inflating: /data/training/Baburam_Bhattari_32.jpg \n inflating: /data/training/Barbara_Bodine_00.jpg \n inflating: /data/training/Barbara_Bodine_01.jpg \n inflating: /data/training/Barbara_Bodine_02.jpg \n inflating: /data/training/Barbara_Bodine_20.jpg \n inflating: /data/training/Barbara_Bodine_21.jpg \n inflating: /data/training/Barbara_Bodine_22.jpg \n inflating: /data/training/Barbara_Bodine_40.jpg \n inflating: /data/training/Barbara_Bodine_41.jpg \n inflating: /data/training/Barbara_Bodine_42.jpg \n inflating: /data/training/Barbara_Bodine_50.jpg \n inflating: /data/training/Barbara_Bodine_51.jpg \n inflating: /data/training/Barbara_Bodine_52.jpg \n inflating: /data/training/Barbara_Boxer_10.jpg \n inflating: /data/training/Barbara_Boxer_11.jpg \n inflating: /data/training/Barbara_Boxer_12.jpg \n inflating: /data/training/Barbara_Boxer_40.jpg \n inflating: /data/training/Barbara_Boxer_41.jpg \n inflating: /data/training/Barbara_Boxer_42.jpg \n inflating: /data/training/Barbara_Boxer_50.jpg \n inflating: /data/training/Barbara_Boxer_51.jpg \n inflating: /data/training/Barbara_Boxer_52.jpg \n inflating: /data/training/Barbara_Walters_00.jpg \n inflating: /data/training/Barbara_Walters_01.jpg \n inflating: /data/training/Barbara_Walters_02.jpg \n inflating: /data/training/Barbara_Walters_20.jpg \n inflating: /data/training/Barbara_Walters_21.jpg \n inflating: /data/training/Barbara_Walters_22.jpg \n inflating: /data/training/Barbara_Walters_40.jpg \n inflating: /data/training/Barbara_Walters_41.jpg \n inflating: /data/training/Barbara_Walters_42.jpg \n inflating: /data/training/Barbara_Walters_50.jpg \n inflating: /data/training/Barbara_Walters_51.jpg \n inflating: /data/training/Barbara_Walters_52.jpg \n inflating: /data/training/Barry_Alvarez_00.jpg \n inflating: /data/training/Barry_Alvarez_01.jpg \n inflating: /data/training/Barry_Alvarez_02.jpg \n inflating: /data/training/Barry_Alvarez_10.jpg \n inflating: /data/training/Barry_Alvarez_11.jpg \n inflating: /data/training/Barry_Alvarez_12.jpg \n inflating: /data/training/Barry_Alvarez_20.jpg \n inflating: /data/training/Barry_Alvarez_21.jpg \n inflating: /data/training/Barry_Alvarez_22.jpg \n inflating: /data/training/Barry_Alvarez_30.jpg \n inflating: /data/training/Barry_Alvarez_31.jpg \n inflating: /data/training/Barry_Alvarez_32.jpg \n inflating: /data/training/Ben_Kingsley_10.jpg \n inflating: /data/training/Ben_Kingsley_11.jpg \n inflating: /data/training/Ben_Kingsley_12.jpg \n inflating: /data/training/Ben_Kingsley_20.jpg \n inflating: /data/training/Ben_Kingsley_21.jpg \n inflating: /data/training/Ben_Kingsley_22.jpg \n inflating: /data/training/Ben_Kingsley_50.jpg \n inflating: /data/training/Ben_Kingsley_51.jpg \n inflating: /data/training/Ben_Kingsley_52.jpg \n inflating: /data/training/Ben_Stein_10.jpg \n inflating: /data/training/Ben_Stein_11.jpg \n inflating: /data/training/Ben_Stein_12.jpg \n inflating: /data/training/Ben_Stein_30.jpg \n inflating: /data/training/Ben_Stein_31.jpg \n inflating: /data/training/Ben_Stein_32.jpg \n inflating: /data/training/Ben_Stein_40.jpg \n inflating: /data/training/Ben_Stein_41.jpg \n inflating: /data/training/Ben_Stein_42.jpg \n inflating: /data/training/Ben_Stein_50.jpg \n inflating: /data/training/Ben_Stein_51.jpg \n inflating: /data/training/Ben_Stein_52.jpg \n inflating: /data/training/Benedita_da_Silva_10.jpg \n inflating: /data/training/Benedita_da_Silva_11.jpg \n inflating: /data/training/Benedita_da_Silva_12.jpg \n inflating: /data/training/Benedita_da_Silva_20.jpg \n inflating: /data/training/Benedita_da_Silva_21.jpg \n inflating: /data/training/Benedita_da_Silva_22.jpg \n inflating: /data/training/Benedita_da_Silva_50.jpg \n inflating: /data/training/Benedita_da_Silva_51.jpg \n inflating: /data/training/Benedita_da_Silva_52.jpg \n inflating: /data/training/Benjamin_McKenzie_30.jpg \n inflating: /data/training/Benjamin_McKenzie_31.jpg \n inflating: /data/training/Benjamin_McKenzie_32.jpg \n inflating: /data/training/Benjamin_McKenzie_40.jpg \n inflating: /data/training/Benjamin_McKenzie_41.jpg \n inflating: /data/training/Benjamin_McKenzie_42.jpg \n inflating: /data/training/Benjamin_McKenzie_50.jpg \n inflating: /data/training/Benjamin_McKenzie_51.jpg \n inflating: /data/training/Benjamin_McKenzie_52.jpg \n inflating: /data/training/Benjamin_Netanyahu_00.jpg \n inflating: /data/training/Benjamin_Netanyahu_01.jpg \n inflating: /data/training/Benjamin_Netanyahu_02.jpg \n inflating: /data/training/Benjamin_Netanyahu_10.jpg \n inflating: /data/training/Benjamin_Netanyahu_11.jpg \n inflating: /data/training/Benjamin_Netanyahu_12.jpg \n inflating: /data/training/Benjamin_Netanyahu_30.jpg \n inflating: /data/training/Benjamin_Netanyahu_31.jpg \n inflating: /data/training/Benjamin_Netanyahu_32.jpg \n inflating: /data/training/Benjamin_Netanyahu_40.jpg \n inflating: /data/training/Benjamin_Netanyahu_41.jpg \n inflating: /data/training/Benjamin_Netanyahu_42.jpg \n inflating: /data/training/Beyonce_Knowles_00.jpg \n inflating: /data/training/Beyonce_Knowles_01.jpg \n inflating: /data/training/Beyonce_Knowles_02.jpg \n inflating: /data/training/Beyonce_Knowles_30.jpg \n inflating: /data/training/Beyonce_Knowles_31.jpg \n inflating: /data/training/Beyonce_Knowles_32.jpg \n inflating: /data/training/Beyonce_Knowles_50.jpg \n inflating: /data/training/Beyonce_Knowles_51.jpg \n inflating: /data/training/Beyonce_Knowles_52.jpg \n inflating: /data/training/Bianca_Jagger_20.jpg \n inflating: /data/training/Bianca_Jagger_21.jpg \n inflating: /data/training/Bianca_Jagger_22.jpg \n inflating: /data/training/Bianca_Jagger_30.jpg \n inflating: /data/training/Bianca_Jagger_31.jpg \n inflating: /data/training/Bianca_Jagger_32.jpg \n inflating: /data/training/Bianca_Jagger_40.jpg \n inflating: /data/training/Bianca_Jagger_41.jpg \n inflating: /data/training/Bianca_Jagger_42.jpg \n inflating: /data/training/Biljana_Plavsic_00.jpg \n inflating: /data/training/Biljana_Plavsic_01.jpg \n inflating: /data/training/Biljana_Plavsic_02.jpg \n inflating: /data/training/Biljana_Plavsic_10.jpg \n inflating: /data/training/Biljana_Plavsic_11.jpg \n inflating: /data/training/Biljana_Plavsic_12.jpg \n inflating: /data/training/Biljana_Plavsic_30.jpg \n inflating: /data/training/Biljana_Plavsic_31.jpg \n inflating: /data/training/Biljana_Plavsic_32.jpg \n inflating: /data/training/Bill_Bradley_10.jpg \n inflating: /data/training/Bill_Bradley_11.jpg \n inflating: /data/training/Bill_Bradley_12.jpg \n inflating: /data/training/Bill_Bradley_30.jpg \n inflating: /data/training/Bill_Bradley_31.jpg \n inflating: /data/training/Bill_Bradley_32.jpg \n inflating: /data/training/Bill_Bradley_50.jpg \n inflating: /data/training/Bill_Bradley_51.jpg \n inflating: /data/training/Bill_Bradley_52.jpg \n inflating: /data/training/Bill_Clinton_00.jpg \n inflating: /data/training/Bill_Clinton_01.jpg \n inflating: /data/training/Bill_Clinton_02.jpg \n inflating: /data/training/Bill_Clinton_10.jpg \n inflating: /data/training/Bill_Clinton_11.jpg \n inflating: /data/training/Bill_Clinton_12.jpg \n inflating: /data/training/Bill_Clinton_50.jpg \n inflating: /data/training/Bill_Clinton_51.jpg \n inflating: /data/training/Bill_Clinton_52.jpg \n inflating: /data/training/Bill_Frist_00.jpg \n inflating: /data/training/Bill_Frist_01.jpg \n inflating: /data/training/Bill_Frist_02.jpg \n inflating: /data/training/Bill_Frist_10.jpg \n inflating: /data/training/Bill_Frist_11.jpg \n inflating: /data/training/Bill_Frist_12.jpg \n inflating: /data/training/Bill_Frist_20.jpg \n inflating: /data/training/Bill_Frist_21.jpg \n inflating: /data/training/Bill_Frist_22.jpg \n inflating: /data/training/Cameron_Diaz_20.jpg \n inflating: /data/training/Cameron_Diaz_21.jpg \n inflating: /data/training/Cameron_Diaz_22.jpg \n inflating: /data/training/Cameron_Diaz_40.jpg \n inflating: /data/training/Cameron_Diaz_41.jpg \n inflating: /data/training/Cameron_Diaz_42.jpg \n inflating: /data/training/Cameron_Diaz_50.jpg \n inflating: /data/training/Cameron_Diaz_51.jpg \n inflating: /data/training/Cameron_Diaz_52.jpg \n inflating: /data/training/Carey_Lowell_00.jpg \n inflating: /data/training/Carey_Lowell_01.jpg \n inflating: /data/training/Carey_Lowell_02.jpg \n inflating: /data/training/Carey_Lowell_20.jpg \n inflating: /data/training/Carey_Lowell_21.jpg \n inflating: /data/training/Carey_Lowell_22.jpg \n inflating: /data/training/Carey_Lowell_30.jpg \n inflating: /data/training/Carey_Lowell_31.jpg \n inflating: /data/training/Carey_Lowell_32.jpg \n inflating: /data/training/Carla_Gugino_10.jpg \n inflating: /data/training/Carla_Gugino_11.jpg \n inflating: /data/training/Carla_Gugino_12.jpg \n inflating: /data/training/Carla_Gugino_30.jpg \n inflating: /data/training/Carla_Gugino_31.jpg \n inflating: /data/training/Carla_Gugino_32.jpg \n inflating: /data/training/Carla_Gugino_50.jpg \n inflating: /data/training/Carla_Gugino_51.jpg \n inflating: /data/training/Carla_Gugino_52.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_00.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_01.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_02.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_10.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_11.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_12.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_20.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_21.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_22.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_50.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_51.jpg \n inflating: /data/training/Carlo_Azeglio_Ciampi_52.jpg \n inflating: /data/training/Carlos_Ghosn_10.jpg \n inflating: /data/training/Carlos_Ghosn_11.jpg \n inflating: /data/training/Carlos_Ghosn_12.jpg \n inflating: /data/training/Carlos_Ghosn_20.jpg \n inflating: /data/training/Carlos_Ghosn_21.jpg \n inflating: /data/training/Carlos_Ghosn_22.jpg \n inflating: /data/training/Carlos_Ghosn_40.jpg \n inflating: /data/training/Carlos_Ghosn_41.jpg \n inflating: /data/training/Carlos_Ghosn_42.jpg \n inflating: /data/training/Carlos_Iturgaitz_10.jpg \n inflating: /data/training/Carlos_Iturgaitz_11.jpg \n inflating: /data/training/Carlos_Iturgaitz_12.jpg \n inflating: /data/training/Carlos_Iturgaitz_20.jpg \n inflating: /data/training/Carlos_Iturgaitz_21.jpg \n inflating: /data/training/Carlos_Iturgaitz_22.jpg \n inflating: /data/training/Carlos_Iturgaitz_30.jpg \n inflating: /data/training/Carlos_Iturgaitz_31.jpg \n inflating: /data/training/Carlos_Iturgaitz_32.jpg \n inflating: /data/training/Carlos_Iturgaitz_50.jpg \n inflating: /data/training/Carlos_Iturgaitz_51.jpg \n inflating: /data/training/Carlos_Iturgaitz_52.jpg \n inflating: /data/training/Carlos_Menem_00.jpg \n inflating: /data/training/Carlos_Menem_01.jpg \n inflating: /data/training/Carlos_Menem_02.jpg \n inflating: /data/training/Carlos_Menem_20.jpg \n inflating: /data/training/Carlos_Menem_21.jpg \n inflating: /data/training/Carlos_Menem_22.jpg \n inflating: /data/training/Carlos_Menem_30.jpg \n inflating: /data/training/Carlos_Menem_31.jpg \n inflating: /data/training/Carlos_Menem_32.jpg \n inflating: /data/training/Carlos_Queiroz_00.jpg \n inflating: /data/training/Carlos_Queiroz_01.jpg \n inflating: /data/training/Carlos_Queiroz_02.jpg \n inflating: /data/training/Carlos_Queiroz_10.jpg \n inflating: /data/training/Carlos_Queiroz_11.jpg \n inflating: /data/training/Carlos_Queiroz_12.jpg \n inflating: /data/training/Carlos_Queiroz_50.jpg \n inflating: /data/training/Carlos_Queiroz_51.jpg \n inflating: /data/training/Carlos_Queiroz_52.jpg \n inflating: /data/training/Carrie-Anne_Moss_00.jpg \n inflating: /data/training/Carrie-Anne_Moss_01.jpg \n inflating: /data/training/Carrie-Anne_Moss_02.jpg \n inflating: /data/training/Carrie-Anne_Moss_10.jpg \n inflating: /data/training/Carrie-Anne_Moss_11.jpg \n inflating: /data/training/Carrie-Anne_Moss_12.jpg \n inflating: /data/training/Carrie-Anne_Moss_50.jpg \n inflating: /data/training/Carrie-Anne_Moss_51.jpg \n inflating: /data/training/Carrie-Anne_Moss_52.jpg \n inflating: /data/training/Catriona_Le_May_Doan_10.jpg \n inflating: /data/training/Catriona_Le_May_Doan_11.jpg \n inflating: /data/training/Catriona_Le_May_Doan_12.jpg \n inflating: /data/training/Catriona_Le_May_Doan_30.jpg \n inflating: /data/training/Catriona_Le_May_Doan_31.jpg \n inflating: /data/training/Catriona_Le_May_Doan_32.jpg \n inflating: /data/training/Catriona_Le_May_Doan_40.jpg \n inflating: /data/training/Catriona_Le_May_Doan_41.jpg \n inflating: /data/training/Catriona_Le_May_Doan_42.jpg \n inflating: /data/training/Cecilia_Cheung_00.jpg \n inflating: /data/training/Cecilia_Cheung_01.jpg \n inflating: /data/training/Cecilia_Cheung_02.jpg \n inflating: /data/training/Cecilia_Cheung_10.jpg \n inflating: /data/training/Cecilia_Cheung_11.jpg \n inflating: /data/training/Cecilia_Cheung_12.jpg \n inflating: /data/training/Cecilia_Cheung_20.jpg \n inflating: /data/training/Cecilia_Cheung_21.jpg \n inflating: /data/training/Cecilia_Cheung_22.jpg \n inflating: /data/training/Cecilia_Cheung_50.jpg \n inflating: /data/training/Cecilia_Cheung_51.jpg \n inflating: /data/training/Cecilia_Cheung_52.jpg \n inflating: /data/training/Celso_Amorim_10.jpg \n inflating: /data/training/Celso_Amorim_11.jpg \n inflating: /data/training/Celso_Amorim_12.jpg \n inflating: /data/training/Celso_Amorim_40.jpg \n inflating: /data/training/Celso_Amorim_41.jpg \n inflating: /data/training/Celso_Amorim_42.jpg \n inflating: /data/training/Celso_Amorim_50.jpg \n inflating: /data/training/Celso_Amorim_51.jpg \n inflating: /data/training/Celso_Amorim_52.jpg \n inflating: /data/training/Celso_Lafer_00.jpg \n inflating: /data/training/Celso_Lafer_01.jpg \n inflating: /data/training/Celso_Lafer_02.jpg \n inflating: /data/training/Celso_Lafer_10.jpg \n inflating: /data/training/Celso_Lafer_11.jpg \n inflating: /data/training/Celso_Lafer_12.jpg \n inflating: /data/training/Celso_Lafer_20.jpg \n inflating: /data/training/Celso_Lafer_21.jpg \n inflating: /data/training/Celso_Lafer_22.jpg \n inflating: /data/training/Chadha_Gurinder_10.jpg \n inflating: /data/training/Chadha_Gurinder_11.jpg \n inflating: /data/training/Chadha_Gurinder_12.jpg \n inflating: /data/training/Chadha_Gurinder_20.jpg \n inflating: /data/training/Chadha_Gurinder_21.jpg \n inflating: /data/training/Chadha_Gurinder_22.jpg \n inflating: /data/training/Chadha_Gurinder_50.jpg \n inflating: /data/training/Chadha_Gurinder_51.jpg \n inflating: /data/training/Chadha_Gurinder_52.jpg \n inflating: /data/training/Charles_Bronson_00.jpg \n inflating: /data/training/Charles_Bronson_01.jpg \n inflating: /data/training/Charles_Bronson_02.jpg \n inflating: /data/training/Charles_Bronson_10.jpg \n inflating: /data/training/Charles_Bronson_11.jpg \n inflating: /data/training/Charles_Bronson_12.jpg \n inflating: /data/training/Charles_Bronson_50.jpg \n inflating: /data/training/Charles_Bronson_51.jpg \n inflating: /data/training/Charles_Bronson_52.jpg \n inflating: /data/training/Charlie_Coles_00.jpg \n inflating: /data/training/Charlie_Coles_01.jpg \n inflating: /data/training/Charlie_Coles_02.jpg \n inflating: /data/training/Charlie_Coles_10.jpg \n inflating: /data/training/Charlie_Coles_11.jpg \n inflating: /data/training/Charlie_Coles_12.jpg \n inflating: /data/training/Charlie_Coles_20.jpg \n inflating: /data/training/Charlie_Coles_21.jpg \n inflating: /data/training/Charlie_Coles_22.jpg \n inflating: /data/training/Charlize_Theron_10.jpg \n inflating: /data/training/Charlize_Theron_11.jpg \n inflating: /data/training/Charlize_Theron_12.jpg \n inflating: /data/training/Charlize_Theron_30.jpg \n inflating: /data/training/Charlize_Theron_31.jpg \n inflating: /data/training/Charlize_Theron_32.jpg \n inflating: /data/training/Charlize_Theron_50.jpg \n inflating: /data/training/Charlize_Theron_51.jpg \n inflating: /data/training/Charlize_Theron_52.jpg \n inflating: /data/training/Charlotte_Casiraghi_00.jpg \n inflating: /data/training/Charlotte_Casiraghi_01.jpg \n inflating: /data/training/Charlotte_Casiraghi_02.jpg \n inflating: /data/training/Charlotte_Casiraghi_10.jpg \n inflating: /data/training/Charlotte_Casiraghi_11.jpg \n inflating: /data/training/Charlotte_Casiraghi_12.jpg \n inflating: /data/training/Charlotte_Casiraghi_20.jpg \n inflating: /data/training/Charlotte_Casiraghi_21.jpg \n inflating: /data/training/Charlotte_Casiraghi_22.jpg \n inflating: /data/training/Charlotte_Rampling_00.jpg \n inflating: /data/training/Charlotte_Rampling_01.jpg \n inflating: /data/training/Charlotte_Rampling_02.jpg \n inflating: /data/training/Charlotte_Rampling_30.jpg \n inflating: /data/training/Charlotte_Rampling_31.jpg \n inflating: /data/training/Charlotte_Rampling_32.jpg \n inflating: /data/training/Charlotte_Rampling_40.jpg \n inflating: /data/training/Charlotte_Rampling_41.jpg \n inflating: /data/training/Charlotte_Rampling_42.jpg \n inflating: /data/training/Charlotte_Rampling_50.jpg \n inflating: /data/training/Charlotte_Rampling_51.jpg \n inflating: /data/training/Charlotte_Rampling_52.jpg \n inflating: /data/training/Cherie_Blair_00.jpg \n inflating: /data/training/Cherie_Blair_01.jpg \n inflating: /data/training/Cherie_Blair_02.jpg \n inflating: /data/training/Cherie_Blair_20.jpg \n inflating: /data/training/Cherie_Blair_21.jpg \n inflating: /data/training/Cherie_Blair_22.jpg \n inflating: /data/training/Cherie_Blair_30.jpg \n inflating: /data/training/Cherie_Blair_31.jpg \n inflating: /data/training/Cherie_Blair_32.jpg \n inflating: /data/training/Cherie_Blair_40.jpg \n inflating: /data/training/Cherie_Blair_41.jpg \n inflating: /data/training/Cherie_Blair_42.jpg \n inflating: /data/training/Chita_Rivera_00.jpg \n inflating: /data/training/Chita_Rivera_01.jpg \n inflating: /data/training/Chita_Rivera_02.jpg \n inflating: /data/training/Chita_Rivera_10.jpg \n inflating: /data/training/Chita_Rivera_11.jpg \n inflating: /data/training/Chita_Rivera_12.jpg \n inflating: /data/training/Chita_Rivera_30.jpg \n inflating: /data/training/Chita_Rivera_31.jpg \n inflating: /data/training/Chita_Rivera_32.jpg \n inflating: /data/training/Chris_Cirino_20.jpg \n inflating: /data/training/Chris_Cirino_21.jpg \n inflating: /data/training/Chris_Cirino_22.jpg \n inflating: /data/training/Chris_Cirino_30.jpg \n inflating: /data/training/Chris_Cirino_31.jpg \n inflating: /data/training/Chris_Cirino_32.jpg \n inflating: /data/training/Chris_Cirino_50.jpg \n inflating: /data/training/Chris_Cirino_51.jpg \n inflating: /data/training/Chris_Cirino_52.jpg \n inflating: /data/training/Chris_Cooper_00.jpg \n inflating: /data/training/Chris_Cooper_01.jpg \n inflating: /data/training/Chris_Cooper_02.jpg \n inflating: /data/training/Chris_Cooper_30.jpg \n inflating: /data/training/Chris_Cooper_31.jpg \n inflating: /data/training/Chris_Cooper_32.jpg \n inflating: /data/training/Chris_Cooper_40.jpg \n inflating: /data/training/Chris_Cooper_41.jpg \n inflating: /data/training/Chris_Cooper_42.jpg \n inflating: /data/training/Chris_Matthews_00.jpg \n inflating: /data/training/Chris_Matthews_01.jpg \n inflating: /data/training/Chris_Matthews_02.jpg \n inflating: /data/training/Chris_Matthews_10.jpg \n inflating: /data/training/Chris_Matthews_11.jpg \n inflating: /data/training/Chris_Matthews_12.jpg \n inflating: /data/training/Chris_Matthews_30.jpg \n inflating: /data/training/Chris_Matthews_31.jpg \n inflating: /data/training/Chris_Matthews_32.jpg \n inflating: /data/training/Chris_Matthews_50.jpg \n inflating: /data/training/Chris_Matthews_51.jpg \n inflating: /data/training/Chris_Matthews_52.jpg \n inflating: /data/training/Chris_Noth_00.jpg \n inflating: /data/training/Chris_Noth_01.jpg \n inflating: /data/training/Chris_Noth_02.jpg \n inflating: /data/training/Chris_Noth_10.jpg \n inflating: /data/training/Chris_Noth_11.jpg \n inflating: /data/training/Chris_Noth_12.jpg \n inflating: /data/training/Chris_Noth_30.jpg \n inflating: /data/training/Chris_Noth_31.jpg \n inflating: /data/training/Chris_Noth_32.jpg \n inflating: /data/training/Chris_Rock_00.jpg \n inflating: /data/training/Chris_Rock_01.jpg \n inflating: /data/training/Chris_Rock_02.jpg \n inflating: /data/training/Chris_Rock_10.jpg \n inflating: /data/training/Chris_Rock_11.jpg \n inflating: /data/training/Chris_Rock_12.jpg \n inflating: /data/training/Chris_Rock_20.jpg \n inflating: /data/training/Chris_Rock_21.jpg \n inflating: /data/training/Chris_Rock_22.jpg \n inflating: /data/training/Christine_Ebersole_00.jpg \n inflating: /data/training/Christine_Ebersole_01.jpg \n inflating: /data/training/Christine_Ebersole_02.jpg \n inflating: /data/training/Christine_Ebersole_20.jpg \n inflating: /data/training/Christine_Ebersole_21.jpg \n inflating: /data/training/Christine_Ebersole_22.jpg \n inflating: /data/training/Christine_Ebersole_50.jpg \n inflating: /data/training/Christine_Ebersole_51.jpg \n inflating: /data/training/Christine_Ebersole_52.jpg \n inflating: /data/training/Christopher_Amolsch_20.jpg \n inflating: /data/training/Christopher_Amolsch_21.jpg \n inflating: /data/training/Christopher_Amolsch_22.jpg \n inflating: /data/training/Christopher_Amolsch_40.jpg \n inflating: /data/training/Christopher_Amolsch_41.jpg \n inflating: /data/training/Christopher_Amolsch_42.jpg \n inflating: /data/training/Christopher_Amolsch_50.jpg \n inflating: /data/training/Christopher_Amolsch_51.jpg \n inflating: /data/training/Christopher_Amolsch_52.jpg \n inflating: /data/training/Christopher_Reeve_10.jpg \n inflating: /data/training/Christopher_Reeve_11.jpg \n inflating: /data/training/Christopher_Reeve_12.jpg \n inflating: /data/training/Christopher_Reeve_20.jpg \n inflating: /data/training/Christopher_Reeve_21.jpg \n inflating: /data/training/Christopher_Reeve_22.jpg \n inflating: /data/training/Christopher_Reeve_40.jpg \n inflating: /data/training/Christopher_Reeve_41.jpg \n inflating: /data/training/Christopher_Reeve_42.jpg \n inflating: /data/training/Christopher_Walken_00.jpg \n inflating: /data/training/Christopher_Walken_01.jpg \n inflating: /data/training/Christopher_Walken_02.jpg \n inflating: /data/training/Christopher_Walken_20.jpg \n inflating: /data/training/Christopher_Walken_21.jpg \n inflating: /data/training/Christopher_Walken_22.jpg \n inflating: /data/training/Christopher_Walken_40.jpg \n inflating: /data/training/Christopher_Walken_41.jpg \n inflating: /data/training/Christopher_Walken_42.jpg \n inflating: /data/training/Christopher_Walken_50.jpg \n inflating: /data/training/Christopher_Walken_51.jpg \n inflating: /data/training/Christopher_Walken_52.jpg \n inflating: /data/training/Chuck_Hagel_20.jpg \n inflating: /data/training/Chuck_Hagel_21.jpg \n inflating: /data/training/Chuck_Hagel_22.jpg \n inflating: /data/training/Chuck_Hagel_30.jpg \n inflating: /data/training/Chuck_Hagel_31.jpg \n inflating: /data/training/Chuck_Hagel_32.jpg \n inflating: /data/training/Chuck_Hagel_40.jpg \n inflating: /data/training/Chuck_Hagel_41.jpg \n inflating: /data/training/Chuck_Hagel_42.jpg \n inflating: /data/training/Chuck_Hagel_50.jpg \n inflating: /data/training/Chuck_Hagel_51.jpg \n inflating: /data/training/Chuck_Hagel_52.jpg \n inflating: /data/training/Chuck_Woolery_10.jpg \n inflating: /data/training/Chuck_Woolery_11.jpg \n inflating: /data/training/Chuck_Woolery_12.jpg \n inflating: /data/training/Chuck_Woolery_20.jpg \n inflating: /data/training/Chuck_Woolery_21.jpg \n inflating: /data/training/Chuck_Woolery_22.jpg \n inflating: /data/training/Chuck_Woolery_40.jpg \n inflating: /data/training/Chuck_Woolery_41.jpg \n inflating: /data/training/Chuck_Woolery_42.jpg \n inflating: /data/training/Cindy_Crawford_00.jpg \n inflating: /data/training/Cindy_Crawford_01.jpg \n inflating: /data/training/Cindy_Crawford_02.jpg \n inflating: /data/training/Cindy_Crawford_30.jpg \n inflating: /data/training/Cindy_Crawford_31.jpg \n inflating: /data/training/Cindy_Crawford_32.jpg \n inflating: /data/training/Cindy_Crawford_50.jpg \n inflating: /data/training/Cindy_Crawford_51.jpg \n inflating: /data/training/Cindy_Crawford_52.jpg \n inflating: /data/training/Cindy_Klassen_00.jpg \n inflating: /data/training/Cindy_Klassen_01.jpg \n inflating: /data/training/Cindy_Klassen_02.jpg \n inflating: /data/training/Cindy_Klassen_30.jpg \n inflating: /data/training/Cindy_Klassen_31.jpg \n inflating: /data/training/Cindy_Klassen_32.jpg \n inflating: /data/training/Cindy_Klassen_40.jpg \n inflating: /data/training/Cindy_Klassen_41.jpg \n inflating: /data/training/Cindy_Klassen_42.jpg \n inflating: /data/training/Claire_Danes_30.jpg \n inflating: /data/training/Claire_Danes_31.jpg \n inflating: /data/training/Claire_Danes_32.jpg \n inflating: /data/training/Claire_Danes_40.jpg \n inflating: /data/training/Claire_Danes_41.jpg \n inflating: /data/training/Claire_Danes_42.jpg \n inflating: /data/training/Claire_Danes_50.jpg \n inflating: /data/training/Claire_Danes_51.jpg \n inflating: /data/training/Claire_Danes_52.jpg \n inflating: /data/training/Clark_Randt_10.jpg \n inflating: /data/training/Clark_Randt_11.jpg \n inflating: /data/training/Clark_Randt_12.jpg \n inflating: /data/training/Clark_Randt_20.jpg \n inflating: /data/training/Clark_Randt_21.jpg \n inflating: /data/training/Clark_Randt_22.jpg \n inflating: /data/training/Clark_Randt_40.jpg \n inflating: /data/training/Clark_Randt_41.jpg \n inflating: /data/training/Clark_Randt_42.jpg \n inflating: /data/training/Clark_Randt_50.jpg \n inflating: /data/training/Clark_Randt_51.jpg \n inflating: /data/training/Clark_Randt_52.jpg \n inflating: /data/training/Clay_Aiken_00.jpg \n inflating: /data/training/Clay_Aiken_01.jpg \n inflating: /data/training/Clay_Aiken_02.jpg \n inflating: /data/training/Clay_Aiken_30.jpg \n inflating: /data/training/Clay_Aiken_31.jpg \n inflating: /data/training/Clay_Aiken_32.jpg \n inflating: /data/training/Clay_Aiken_40.jpg \n inflating: /data/training/Clay_Aiken_41.jpg \n inflating: /data/training/Clay_Aiken_42.jpg \n inflating: /data/training/Clay_Aiken_50.jpg \n inflating: /data/training/Clay_Aiken_51.jpg \n inflating: /data/training/Clay_Aiken_52.jpg \n inflating: /data/training/Clint_Howard_00.jpg \n inflating: /data/training/Clint_Howard_01.jpg \n inflating: /data/training/Clint_Howard_02.jpg \n inflating: /data/training/Clint_Howard_10.jpg \n inflating: /data/training/Clint_Howard_11.jpg \n inflating: /data/training/Clint_Howard_12.jpg \n inflating: /data/training/Clint_Howard_20.jpg \n inflating: /data/training/Clint_Howard_21.jpg \n inflating: /data/training/Clint_Howard_22.jpg \n inflating: /data/training/Clint_Howard_30.jpg \n inflating: /data/training/Clint_Howard_31.jpg \n inflating: /data/training/Clint_Howard_32.jpg \n inflating: /data/training/Clive_Lloyd_30.jpg \n inflating: /data/training/Clive_Lloyd_31.jpg \n inflating: /data/training/Clive_Lloyd_32.jpg \n inflating: /data/training/Clive_Lloyd_40.jpg \n inflating: /data/training/Clive_Lloyd_41.jpg \n inflating: /data/training/Clive_Lloyd_42.jpg \n inflating: /data/training/Clive_Lloyd_50.jpg \n inflating: /data/training/Clive_Lloyd_51.jpg \n inflating: /data/training/Clive_Lloyd_52.jpg \n inflating: /data/training/Colin_Powell_10.jpg \n inflating: /data/training/Colin_Powell_11.jpg \n inflating: /data/training/Colin_Powell_12.jpg \n inflating: /data/training/Colin_Powell_40.jpg \n inflating: /data/training/Colin_Powell_41.jpg \n inflating: /data/training/Colin_Powell_42.jpg \n inflating: /data/training/Colin_Powell_50.jpg \n inflating: /data/training/Colin_Powell_51.jpg \n inflating: /data/training/Colin_Powell_52.jpg \n inflating: /data/training/Conan_OBrien_00.jpg \n inflating: /data/training/Conan_OBrien_01.jpg \n inflating: /data/training/Conan_OBrien_02.jpg \n inflating: /data/training/Conan_OBrien_10.jpg \n inflating: /data/training/Conan_OBrien_11.jpg \n inflating: /data/training/Conan_OBrien_12.jpg \n inflating: /data/training/Conan_OBrien_50.jpg \n inflating: /data/training/Conan_OBrien_51.jpg \n inflating: /data/training/Conan_OBrien_52.jpg \n inflating: /data/training/Condoleezza_Rice_10.jpg \n inflating: /data/training/Condoleezza_Rice_11.jpg \n inflating: /data/training/Condoleezza_Rice_12.jpg \n inflating: /data/training/Condoleezza_Rice_20.jpg \n inflating: /data/training/Condoleezza_Rice_21.jpg \n inflating: /data/training/Condoleezza_Rice_22.jpg \n inflating: /data/training/Condoleezza_Rice_30.jpg \n inflating: /data/training/Condoleezza_Rice_31.jpg \n inflating: /data/training/Condoleezza_Rice_32.jpg \n inflating: /data/training/Condoleezza_Rice_40.jpg \n inflating: /data/training/Condoleezza_Rice_41.jpg \n inflating: /data/training/Condoleezza_Rice_42.jpg \n inflating: /data/training/Connie_Chung_20.jpg \n inflating: /data/training/Connie_Chung_21.jpg \n inflating: /data/training/Connie_Chung_22.jpg \n inflating: /data/training/Connie_Chung_30.jpg \n inflating: /data/training/Connie_Chung_31.jpg \n inflating: /data/training/Connie_Chung_32.jpg \n inflating: /data/training/Connie_Chung_40.jpg \n inflating: /data/training/Connie_Chung_41.jpg \n inflating: /data/training/Connie_Chung_42.jpg \n inflating: /data/training/Connie_Chung_50.jpg \n inflating: /data/training/Connie_Chung_51.jpg \n inflating: /data/training/Connie_Chung_52.jpg \n inflating: /data/training/Craig_David_10.jpg \n inflating: /data/training/Craig_David_11.jpg \n inflating: /data/training/Craig_David_12.jpg \n inflating: /data/training/Craig_David_20.jpg \n inflating: /data/training/Craig_David_21.jpg \n inflating: /data/training/Craig_David_22.jpg \n inflating: /data/training/Craig_David_30.jpg \n inflating: /data/training/Craig_David_31.jpg \n inflating: /data/training/Craig_David_32.jpg \n inflating: /data/training/Craig_David_50.jpg \n inflating: /data/training/Craig_David_51.jpg \n inflating: /data/training/Craig_David_52.jpg \n inflating: /data/training/Cristina_Fernandez_30.jpg \n inflating: /data/training/Cristina_Fernandez_31.jpg \n inflating: /data/training/Cristina_Fernandez_32.jpg \n inflating: /data/training/Cristina_Fernandez_40.jpg \n inflating: /data/training/Cristina_Fernandez_41.jpg \n inflating: /data/training/Cristina_Fernandez_42.jpg \n inflating: /data/training/Cristina_Fernandez_50.jpg \n inflating: /data/training/Cristina_Fernandez_51.jpg \n inflating: /data/training/Cristina_Fernandez_52.jpg \n inflating: /data/training/Cristina_Saralegui_00.jpg \n inflating: /data/training/Cristina_Saralegui_01.jpg \n inflating: /data/training/Cristina_Saralegui_02.jpg \n inflating: /data/training/Cristina_Saralegui_30.jpg \n inflating: /data/training/Cristina_Saralegui_31.jpg \n inflating: /data/training/Cristina_Saralegui_32.jpg \n inflating: /data/training/Cristina_Saralegui_50.jpg \n inflating: /data/training/Cristina_Saralegui_51.jpg \n inflating: /data/training/Cristina_Saralegui_52.jpg \n inflating: /data/training/Dai_Bachtiar_00.jpg \n inflating: /data/training/Dai_Bachtiar_01.jpg \n inflating: /data/training/Dai_Bachtiar_02.jpg \n inflating: /data/training/Dai_Bachtiar_10.jpg \n inflating: /data/training/Dai_Bachtiar_11.jpg \n inflating: /data/training/Dai_Bachtiar_12.jpg \n inflating: /data/training/Dai_Bachtiar_20.jpg \n inflating: /data/training/Dai_Bachtiar_21.jpg \n inflating: /data/training/Dai_Bachtiar_22.jpg \n inflating: /data/training/Dai_Bachtiar_50.jpg \n inflating: /data/training/Dai_Bachtiar_51.jpg \n inflating: /data/training/Dai_Bachtiar_52.jpg \n inflating: /data/training/Daisy_Fuentes_20.jpg \n inflating: /data/training/Daisy_Fuentes_21.jpg \n inflating: /data/training/Daisy_Fuentes_22.jpg \n inflating: /data/training/Daisy_Fuentes_30.jpg \n inflating: /data/training/Daisy_Fuentes_31.jpg \n inflating: /data/training/Daisy_Fuentes_32.jpg \n inflating: /data/training/Daisy_Fuentes_40.jpg \n inflating: /data/training/Daisy_Fuentes_41.jpg \n inflating: /data/training/Daisy_Fuentes_42.jpg \n inflating: /data/training/Dan_Ackroyd_00.jpg \n inflating: /data/training/Dan_Ackroyd_01.jpg \n inflating: /data/training/Dan_Ackroyd_02.jpg \n inflating: /data/training/Dan_Ackroyd_20.jpg \n inflating: /data/training/Dan_Ackroyd_21.jpg \n inflating: /data/training/Dan_Ackroyd_22.jpg \n inflating: /data/training/Dan_Ackroyd_30.jpg \n inflating: /data/training/Dan_Ackroyd_31.jpg \n inflating: /data/training/Dan_Ackroyd_32.jpg \n inflating: /data/training/Daniel_Radcliffe_00.jpg \n inflating: /data/training/Daniel_Radcliffe_01.jpg \n inflating: /data/training/Daniel_Radcliffe_02.jpg \n inflating: /data/training/Daniel_Radcliffe_20.jpg \n inflating: /data/training/Daniel_Radcliffe_21.jpg \n inflating: /data/training/Daniel_Radcliffe_22.jpg \n inflating: /data/training/Daniel_Radcliffe_50.jpg \n inflating: /data/training/Daniel_Radcliffe_51.jpg \n inflating: /data/training/Daniel_Radcliffe_52.jpg \n inflating: /data/training/Daniel_Rouse_00.jpg \n inflating: /data/training/Daniel_Rouse_01.jpg \n inflating: /data/training/Daniel_Rouse_02.jpg \n inflating: /data/training/Daniel_Rouse_10.jpg \n inflating: /data/training/Daniel_Rouse_11.jpg \n inflating: /data/training/Daniel_Rouse_12.jpg \n inflating: /data/training/Daniel_Rouse_20.jpg \n inflating: /data/training/Daniel_Rouse_21.jpg \n inflating: /data/training/Daniel_Rouse_22.jpg \n inflating: /data/training/Daniel_Rouse_30.jpg \n inflating: /data/training/Daniel_Rouse_31.jpg \n inflating: /data/training/Daniel_Rouse_32.jpg \n inflating: /data/training/Daniell_Sunjata_10.jpg \n inflating: /data/training/Daniell_Sunjata_11.jpg \n inflating: /data/training/Daniell_Sunjata_12.jpg \n inflating: /data/training/Daniell_Sunjata_20.jpg \n inflating: /data/training/Daniell_Sunjata_21.jpg \n inflating: /data/training/Daniell_Sunjata_22.jpg \n inflating: /data/training/Daniell_Sunjata_40.jpg \n inflating: /data/training/Daniell_Sunjata_41.jpg \n inflating: /data/training/Daniell_Sunjata_42.jpg \n inflating: /data/training/Danny_Glover_10.jpg \n inflating: /data/training/Danny_Glover_11.jpg \n inflating: /data/training/Danny_Glover_12.jpg \n inflating: /data/training/Danny_Glover_30.jpg \n inflating: /data/training/Danny_Glover_31.jpg \n inflating: /data/training/Danny_Glover_32.jpg \n inflating: /data/training/Danny_Glover_50.jpg \n inflating: /data/training/Danny_Glover_51.jpg \n inflating: /data/training/Danny_Glover_52.jpg \n inflating: /data/training/Darrell_Issa_00.jpg \n inflating: /data/training/Darrell_Issa_01.jpg \n inflating: /data/training/Darrell_Issa_02.jpg \n inflating: /data/training/Darrell_Issa_20.jpg \n inflating: /data/training/Darrell_Issa_21.jpg \n inflating: /data/training/Darrell_Issa_22.jpg \n inflating: /data/training/Darrell_Issa_30.jpg \n inflating: /data/training/Darrell_Issa_31.jpg \n inflating: /data/training/Darrell_Issa_32.jpg \n inflating: /data/training/Darrell_Issa_40.jpg \n inflating: /data/training/Darrell_Issa_41.jpg \n inflating: /data/training/Darrell_Issa_42.jpg \n inflating: /data/training/Dave_Campo_10.jpg \n inflating: /data/training/Dave_Campo_11.jpg \n inflating: /data/training/Dave_Campo_12.jpg \n inflating: /data/training/Dave_Campo_20.jpg \n inflating: /data/training/Dave_Campo_21.jpg \n inflating: /data/training/Dave_Campo_22.jpg \n inflating: /data/training/Dave_Campo_30.jpg \n inflating: /data/training/Dave_Campo_31.jpg \n inflating: /data/training/Dave_Campo_32.jpg \n inflating: /data/training/David_Brent_00.jpg \n inflating: /data/training/David_Brent_01.jpg \n inflating: /data/training/David_Brent_02.jpg \n inflating: /data/training/David_Brent_10.jpg \n inflating: /data/training/David_Brent_11.jpg \n inflating: /data/training/David_Brent_12.jpg \n inflating: /data/training/David_Brent_20.jpg \n inflating: /data/training/David_Brent_21.jpg \n inflating: /data/training/David_Brent_22.jpg \n inflating: /data/training/David_Brent_30.jpg \n inflating: /data/training/David_Brent_31.jpg \n inflating: /data/training/David_Brent_32.jpg \n inflating: /data/training/David_Caruso_00.jpg \n inflating: /data/training/David_Caruso_01.jpg \n inflating: /data/training/David_Caruso_02.jpg \n inflating: /data/training/David_Caruso_10.jpg \n inflating: /data/training/David_Caruso_11.jpg \n inflating: /data/training/David_Caruso_12.jpg \n inflating: /data/training/David_Caruso_30.jpg \n inflating: /data/training/David_Caruso_31.jpg \n inflating: /data/training/David_Caruso_32.jpg \n inflating: /data/training/David_Caruso_40.jpg \n inflating: /data/training/David_Caruso_41.jpg \n inflating: /data/training/David_Caruso_42.jpg \n inflating: /data/training/Ed_Rendell_00.jpg \n inflating: /data/training/Ed_Rendell_01.jpg \n inflating: /data/training/Ed_Rendell_02.jpg \n inflating: /data/training/Ed_Rendell_20.jpg \n inflating: /data/training/Ed_Rendell_21.jpg \n inflating: /data/training/Ed_Rendell_22.jpg \n inflating: /data/training/Ed_Rendell_50.jpg \n inflating: /data/training/Ed_Rendell_51.jpg \n inflating: /data/training/Ed_Rendell_52.jpg \n inflating: /data/training/Ed_Smart_10.jpg \n inflating: /data/training/Ed_Smart_11.jpg \n inflating: /data/training/Ed_Smart_12.jpg \n inflating: /data/training/Ed_Smart_30.jpg \n inflating: /data/training/Ed_Smart_31.jpg \n inflating: /data/training/Ed_Smart_32.jpg \n inflating: /data/training/Ed_Smart_50.jpg \n inflating: /data/training/Ed_Smart_51.jpg \n inflating: /data/training/Ed_Smart_52.jpg \n inflating: /data/training/Edie_Falco_20.jpg \n inflating: /data/training/Edie_Falco_21.jpg \n inflating: /data/training/Edie_Falco_22.jpg \n inflating: /data/training/Edie_Falco_30.jpg \n inflating: /data/training/Edie_Falco_31.jpg \n inflating: /data/training/Edie_Falco_32.jpg \n inflating: /data/training/Edie_Falco_40.jpg \n inflating: /data/training/Edie_Falco_41.jpg \n inflating: /data/training/Edie_Falco_42.jpg \n inflating: /data/training/Edie_Falco_50.jpg \n inflating: /data/training/Edie_Falco_51.jpg \n inflating: /data/training/Edie_Falco_52.jpg \n inflating: /data/training/Eduardo_Duhalde_00.jpg \n inflating: /data/training/Eduardo_Duhalde_01.jpg \n inflating: /data/training/Eduardo_Duhalde_02.jpg \n inflating: /data/training/Eduardo_Duhalde_10.jpg \n inflating: /data/training/Eduardo_Duhalde_11.jpg \n inflating: /data/training/Eduardo_Duhalde_12.jpg \n inflating: /data/training/Eduardo_Duhalde_30.jpg \n inflating: /data/training/Eduardo_Duhalde_31.jpg \n inflating: /data/training/Eduardo_Duhalde_32.jpg \n inflating: /data/training/Edward_Burns_10.jpg \n inflating: /data/training/Edward_Burns_11.jpg \n inflating: /data/training/Edward_Burns_12.jpg \n inflating: /data/training/Edward_Burns_20.jpg \n inflating: /data/training/Edward_Burns_21.jpg \n inflating: /data/training/Edward_Burns_22.jpg \n inflating: /data/training/Edward_Burns_30.jpg \n inflating: /data/training/Edward_Burns_31.jpg \n inflating: /data/training/Edward_Burns_32.jpg \n inflating: /data/training/Edward_Burns_50.jpg \n inflating: /data/training/Edward_Burns_51.jpg \n inflating: /data/training/Edward_Burns_52.jpg \n inflating: /data/training/Edward_Norton_10.jpg \n inflating: /data/training/Edward_Norton_11.jpg \n inflating: /data/training/Edward_Norton_12.jpg \n inflating: /data/training/Edward_Norton_30.jpg \n inflating: /data/training/Edward_Norton_31.jpg \n inflating: /data/training/Edward_Norton_32.jpg \n inflating: /data/training/Edward_Norton_40.jpg \n inflating: /data/training/Edward_Norton_41.jpg \n inflating: /data/training/Edward_Norton_42.jpg \n inflating: /data/training/Edward_Norton_50.jpg \n inflating: /data/training/Edward_Norton_51.jpg \n inflating: /data/training/Edward_Norton_52.jpg \n inflating: /data/training/Elaine_Chao_00.jpg \n inflating: /data/training/Elaine_Chao_01.jpg \n inflating: /data/training/Elaine_Chao_02.jpg \n inflating: /data/training/Elaine_Chao_20.jpg \n inflating: /data/training/Elaine_Chao_21.jpg \n inflating: /data/training/Elaine_Chao_22.jpg \n inflating: /data/training/Elaine_Chao_50.jpg \n inflating: /data/training/Elaine_Chao_51.jpg \n inflating: /data/training/Elaine_Chao_52.jpg \n inflating: /data/training/Elaine_Stritch_10.jpg \n inflating: /data/training/Elaine_Stritch_11.jpg \n inflating: /data/training/Elaine_Stritch_12.jpg \n inflating: /data/training/Elaine_Stritch_40.jpg \n inflating: /data/training/Elaine_Stritch_41.jpg \n inflating: /data/training/Elaine_Stritch_42.jpg \n inflating: /data/training/Elaine_Stritch_50.jpg \n inflating: /data/training/Elaine_Stritch_51.jpg \n inflating: /data/training/Elaine_Stritch_52.jpg \n inflating: /data/training/Eliane_Karp_00.jpg \n inflating: /data/training/Eliane_Karp_01.jpg \n inflating: /data/training/Eliane_Karp_02.jpg \n inflating: /data/training/Eliane_Karp_10.jpg \n inflating: /data/training/Eliane_Karp_11.jpg \n inflating: /data/training/Eliane_Karp_12.jpg \n inflating: /data/training/Eliane_Karp_30.jpg \n inflating: /data/training/Eliane_Karp_31.jpg \n inflating: /data/training/Eliane_Karp_32.jpg \n inflating: /data/training/Eliane_Karp_40.jpg \n inflating: /data/training/Eliane_Karp_41.jpg \n inflating: /data/training/Eliane_Karp_42.jpg \n inflating: /data/training/Elijah_Wood_00.jpg \n inflating: /data/training/Elijah_Wood_01.jpg \n inflating: /data/training/Elijah_Wood_02.jpg \n inflating: /data/training/Elijah_Wood_10.jpg \n inflating: /data/training/Elijah_Wood_11.jpg \n inflating: /data/training/Elijah_Wood_12.jpg \n inflating: /data/training/Elijah_Wood_30.jpg \n inflating: /data/training/Elijah_Wood_31.jpg \n inflating: /data/training/Elijah_Wood_32.jpg \n inflating: /data/training/Eliza_Dushku_00.jpg \n inflating: /data/training/Eliza_Dushku_01.jpg \n inflating: /data/training/Eliza_Dushku_02.jpg \n inflating: /data/training/Eliza_Dushku_10.jpg \n inflating: /data/training/Eliza_Dushku_11.jpg \n inflating: /data/training/Eliza_Dushku_12.jpg \n inflating: /data/training/Eliza_Dushku_20.jpg \n inflating: /data/training/Eliza_Dushku_21.jpg \n inflating: /data/training/Eliza_Dushku_22.jpg \n inflating: /data/training/Eliza_Dushku_30.jpg \n inflating: /data/training/Eliza_Dushku_31.jpg \n inflating: /data/training/Eliza_Dushku_32.jpg \n inflating: /data/training/Elizabeth_Dole_00.jpg \n inflating: /data/training/Elizabeth_Dole_01.jpg \n inflating: /data/training/Elizabeth_Dole_02.jpg \n inflating: /data/training/Elizabeth_Dole_10.jpg \n inflating: /data/training/Elizabeth_Dole_11.jpg \n inflating: /data/training/Elizabeth_Dole_12.jpg \n inflating: /data/training/Elizabeth_Dole_30.jpg \n inflating: /data/training/Elizabeth_Dole_31.jpg \n inflating: /data/training/Elizabeth_Dole_32.jpg \n inflating: /data/training/Elizabeth_Shue_00.jpg \n inflating: /data/training/Elizabeth_Shue_01.jpg \n inflating: /data/training/Elizabeth_Shue_02.jpg \n inflating: /data/training/Elizabeth_Shue_20.jpg \n inflating: /data/training/Elizabeth_Shue_21.jpg \n inflating: /data/training/Elizabeth_Shue_22.jpg \n inflating: /data/training/Elizabeth_Shue_40.jpg \n inflating: /data/training/Elizabeth_Shue_41.jpg \n inflating: /data/training/Elizabeth_Shue_42.jpg \n inflating: /data/training/Ellen_DeGeneres_10.jpg \n inflating: /data/training/Ellen_DeGeneres_11.jpg \n inflating: /data/training/Ellen_DeGeneres_12.jpg \n inflating: /data/training/Ellen_DeGeneres_40.jpg \n inflating: /data/training/Ellen_DeGeneres_41.jpg \n inflating: /data/training/Ellen_DeGeneres_42.jpg \n inflating: /data/training/Ellen_DeGeneres_50.jpg \n inflating: /data/training/Ellen_DeGeneres_51.jpg \n inflating: /data/training/Ellen_DeGeneres_52.jpg \n inflating: /data/training/Elmar_Brok_00.jpg \n inflating: /data/training/Elmar_Brok_01.jpg \n inflating: /data/training/Elmar_Brok_02.jpg \n inflating: /data/training/Elmar_Brok_20.jpg \n inflating: /data/training/Elmar_Brok_21.jpg \n inflating: /data/training/Elmar_Brok_22.jpg \n inflating: /data/training/Elmar_Brok_30.jpg \n inflating: /data/training/Elmar_Brok_31.jpg \n inflating: /data/training/Elmar_Brok_32.jpg \n inflating: /data/training/Elsa_Zylberstein_00.jpg \n inflating: /data/training/Elsa_Zylberstein_01.jpg \n inflating: /data/training/Elsa_Zylberstein_02.jpg \n inflating: /data/training/Elsa_Zylberstein_10.jpg \n inflating: /data/training/Elsa_Zylberstein_11.jpg \n inflating: /data/training/Elsa_Zylberstein_12.jpg \n inflating: /data/training/Elsa_Zylberstein_40.jpg \n inflating: /data/training/Elsa_Zylberstein_41.jpg \n inflating: /data/training/Elsa_Zylberstein_42.jpg \n inflating: /data/training/Elton_John_10.jpg \n inflating: /data/training/Elton_John_11.jpg \n inflating: /data/training/Elton_John_12.jpg \n inflating: /data/training/Elton_John_20.jpg \n inflating: /data/training/Elton_John_21.jpg \n inflating: /data/training/Elton_John_22.jpg \n inflating: /data/training/Elton_John_30.jpg \n inflating: /data/training/Elton_John_31.jpg \n inflating: /data/training/Elton_John_32.jpg \n inflating: /data/training/Elton_John_40.jpg \n inflating: /data/training/Elton_John_41.jpg \n inflating: /data/training/Elton_John_42.jpg \n inflating: /data/training/Emile_Lahoud_00.jpg \n inflating: /data/training/Emile_Lahoud_01.jpg \n inflating: /data/training/Emile_Lahoud_02.jpg \n inflating: /data/training/Emile_Lahoud_30.jpg \n inflating: /data/training/Emile_Lahoud_31.jpg \n inflating: /data/training/Emile_Lahoud_32.jpg \n inflating: /data/training/Emile_Lahoud_40.jpg \n inflating: /data/training/Emile_Lahoud_41.jpg \n inflating: /data/training/Emile_Lahoud_42.jpg \n inflating: /data/training/Emilio_Botin_00.jpg \n inflating: /data/training/Emilio_Botin_01.jpg \n inflating: /data/training/Emilio_Botin_02.jpg \n inflating: /data/training/Emilio_Botin_10.jpg \n inflating: /data/training/Emilio_Botin_11.jpg \n inflating: /data/training/Emilio_Botin_12.jpg \n inflating: /data/training/Emilio_Botin_20.jpg \n inflating: /data/training/Emilio_Botin_21.jpg \n inflating: /data/training/Emilio_Botin_22.jpg \n inflating: /data/training/Emilio_Botin_40.jpg \n inflating: /data/training/Emilio_Botin_41.jpg \n inflating: /data/training/Emilio_Botin_42.jpg \n inflating: /data/training/Emma_Nicholson_10.jpg \n inflating: /data/training/Emma_Nicholson_11.jpg \n inflating: /data/training/Emma_Nicholson_12.jpg \n inflating: /data/training/Emma_Nicholson_20.jpg \n inflating: /data/training/Emma_Nicholson_21.jpg \n inflating: /data/training/Emma_Nicholson_22.jpg \n inflating: /data/training/Emma_Nicholson_30.jpg \n inflating: /data/training/Emma_Nicholson_31.jpg \n inflating: /data/training/Emma_Nicholson_32.jpg \n inflating: /data/training/Emma_Thompson_20.jpg \n inflating: /data/training/Emma_Thompson_21.jpg \n inflating: /data/training/Emma_Thompson_22.jpg \n inflating: /data/training/Emma_Thompson_30.jpg \n inflating: /data/training/Emma_Thompson_31.jpg \n inflating: /data/training/Emma_Thompson_32.jpg \n inflating: /data/training/Emma_Thompson_40.jpg \n inflating: /data/training/Emma_Thompson_41.jpg \n inflating: /data/training/Emma_Thompson_42.jpg \n inflating: /data/training/Emma_Thompson_50.jpg \n inflating: /data/training/Emma_Thompson_51.jpg \n inflating: /data/training/Emma_Thompson_52.jpg \n inflating: /data/training/Emmy_Rossum_20.jpg \n inflating: /data/training/Emmy_Rossum_21.jpg \n inflating: /data/training/Emmy_Rossum_22.jpg \n inflating: /data/training/Emmy_Rossum_30.jpg \n inflating: /data/training/Emmy_Rossum_31.jpg \n inflating: /data/training/Emmy_Rossum_32.jpg \n inflating: /data/training/Emmy_Rossum_40.jpg \n inflating: /data/training/Emmy_Rossum_41.jpg \n inflating: /data/training/Emmy_Rossum_42.jpg \n inflating: /data/training/Emmy_Rossum_50.jpg \n inflating: /data/training/Emmy_Rossum_51.jpg \n inflating: /data/training/Emmy_Rossum_52.jpg \n inflating: /data/training/Eric_Benet_00.jpg \n inflating: /data/training/Eric_Benet_01.jpg \n inflating: /data/training/Eric_Benet_02.jpg \n inflating: /data/training/Eric_Benet_10.jpg \n inflating: /data/training/Eric_Benet_11.jpg \n inflating: /data/training/Eric_Benet_12.jpg \n inflating: /data/training/Eric_Benet_30.jpg \n inflating: /data/training/Eric_Benet_31.jpg \n inflating: /data/training/Eric_Benet_32.jpg \n inflating: /data/training/Erin_Hershey_Presley_10.jpg \n inflating: /data/training/Erin_Hershey_Presley_11.jpg \n inflating: /data/training/Erin_Hershey_Presley_12.jpg \n inflating: /data/training/Erin_Hershey_Presley_30.jpg \n inflating: /data/training/Erin_Hershey_Presley_31.jpg \n inflating: /data/training/Erin_Hershey_Presley_32.jpg \n inflating: /data/training/Erin_Hershey_Presley_40.jpg \n inflating: /data/training/Erin_Hershey_Presley_41.jpg \n inflating: /data/training/Erin_Hershey_Presley_42.jpg \n inflating: /data/training/Ernest_Hollings_00.jpg \n inflating: /data/training/Ernest_Hollings_01.jpg \n inflating: /data/training/Ernest_Hollings_02.jpg \n inflating: /data/training/Ernest_Hollings_10.jpg \n inflating: /data/training/Ernest_Hollings_11.jpg \n inflating: /data/training/Ernest_Hollings_12.jpg \n inflating: /data/training/Ernest_Hollings_20.jpg \n inflating: /data/training/Ernest_Hollings_21.jpg \n inflating: /data/training/Ernest_Hollings_22.jpg \n inflating: /data/training/Ernesto_Zedillo_10.jpg \n inflating: /data/training/Ernesto_Zedillo_11.jpg \n inflating: /data/training/Ernesto_Zedillo_12.jpg \n inflating: /data/training/Ernesto_Zedillo_20.jpg \n inflating: /data/training/Ernesto_Zedillo_21.jpg \n inflating: /data/training/Ernesto_Zedillo_22.jpg \n inflating: /data/training/Ernesto_Zedillo_30.jpg \n inflating: /data/training/Ernesto_Zedillo_31.jpg \n inflating: /data/training/Ernesto_Zedillo_32.jpg \n inflating: /data/training/Ernesto_Zedillo_40.jpg \n inflating: /data/training/Ernesto_Zedillo_41.jpg \n inflating: /data/training/Ernesto_Zedillo_42.jpg \n inflating: /data/training/Ernie_Grunfeld_20.jpg \n inflating: /data/training/Ernie_Grunfeld_21.jpg \n inflating: /data/training/Ernie_Grunfeld_22.jpg \n inflating: /data/training/Ernie_Grunfeld_30.jpg \n inflating: /data/training/Ernie_Grunfeld_31.jpg \n inflating: /data/training/Ernie_Grunfeld_32.jpg \n inflating: /data/training/Ernie_Grunfeld_40.jpg \n inflating: /data/training/Ernie_Grunfeld_41.jpg \n inflating: /data/training/Ernie_Grunfeld_42.jpg \n inflating: /data/training/Ernie_Grunfeld_50.jpg \n inflating: /data/training/Ernie_Grunfeld_51.jpg \n inflating: /data/training/Ernie_Grunfeld_52.jpg \n inflating: /data/training/Estelle_Morris_10.jpg \n inflating: /data/training/Estelle_Morris_11.jpg \n inflating: /data/training/Estelle_Morris_12.jpg \n inflating: /data/training/Estelle_Morris_20.jpg \n inflating: /data/training/Estelle_Morris_21.jpg \n inflating: /data/training/Estelle_Morris_22.jpg \n inflating: /data/training/Estelle_Morris_30.jpg \n inflating: /data/training/Estelle_Morris_31.jpg \n inflating: /data/training/Estelle_Morris_32.jpg \n inflating: /data/training/Ethan_Hawke_00.jpg \n inflating: /data/training/Ethan_Hawke_01.jpg \n inflating: /data/training/Ethan_Hawke_02.jpg \n inflating: /data/training/Ethan_Hawke_10.jpg \n inflating: /data/training/Ethan_Hawke_11.jpg \n inflating: /data/training/Ethan_Hawke_12.jpg \n inflating: /data/training/Ethan_Hawke_30.jpg \n inflating: /data/training/Ethan_Hawke_31.jpg \n inflating: /data/training/Ethan_Hawke_32.jpg \n inflating: /data/training/Ethan_Hawke_40.jpg \n inflating: /data/training/Ethan_Hawke_41.jpg \n inflating: /data/training/Ethan_Hawke_42.jpg \n inflating: /data/training/Eunice_Barber_00.jpg \n inflating: /data/training/Eunice_Barber_01.jpg \n inflating: /data/training/Eunice_Barber_02.jpg \n inflating: /data/training/Eunice_Barber_10.jpg \n inflating: /data/training/Eunice_Barber_11.jpg \n inflating: /data/training/Eunice_Barber_12.jpg \n inflating: /data/training/Eunice_Barber_50.jpg \n inflating: /data/training/Eunice_Barber_51.jpg \n inflating: /data/training/Eunice_Barber_52.jpg \n inflating: /data/training/Fernando_Henrique_Cardoso_00.jpg \n inflating: /data/training/Fernando_Henrique_Cardoso_01.jpg \n inflating: /data/training/Fernando_Henrique_Cardoso_02.jpg \n inflating: /data/training/Fernando_Henrique_Cardoso_20.jpg \n inflating: /data/training/Fernando_Henrique_Cardoso_21.jpg \n inflating: /data/training/Fernando_Henrique_Cardoso_22.jpg \n inflating: /data/training/Fernando_Henrique_Cardoso_30.jpg \n inflating: /data/training/Fernando_Henrique_Cardoso_31.jpg \n inflating: /data/training/Fernando_Henrique_Cardoso_32.jpg \n inflating: /data/training/Fernando_Sanz_30.jpg \n inflating: /data/training/Fernando_Sanz_31.jpg \n inflating: /data/training/Fernando_Sanz_32.jpg \n inflating: /data/training/Fernando_Sanz_40.jpg \n inflating: /data/training/Fernando_Sanz_41.jpg \n inflating: /data/training/Fernando_Sanz_42.jpg \n inflating: /data/training/Fernando_Sanz_50.jpg \n inflating: /data/training/Fernando_Sanz_51.jpg \n inflating: /data/training/Fernando_Sanz_52.jpg \n inflating: /data/training/Fidel_Castro_Daiz-Balart_10.jpg \n inflating: /data/training/Fidel_Castro_Daiz-Balart_11.jpg \n inflating: /data/training/Fidel_Castro_Daiz-Balart_12.jpg \n inflating: /data/training/Fidel_Castro_Daiz-Balart_30.jpg \n inflating: /data/training/Fidel_Castro_Daiz-Balart_31.jpg \n inflating: /data/training/Fidel_Castro_Daiz-Balart_32.jpg \n inflating: /data/training/Fidel_Castro_Daiz-Balart_40.jpg \n inflating: /data/training/Fidel_Castro_Daiz-Balart_41.jpg \n inflating: /data/training/Fidel_Castro_Daiz-Balart_42.jpg \n inflating: /data/training/Flavia_Pennetta_30.jpg \n inflating: /data/training/Flavia_Pennetta_31.jpg \n inflating: /data/training/Flavia_Pennetta_32.jpg \n inflating: /data/training/Flavia_Pennetta_40.jpg \n inflating: /data/training/Flavia_Pennetta_41.jpg \n inflating: /data/training/Flavia_Pennetta_42.jpg \n inflating: /data/training/Flavia_Pennetta_50.jpg \n inflating: /data/training/Flavia_Pennetta_51.jpg \n inflating: /data/training/Flavia_Pennetta_52.jpg \n inflating: /data/training/Florecita_Cobian_00.jpg \n inflating: /data/training/Florecita_Cobian_01.jpg \n inflating: /data/training/Florecita_Cobian_02.jpg \n inflating: /data/training/Florecita_Cobian_10.jpg \n inflating: /data/training/Florecita_Cobian_11.jpg \n inflating: /data/training/Florecita_Cobian_12.jpg \n inflating: /data/training/Florecita_Cobian_20.jpg \n inflating: /data/training/Florecita_Cobian_21.jpg \n inflating: /data/training/Florecita_Cobian_22.jpg \n inflating: /data/training/Frances_Fisher_20.jpg \n inflating: /data/training/Frances_Fisher_21.jpg \n inflating: /data/training/Frances_Fisher_22.jpg \n inflating: /data/training/Frances_Fisher_30.jpg \n inflating: /data/training/Frances_Fisher_31.jpg \n inflating: /data/training/Frances_Fisher_32.jpg \n inflating: /data/training/Frances_Fisher_40.jpg \n inflating: /data/training/Frances_Fisher_41.jpg \n inflating: /data/training/Frances_Fisher_42.jpg \n inflating: /data/training/Francis_Collins_00.jpg \n inflating: /data/training/Francis_Collins_01.jpg \n inflating: /data/training/Francis_Collins_02.jpg \n inflating: /data/training/Francis_Collins_10.jpg \n inflating: /data/training/Francis_Collins_11.jpg \n inflating: /data/training/Francis_Collins_12.jpg \n inflating: /data/training/Francis_Collins_20.jpg \n inflating: /data/training/Francis_Collins_21.jpg \n inflating: /data/training/Francis_Collins_22.jpg \n inflating: /data/training/Francis_Collins_40.jpg \n inflating: /data/training/Francis_Collins_41.jpg \n inflating: /data/training/Francis_Collins_42.jpg \n inflating: /data/training/Frank_Beamer_00.jpg \n inflating: /data/training/Frank_Beamer_01.jpg \n inflating: /data/training/Frank_Beamer_02.jpg \n inflating: /data/training/Frank_Beamer_20.jpg \n inflating: /data/training/Frank_Beamer_21.jpg \n inflating: /data/training/Frank_Beamer_22.jpg \n inflating: /data/training/Frank_Beamer_30.jpg \n inflating: /data/training/Frank_Beamer_31.jpg \n inflating: /data/training/Frank_Beamer_32.jpg \n inflating: /data/training/Frank_Caliendo_10.jpg \n inflating: /data/training/Frank_Caliendo_11.jpg \n inflating: /data/training/Frank_Caliendo_12.jpg \n inflating: /data/training/Frank_Caliendo_30.jpg \n inflating: /data/training/Frank_Caliendo_31.jpg \n inflating: /data/training/Frank_Caliendo_32.jpg \n inflating: /data/training/Frank_Caliendo_40.jpg \n inflating: /data/training/Frank_Caliendo_41.jpg \n inflating: /data/training/Frank_Caliendo_42.jpg \n inflating: /data/training/Frank_Caliendo_50.jpg \n inflating: /data/training/Frank_Caliendo_51.jpg \n inflating: /data/training/Frank_Caliendo_52.jpg \n inflating: /data/training/Frank_Keating_30.jpg \n inflating: /data/training/Frank_Keating_31.jpg \n inflating: /data/training/Frank_Keating_32.jpg \n inflating: /data/training/Frank_Keating_40.jpg \n inflating: /data/training/Frank_Keating_41.jpg \n inflating: /data/training/Frank_Keating_42.jpg \n inflating: /data/training/Frank_Keating_50.jpg \n inflating: /data/training/Frank_Keating_51.jpg \n inflating: /data/training/Frank_Keating_52.jpg \n inflating: /data/training/Frank_Solich_10.jpg \n inflating: /data/training/Frank_Solich_11.jpg \n inflating: /data/training/Frank_Solich_12.jpg \n inflating: /data/training/Frank_Solich_20.jpg \n inflating: /data/training/Frank_Solich_21.jpg \n inflating: /data/training/Frank_Solich_22.jpg \n inflating: /data/training/Frank_Solich_30.jpg \n inflating: /data/training/Frank_Solich_31.jpg \n inflating: /data/training/Frank_Solich_32.jpg \n inflating: /data/training/Franz_Fischler_00.jpg \n inflating: /data/training/Franz_Fischler_01.jpg \n inflating: /data/training/Franz_Fischler_02.jpg \n inflating: /data/training/Franz_Fischler_30.jpg \n inflating: /data/training/Franz_Fischler_31.jpg \n inflating: /data/training/Franz_Fischler_32.jpg \n inflating: /data/training/Franz_Fischler_40.jpg \n inflating: /data/training/Franz_Fischler_41.jpg \n inflating: /data/training/Franz_Fischler_42.jpg \n inflating: /data/training/Franz_Fischler_50.jpg \n inflating: /data/training/Franz_Fischler_51.jpg \n inflating: /data/training/Franz_Fischler_52.jpg \n inflating: /data/training/Gabi_Zimmer_00.jpg \n inflating: /data/training/Gabi_Zimmer_01.jpg \n inflating: /data/training/Gabi_Zimmer_02.jpg \n inflating: /data/training/Gabi_Zimmer_10.jpg \n inflating: /data/training/Gabi_Zimmer_11.jpg \n inflating: /data/training/Gabi_Zimmer_12.jpg \n inflating: /data/training/Gabi_Zimmer_20.jpg \n inflating: /data/training/Gabi_Zimmer_21.jpg \n inflating: /data/training/Gabi_Zimmer_22.jpg \n inflating: /data/training/Gabi_Zimmer_50.jpg \n inflating: /data/training/Gabi_Zimmer_51.jpg \n inflating: /data/training/Gabi_Zimmer_52.jpg \n inflating: /data/training/Gary_Bettman_10.jpg \n inflating: /data/training/Gary_Bettman_11.jpg \n inflating: /data/training/Gary_Bettman_12.jpg \n inflating: /data/training/Gary_Bettman_30.jpg \n inflating: /data/training/Gary_Bettman_31.jpg \n inflating: /data/training/Gary_Bettman_32.jpg \n inflating: /data/training/Gary_Bettman_40.jpg \n inflating: /data/training/Gary_Bettman_41.jpg \n inflating: /data/training/Gary_Bettman_42.jpg \n inflating: /data/training/Gary_Coleman_30.jpg \n inflating: /data/training/Gary_Coleman_31.jpg \n inflating: /data/training/Gary_Coleman_32.jpg \n inflating: /data/training/Gary_Coleman_40.jpg \n inflating: /data/training/Gary_Coleman_41.jpg \n inflating: /data/training/Gary_Coleman_42.jpg \n inflating: /data/training/Gary_Coleman_50.jpg \n inflating: /data/training/Gary_Coleman_51.jpg \n inflating: /data/training/Gary_Coleman_52.jpg \n inflating: /data/training/Gary_Condit_00.jpg \n inflating: /data/training/Gary_Condit_01.jpg \n inflating: /data/training/Gary_Condit_02.jpg \n inflating: /data/training/Gary_Condit_10.jpg \n inflating: /data/training/Gary_Condit_11.jpg \n inflating: /data/training/Gary_Condit_12.jpg \n inflating: /data/training/Gary_Condit_30.jpg \n inflating: /data/training/Gary_Condit_31.jpg \n inflating: /data/training/Gary_Condit_32.jpg \n inflating: /data/training/Gene_Hackman_20.jpg \n inflating: /data/training/Gene_Hackman_21.jpg \n inflating: /data/training/Gene_Hackman_22.jpg \n inflating: /data/training/Gene_Hackman_30.jpg \n inflating: /data/training/Gene_Hackman_31.jpg \n inflating: /data/training/Gene_Hackman_32.jpg \n inflating: /data/training/Gene_Hackman_40.jpg \n inflating: /data/training/Gene_Hackman_41.jpg \n inflating: /data/training/Gene_Hackman_42.jpg \n inflating: /data/training/Geoffrey_Rush_00.jpg \n inflating: /data/training/Geoffrey_Rush_01.jpg \n inflating: /data/training/Geoffrey_Rush_02.jpg \n inflating: /data/training/Geoffrey_Rush_10.jpg \n inflating: /data/training/Geoffrey_Rush_11.jpg \n inflating: /data/training/Geoffrey_Rush_12.jpg \n inflating: /data/training/Geoffrey_Rush_20.jpg \n inflating: /data/training/Geoffrey_Rush_21.jpg \n inflating: /data/training/Geoffrey_Rush_22.jpg \n inflating: /data/training/George_Galloway_00.jpg \n inflating: /data/training/George_Galloway_01.jpg \n inflating: /data/training/George_Galloway_02.jpg \n inflating: /data/training/George_Galloway_20.jpg \n inflating: /data/training/George_Galloway_21.jpg \n inflating: /data/training/George_Galloway_22.jpg \n inflating: /data/training/George_Galloway_40.jpg \n inflating: /data/training/George_Galloway_41.jpg \n inflating: /data/training/George_Galloway_42.jpg \n inflating: /data/training/George_Galloway_50.jpg \n inflating: /data/training/George_Galloway_51.jpg \n inflating: /data/training/George_Galloway_52.jpg \n inflating: /data/training/George_Karl_10.jpg \n inflating: /data/training/George_Karl_11.jpg \n inflating: /data/training/George_Karl_12.jpg \n inflating: /data/training/George_Karl_20.jpg \n inflating: /data/training/George_Karl_21.jpg \n inflating: /data/training/George_Karl_22.jpg \n inflating: /data/training/George_Karl_50.jpg \n inflating: /data/training/George_Karl_51.jpg \n inflating: /data/training/George_Karl_52.jpg \n inflating: /data/training/GL_Peiris_00.jpg \n inflating: /data/training/GL_Peiris_01.jpg \n inflating: /data/training/GL_Peiris_02.jpg \n inflating: /data/training/GL_Peiris_10.jpg \n inflating: /data/training/GL_Peiris_11.jpg \n inflating: /data/training/GL_Peiris_12.jpg \n inflating: /data/training/GL_Peiris_30.jpg \n inflating: /data/training/GL_Peiris_31.jpg \n inflating: /data/training/GL_Peiris_32.jpg \n inflating: /data/training/Hanan_Ashrawi_10.jpg \n inflating: /data/training/Hanan_Ashrawi_11.jpg \n inflating: /data/training/Hanan_Ashrawi_12.jpg \n inflating: /data/training/Hanan_Ashrawi_20.jpg \n inflating: /data/training/Hanan_Ashrawi_21.jpg \n inflating: /data/training/Hanan_Ashrawi_22.jpg \n inflating: /data/training/Hanan_Ashrawi_40.jpg \n inflating: /data/training/Hanan_Ashrawi_41.jpg \n inflating: /data/training/Hanan_Ashrawi_42.jpg \n inflating: /data/training/Harrison_Ford_10.jpg \n inflating: /data/training/Harrison_Ford_11.jpg \n inflating: /data/training/Harrison_Ford_12.jpg \n inflating: /data/training/Harrison_Ford_20.jpg \n inflating: /data/training/Harrison_Ford_21.jpg \n inflating: /data/training/Harrison_Ford_22.jpg \n inflating: /data/training/Harrison_Ford_50.jpg \n inflating: /data/training/Harrison_Ford_51.jpg \n inflating: /data/training/Harrison_Ford_52.jpg \n inflating: /data/training/Hassan_Nasrallah_30.jpg \n inflating: /data/training/Hassan_Nasrallah_31.jpg \n inflating: /data/training/Hassan_Nasrallah_32.jpg \n inflating: /data/training/Hassan_Nasrallah_40.jpg \n inflating: /data/training/Hassan_Nasrallah_41.jpg \n inflating: /data/training/Hassan_Nasrallah_42.jpg \n inflating: /data/training/Hassan_Nasrallah_50.jpg \n inflating: /data/training/Hassan_Nasrallah_51.jpg \n inflating: /data/training/Hassan_Nasrallah_52.jpg \n inflating: /data/training/Irene_Kahn_00.jpg \n inflating: /data/training/Irene_Kahn_01.jpg \n inflating: /data/training/Irene_Kahn_02.jpg \n inflating: /data/training/Irene_Kahn_30.jpg \n inflating: /data/training/Irene_Kahn_31.jpg \n inflating: /data/training/Irene_Kahn_32.jpg \n inflating: /data/training/Irene_Kahn_40.jpg \n inflating: /data/training/Irene_Kahn_41.jpg \n inflating: /data/training/Irene_Kahn_42.jpg \n inflating: /data/training/Isabella_Rossellini_00.jpg \n inflating: /data/training/Isabella_Rossellini_01.jpg \n inflating: /data/training/Isabella_Rossellini_02.jpg \n inflating: /data/training/Isabella_Rossellini_10.jpg \n inflating: /data/training/Isabella_Rossellini_11.jpg \n inflating: /data/training/Isabella_Rossellini_12.jpg \n inflating: /data/training/Isabella_Rossellini_20.jpg \n inflating: /data/training/Isabella_Rossellini_21.jpg \n inflating: /data/training/Isabella_Rossellini_22.jpg \n inflating: /data/training/Isabelle_Huppert_20.jpg \n inflating: /data/training/Isabelle_Huppert_21.jpg \n inflating: /data/training/Isabelle_Huppert_22.jpg \n inflating: /data/training/Isabelle_Huppert_30.jpg \n inflating: /data/training/Isabelle_Huppert_31.jpg \n inflating: /data/training/Isabelle_Huppert_32.jpg \n inflating: /data/training/Isabelle_Huppert_40.jpg \n inflating: /data/training/Isabelle_Huppert_41.jpg \n inflating: /data/training/Isabelle_Huppert_42.jpg \n inflating: /data/training/Itzhak_Perlman_10.jpg \n inflating: /data/training/Itzhak_Perlman_11.jpg \n inflating: /data/training/Itzhak_Perlman_12.jpg \n inflating: /data/training/Itzhak_Perlman_30.jpg \n inflating: /data/training/Itzhak_Perlman_31.jpg \n inflating: /data/training/Itzhak_Perlman_32.jpg \n inflating: /data/training/Itzhak_Perlman_40.jpg \n inflating: /data/training/Itzhak_Perlman_41.jpg \n inflating: /data/training/Itzhak_Perlman_42.jpg \n inflating: /data/training/Jack_Welch_10.jpg \n inflating: /data/training/Jack_Welch_11.jpg \n inflating: /data/training/Jack_Welch_12.jpg \n inflating: /data/training/Jack_Welch_30.jpg \n inflating: /data/training/Jack_Welch_31.jpg \n inflating: /data/training/Jack_Welch_32.jpg \n inflating: /data/training/Jack_Welch_40.jpg \n inflating: /data/training/Jack_Welch_41.jpg \n inflating: /data/training/Jack_Welch_42.jpg \n inflating: /data/training/Jack_Welch_50.jpg \n inflating: /data/training/Jack_Welch_51.jpg \n inflating: /data/training/Jack_Welch_52.jpg \n inflating: /data/training/Jackie_Sherrill_20.jpg \n inflating: /data/training/Jackie_Sherrill_21.jpg \n inflating: /data/training/Jackie_Sherrill_22.jpg \n inflating: /data/training/Jackie_Sherrill_40.jpg \n inflating: /data/training/Jackie_Sherrill_41.jpg \n inflating: /data/training/Jackie_Sherrill_42.jpg \n inflating: /data/training/Jackie_Sherrill_50.jpg \n inflating: /data/training/Jackie_Sherrill_51.jpg \n inflating: /data/training/Jackie_Sherrill_52.jpg \n inflating: /data/training/Jacqueline_Gold_00.jpg \n inflating: /data/training/Jacqueline_Gold_01.jpg \n inflating: /data/training/Jacqueline_Gold_02.jpg \n inflating: /data/training/Jacqueline_Gold_20.jpg \n inflating: /data/training/Jacqueline_Gold_21.jpg \n inflating: /data/training/Jacqueline_Gold_22.jpg \n inflating: /data/training/Jacqueline_Gold_30.jpg \n inflating: /data/training/Jacqueline_Gold_31.jpg \n inflating: /data/training/Jacqueline_Gold_32.jpg \n inflating: /data/training/Jafar_Umar_Thalib_00.jpg \n inflating: /data/training/Jafar_Umar_Thalib_01.jpg \n inflating: /data/training/Jafar_Umar_Thalib_02.jpg \n inflating: /data/training/Jafar_Umar_Thalib_20.jpg \n inflating: /data/training/Jafar_Umar_Thalib_21.jpg \n inflating: /data/training/Jafar_Umar_Thalib_22.jpg \n inflating: /data/training/Jafar_Umar_Thalib_30.jpg \n inflating: /data/training/Jafar_Umar_Thalib_31.jpg \n inflating: /data/training/Jafar_Umar_Thalib_32.jpg \n inflating: /data/training/Jafar_Umar_Thalib_50.jpg \n inflating: /data/training/Jafar_Umar_Thalib_51.jpg \n inflating: /data/training/Jafar_Umar_Thalib_52.jpg \n inflating: /data/training/Jaime_Pressly_00.jpg \n inflating: /data/training/Jaime_Pressly_01.jpg \n inflating: /data/training/Jaime_Pressly_02.jpg \n inflating: /data/training/Jaime_Pressly_10.jpg \n inflating: /data/training/Jaime_Pressly_11.jpg \n inflating: /data/training/Jaime_Pressly_12.jpg \n inflating: /data/training/Jaime_Pressly_40.jpg \n inflating: /data/training/Jaime_Pressly_41.jpg \n inflating: /data/training/Jaime_Pressly_42.jpg \n inflating: /data/training/Jake_Gyllenhaal_00.jpg \n inflating: /data/training/Jake_Gyllenhaal_01.jpg \n inflating: /data/training/Jake_Gyllenhaal_02.jpg \n inflating: /data/training/Jake_Gyllenhaal_40.jpg \n inflating: /data/training/Jake_Gyllenhaal_41.jpg \n inflating: /data/training/Jake_Gyllenhaal_42.jpg \n inflating: /data/training/Jake_Gyllenhaal_50.jpg \n inflating: /data/training/Jake_Gyllenhaal_51.jpg \n inflating: /data/training/Jake_Gyllenhaal_52.jpg \n inflating: /data/training/Jake_Plummer_20.jpg \n inflating: /data/training/Jake_Plummer_21.jpg \n inflating: /data/training/Jake_Plummer_22.jpg \n inflating: /data/training/Jake_Plummer_40.jpg \n inflating: /data/training/Jake_Plummer_41.jpg \n inflating: /data/training/Jake_Plummer_42.jpg \n inflating: /data/training/Jake_Plummer_50.jpg \n inflating: /data/training/Jake_Plummer_51.jpg \n inflating: /data/training/Jake_Plummer_52.jpg \n inflating: /data/training/James_Carville_00.jpg \n inflating: /data/training/James_Carville_01.jpg \n inflating: /data/training/James_Carville_02.jpg \n inflating: /data/training/James_Carville_10.jpg \n inflating: /data/training/James_Carville_11.jpg \n inflating: /data/training/James_Carville_12.jpg \n inflating: /data/training/James_Carville_30.jpg \n inflating: /data/training/James_Carville_31.jpg \n inflating: /data/training/James_Carville_32.jpg \n inflating: /data/training/James_Carville_50.jpg \n inflating: /data/training/James_Carville_51.jpg \n inflating: /data/training/James_Carville_52.jpg \n inflating: /data/training/James_Cunningham_00.jpg \n inflating: /data/training/James_Cunningham_01.jpg \n inflating: /data/training/James_Cunningham_02.jpg \n inflating: /data/training/James_Cunningham_20.jpg \n inflating: /data/training/James_Cunningham_21.jpg \n inflating: /data/training/James_Cunningham_22.jpg \n inflating: /data/training/James_Cunningham_30.jpg \n inflating: /data/training/James_Cunningham_31.jpg \n inflating: /data/training/James_Cunningham_32.jpg \n inflating: /data/training/James_Cunningham_40.jpg \n inflating: /data/training/James_Cunningham_41.jpg \n inflating: /data/training/James_Cunningham_42.jpg \n inflating: /data/training/James_Hoffa_10.jpg \n inflating: /data/training/James_Hoffa_11.jpg \n inflating: /data/training/James_Hoffa_12.jpg \n inflating: /data/training/James_Hoffa_20.jpg \n inflating: /data/training/James_Hoffa_21.jpg \n inflating: /data/training/James_Hoffa_22.jpg \n inflating: /data/training/James_Hoffa_40.jpg \n inflating: /data/training/James_Hoffa_41.jpg \n inflating: /data/training/James_Hoffa_42.jpg \n inflating: /data/training/James_Hoffa_50.jpg \n inflating: /data/training/James_Hoffa_51.jpg \n inflating: /data/training/James_Hoffa_52.jpg \n inflating: /data/training/James_Lockhart_00.jpg \n inflating: /data/training/James_Lockhart_01.jpg \n inflating: /data/training/James_Lockhart_02.jpg \n inflating: /data/training/James_Lockhart_10.jpg \n inflating: /data/training/James_Lockhart_11.jpg \n inflating: /data/training/James_Lockhart_12.jpg \n inflating: /data/training/James_Lockhart_50.jpg \n inflating: /data/training/James_Lockhart_51.jpg \n inflating: /data/training/James_Lockhart_52.jpg \n inflating: /data/training/James_McPherson_00.jpg \n inflating: /data/training/James_McPherson_01.jpg \n inflating: /data/training/James_McPherson_02.jpg \n inflating: /data/training/James_McPherson_10.jpg \n inflating: /data/training/James_McPherson_11.jpg \n inflating: /data/training/James_McPherson_12.jpg \n inflating: /data/training/James_McPherson_20.jpg \n inflating: /data/training/James_McPherson_21.jpg \n inflating: /data/training/James_McPherson_22.jpg \n inflating: /data/training/James_Wolfensohn_00.jpg \n inflating: /data/training/James_Wolfensohn_01.jpg \n inflating: /data/training/James_Wolfensohn_02.jpg \n inflating: /data/training/James_Wolfensohn_20.jpg \n inflating: /data/training/James_Wolfensohn_21.jpg \n inflating: /data/training/James_Wolfensohn_22.jpg \n inflating: /data/training/James_Wolfensohn_30.jpg \n inflating: /data/training/James_Wolfensohn_31.jpg \n inflating: /data/training/James_Wolfensohn_32.jpg \n inflating: /data/training/James_Wolfensohn_50.jpg \n inflating: /data/training/James_Wolfensohn_51.jpg \n inflating: /data/training/James_Wolfensohn_52.jpg \n inflating: /data/training/Jan_Peter_Balkenende_00.jpg \n inflating: /data/training/Jan_Peter_Balkenende_01.jpg \n inflating: /data/training/Jan_Peter_Balkenende_02.jpg \n inflating: /data/training/Jan_Peter_Balkenende_10.jpg \n inflating: /data/training/Jan_Peter_Balkenende_11.jpg \n inflating: /data/training/Jan_Peter_Balkenende_12.jpg \n inflating: /data/training/Jan_Peter_Balkenende_30.jpg \n inflating: /data/training/Jan_Peter_Balkenende_31.jpg \n inflating: /data/training/Jan_Peter_Balkenende_32.jpg \n inflating: /data/training/Jan_Peter_Balkenende_50.jpg \n inflating: /data/training/Jan_Peter_Balkenende_51.jpg \n inflating: /data/training/Jan_Peter_Balkenende_52.jpg \n inflating: /data/training/Jane_Krakowski_00.jpg \n inflating: /data/training/Jane_Krakowski_01.jpg \n inflating: /data/training/Jane_Krakowski_02.jpg \n inflating: /data/training/Jane_Krakowski_10.jpg \n inflating: /data/training/Jane_Krakowski_11.jpg \n inflating: /data/training/Jane_Krakowski_12.jpg \n inflating: /data/training/Jane_Krakowski_40.jpg \n inflating: /data/training/Jane_Krakowski_41.jpg \n inflating: /data/training/Jane_Krakowski_42.jpg \n inflating: /data/training/Jane_Krakowski_50.jpg \n inflating: /data/training/Jane_Krakowski_51.jpg \n inflating: /data/training/Jane_Krakowski_52.jpg \n inflating: /data/training/Jane_Pauley_10.jpg \n inflating: /data/training/Jane_Pauley_11.jpg \n inflating: /data/training/Jane_Pauley_12.jpg \n inflating: /data/training/Jane_Pauley_30.jpg \n inflating: /data/training/Jane_Pauley_31.jpg \n inflating: /data/training/Jane_Pauley_32.jpg \n inflating: /data/training/Jane_Pauley_40.jpg \n inflating: /data/training/Jane_Pauley_41.jpg \n inflating: /data/training/Jane_Pauley_42.jpg \n inflating: /data/training/Jane_Rooney_00.jpg \n inflating: /data/training/Jane_Rooney_01.jpg \n inflating: /data/training/Jane_Rooney_02.jpg \n inflating: /data/training/Jane_Rooney_10.jpg \n inflating: /data/training/Jane_Rooney_11.jpg \n inflating: /data/training/Jane_Rooney_12.jpg \n inflating: /data/training/Jane_Rooney_20.jpg \n inflating: /data/training/Jane_Rooney_21.jpg \n inflating: /data/training/Jane_Rooney_22.jpg \n inflating: /data/training/Janis_Ruth_Coulter_00.jpg \n inflating: /data/training/Janis_Ruth_Coulter_01.jpg \n inflating: /data/training/Janis_Ruth_Coulter_02.jpg \n inflating: /data/training/Janis_Ruth_Coulter_20.jpg \n inflating: /data/training/Janis_Ruth_Coulter_21.jpg \n inflating: /data/training/Janis_Ruth_Coulter_22.jpg \n inflating: /data/training/Janis_Ruth_Coulter_40.jpg \n inflating: /data/training/Janis_Ruth_Coulter_41.jpg \n inflating: /data/training/Janis_Ruth_Coulter_42.jpg \n inflating: /data/training/Janis_Ruth_Coulter_50.jpg \n inflating: /data/training/Janis_Ruth_Coulter_51.jpg \n inflating: /data/training/Janis_Ruth_Coulter_52.jpg \n inflating: /data/training/JK_Rowling_20.jpg \n inflating: /data/training/JK_Rowling_21.jpg \n inflating: /data/training/JK_Rowling_22.jpg \n inflating: /data/training/JK_Rowling_30.jpg \n inflating: /data/training/JK_Rowling_31.jpg \n inflating: /data/training/JK_Rowling_32.jpg \n inflating: /data/training/JK_Rowling_40.jpg \n inflating: /data/training/JK_Rowling_41.jpg \n inflating: /data/training/JK_Rowling_42.jpg \n inflating: /data/training/JK_Rowling_50.jpg \n inflating: /data/training/JK_Rowling_51.jpg \n inflating: /data/training/JK_Rowling_52.jpg \n inflating: /data/training/Kate_Capshaw_10.jpg \n inflating: /data/training/Kate_Capshaw_11.jpg \n inflating: /data/training/Kate_Capshaw_12.jpg \n inflating: /data/training/Kate_Capshaw_20.jpg \n inflating: /data/training/Kate_Capshaw_21.jpg \n inflating: /data/training/Kate_Capshaw_22.jpg \n inflating: /data/training/Kate_Capshaw_40.jpg \n inflating: /data/training/Kate_Capshaw_41.jpg \n inflating: /data/training/Kate_Capshaw_42.jpg \n inflating: /data/training/Kate_Winslet_00.jpg \n inflating: /data/training/Kate_Winslet_01.jpg \n inflating: /data/training/Kate_Winslet_02.jpg \n inflating: /data/training/Kate_Winslet_10.jpg \n inflating: /data/training/Kate_Winslet_11.jpg \n inflating: /data/training/Kate_Winslet_12.jpg \n inflating: /data/training/Kate_Winslet_50.jpg \n inflating: /data/training/Kate_Winslet_51.jpg \n inflating: /data/training/Kate_Winslet_52.jpg \n inflating: /data/training/Katharine_Hepburn_10.jpg \n inflating: /data/training/Katharine_Hepburn_11.jpg \n inflating: /data/training/Katharine_Hepburn_12.jpg \n inflating: /data/training/Katharine_Hepburn_30.jpg \n inflating: /data/training/Katharine_Hepburn_31.jpg \n inflating: /data/training/Katharine_Hepburn_32.jpg \n inflating: /data/training/Katharine_Hepburn_40.jpg \n inflating: /data/training/Katharine_Hepburn_41.jpg \n inflating: /data/training/Katharine_Hepburn_42.jpg \n inflating: /data/training/Kathryn_Morris_10.jpg \n inflating: /data/training/Kathryn_Morris_11.jpg \n inflating: /data/training/Kathryn_Morris_12.jpg \n inflating: /data/training/Kathryn_Morris_20.jpg \n inflating: /data/training/Kathryn_Morris_21.jpg \n inflating: /data/training/Kathryn_Morris_22.jpg \n inflating: /data/training/Kathryn_Morris_40.jpg \n inflating: /data/training/Kathryn_Morris_41.jpg \n inflating: /data/training/Kathryn_Morris_42.jpg \n inflating: /data/training/Kathryn_Morris_50.jpg \n inflating: /data/training/Kathryn_Morris_51.jpg \n inflating: /data/training/Kathryn_Morris_52.jpg \n inflating: /data/training/Katja_Riemann_00.jpg \n inflating: /data/training/Katja_Riemann_01.jpg \n inflating: /data/training/Katja_Riemann_02.jpg \n inflating: /data/training/Katja_Riemann_10.jpg \n inflating: /data/training/Katja_Riemann_11.jpg \n inflating: /data/training/Katja_Riemann_12.jpg \n inflating: /data/training/Katja_Riemann_20.jpg \n inflating: /data/training/Katja_Riemann_21.jpg \n inflating: /data/training/Katja_Riemann_22.jpg \n inflating: /data/training/Keith_Olbermann_00.jpg \n inflating: /data/training/Keith_Olbermann_01.jpg \n inflating: /data/training/Keith_Olbermann_02.jpg \n inflating: /data/training/Keith_Olbermann_10.jpg \n inflating: /data/training/Keith_Olbermann_11.jpg \n inflating: /data/training/Keith_Olbermann_12.jpg \n inflating: /data/training/Keith_Olbermann_20.jpg \n inflating: /data/training/Keith_Olbermann_21.jpg \n inflating: /data/training/Keith_Olbermann_22.jpg \n inflating: /data/training/Keith_Olbermann_50.jpg \n inflating: /data/training/Keith_Olbermann_51.jpg \n inflating: /data/training/Keith_Olbermann_52.jpg \n inflating: /data/training/Keith_Tyson_00.jpg \n inflating: /data/training/Keith_Tyson_01.jpg \n inflating: /data/training/Keith_Tyson_02.jpg \n inflating: /data/training/Keith_Tyson_10.jpg \n inflating: /data/training/Keith_Tyson_11.jpg \n inflating: /data/training/Keith_Tyson_12.jpg \n inflating: /data/training/Keith_Tyson_20.jpg \n inflating: /data/training/Keith_Tyson_21.jpg \n inflating: /data/training/Keith_Tyson_22.jpg \n inflating: /data/training/Kemal_Dervis_00.jpg \n inflating: /data/training/Kemal_Dervis_01.jpg \n inflating: /data/training/Kemal_Dervis_02.jpg \n inflating: /data/training/Kemal_Dervis_10.jpg \n inflating: /data/training/Kemal_Dervis_11.jpg \n inflating: /data/training/Kemal_Dervis_12.jpg \n inflating: /data/training/Kemal_Dervis_30.jpg \n inflating: /data/training/Kemal_Dervis_31.jpg \n inflating: /data/training/Kemal_Dervis_32.jpg \n inflating: /data/training/Kevin_Satterfield_00.jpg \n inflating: /data/training/Kevin_Satterfield_01.jpg \n inflating: /data/training/Kevin_Satterfield_02.jpg \n inflating: /data/training/Kevin_Satterfield_10.jpg \n inflating: /data/training/Kevin_Satterfield_11.jpg \n inflating: /data/training/Kevin_Satterfield_12.jpg \n inflating: /data/training/Kevin_Satterfield_20.jpg \n inflating: /data/training/Kevin_Satterfield_21.jpg \n inflating: /data/training/Kevin_Satterfield_22.jpg \n inflating: /data/training/Kieran_Culkin_00.jpg \n inflating: /data/training/Kieran_Culkin_01.jpg \n inflating: /data/training/Kieran_Culkin_02.jpg \n inflating: /data/training/Kieran_Culkin_10.jpg \n inflating: /data/training/Kieran_Culkin_11.jpg \n inflating: /data/training/Kieran_Culkin_12.jpg \n inflating: /data/training/Kieran_Culkin_20.jpg \n inflating: /data/training/Kieran_Culkin_21.jpg \n inflating: /data/training/Kieran_Culkin_22.jpg \n inflating: /data/training/Kirk_Ferentz_00.jpg \n inflating: /data/training/Kirk_Ferentz_01.jpg \n inflating: /data/training/Kirk_Ferentz_02.jpg \n inflating: /data/training/Kirk_Ferentz_20.jpg \n inflating: /data/training/Kirk_Ferentz_21.jpg \n inflating: /data/training/Kirk_Ferentz_22.jpg \n inflating: /data/training/Kirk_Ferentz_40.jpg \n inflating: /data/training/Kirk_Ferentz_41.jpg \n inflating: /data/training/Kirk_Ferentz_42.jpg \n inflating: /data/training/Kirk_Ferentz_50.jpg \n inflating: /data/training/Kirk_Ferentz_51.jpg \n inflating: /data/training/Kirk_Ferentz_52.jpg \n inflating: /data/training/Kirsten_Dunst_00.jpg \n inflating: /data/training/Kirsten_Dunst_01.jpg \n inflating: /data/training/Kirsten_Dunst_02.jpg \n inflating: /data/training/Kirsten_Dunst_20.jpg \n inflating: /data/training/Kirsten_Dunst_21.jpg \n inflating: /data/training/Kirsten_Dunst_22.jpg \n inflating: /data/training/Kirsten_Dunst_30.jpg \n inflating: /data/training/Kirsten_Dunst_31.jpg \n inflating: /data/training/Kirsten_Dunst_32.jpg \n inflating: /data/training/Kit_Bond_10.jpg \n inflating: /data/training/Kit_Bond_11.jpg \n inflating: /data/training/Kit_Bond_12.jpg \n inflating: /data/training/Kit_Bond_20.jpg \n inflating: /data/training/Kit_Bond_21.jpg \n inflating: /data/training/Kit_Bond_22.jpg \n inflating: /data/training/Kit_Bond_30.jpg \n inflating: /data/training/Kit_Bond_31.jpg \n inflating: /data/training/Kit_Bond_32.jpg \n inflating: /data/training/Kit_Bond_50.jpg \n inflating: /data/training/Kit_Bond_51.jpg \n inflating: /data/training/Kit_Bond_52.jpg \n inflating: /data/training/Kristen_Breitweiser_00.jpg \n inflating: /data/training/Kristen_Breitweiser_01.jpg \n inflating: /data/training/Kristen_Breitweiser_02.jpg \n inflating: /data/training/Kristen_Breitweiser_10.jpg \n inflating: /data/training/Kristen_Breitweiser_11.jpg \n inflating: /data/training/Kristen_Breitweiser_12.jpg \n inflating: /data/training/Kristen_Breitweiser_20.jpg \n inflating: /data/training/Kristen_Breitweiser_21.jpg \n inflating: /data/training/Kristen_Breitweiser_22.jpg \n inflating: /data/training/Kristen_Breitweiser_50.jpg \n inflating: /data/training/Kristen_Breitweiser_51.jpg \n inflating: /data/training/Kristen_Breitweiser_52.jpg \n inflating: /data/training/Kristin_Chenoweth_10.jpg \n inflating: /data/training/Kristin_Chenoweth_11.jpg \n inflating: /data/training/Kristin_Chenoweth_12.jpg \n inflating: /data/training/Kristin_Chenoweth_40.jpg \n inflating: /data/training/Kristin_Chenoweth_41.jpg \n inflating: /data/training/Kristin_Chenoweth_42.jpg \n inflating: /data/training/Kristin_Chenoweth_50.jpg \n inflating: /data/training/Kristin_Chenoweth_51.jpg \n inflating: /data/training/Kristin_Chenoweth_52.jpg \n inflating: /data/training/Kristin_Scott_10.jpg \n inflating: /data/training/Kristin_Scott_11.jpg \n inflating: /data/training/Kristin_Scott_12.jpg \n inflating: /data/training/Kristin_Scott_40.jpg \n inflating: /data/training/Kristin_Scott_41.jpg \n inflating: /data/training/Kristin_Scott_42.jpg \n inflating: /data/training/Kristin_Scott_50.jpg \n inflating: /data/training/Kristin_Scott_51.jpg \n inflating: /data/training/Kristin_Scott_52.jpg \n inflating: /data/training/Kristy_Curry_00.jpg \n inflating: /data/training/Kristy_Curry_01.jpg \n inflating: /data/training/Kristy_Curry_02.jpg \n inflating: /data/training/Kristy_Curry_20.jpg \n inflating: /data/training/Kristy_Curry_21.jpg \n inflating: /data/training/Kristy_Curry_22.jpg \n inflating: /data/training/Kristy_Curry_30.jpg \n inflating: /data/training/Kristy_Curry_31.jpg \n inflating: /data/training/Kristy_Curry_32.jpg \n inflating: /data/training/Kurt_Warner_00.jpg \n inflating: /data/training/Kurt_Warner_01.jpg \n inflating: /data/training/Kurt_Warner_02.jpg \n inflating: /data/training/Kurt_Warner_10.jpg \n inflating: /data/training/Kurt_Warner_11.jpg \n inflating: /data/training/Kurt_Warner_12.jpg \n inflating: /data/training/Kurt_Warner_40.jpg \n inflating: /data/training/Kurt_Warner_41.jpg \n inflating: /data/training/Kurt_Warner_42.jpg \n inflating: /data/training/Kweisi_Mfume_00.jpg \n inflating: /data/training/Kweisi_Mfume_01.jpg \n inflating: /data/training/Kweisi_Mfume_02.jpg \n inflating: /data/training/Kweisi_Mfume_10.jpg \n inflating: /data/training/Kweisi_Mfume_11.jpg \n inflating: /data/training/Kweisi_Mfume_12.jpg \n inflating: /data/training/Kweisi_Mfume_40.jpg \n inflating: /data/training/Kweisi_Mfume_41.jpg \n inflating: /data/training/Kweisi_Mfume_42.jpg \n inflating: /data/training/Kweisi_Mfume_50.jpg \n inflating: /data/training/Kweisi_Mfume_51.jpg \n inflating: /data/training/Kweisi_Mfume_52.jpg \n inflating: /data/training/Kyle_Shewfelt_00.jpg \n inflating: /data/training/Kyle_Shewfelt_01.jpg \n inflating: /data/training/Kyle_Shewfelt_02.jpg \n inflating: /data/training/Kyle_Shewfelt_10.jpg \n inflating: /data/training/Kyle_Shewfelt_11.jpg \n inflating: /data/training/Kyle_Shewfelt_12.jpg \n inflating: /data/training/Kyle_Shewfelt_20.jpg \n inflating: /data/training/Kyle_Shewfelt_21.jpg \n inflating: /data/training/Kyle_Shewfelt_22.jpg \n inflating: /data/training/Kyle_Shewfelt_40.jpg \n inflating: /data/training/Kyle_Shewfelt_41.jpg \n inflating: /data/training/Kyle_Shewfelt_42.jpg \n inflating: /data/training/Larry_Flynt_00.jpg \n inflating: /data/training/Larry_Flynt_01.jpg \n inflating: /data/training/Larry_Flynt_02.jpg \n inflating: /data/training/Larry_Flynt_10.jpg \n inflating: /data/training/Larry_Flynt_11.jpg \n inflating: /data/training/Larry_Flynt_12.jpg \n inflating: /data/training/Larry_Flynt_20.jpg \n inflating: /data/training/Larry_Flynt_21.jpg \n inflating: /data/training/Larry_Flynt_22.jpg \n inflating: /data/training/Laura_Bozzo_00.jpg \n inflating: /data/training/Laura_Bozzo_01.jpg \n inflating: /data/training/Laura_Bozzo_02.jpg \n inflating: /data/training/Laura_Bozzo_10.jpg \n inflating: /data/training/Laura_Bozzo_11.jpg \n inflating: /data/training/Laura_Bozzo_12.jpg \n inflating: /data/training/Laura_Bozzo_40.jpg \n inflating: /data/training/Laura_Bozzo_41.jpg \n inflating: /data/training/Laura_Bozzo_42.jpg \n inflating: /data/training/Laura_Bush_10.jpg \n inflating: /data/training/Laura_Bush_11.jpg \n inflating: /data/training/Laura_Bush_12.jpg \n inflating: /data/training/Laura_Bush_20.jpg \n inflating: /data/training/Laura_Bush_21.jpg \n inflating: /data/training/Laura_Bush_22.jpg \n inflating: /data/training/Laura_Bush_40.jpg \n inflating: /data/training/Laura_Bush_41.jpg \n inflating: /data/training/Laura_Bush_42.jpg \n inflating: /data/training/Laura_Bush_50.jpg \n inflating: /data/training/Laura_Bush_51.jpg \n inflating: /data/training/Laura_Bush_52.jpg \n inflating: /data/training/Laura_Elena_Harring_00.jpg \n inflating: /data/training/Laura_Elena_Harring_01.jpg \n inflating: /data/training/Laura_Elena_Harring_02.jpg \n inflating: /data/training/Laura_Elena_Harring_20.jpg \n inflating: /data/training/Laura_Elena_Harring_21.jpg \n inflating: /data/training/Laura_Elena_Harring_22.jpg \n inflating: /data/training/Laura_Elena_Harring_40.jpg \n inflating: /data/training/Laura_Elena_Harring_41.jpg \n inflating: /data/training/Laura_Elena_Harring_42.jpg \n inflating: /data/training/Laura_Elena_Harring_50.jpg \n inflating: /data/training/Laura_Elena_Harring_51.jpg \n inflating: /data/training/Laura_Elena_Harring_52.jpg \n inflating: /data/training/Laurence_Fishburne_20.jpg \n inflating: /data/training/Laurence_Fishburne_21.jpg \n inflating: /data/training/Laurence_Fishburne_22.jpg \n inflating: /data/training/Laurence_Fishburne_40.jpg \n inflating: /data/training/Laurence_Fishburne_41.jpg \n inflating: /data/training/Laurence_Fishburne_42.jpg \n inflating: /data/training/Laurence_Fishburne_50.jpg \n inflating: /data/training/Laurence_Fishburne_51.jpg \n inflating: /data/training/Laurence_Fishburne_52.jpg \n inflating: /data/training/Lee_Baca_00.jpg \n inflating: /data/training/Lee_Baca_01.jpg \n inflating: /data/training/Lee_Baca_02.jpg \n inflating: /data/training/Lee_Baca_10.jpg \n inflating: /data/training/Lee_Baca_11.jpg \n inflating: /data/training/Lee_Baca_12.jpg \n inflating: /data/training/Lee_Baca_40.jpg \n inflating: /data/training/Lee_Baca_41.jpg \n inflating: /data/training/Lee_Baca_42.jpg \n inflating: /data/training/Lene_Espersen_10.jpg \n inflating: /data/training/Lene_Espersen_11.jpg \n inflating: /data/training/Lene_Espersen_12.jpg \n inflating: /data/training/Lene_Espersen_20.jpg \n inflating: /data/training/Lene_Espersen_21.jpg \n inflating: /data/training/Lene_Espersen_22.jpg \n inflating: /data/training/Lene_Espersen_40.jpg \n inflating: /data/training/Lene_Espersen_41.jpg \n inflating: /data/training/Lene_Espersen_42.jpg \n inflating: /data/training/Lesia_Burlak_00.jpg \n inflating: /data/training/Lesia_Burlak_01.jpg \n inflating: /data/training/Lesia_Burlak_02.jpg \n inflating: /data/training/Lesia_Burlak_20.jpg \n inflating: /data/training/Lesia_Burlak_21.jpg \n inflating: /data/training/Lesia_Burlak_22.jpg \n inflating: /data/training/Lesia_Burlak_30.jpg \n inflating: /data/training/Lesia_Burlak_31.jpg \n inflating: /data/training/Lesia_Burlak_32.jpg \n inflating: /data/training/Lester_Holt_00.jpg \n inflating: /data/training/Lester_Holt_01.jpg \n inflating: /data/training/Lester_Holt_02.jpg \n inflating: /data/training/Lester_Holt_30.jpg \n inflating: /data/training/Lester_Holt_31.jpg \n inflating: /data/training/Lester_Holt_32.jpg \n inflating: /data/training/Lester_Holt_40.jpg \n inflating: /data/training/Lester_Holt_41.jpg \n inflating: /data/training/Lester_Holt_42.jpg \n inflating: /data/training/Leszek_Miller_00.jpg \n inflating: /data/training/Leszek_Miller_01.jpg \n inflating: /data/training/Leszek_Miller_02.jpg \n inflating: /data/training/Leszek_Miller_10.jpg \n inflating: /data/training/Leszek_Miller_11.jpg \n inflating: /data/training/Leszek_Miller_12.jpg \n inflating: /data/training/Leszek_Miller_30.jpg \n inflating: /data/training/Leszek_Miller_31.jpg \n inflating: /data/training/Leszek_Miller_32.jpg \n inflating: /data/training/Leticia_Van_de_Putte_00.jpg \n inflating: /data/training/Leticia_Van_de_Putte_01.jpg \n inflating: /data/training/Leticia_Van_de_Putte_02.jpg \n inflating: /data/training/Leticia_Van_de_Putte_10.jpg \n inflating: /data/training/Leticia_Van_de_Putte_11.jpg \n inflating: /data/training/Leticia_Van_de_Putte_12.jpg \n inflating: /data/training/Leticia_Van_de_Putte_40.jpg \n inflating: /data/training/Leticia_Van_de_Putte_41.jpg \n inflating: /data/training/Leticia_Van_de_Putte_42.jpg \n inflating: /data/training/Leuris_Pupo_00.jpg \n inflating: /data/training/Leuris_Pupo_01.jpg \n inflating: /data/training/Leuris_Pupo_02.jpg \n inflating: /data/training/Leuris_Pupo_20.jpg \n inflating: /data/training/Leuris_Pupo_21.jpg \n inflating: /data/training/Leuris_Pupo_22.jpg \n inflating: /data/training/Leuris_Pupo_30.jpg \n inflating: /data/training/Leuris_Pupo_31.jpg \n inflating: /data/training/Leuris_Pupo_32.jpg \n inflating: /data/training/Leuris_Pupo_40.jpg \n inflating: /data/training/Leuris_Pupo_41.jpg \n inflating: /data/training/Leuris_Pupo_42.jpg \n inflating: /data/training/Li_Zhaoxing_00.jpg \n inflating: /data/training/Li_Zhaoxing_01.jpg \n inflating: /data/training/Li_Zhaoxing_02.jpg \n inflating: /data/training/Li_Zhaoxing_30.jpg \n inflating: /data/training/Li_Zhaoxing_31.jpg \n inflating: /data/training/Li_Zhaoxing_32.jpg \n inflating: /data/training/Li_Zhaoxing_40.jpg \n inflating: /data/training/Li_Zhaoxing_41.jpg \n inflating: /data/training/Li_Zhaoxing_42.jpg \n inflating: /data/training/Lincoln_Chafee_20.jpg \n inflating: /data/training/Lincoln_Chafee_21.jpg \n inflating: /data/training/Lincoln_Chafee_22.jpg \n inflating: /data/training/Lincoln_Chafee_30.jpg \n inflating: /data/training/Lincoln_Chafee_31.jpg \n inflating: /data/training/Lincoln_Chafee_32.jpg \n inflating: /data/training/Lincoln_Chafee_50.jpg \n inflating: /data/training/Lincoln_Chafee_51.jpg \n inflating: /data/training/Lincoln_Chafee_52.jpg \n inflating: /data/training/Linda_Dano_00.jpg \n inflating: /data/training/Linda_Dano_01.jpg \n inflating: /data/training/Linda_Dano_02.jpg \n inflating: /data/training/Linda_Dano_20.jpg \n inflating: /data/training/Linda_Dano_21.jpg \n inflating: /data/training/Linda_Dano_22.jpg \n inflating: /data/training/Linda_Dano_30.jpg \n inflating: /data/training/Linda_Dano_31.jpg \n inflating: /data/training/Linda_Dano_32.jpg \n inflating: /data/training/Linda_Dano_50.jpg \n inflating: /data/training/Linda_Dano_51.jpg \n inflating: /data/training/Linda_Dano_52.jpg \n inflating: /data/training/Linda_Franklin_00.jpg \n inflating: /data/training/Linda_Franklin_01.jpg \n inflating: /data/training/Linda_Franklin_02.jpg \n inflating: /data/training/Linda_Franklin_10.jpg \n inflating: /data/training/Linda_Franklin_11.jpg \n inflating: /data/training/Linda_Franklin_12.jpg \n inflating: /data/training/Linda_Franklin_20.jpg \n inflating: /data/training/Linda_Franklin_21.jpg \n inflating: /data/training/Linda_Franklin_22.jpg \n inflating: /data/training/Linda_Franklin_40.jpg \n inflating: /data/training/Linda_Franklin_41.jpg \n inflating: /data/training/Linda_Franklin_42.jpg \n inflating: /data/training/Linda_Sanchez_00.jpg \n inflating: /data/training/Linda_Sanchez_01.jpg \n inflating: /data/training/Linda_Sanchez_02.jpg \n inflating: /data/training/Linda_Sanchez_10.jpg \n inflating: /data/training/Linda_Sanchez_11.jpg \n inflating: /data/training/Linda_Sanchez_12.jpg \n inflating: /data/training/Linda_Sanchez_20.jpg \n inflating: /data/training/Linda_Sanchez_21.jpg \n inflating: /data/training/Linda_Sanchez_22.jpg \n inflating: /data/training/Linda_Sanchez_40.jpg \n inflating: /data/training/Linda_Sanchez_41.jpg \n inflating: /data/training/Linda_Sanchez_42.jpg \n inflating: /data/training/Lindsey_Graham_00.jpg \n inflating: /data/training/Lindsey_Graham_01.jpg \n inflating: /data/training/Lindsey_Graham_02.jpg \n inflating: /data/training/Lindsey_Graham_10.jpg \n inflating: /data/training/Lindsey_Graham_11.jpg \n inflating: /data/training/Lindsey_Graham_12.jpg \n inflating: /data/training/Lindsey_Graham_20.jpg \n inflating: /data/training/Lindsey_Graham_21.jpg \n inflating: /data/training/Lindsey_Graham_22.jpg \n inflating: /data/training/Lindsey_Graham_30.jpg \n inflating: /data/training/Lindsey_Graham_31.jpg \n inflating: /data/training/Lindsey_Graham_32.jpg \n inflating: /data/training/Lino_Oviedo_00.jpg \n inflating: /data/training/Lino_Oviedo_01.jpg \n inflating: /data/training/Lino_Oviedo_02.jpg \n inflating: /data/training/Lino_Oviedo_30.jpg \n inflating: /data/training/Lino_Oviedo_31.jpg \n inflating: /data/training/Lino_Oviedo_32.jpg \n inflating: /data/training/Lino_Oviedo_50.jpg \n inflating: /data/training/Lino_Oviedo_51.jpg \n inflating: /data/training/Lino_Oviedo_52.jpg \n inflating: /data/training/Lisa_Ling_00.jpg \n inflating: /data/training/Lisa_Ling_01.jpg \n inflating: /data/training/Lisa_Ling_02.jpg \n inflating: /data/training/Lisa_Ling_10.jpg \n inflating: /data/training/Lisa_Ling_11.jpg \n inflating: /data/training/Lisa_Ling_12.jpg \n inflating: /data/training/Lisa_Ling_20.jpg \n inflating: /data/training/Lisa_Ling_21.jpg \n inflating: /data/training/Lisa_Ling_22.jpg \n inflating: /data/training/Liu_Ye_00.jpg \n inflating: /data/training/Liu_Ye_01.jpg \n inflating: /data/training/Liu_Ye_02.jpg \n inflating: /data/training/Liu_Ye_10.jpg \n inflating: /data/training/Liu_Ye_11.jpg \n inflating: /data/training/Liu_Ye_12.jpg \n inflating: /data/training/Liu_Ye_20.jpg \n inflating: /data/training/Liu_Ye_21.jpg \n inflating: /data/training/Liu_Ye_22.jpg \n inflating: /data/training/Liu_Ye_50.jpg \n inflating: /data/training/Liu_Ye_51.jpg \n inflating: /data/training/Liu_Ye_52.jpg \n inflating: /data/training/Loretta_Lynn_Harper_00.jpg \n inflating: /data/training/Loretta_Lynn_Harper_01.jpg \n inflating: /data/training/Loretta_Lynn_Harper_02.jpg \n inflating: /data/training/Loretta_Lynn_Harper_30.jpg \n inflating: /data/training/Loretta_Lynn_Harper_31.jpg \n inflating: /data/training/Loretta_Lynn_Harper_32.jpg \n inflating: /data/training/Loretta_Lynn_Harper_40.jpg \n inflating: /data/training/Loretta_Lynn_Harper_41.jpg \n inflating: /data/training/Loretta_Lynn_Harper_42.jpg \n inflating: /data/training/Loretta_Lynn_Harper_50.jpg \n inflating: /data/training/Loretta_Lynn_Harper_51.jpg \n inflating: /data/training/Loretta_Lynn_Harper_52.jpg \n inflating: /data/training/Louis_Van_Gaal_00.jpg \n inflating: /data/training/Louis_Van_Gaal_01.jpg \n inflating: /data/training/Louis_Van_Gaal_02.jpg \n inflating: /data/training/Louis_Van_Gaal_10.jpg \n inflating: /data/training/Louis_Van_Gaal_11.jpg \n inflating: /data/training/Louis_Van_Gaal_12.jpg \n inflating: /data/training/Louis_Van_Gaal_40.jpg \n inflating: /data/training/Louis_Van_Gaal_41.jpg \n inflating: /data/training/Louis_Van_Gaal_42.jpg \n inflating: /data/training/Louisa_Baileche_00.jpg \n inflating: /data/training/Louisa_Baileche_01.jpg \n inflating: /data/training/Louisa_Baileche_02.jpg \n inflating: /data/training/Louisa_Baileche_10.jpg \n inflating: /data/training/Louisa_Baileche_11.jpg \n inflating: /data/training/Louisa_Baileche_12.jpg \n inflating: /data/training/Louisa_Baileche_20.jpg \n inflating: /data/training/Louisa_Baileche_21.jpg \n inflating: /data/training/Louisa_Baileche_22.jpg \n inflating: /data/training/Luc_Montagnier_20.jpg \n inflating: /data/training/Luc_Montagnier_21.jpg \n inflating: /data/training/Luc_Montagnier_22.jpg \n inflating: /data/training/Luc_Montagnier_40.jpg \n inflating: /data/training/Luc_Montagnier_41.jpg \n inflating: /data/training/Luc_Montagnier_42.jpg \n inflating: /data/training/Luc_Montagnier_50.jpg \n inflating: /data/training/Luc_Montagnier_51.jpg \n inflating: /data/training/Luc_Montagnier_52.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_00.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_01.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_02.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_10.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_11.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_12.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_40.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_41.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_42.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_50.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_51.jpg \n inflating: /data/training/Lucia_Kenny_Anthony_52.jpg \n inflating: /data/training/Lucio_Stanca_00.jpg \n inflating: /data/training/Lucio_Stanca_01.jpg \n inflating: /data/training/Lucio_Stanca_02.jpg \n inflating: /data/training/Lucio_Stanca_20.jpg \n inflating: /data/training/Lucio_Stanca_21.jpg \n inflating: /data/training/Lucio_Stanca_22.jpg \n inflating: /data/training/Lucio_Stanca_30.jpg \n inflating: /data/training/Lucio_Stanca_31.jpg \n inflating: /data/training/Lucio_Stanca_32.jpg \n inflating: /data/training/Lucio_Stanca_40.jpg \n inflating: /data/training/Lucio_Stanca_41.jpg \n inflating: /data/training/Lucio_Stanca_42.jpg \n inflating: /data/training/Luis_Ernesto_Derbez_Bautista_00.jpg \n inflating: /data/training/Luis_Ernesto_Derbez_Bautista_01.jpg \n inflating: /data/training/Luis_Ernesto_Derbez_Bautista_02.jpg \n inflating: /data/training/Luis_Ernesto_Derbez_Bautista_10.jpg \n inflating: /data/training/Luis_Ernesto_Derbez_Bautista_11.jpg \n inflating: /data/training/Luis_Ernesto_Derbez_Bautista_12.jpg \n inflating: /data/training/Luis_Ernesto_Derbez_Bautista_50.jpg \n inflating: /data/training/Luis_Ernesto_Derbez_Bautista_51.jpg \n inflating: /data/training/Luis_Ernesto_Derbez_Bautista_52.jpg \n inflating: /data/training/Luis_Fonsi_20.jpg \n inflating: /data/training/Luis_Fonsi_21.jpg \n inflating: /data/training/Luis_Fonsi_22.jpg \n inflating: /data/training/Luis_Fonsi_40.jpg \n inflating: /data/training/Luis_Fonsi_41.jpg \n inflating: /data/training/Luis_Fonsi_42.jpg \n inflating: /data/training/Luis_Fonsi_50.jpg \n inflating: /data/training/Luis_Fonsi_51.jpg \n inflating: /data/training/Luis_Fonsi_52.jpg \n inflating: /data/training/Lyle_Lovett_20.jpg \n inflating: /data/training/Lyle_Lovett_21.jpg \n inflating: /data/training/Lyle_Lovett_22.jpg \n inflating: /data/training/Lyle_Lovett_40.jpg \n inflating: /data/training/Lyle_Lovett_41.jpg \n inflating: /data/training/Lyle_Lovett_42.jpg \n inflating: /data/training/Lyle_Lovett_50.jpg \n inflating: /data/training/Lyle_Lovett_51.jpg \n inflating: /data/training/Lyle_Lovett_52.jpg \n inflating: /data/training/Mack_Brown_00.jpg \n inflating: /data/training/Mack_Brown_01.jpg \n inflating: /data/training/Mack_Brown_02.jpg \n inflating: /data/training/Mack_Brown_40.jpg \n inflating: /data/training/Mack_Brown_41.jpg \n inflating: /data/training/Mack_Brown_42.jpg \n inflating: /data/training/Mack_Brown_50.jpg \n inflating: /data/training/Mack_Brown_51.jpg \n inflating: /data/training/Mack_Brown_52.jpg \n inflating: /data/training/Maggie_Cheung_00.jpg \n inflating: /data/training/Maggie_Cheung_01.jpg \n inflating: /data/training/Maggie_Cheung_02.jpg \n inflating: /data/training/Maggie_Cheung_30.jpg \n inflating: /data/training/Maggie_Cheung_31.jpg \n inflating: /data/training/Maggie_Cheung_32.jpg \n inflating: /data/training/Maggie_Cheung_50.jpg \n inflating: /data/training/Maggie_Cheung_51.jpg \n inflating: /data/training/Maggie_Cheung_52.jpg \n inflating: /data/training/Maggie_Smith_00.jpg \n inflating: /data/training/Maggie_Smith_01.jpg \n inflating: /data/training/Maggie_Smith_02.jpg \n inflating: /data/training/Maggie_Smith_30.jpg \n inflating: /data/training/Maggie_Smith_31.jpg \n inflating: /data/training/Maggie_Smith_32.jpg \n inflating: /data/training/Maggie_Smith_40.jpg \n inflating: /data/training/Maggie_Smith_41.jpg \n inflating: /data/training/Maggie_Smith_42.jpg \n inflating: /data/training/Mahathir_Mohamad_00.jpg \n inflating: /data/training/Mahathir_Mohamad_01.jpg \n inflating: /data/training/Mahathir_Mohamad_02.jpg \n inflating: /data/training/Mahathir_Mohamad_10.jpg \n inflating: /data/training/Mahathir_Mohamad_11.jpg \n inflating: /data/training/Mahathir_Mohamad_12.jpg \n inflating: /data/training/Mahathir_Mohamad_20.jpg \n inflating: /data/training/Mahathir_Mohamad_21.jpg \n inflating: /data/training/Mahathir_Mohamad_22.jpg \n inflating: /data/training/Mahathir_Mohamad_30.jpg \n inflating: /data/training/Mahathir_Mohamad_31.jpg \n inflating: /data/training/Mahathir_Mohamad_32.jpg \n inflating: /data/training/Malcolm_Jamal_Warner_00.jpg \n inflating: /data/training/Malcolm_Jamal_Warner_01.jpg \n inflating: /data/training/Malcolm_Jamal_Warner_02.jpg \n inflating: /data/training/Malcolm_Jamal_Warner_10.jpg \n inflating: /data/training/Malcolm_Jamal_Warner_11.jpg \n inflating: /data/training/Malcolm_Jamal_Warner_12.jpg \n inflating: /data/training/Malcolm_Jamal_Warner_20.jpg \n inflating: /data/training/Malcolm_Jamal_Warner_21.jpg \n inflating: /data/training/Malcolm_Jamal_Warner_22.jpg \n inflating: /data/training/Manuel_Pellegrini_10.jpg \n inflating: /data/training/Manuel_Pellegrini_11.jpg \n inflating: /data/training/Manuel_Pellegrini_12.jpg \n inflating: /data/training/Manuel_Pellegrini_20.jpg \n inflating: /data/training/Manuel_Pellegrini_21.jpg \n inflating: /data/training/Manuel_Pellegrini_22.jpg \n inflating: /data/training/Manuel_Pellegrini_30.jpg \n inflating: /data/training/Manuel_Pellegrini_31.jpg \n inflating: /data/training/Manuel_Pellegrini_32.jpg \n inflating: /data/training/Marc_Anthony_10.jpg \n inflating: /data/training/Marc_Anthony_11.jpg \n inflating: /data/training/Marc_Anthony_12.jpg \n inflating: /data/training/Marc_Anthony_20.jpg \n inflating: /data/training/Marc_Anthony_21.jpg \n inflating: /data/training/Marc_Anthony_22.jpg \n inflating: /data/training/Marc_Anthony_50.jpg \n inflating: /data/training/Marc_Anthony_51.jpg \n inflating: /data/training/Marc_Anthony_52.jpg \n inflating: /data/training/Marc_Racicot_00.jpg \n inflating: /data/training/Marc_Racicot_01.jpg \n inflating: /data/training/Marc_Racicot_02.jpg \n inflating: /data/training/Marc_Racicot_20.jpg \n inflating: /data/training/Marc_Racicot_21.jpg \n inflating: /data/training/Marc_Racicot_22.jpg \n inflating: /data/training/Marc_Racicot_40.jpg \n inflating: /data/training/Marc_Racicot_41.jpg \n inflating: /data/training/Marc_Racicot_42.jpg \n inflating: /data/training/Marc_Racicot_50.jpg \n inflating: /data/training/Marc_Racicot_51.jpg \n inflating: /data/training/Marc_Racicot_52.jpg \n inflating: /data/training/Marc_Shaiman_10.jpg \n inflating: /data/training/Marc_Shaiman_11.jpg \n inflating: /data/training/Marc_Shaiman_12.jpg \n inflating: /data/training/Marc_Shaiman_20.jpg \n inflating: /data/training/Marc_Shaiman_21.jpg \n inflating: /data/training/Marc_Shaiman_22.jpg \n inflating: /data/training/Marc_Shaiman_30.jpg \n inflating: /data/training/Marc_Shaiman_31.jpg \n inflating: /data/training/Marc_Shaiman_32.jpg \n inflating: /data/training/Margaret_Thatcher_10.jpg \n inflating: /data/training/Margaret_Thatcher_11.jpg \n inflating: /data/training/Margaret_Thatcher_12.jpg \n inflating: /data/training/Margaret_Thatcher_30.jpg \n inflating: /data/training/Margaret_Thatcher_31.jpg \n inflating: /data/training/Margaret_Thatcher_32.jpg \n inflating: /data/training/Margaret_Thatcher_40.jpg \n inflating: /data/training/Margaret_Thatcher_41.jpg \n inflating: /data/training/Margaret_Thatcher_42.jpg \n inflating: /data/training/Margaret_Thatcher_50.jpg \n inflating: /data/training/Margaret_Thatcher_51.jpg \n inflating: /data/training/Margaret_Thatcher_52.jpg \n inflating: /data/training/Maria_Soledad_Alvear_Valenzuela_10.jpg \n inflating: /data/training/Maria_Soledad_Alvear_Valenzuela_11.jpg \n inflating: /data/training/Maria_Soledad_Alvear_Valenzuela_12.jpg \n inflating: /data/training/Maria_Soledad_Alvear_Valenzuela_30.jpg \n inflating: /data/training/Maria_Soledad_Alvear_Valenzuela_31.jpg \n inflating: /data/training/Maria_Soledad_Alvear_Valenzuela_32.jpg \n inflating: /data/training/Maria_Soledad_Alvear_Valenzuela_40.jpg \n inflating: /data/training/Maria_Soledad_Alvear_Valenzuela_41.jpg \n inflating: /data/training/Maria_Soledad_Alvear_Valenzuela_42.jpg \n inflating: /data/training/Mariana_Ohata_00.jpg \n inflating: /data/training/Mariana_Ohata_01.jpg \n inflating: /data/training/Mariana_Ohata_02.jpg \n inflating: /data/training/Mariana_Ohata_20.jpg \n inflating: /data/training/Mariana_Ohata_21.jpg \n inflating: /data/training/Mariana_Ohata_22.jpg \n inflating: /data/training/Mariana_Ohata_30.jpg \n inflating: /data/training/Mariana_Ohata_31.jpg \n inflating: /data/training/Mariana_Ohata_32.jpg \n inflating: /data/training/Marieta_Chrousala_00.jpg \n inflating: /data/training/Marieta_Chrousala_01.jpg \n inflating: /data/training/Marieta_Chrousala_02.jpg \n inflating: /data/training/Marieta_Chrousala_10.jpg \n inflating: /data/training/Marieta_Chrousala_11.jpg \n inflating: /data/training/Marieta_Chrousala_12.jpg \n inflating: /data/training/Marieta_Chrousala_40.jpg \n inflating: /data/training/Marieta_Chrousala_41.jpg \n inflating: /data/training/Marieta_Chrousala_42.jpg \n inflating: /data/training/Marina_Silva_10.jpg \n inflating: /data/training/Marina_Silva_11.jpg \n inflating: /data/training/Marina_Silva_12.jpg \n inflating: /data/training/Marina_Silva_20.jpg \n inflating: /data/training/Marina_Silva_21.jpg \n inflating: /data/training/Marina_Silva_22.jpg \n inflating: /data/training/Marina_Silva_40.jpg \n inflating: /data/training/Marina_Silva_41.jpg \n inflating: /data/training/Marina_Silva_42.jpg \n inflating: /data/training/Marina_Silva_50.jpg \n inflating: /data/training/Marina_Silva_51.jpg \n inflating: /data/training/Marina_Silva_52.jpg \n inflating: /data/training/Mario_Kreutzberger_20.jpg \n inflating: /data/training/Mario_Kreutzberger_21.jpg \n inflating: /data/training/Mario_Kreutzberger_22.jpg \n inflating: /data/training/Mario_Kreutzberger_30.jpg \n inflating: /data/training/Mario_Kreutzberger_31.jpg \n inflating: /data/training/Mario_Kreutzberger_32.jpg \n inflating: /data/training/Mario_Kreutzberger_40.jpg \n inflating: /data/training/Mario_Kreutzberger_41.jpg \n inflating: /data/training/Mario_Kreutzberger_42.jpg \n inflating: /data/training/Marisa_Tomei_10.jpg \n inflating: /data/training/Marisa_Tomei_11.jpg \n inflating: /data/training/Marisa_Tomei_12.jpg \n inflating: /data/training/Marisa_Tomei_20.jpg \n inflating: /data/training/Marisa_Tomei_21.jpg \n inflating: /data/training/Marisa_Tomei_22.jpg \n inflating: /data/training/Marisa_Tomei_40.jpg \n inflating: /data/training/Marisa_Tomei_41.jpg \n inflating: /data/training/Marisa_Tomei_42.jpg \n inflating: /data/training/Marissa_Jaret_Winokur_00.jpg \n inflating: /data/training/Marissa_Jaret_Winokur_01.jpg \n inflating: /data/training/Marissa_Jaret_Winokur_02.jpg \n inflating: /data/training/Marissa_Jaret_Winokur_30.jpg \n inflating: /data/training/Marissa_Jaret_Winokur_31.jpg \n inflating: /data/training/Marissa_Jaret_Winokur_32.jpg \n inflating: /data/training/Marissa_Jaret_Winokur_40.jpg \n inflating: /data/training/Marissa_Jaret_Winokur_41.jpg \n inflating: /data/training/Marissa_Jaret_Winokur_42.jpg \n inflating: /data/training/Mark_Foley_10.jpg \n inflating: /data/training/Mark_Foley_11.jpg \n inflating: /data/training/Mark_Foley_12.jpg \n inflating: /data/training/Mark_Foley_40.jpg \n inflating: /data/training/Mark_Foley_41.jpg \n inflating: /data/training/Mark_Foley_42.jpg \n inflating: /data/training/Mark_Foley_50.jpg \n inflating: /data/training/Mark_Foley_51.jpg \n inflating: /data/training/Mark_Foley_52.jpg \n inflating: /data/training/Mark_Leno_10.jpg \n inflating: /data/training/Mark_Leno_11.jpg \n inflating: /data/training/Mark_Leno_12.jpg \n inflating: /data/training/Mark_Leno_20.jpg \n inflating: /data/training/Mark_Leno_21.jpg \n inflating: /data/training/Mark_Leno_22.jpg \n inflating: /data/training/Mark_Leno_30.jpg \n inflating: /data/training/Mark_Leno_31.jpg \n inflating: /data/training/Mark_Leno_32.jpg \n inflating: /data/training/Martin_Luther_King_III_00.jpg \n inflating: /data/training/Martin_Luther_King_III_01.jpg \n inflating: /data/training/Martin_Luther_King_III_02.jpg \n inflating: /data/training/Martin_Luther_King_III_30.jpg \n inflating: /data/training/Martin_Luther_King_III_31.jpg \n inflating: /data/training/Martin_Luther_King_III_32.jpg \n inflating: /data/training/Martin_Luther_King_III_50.jpg \n inflating: /data/training/Martin_Luther_King_III_51.jpg \n inflating: /data/training/Martin_Luther_King_III_52.jpg \n inflating: /data/training/Martin_Sheen_00.jpg \n inflating: /data/training/Martin_Sheen_01.jpg \n inflating: /data/training/Martin_Sheen_02.jpg \n inflating: /data/training/Martin_Sheen_30.jpg \n inflating: /data/training/Martin_Sheen_31.jpg \n inflating: /data/training/Martin_Sheen_32.jpg \n inflating: /data/training/Martin_Sheen_40.jpg \n inflating: /data/training/Martin_Sheen_41.jpg \n inflating: /data/training/Martin_Sheen_42.jpg \n inflating: /data/training/Martin_Sheen_50.jpg \n inflating: /data/training/Martin_Sheen_51.jpg \n inflating: /data/training/Martin_Sheen_52.jpg \n inflating: /data/training/Mary_Landrieu_00.jpg \n inflating: /data/training/Mary_Landrieu_01.jpg \n inflating: /data/training/Mary_Landrieu_02.jpg \n inflating: /data/training/Mary_Landrieu_20.jpg \n inflating: /data/training/Mary_Landrieu_21.jpg \n inflating: /data/training/Mary_Landrieu_22.jpg \n inflating: /data/training/Mary_Landrieu_30.jpg \n inflating: /data/training/Mary_Landrieu_31.jpg \n inflating: /data/training/Mary_Landrieu_32.jpg \n inflating: /data/training/Mary_Robinson_10.jpg \n inflating: /data/training/Mary_Robinson_11.jpg \n inflating: /data/training/Mary_Robinson_12.jpg \n inflating: /data/training/Mary_Robinson_20.jpg \n inflating: /data/training/Mary_Robinson_21.jpg \n inflating: /data/training/Mary_Robinson_22.jpg \n inflating: /data/training/Mary_Robinson_40.jpg \n inflating: /data/training/Mary_Robinson_41.jpg \n inflating: /data/training/Mary_Robinson_42.jpg \n inflating: /data/training/Mary_Robinson_50.jpg \n inflating: /data/training/Mary_Robinson_51.jpg \n inflating: /data/training/Mary_Robinson_52.jpg \n inflating: /data/training/Massoud_Barzani_00.jpg \n inflating: /data/training/Massoud_Barzani_01.jpg \n inflating: /data/training/Massoud_Barzani_02.jpg \n inflating: /data/training/Massoud_Barzani_10.jpg \n inflating: /data/training/Massoud_Barzani_11.jpg \n inflating: /data/training/Massoud_Barzani_12.jpg \n inflating: /data/training/Massoud_Barzani_20.jpg \n inflating: /data/training/Massoud_Barzani_21.jpg \n inflating: /data/training/Massoud_Barzani_22.jpg \n inflating: /data/training/Massoud_Barzani_40.jpg \n inflating: /data/training/Massoud_Barzani_41.jpg \n inflating: /data/training/Massoud_Barzani_42.jpg \n inflating: /data/training/Matt_LeBlanc_00.jpg \n inflating: /data/training/Matt_LeBlanc_01.jpg \n inflating: /data/training/Matt_LeBlanc_02.jpg \n inflating: /data/training/Matt_LeBlanc_20.jpg \n inflating: /data/training/Matt_LeBlanc_21.jpg \n inflating: /data/training/Matt_LeBlanc_22.jpg \n inflating: /data/training/Matt_LeBlanc_30.jpg \n inflating: /data/training/Matt_LeBlanc_31.jpg \n inflating: /data/training/Matt_LeBlanc_32.jpg \n inflating: /data/training/Nancy_Kerrigan_00.jpg \n inflating: /data/training/Nancy_Kerrigan_01.jpg \n inflating: /data/training/Nancy_Kerrigan_02.jpg \n inflating: /data/training/Nancy_Kerrigan_20.jpg \n inflating: /data/training/Nancy_Kerrigan_21.jpg \n inflating: /data/training/Nancy_Kerrigan_22.jpg \n inflating: /data/training/Nancy_Kerrigan_30.jpg \n inflating: /data/training/Nancy_Kerrigan_31.jpg \n inflating: /data/training/Nancy_Kerrigan_32.jpg \n inflating: /data/training/Nancy_Kerrigan_40.jpg \n inflating: /data/training/Nancy_Kerrigan_41.jpg \n inflating: /data/training/Nancy_Kerrigan_42.jpg \n inflating: /data/training/Nancy_Reagan_00.jpg \n inflating: /data/training/Nancy_Reagan_01.jpg \n inflating: /data/training/Nancy_Reagan_02.jpg \n inflating: /data/training/Nancy_Reagan_10.jpg \n inflating: /data/training/Nancy_Reagan_11.jpg \n inflating: /data/training/Nancy_Reagan_12.jpg \n inflating: /data/training/Nancy_Reagan_30.jpg \n inflating: /data/training/Nancy_Reagan_31.jpg \n inflating: /data/training/Nancy_Reagan_32.jpg \n inflating: /data/training/Nancy_Reagan_40.jpg \n inflating: /data/training/Nancy_Reagan_41.jpg \n inflating: /data/training/Nancy_Reagan_42.jpg \n inflating: /data/training/Nanni_Moretti_10.jpg \n inflating: /data/training/Nanni_Moretti_11.jpg \n inflating: /data/training/Nanni_Moretti_12.jpg \n inflating: /data/training/Nanni_Moretti_20.jpg \n inflating: /data/training/Nanni_Moretti_21.jpg \n inflating: /data/training/Nanni_Moretti_22.jpg \n inflating: /data/training/Nanni_Moretti_40.jpg \n inflating: /data/training/Nanni_Moretti_41.jpg \n inflating: /data/training/Nanni_Moretti_42.jpg \n inflating: /data/training/Natalia_Vodonova_00.jpg \n inflating: /data/training/Natalia_Vodonova_01.jpg \n inflating: /data/training/Natalia_Vodonova_02.jpg \n inflating: /data/training/Natalia_Vodonova_10.jpg \n inflating: /data/training/Natalia_Vodonova_11.jpg \n inflating: /data/training/Natalia_Vodonova_12.jpg \n inflating: /data/training/Natalia_Vodonova_20.jpg \n inflating: /data/training/Natalia_Vodonova_21.jpg \n inflating: /data/training/Natalia_Vodonova_22.jpg \n inflating: /data/training/Natasha_Lyonne_00.jpg \n inflating: /data/training/Natasha_Lyonne_01.jpg \n inflating: /data/training/Natasha_Lyonne_02.jpg \n inflating: /data/training/Natasha_Lyonne_10.jpg \n inflating: /data/training/Natasha_Lyonne_11.jpg \n inflating: /data/training/Natasha_Lyonne_12.jpg \n inflating: /data/training/Natasha_Lyonne_40.jpg \n inflating: /data/training/Natasha_Lyonne_41.jpg \n inflating: /data/training/Natasha_Lyonne_42.jpg \n inflating: /data/training/Nick_Reilly_10.jpg \n inflating: /data/training/Nick_Reilly_11.jpg \n inflating: /data/training/Nick_Reilly_12.jpg \n inflating: /data/training/Nick_Reilly_40.jpg \n inflating: /data/training/Nick_Reilly_41.jpg \n inflating: /data/training/Nick_Reilly_42.jpg \n inflating: /data/training/Nick_Reilly_50.jpg \n inflating: /data/training/Nick_Reilly_51.jpg \n inflating: /data/training/Nick_Reilly_52.jpg \n inflating: /data/training/Nicolas_Eyzaguirre_00.jpg \n inflating: /data/training/Nicolas_Eyzaguirre_01.jpg \n inflating: /data/training/Nicolas_Eyzaguirre_02.jpg \n inflating: /data/training/Nicolas_Eyzaguirre_10.jpg \n inflating: /data/training/Nicolas_Eyzaguirre_11.jpg \n inflating: /data/training/Nicolas_Eyzaguirre_12.jpg \n inflating: /data/training/Nicolas_Eyzaguirre_20.jpg \n inflating: /data/training/Nicolas_Eyzaguirre_21.jpg \n inflating: /data/training/Nicolas_Eyzaguirre_22.jpg \n inflating: /data/training/Nicolas_Sarkozy_00.jpg \n inflating: /data/training/Nicolas_Sarkozy_01.jpg \n inflating: /data/training/Nicolas_Sarkozy_02.jpg \n inflating: /data/training/Nicolas_Sarkozy_10.jpg \n inflating: /data/training/Nicolas_Sarkozy_11.jpg \n inflating: /data/training/Nicolas_Sarkozy_12.jpg \n inflating: /data/training/Nicolas_Sarkozy_20.jpg \n inflating: /data/training/Nicolas_Sarkozy_21.jpg \n inflating: /data/training/Nicolas_Sarkozy_22.jpg \n inflating: /data/training/Nicolas_Sarkozy_50.jpg \n inflating: /data/training/Nicolas_Sarkozy_51.jpg \n inflating: /data/training/Nicolas_Sarkozy_52.jpg \n inflating: /data/training/Nina_Jacobson_00.jpg \n inflating: /data/training/Nina_Jacobson_01.jpg \n inflating: /data/training/Nina_Jacobson_02.jpg \n inflating: /data/training/Nina_Jacobson_10.jpg \n inflating: /data/training/Nina_Jacobson_11.jpg \n inflating: /data/training/Nina_Jacobson_12.jpg \n inflating: /data/training/Nina_Jacobson_30.jpg \n inflating: /data/training/Nina_Jacobson_31.jpg \n inflating: /data/training/Nina_Jacobson_32.jpg \n inflating: /data/training/Norah_Jones_10.jpg \n inflating: /data/training/Norah_Jones_11.jpg \n inflating: /data/training/Norah_Jones_12.jpg \n inflating: /data/training/Norah_Jones_20.jpg \n inflating: /data/training/Norah_Jones_21.jpg \n inflating: /data/training/Norah_Jones_22.jpg \n inflating: /data/training/Norah_Jones_40.jpg \n inflating: /data/training/Norah_Jones_41.jpg \n inflating: /data/training/Norah_Jones_42.jpg \n inflating: /data/training/Norah_Jones_50.jpg \n inflating: /data/training/Norah_Jones_51.jpg \n inflating: /data/training/Norah_Jones_52.jpg \n inflating: /data/training/Norman_Mineta_00.jpg \n inflating: /data/training/Norman_Mineta_01.jpg \n inflating: /data/training/Norman_Mineta_02.jpg \n inflating: /data/training/Norman_Mineta_30.jpg \n inflating: /data/training/Norman_Mineta_31.jpg \n inflating: /data/training/Norman_Mineta_32.jpg \n inflating: /data/training/Norman_Mineta_50.jpg \n inflating: /data/training/Norman_Mineta_51.jpg \n inflating: /data/training/Norman_Mineta_52.jpg \n inflating: /data/training/Olene_Walker_00.jpg \n inflating: /data/training/Olene_Walker_01.jpg \n inflating: /data/training/Olene_Walker_02.jpg \n inflating: /data/training/Olene_Walker_10.jpg \n inflating: /data/training/Olene_Walker_11.jpg \n inflating: /data/training/Olene_Walker_12.jpg \n inflating: /data/training/Olene_Walker_30.jpg \n inflating: /data/training/Olene_Walker_31.jpg \n inflating: /data/training/Olene_Walker_32.jpg \n inflating: /data/training/Olene_Walker_40.jpg \n inflating: /data/training/Olene_Walker_41.jpg \n inflating: /data/training/Olene_Walker_42.jpg \n inflating: /data/training/Olivia_Newton-John_00.jpg \n inflating: /data/training/Olivia_Newton-John_01.jpg \n inflating: /data/training/Olivia_Newton-John_02.jpg \n inflating: /data/training/Olivia_Newton-John_10.jpg \n inflating: /data/training/Olivia_Newton-John_11.jpg \n inflating: /data/training/Olivia_Newton-John_12.jpg \n inflating: /data/training/Olivia_Newton-John_40.jpg \n inflating: /data/training/Olivia_Newton-John_41.jpg \n inflating: /data/training/Olivia_Newton-John_42.jpg \n inflating: /data/training/Orlando_Bloom_00.jpg \n inflating: /data/training/Orlando_Bloom_01.jpg \n inflating: /data/training/Orlando_Bloom_02.jpg \n inflating: /data/training/Orlando_Bloom_30.jpg \n inflating: /data/training/Orlando_Bloom_31.jpg \n inflating: /data/training/Orlando_Bloom_32.jpg \n inflating: /data/training/Orlando_Bloom_40.jpg \n inflating: /data/training/Orlando_Bloom_41.jpg \n inflating: /data/training/Orlando_Bloom_42.jpg \n inflating: /data/training/Orlando_Bloom_50.jpg \n inflating: /data/training/Orlando_Bloom_51.jpg \n inflating: /data/training/Orlando_Bloom_52.jpg \n inflating: /data/training/Otto_Reich_00.jpg \n inflating: /data/training/Otto_Reich_01.jpg \n inflating: /data/training/Otto_Reich_02.jpg \n inflating: /data/training/Otto_Reich_10.jpg \n inflating: /data/training/Otto_Reich_11.jpg \n inflating: /data/training/Otto_Reich_12.jpg \n inflating: /data/training/Otto_Reich_30.jpg \n inflating: /data/training/Otto_Reich_31.jpg \n inflating: /data/training/Otto_Reich_32.jpg \n inflating: /data/training/Otto_Reich_40.jpg \n inflating: /data/training/Otto_Reich_41.jpg \n inflating: /data/training/Otto_Reich_42.jpg \n inflating: /data/training/Pat_Riley_00.jpg \n inflating: /data/training/Pat_Riley_01.jpg \n inflating: /data/training/Pat_Riley_02.jpg \n inflating: /data/training/Pat_Riley_20.jpg \n inflating: /data/training/Pat_Riley_21.jpg \n inflating: /data/training/Pat_Riley_22.jpg \n inflating: /data/training/Pat_Riley_50.jpg \n inflating: /data/training/Pat_Riley_51.jpg \n inflating: /data/training/Pat_Riley_52.jpg \n inflating: /data/training/Patrick_Leahy_10.jpg \n inflating: /data/training/Patrick_Leahy_11.jpg \n inflating: /data/training/Patrick_Leahy_12.jpg \n inflating: /data/training/Patrick_Leahy_20.jpg \n inflating: /data/training/Patrick_Leahy_21.jpg \n inflating: /data/training/Patrick_Leahy_22.jpg \n inflating: /data/training/Patrick_Leahy_30.jpg \n inflating: /data/training/Patrick_Leahy_31.jpg \n inflating: /data/training/Patrick_Leahy_32.jpg \n inflating: /data/training/Paul_Otellini_00.jpg \n inflating: /data/training/Paul_Otellini_01.jpg \n inflating: /data/training/Paul_Otellini_02.jpg \n inflating: /data/training/Paul_Otellini_10.jpg \n inflating: /data/training/Paul_Otellini_11.jpg \n inflating: /data/training/Paul_Otellini_12.jpg \n inflating: /data/training/Paul_Otellini_20.jpg \n inflating: /data/training/Paul_Otellini_21.jpg \n inflating: /data/training/Paul_Otellini_22.jpg \n inflating: /data/training/Paul_Reiser_00.jpg \n inflating: /data/training/Paul_Reiser_01.jpg \n inflating: /data/training/Paul_Reiser_02.jpg \n inflating: /data/training/Paul_Reiser_20.jpg \n inflating: /data/training/Paul_Reiser_21.jpg \n inflating: /data/training/Paul_Reiser_22.jpg \n inflating: /data/training/Paul_Reiser_30.jpg \n inflating: /data/training/Paul_Reiser_31.jpg \n inflating: /data/training/Paul_Reiser_32.jpg \n inflating: /data/training/Pedro_Solbes_00.jpg \n inflating: /data/training/Pedro_Solbes_01.jpg \n inflating: /data/training/Pedro_Solbes_02.jpg \n inflating: /data/training/Pedro_Solbes_20.jpg \n inflating: /data/training/Pedro_Solbes_21.jpg \n inflating: /data/training/Pedro_Solbes_22.jpg \n inflating: /data/training/Pedro_Solbes_30.jpg \n inflating: /data/training/Pedro_Solbes_31.jpg \n inflating: /data/training/Pedro_Solbes_32.jpg \n inflating: /data/training/Penelope_Ann_Miller_00.jpg \n inflating: /data/training/Penelope_Ann_Miller_01.jpg \n inflating: /data/training/Penelope_Ann_Miller_02.jpg \n inflating: /data/training/Penelope_Ann_Miller_20.jpg \n inflating: /data/training/Penelope_Ann_Miller_21.jpg \n inflating: /data/training/Penelope_Ann_Miller_22.jpg \n inflating: /data/training/Penelope_Ann_Miller_50.jpg \n inflating: /data/training/Penelope_Ann_Miller_51.jpg \n inflating: /data/training/Penelope_Ann_Miller_52.jpg \n inflating: /data/training/Peter_Goldmark_10.jpg \n inflating: /data/training/Peter_Goldmark_11.jpg \n inflating: /data/training/Peter_Goldmark_12.jpg \n inflating: /data/training/Peter_Goldmark_40.jpg \n inflating: /data/training/Peter_Goldmark_41.jpg \n inflating: /data/training/Peter_Goldmark_42.jpg \n inflating: /data/training/Peter_Goldmark_50.jpg \n inflating: /data/training/Peter_Goldmark_51.jpg \n inflating: /data/training/Peter_Goldmark_52.jpg \n inflating: /data/training/Peter_Medgyessy_10.jpg \n inflating: /data/training/Peter_Medgyessy_11.jpg \n inflating: /data/training/Peter_Medgyessy_12.jpg \n inflating: /data/training/Peter_Medgyessy_30.jpg \n inflating: /data/training/Peter_Medgyessy_31.jpg \n inflating: /data/training/Peter_Medgyessy_32.jpg \n inflating: /data/training/Peter_Medgyessy_40.jpg \n inflating: /data/training/Peter_Medgyessy_41.jpg \n inflating: /data/training/Peter_Medgyessy_42.jpg \n inflating: /data/training/Peter_Medgyessy_50.jpg \n inflating: /data/training/Peter_Medgyessy_51.jpg \n inflating: /data/training/Peter_Medgyessy_52.jpg \n inflating: /data/training/Philippe_Gagnon_00.jpg \n inflating: /data/training/Philippe_Gagnon_01.jpg \n inflating: /data/training/Philippe_Gagnon_02.jpg \n inflating: /data/training/Philippe_Gagnon_10.jpg \n inflating: /data/training/Philippe_Gagnon_11.jpg \n inflating: /data/training/Philippe_Gagnon_12.jpg \n inflating: /data/training/Philippe_Gagnon_20.jpg \n inflating: /data/training/Philippe_Gagnon_21.jpg \n inflating: /data/training/Philippe_Gagnon_22.jpg \n inflating: /data/training/Philippe_Gagnon_30.jpg \n inflating: /data/training/Philippe_Gagnon_31.jpg \n inflating: /data/training/Philippe_Gagnon_32.jpg \n inflating: /data/training/Philippe_Noiret_10.jpg \n inflating: /data/training/Philippe_Noiret_11.jpg \n inflating: /data/training/Philippe_Noiret_12.jpg \n inflating: /data/training/Philippe_Noiret_30.jpg \n inflating: /data/training/Philippe_Noiret_31.jpg \n inflating: /data/training/Philippe_Noiret_32.jpg \n inflating: /data/training/Philippe_Noiret_50.jpg \n inflating: /data/training/Philippe_Noiret_51.jpg \n inflating: /data/training/Philippe_Noiret_52.jpg \n inflating: /data/training/Picabo_Street_00.jpg \n inflating: /data/training/Picabo_Street_01.jpg \n inflating: /data/training/Picabo_Street_02.jpg \n inflating: /data/training/Picabo_Street_20.jpg \n inflating: /data/training/Picabo_Street_21.jpg \n inflating: /data/training/Picabo_Street_22.jpg \n inflating: /data/training/Picabo_Street_40.jpg \n inflating: /data/training/Picabo_Street_41.jpg \n inflating: /data/training/Picabo_Street_42.jpg \n inflating: /data/training/Pilar_Montenegro_10.jpg \n inflating: /data/training/Pilar_Montenegro_11.jpg \n inflating: /data/training/Pilar_Montenegro_12.jpg \n inflating: /data/training/Pilar_Montenegro_20.jpg \n inflating: /data/training/Pilar_Montenegro_21.jpg \n inflating: /data/training/Pilar_Montenegro_22.jpg \n inflating: /data/training/Pilar_Montenegro_50.jpg \n inflating: /data/training/Pilar_Montenegro_51.jpg \n inflating: /data/training/Pilar_Montenegro_52.jpg \n inflating: /data/training/Piotr_Anderszewski_20.jpg \n inflating: /data/training/Piotr_Anderszewski_21.jpg \n inflating: /data/training/Piotr_Anderszewski_22.jpg \n inflating: /data/training/Piotr_Anderszewski_30.jpg \n inflating: /data/training/Piotr_Anderszewski_31.jpg \n inflating: /data/training/Piotr_Anderszewski_32.jpg \n inflating: /data/training/Piotr_Anderszewski_50.jpg \n inflating: /data/training/Piotr_Anderszewski_51.jpg \n inflating: /data/training/Piotr_Anderszewski_52.jpg \n inflating: /data/training/Poala_Suarez_30.jpg \n inflating: /data/training/Poala_Suarez_31.jpg \n inflating: /data/training/Poala_Suarez_32.jpg \n inflating: /data/training/Poala_Suarez_40.jpg \n inflating: /data/training/Poala_Suarez_41.jpg \n inflating: /data/training/Poala_Suarez_42.jpg \n inflating: /data/training/Poala_Suarez_50.jpg \n inflating: /data/training/Poala_Suarez_51.jpg \n inflating: /data/training/Poala_Suarez_52.jpg \n inflating: /data/training/Prince_Harry_10.jpg \n inflating: /data/training/Prince_Harry_11.jpg \n inflating: /data/training/Prince_Harry_12.jpg \n inflating: /data/training/Prince_Harry_20.jpg \n inflating: /data/training/Prince_Harry_21.jpg \n inflating: /data/training/Prince_Harry_22.jpg \n inflating: /data/training/Prince_Harry_40.jpg \n inflating: /data/training/Prince_Harry_41.jpg \n inflating: /data/training/Prince_Harry_42.jpg \n inflating: /data/training/Princess_Stephanie_00.jpg \n inflating: /data/training/Princess_Stephanie_01.jpg \n inflating: /data/training/Princess_Stephanie_02.jpg \n inflating: /data/training/Princess_Stephanie_20.jpg \n inflating: /data/training/Princess_Stephanie_21.jpg \n inflating: /data/training/Princess_Stephanie_22.jpg \n inflating: /data/training/Princess_Stephanie_40.jpg \n inflating: /data/training/Princess_Stephanie_41.jpg \n inflating: /data/training/Princess_Stephanie_42.jpg \n inflating: /data/training/Princess_Stephanie_50.jpg \n inflating: /data/training/Princess_Stephanie_51.jpg \n inflating: /data/training/Princess_Stephanie_52.jpg \n inflating: /data/training/Priyanka_Chopra_10.jpg \n inflating: /data/training/Priyanka_Chopra_11.jpg \n inflating: /data/training/Priyanka_Chopra_12.jpg \n inflating: /data/training/Priyanka_Chopra_40.jpg \n inflating: /data/training/Priyanka_Chopra_41.jpg \n inflating: /data/training/Priyanka_Chopra_42.jpg \n inflating: /data/training/Priyanka_Chopra_50.jpg \n inflating: /data/training/Priyanka_Chopra_51.jpg \n inflating: /data/training/Priyanka_Chopra_52.jpg \n inflating: /data/training/Queen_Noor_10.jpg \n inflating: /data/training/Queen_Noor_11.jpg \n inflating: /data/training/Queen_Noor_12.jpg \n inflating: /data/training/Queen_Noor_30.jpg \n inflating: /data/training/Queen_Noor_31.jpg \n inflating: /data/training/Queen_Noor_32.jpg \n inflating: /data/training/Queen_Noor_50.jpg \n inflating: /data/training/Queen_Noor_51.jpg \n inflating: /data/training/Queen_Noor_52.jpg \n inflating: /data/training/Queen_Rania_10.jpg \n inflating: /data/training/Queen_Rania_11.jpg \n inflating: /data/training/Queen_Rania_12.jpg \n inflating: /data/training/Queen_Rania_30.jpg \n inflating: /data/training/Queen_Rania_31.jpg \n inflating: /data/training/Queen_Rania_32.jpg \n inflating: /data/training/Queen_Rania_50.jpg \n inflating: /data/training/Queen_Rania_51.jpg \n inflating: /data/training/Queen_Rania_52.jpg \n inflating: /data/training/Rachel_Hunter_30.jpg \n inflating: /data/training/Rachel_Hunter_31.jpg \n inflating: /data/training/Rachel_Hunter_32.jpg \n inflating: /data/training/Rachel_Hunter_40.jpg \n inflating: /data/training/Rachel_Hunter_41.jpg \n inflating: /data/training/Rachel_Hunter_42.jpg \n inflating: /data/training/Rachel_Hunter_50.jpg \n inflating: /data/training/Rachel_Hunter_51.jpg \n inflating: /data/training/Rachel_Hunter_52.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_00.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_01.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_02.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_10.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_11.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_12.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_20.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_21.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_22.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_30.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_31.jpg \n inflating: /data/training/Raja_Zafar-ul-Haq_32.jpg \n inflating: /data/training/Ralph_Klein_00.jpg \n inflating: /data/training/Ralph_Klein_01.jpg \n inflating: /data/training/Ralph_Klein_02.jpg \n inflating: /data/training/Ralph_Klein_10.jpg \n inflating: /data/training/Ralph_Klein_11.jpg \n inflating: /data/training/Ralph_Klein_12.jpg \n inflating: /data/training/Ralph_Klein_30.jpg \n inflating: /data/training/Ralph_Klein_31.jpg \n inflating: /data/training/Ralph_Klein_32.jpg \n inflating: /data/training/Raza_Rabbani_20.jpg \n inflating: /data/training/Raza_Rabbani_21.jpg \n inflating: /data/training/Raza_Rabbani_22.jpg \n inflating: /data/training/Raza_Rabbani_30.jpg \n inflating: /data/training/Raza_Rabbani_31.jpg \n inflating: /data/training/Raza_Rabbani_32.jpg \n inflating: /data/training/Raza_Rabbani_50.jpg \n inflating: /data/training/Raza_Rabbani_51.jpg \n inflating: /data/training/Raza_Rabbani_52.jpg \n inflating: /data/training/Recep_Tayyip_Erdogan_00.jpg \n inflating: /data/training/Recep_Tayyip_Erdogan_01.jpg \n inflating: /data/training/Recep_Tayyip_Erdogan_02.jpg \n inflating: /data/training/Recep_Tayyip_Erdogan_20.jpg \n inflating: /data/training/Recep_Tayyip_Erdogan_21.jpg \n inflating: /data/training/Recep_Tayyip_Erdogan_22.jpg \n inflating: /data/training/Recep_Tayyip_Erdogan_40.jpg \n inflating: /data/training/Recep_Tayyip_Erdogan_41.jpg \n inflating: /data/training/Recep_Tayyip_Erdogan_42.jpg \n inflating: /data/training/Reese_Witherspoon_00.jpg \n inflating: /data/training/Reese_Witherspoon_01.jpg \n inflating: /data/training/Reese_Witherspoon_02.jpg \n inflating: /data/training/Reese_Witherspoon_10.jpg \n inflating: /data/training/Reese_Witherspoon_11.jpg \n inflating: /data/training/Reese_Witherspoon_12.jpg \n inflating: /data/training/Reese_Witherspoon_40.jpg \n inflating: /data/training/Reese_Witherspoon_41.jpg \n inflating: /data/training/Reese_Witherspoon_42.jpg \n inflating: /data/training/Ricardo_Lopez_Murphy_10.jpg \n inflating: /data/training/Ricardo_Lopez_Murphy_11.jpg \n inflating: /data/training/Ricardo_Lopez_Murphy_12.jpg \n inflating: /data/training/Ricardo_Lopez_Murphy_30.jpg \n inflating: /data/training/Ricardo_Lopez_Murphy_31.jpg \n inflating: /data/training/Ricardo_Lopez_Murphy_32.jpg \n inflating: /data/training/Ricardo_Lopez_Murphy_40.jpg \n inflating: /data/training/Ricardo_Lopez_Murphy_41.jpg \n inflating: /data/training/Ricardo_Lopez_Murphy_42.jpg \n inflating: /data/training/Ricardo_Sanchez_20.jpg \n inflating: /data/training/Ricardo_Sanchez_21.jpg \n inflating: /data/training/Ricardo_Sanchez_22.jpg \n inflating: /data/training/Ricardo_Sanchez_30.jpg \n inflating: /data/training/Ricardo_Sanchez_31.jpg \n inflating: /data/training/Ricardo_Sanchez_32.jpg \n inflating: /data/training/Ricardo_Sanchez_40.jpg \n inflating: /data/training/Ricardo_Sanchez_41.jpg \n inflating: /data/training/Ricardo_Sanchez_42.jpg \n inflating: /data/training/Richard_Branson_00.jpg \n inflating: /data/training/Richard_Branson_01.jpg \n inflating: /data/training/Richard_Branson_02.jpg \n inflating: /data/training/Richard_Branson_10.jpg \n inflating: /data/training/Richard_Branson_11.jpg \n inflating: /data/training/Richard_Branson_12.jpg \n inflating: /data/training/Richard_Branson_50.jpg \n inflating: /data/training/Richard_Branson_51.jpg \n inflating: /data/training/Richard_Branson_52.jpg \n inflating: /data/training/Richard_Lennon_00.jpg \n inflating: /data/training/Richard_Lennon_01.jpg \n inflating: /data/training/Richard_Lennon_02.jpg \n inflating: /data/training/Richard_Lennon_30.jpg \n inflating: /data/training/Richard_Lennon_31.jpg \n inflating: /data/training/Richard_Lennon_32.jpg \n inflating: /data/training/Richard_Lennon_40.jpg \n inflating: /data/training/Richard_Lennon_41.jpg \n inflating: /data/training/Richard_Lennon_42.jpg \n inflating: /data/training/Richard_Lugar_00.jpg \n inflating: /data/training/Richard_Lugar_01.jpg \n inflating: /data/training/Richard_Lugar_02.jpg \n inflating: /data/training/Richard_Lugar_10.jpg \n inflating: /data/training/Richard_Lugar_11.jpg \n inflating: /data/training/Richard_Lugar_12.jpg \n inflating: /data/training/Richard_Lugar_20.jpg \n inflating: /data/training/Richard_Lugar_21.jpg \n inflating: /data/training/Richard_Lugar_22.jpg \n inflating: /data/training/Richard_Lugar_50.jpg \n inflating: /data/training/Richard_Lugar_51.jpg \n inflating: /data/training/Richard_Lugar_52.jpg \n inflating: /data/training/Richard_Paul_Evans_00.jpg \n inflating: /data/training/Richard_Paul_Evans_01.jpg \n inflating: /data/training/Richard_Paul_Evans_02.jpg \n inflating: /data/training/Richard_Paul_Evans_20.jpg \n inflating: /data/training/Richard_Paul_Evans_21.jpg \n inflating: /data/training/Richard_Paul_Evans_22.jpg \n inflating: /data/training/Richard_Paul_Evans_40.jpg \n inflating: /data/training/Richard_Paul_Evans_41.jpg \n inflating: /data/training/Richard_Paul_Evans_42.jpg \n inflating: /data/training/Richard_Paul_Evans_50.jpg \n inflating: /data/training/Richard_Paul_Evans_51.jpg \n inflating: /data/training/Richard_Paul_Evans_52.jpg \n inflating: /data/training/Rick_Bragg_20.jpg \n inflating: /data/training/Rick_Bragg_21.jpg \n inflating: /data/training/Rick_Bragg_22.jpg \n inflating: /data/training/Rick_Bragg_30.jpg \n inflating: /data/training/Rick_Bragg_31.jpg \n inflating: /data/training/Rick_Bragg_32.jpg \n inflating: /data/training/Rick_Bragg_50.jpg \n inflating: /data/training/Rick_Bragg_51.jpg \n inflating: /data/training/Rick_Bragg_52.jpg \n inflating: /data/training/Ridley_Scott_10.jpg \n inflating: /data/training/Ridley_Scott_11.jpg \n inflating: /data/training/Ridley_Scott_12.jpg \n inflating: /data/training/Ridley_Scott_20.jpg \n inflating: /data/training/Ridley_Scott_21.jpg \n inflating: /data/training/Ridley_Scott_22.jpg \n inflating: /data/training/Ridley_Scott_30.jpg \n inflating: /data/training/Ridley_Scott_31.jpg \n inflating: /data/training/Ridley_Scott_32.jpg \n inflating: /data/training/Robbie_Coltrane_00.jpg \n inflating: /data/training/Robbie_Coltrane_01.jpg \n inflating: /data/training/Robbie_Coltrane_02.jpg \n inflating: /data/training/Robbie_Coltrane_10.jpg \n inflating: /data/training/Robbie_Coltrane_11.jpg \n inflating: /data/training/Robbie_Coltrane_12.jpg \n inflating: /data/training/Robbie_Coltrane_20.jpg \n inflating: /data/training/Robbie_Coltrane_21.jpg \n inflating: /data/training/Robbie_Coltrane_22.jpg \n inflating: /data/training/Robbie_Coltrane_50.jpg \n inflating: /data/training/Robbie_Coltrane_51.jpg \n inflating: /data/training/Robbie_Coltrane_52.jpg \n inflating: /data/training/Robert_Altman_10.jpg \n inflating: /data/training/Robert_Altman_11.jpg \n inflating: /data/training/Robert_Altman_12.jpg \n inflating: /data/training/Robert_Altman_20.jpg \n inflating: /data/training/Robert_Altman_21.jpg \n inflating: /data/training/Robert_Altman_22.jpg \n inflating: /data/training/Robert_Altman_50.jpg \n inflating: /data/training/Robert_Altman_51.jpg \n inflating: /data/training/Robert_Altman_52.jpg \n inflating: /data/training/Roberto_Benigni_00.jpg \n inflating: /data/training/Roberto_Benigni_01.jpg \n inflating: /data/training/Roberto_Benigni_02.jpg \n inflating: /data/training/Roberto_Benigni_10.jpg \n inflating: /data/training/Roberto_Benigni_11.jpg \n inflating: /data/training/Roberto_Benigni_12.jpg \n inflating: /data/training/Roberto_Benigni_50.jpg \n inflating: /data/training/Roberto_Benigni_51.jpg \n inflating: /data/training/Roberto_Benigni_52.jpg \n inflating: /data/training/Rocco_Buttiglione_00.jpg \n inflating: /data/training/Rocco_Buttiglione_01.jpg \n inflating: /data/training/Rocco_Buttiglione_02.jpg \n inflating: /data/training/Rocco_Buttiglione_10.jpg \n inflating: /data/training/Rocco_Buttiglione_11.jpg \n inflating: /data/training/Rocco_Buttiglione_12.jpg \n inflating: /data/training/Rocco_Buttiglione_30.jpg \n inflating: /data/training/Rocco_Buttiglione_31.jpg \n inflating: /data/training/Rocco_Buttiglione_32.jpg \n inflating: /data/training/Rocco_Buttiglione_40.jpg \n inflating: /data/training/Rocco_Buttiglione_41.jpg \n inflating: /data/training/Rocco_Buttiglione_42.jpg \n inflating: /data/training/Rodrigo_Borja_00.jpg \n inflating: /data/training/Rodrigo_Borja_01.jpg \n inflating: /data/training/Rodrigo_Borja_02.jpg \n inflating: /data/training/Rodrigo_Borja_40.jpg \n inflating: /data/training/Rodrigo_Borja_41.jpg \n inflating: /data/training/Rodrigo_Borja_42.jpg \n inflating: /data/training/Rodrigo_Borja_50.jpg \n inflating: /data/training/Rodrigo_Borja_51.jpg \n inflating: /data/training/Rodrigo_Borja_52.jpg \n inflating: /data/training/Saeed_Mortazavi_10.jpg \n inflating: /data/training/Saeed_Mortazavi_11.jpg \n inflating: /data/training/Saeed_Mortazavi_12.jpg \n inflating: /data/training/Saeed_Mortazavi_20.jpg \n inflating: /data/training/Saeed_Mortazavi_21.jpg \n inflating: /data/training/Saeed_Mortazavi_22.jpg \n inflating: /data/training/Saeed_Mortazavi_50.jpg \n inflating: /data/training/Saeed_Mortazavi_51.jpg \n inflating: /data/training/Saeed_Mortazavi_52.jpg \n inflating: /data/training/Sally_Ride_00.jpg \n inflating: /data/training/Sally_Ride_01.jpg \n inflating: /data/training/Sally_Ride_02.jpg \n inflating: /data/training/Sally_Ride_40.jpg \n inflating: /data/training/Sally_Ride_41.jpg \n inflating: /data/training/Sally_Ride_42.jpg \n inflating: /data/training/Sally_Ride_50.jpg \n inflating: /data/training/Sally_Ride_51.jpg \n inflating: /data/training/Sally_Ride_52.jpg \n inflating: /data/training/Sanjay_Gupta_10.jpg \n inflating: /data/training/Sanjay_Gupta_11.jpg \n inflating: /data/training/Sanjay_Gupta_12.jpg \n inflating: /data/training/Sanjay_Gupta_20.jpg \n inflating: /data/training/Sanjay_Gupta_21.jpg \n inflating: /data/training/Sanjay_Gupta_22.jpg \n inflating: /data/training/Sanjay_Gupta_40.jpg \n inflating: /data/training/Sanjay_Gupta_41.jpg \n inflating: /data/training/Sanjay_Gupta_42.jpg \n inflating: /data/training/Sara_Silverman_10.jpg \n inflating: /data/training/Sara_Silverman_11.jpg \n inflating: /data/training/Sara_Silverman_12.jpg \n inflating: /data/training/Sara_Silverman_20.jpg \n inflating: /data/training/Sara_Silverman_21.jpg \n inflating: /data/training/Sara_Silverman_22.jpg \n inflating: /data/training/Sara_Silverman_40.jpg \n inflating: /data/training/Sara_Silverman_41.jpg \n inflating: /data/training/Sara_Silverman_42.jpg \n inflating: /data/training/Sara_Silverman_50.jpg \n inflating: /data/training/Sara_Silverman_51.jpg \n inflating: /data/training/Sara_Silverman_52.jpg \n inflating: /data/training/Sarah_Wynter_00.jpg \n inflating: /data/training/Sarah_Wynter_01.jpg \n inflating: /data/training/Sarah_Wynter_02.jpg \n inflating: /data/training/Sarah_Wynter_40.jpg \n inflating: /data/training/Sarah_Wynter_41.jpg \n inflating: /data/training/Sarah_Wynter_42.jpg \n inflating: /data/training/Sarah_Wynter_50.jpg \n inflating: /data/training/Sarah_Wynter_51.jpg \n inflating: /data/training/Sarah_Wynter_52.jpg \n inflating: /data/training/Sasha_Cohen_20.jpg \n inflating: /data/training/Sasha_Cohen_21.jpg \n inflating: /data/training/Sasha_Cohen_22.jpg \n inflating: /data/training/Sasha_Cohen_40.jpg \n inflating: /data/training/Sasha_Cohen_41.jpg \n inflating: /data/training/Sasha_Cohen_42.jpg \n inflating: /data/training/Sasha_Cohen_50.jpg \n inflating: /data/training/Sasha_Cohen_51.jpg \n inflating: /data/training/Sasha_Cohen_52.jpg \n inflating: /data/training/T_Boone_Pickens_10.jpg \n inflating: /data/training/T_Boone_Pickens_11.jpg \n inflating: /data/training/T_Boone_Pickens_12.jpg \n inflating: /data/training/T_Boone_Pickens_20.jpg \n inflating: /data/training/T_Boone_Pickens_21.jpg \n inflating: /data/training/T_Boone_Pickens_22.jpg \n inflating: /data/training/T_Boone_Pickens_30.jpg \n inflating: /data/training/T_Boone_Pickens_31.jpg \n inflating: /data/training/T_Boone_Pickens_32.jpg \n inflating: /data/training/T_Boone_Pickens_50.jpg \n inflating: /data/training/T_Boone_Pickens_51.jpg \n inflating: /data/training/T_Boone_Pickens_52.jpg \n inflating: /data/training/Takeo_Hiranuma_00.jpg \n inflating: /data/training/Takeo_Hiranuma_01.jpg \n inflating: /data/training/Takeo_Hiranuma_02.jpg \n inflating: /data/training/Takeo_Hiranuma_10.jpg \n inflating: /data/training/Takeo_Hiranuma_11.jpg \n inflating: /data/training/Takeo_Hiranuma_12.jpg \n inflating: /data/training/Takeo_Hiranuma_30.jpg \n inflating: /data/training/Takeo_Hiranuma_31.jpg \n inflating: /data/training/Takeo_Hiranuma_32.jpg \n inflating: /data/training/Ted_Turner_20.jpg \n inflating: /data/training/Ted_Turner_21.jpg \n inflating: /data/training/Ted_Turner_22.jpg \n inflating: /data/training/Ted_Turner_30.jpg \n inflating: /data/training/Ted_Turner_31.jpg \n inflating: /data/training/Ted_Turner_32.jpg \n inflating: /data/training/Ted_Turner_50.jpg \n inflating: /data/training/Ted_Turner_51.jpg \n inflating: /data/training/Ted_Turner_52.jpg \n inflating: /data/training/Teresa_Heinz_Kerry_00.jpg \n inflating: /data/training/Teresa_Heinz_Kerry_01.jpg \n inflating: /data/training/Teresa_Heinz_Kerry_02.jpg \n inflating: /data/training/Teresa_Heinz_Kerry_10.jpg \n inflating: /data/training/Teresa_Heinz_Kerry_11.jpg \n inflating: /data/training/Teresa_Heinz_Kerry_12.jpg \n inflating: /data/training/Teresa_Heinz_Kerry_20.jpg \n inflating: /data/training/Teresa_Heinz_Kerry_21.jpg \n inflating: /data/training/Teresa_Heinz_Kerry_22.jpg \n inflating: /data/training/Terje_Roed-Larsen_00.jpg \n inflating: /data/training/Terje_Roed-Larsen_01.jpg \n inflating: /data/training/Terje_Roed-Larsen_02.jpg \n inflating: /data/training/Terje_Roed-Larsen_20.jpg \n inflating: /data/training/Terje_Roed-Larsen_21.jpg \n inflating: /data/training/Terje_Roed-Larsen_22.jpg \n inflating: /data/training/Terje_Roed-Larsen_30.jpg \n inflating: /data/training/Terje_Roed-Larsen_31.jpg \n inflating: /data/training/Terje_Roed-Larsen_32.jpg \n inflating: /data/training/Tessa_Jowell_00.jpg \n inflating: /data/training/Tessa_Jowell_01.jpg \n inflating: /data/training/Tessa_Jowell_02.jpg \n inflating: /data/training/Tessa_Jowell_20.jpg \n inflating: /data/training/Tessa_Jowell_21.jpg \n inflating: /data/training/Tessa_Jowell_22.jpg \n inflating: /data/training/Tessa_Jowell_30.jpg \n inflating: /data/training/Tessa_Jowell_31.jpg \n inflating: /data/training/Tessa_Jowell_32.jpg \n inflating: /data/training/Tessa_Jowell_50.jpg \n inflating: /data/training/Tessa_Jowell_51.jpg \n inflating: /data/training/Tessa_Jowell_52.jpg \n inflating: /data/training/Thomas_Ferguson_00.jpg \n inflating: /data/training/Thomas_Ferguson_01.jpg \n inflating: /data/training/Thomas_Ferguson_02.jpg \n inflating: /data/training/Thomas_Ferguson_10.jpg \n inflating: /data/training/Thomas_Ferguson_11.jpg \n inflating: /data/training/Thomas_Ferguson_12.jpg \n inflating: /data/training/Thomas_Ferguson_50.jpg \n inflating: /data/training/Thomas_Ferguson_51.jpg \n inflating: /data/training/Thomas_Ferguson_52.jpg \n inflating: /data/training/Tim_Howard_00.jpg \n inflating: /data/training/Tim_Howard_01.jpg \n inflating: /data/training/Tim_Howard_02.jpg \n inflating: /data/training/Tim_Howard_10.jpg \n inflating: /data/training/Tim_Howard_11.jpg \n inflating: /data/training/Tim_Howard_12.jpg \n inflating: /data/training/Tim_Howard_30.jpg \n inflating: /data/training/Tim_Howard_31.jpg \n inflating: /data/training/Tim_Howard_32.jpg \n inflating: /data/training/Tim_Pawlenty_00.jpg \n inflating: /data/training/Tim_Pawlenty_01.jpg \n inflating: /data/training/Tim_Pawlenty_02.jpg \n inflating: /data/training/Tim_Pawlenty_30.jpg \n inflating: /data/training/Tim_Pawlenty_31.jpg \n inflating: /data/training/Tim_Pawlenty_32.jpg \n inflating: /data/training/Tim_Pawlenty_40.jpg \n inflating: /data/training/Tim_Pawlenty_41.jpg \n inflating: /data/training/Tim_Pawlenty_42.jpg \n inflating: /data/training/Tim_Pawlenty_50.jpg \n inflating: /data/training/Tim_Pawlenty_51.jpg \n inflating: /data/training/Tim_Pawlenty_52.jpg \n inflating: /data/training/Timothy_Goebel_00.jpg \n inflating: /data/training/Timothy_Goebel_01.jpg \n inflating: /data/training/Timothy_Goebel_02.jpg \n inflating: /data/training/Timothy_Goebel_30.jpg \n inflating: /data/training/Timothy_Goebel_31.jpg \n inflating: /data/training/Timothy_Goebel_32.jpg \n inflating: /data/training/Timothy_Goebel_40.jpg \n inflating: /data/training/Timothy_Goebel_41.jpg \n inflating: /data/training/Timothy_Goebel_42.jpg \n inflating: /data/training/Timothy_Goebel_50.jpg \n inflating: /data/training/Timothy_Goebel_51.jpg \n inflating: /data/training/Timothy_Goebel_52.jpg \n inflating: /data/training/Tina_Brown_20.jpg \n inflating: /data/training/Tina_Brown_21.jpg \n inflating: /data/training/Tina_Brown_22.jpg \n inflating: /data/training/Tina_Brown_40.jpg \n inflating: /data/training/Tina_Brown_41.jpg \n inflating: /data/training/Tina_Brown_42.jpg \n inflating: /data/training/Tina_Brown_50.jpg \n inflating: /data/training/Tina_Brown_51.jpg \n inflating: /data/training/Tina_Brown_52.jpg \n inflating: /data/training/Tom_Coughlin_20.jpg \n inflating: /data/training/Tom_Coughlin_21.jpg \n inflating: /data/training/Tom_Coughlin_22.jpg \n inflating: /data/training/Tom_Coughlin_30.jpg \n inflating: /data/training/Tom_Coughlin_31.jpg \n inflating: /data/training/Tom_Coughlin_32.jpg \n inflating: /data/training/Tom_Coughlin_50.jpg \n inflating: /data/training/Tom_Coughlin_51.jpg \n inflating: /data/training/Tom_Coughlin_52.jpg \n inflating: /data/training/Tom_Hanks_30.jpg \n inflating: /data/training/Tom_Hanks_31.jpg \n inflating: /data/training/Tom_Hanks_32.jpg \n inflating: /data/training/Tom_Hanks_40.jpg \n inflating: /data/training/Tom_Hanks_41.jpg \n inflating: /data/training/Tom_Hanks_42.jpg \n inflating: /data/training/Tom_Hanks_50.jpg \n inflating: /data/training/Tom_Hanks_51.jpg \n inflating: /data/training/Tom_Hanks_52.jpg \n inflating: /data/training/Tom_Harkin_00.jpg \n inflating: /data/training/Tom_Harkin_01.jpg \n inflating: /data/training/Tom_Harkin_02.jpg \n inflating: /data/training/Tom_Harkin_30.jpg \n inflating: /data/training/Tom_Harkin_31.jpg \n inflating: /data/training/Tom_Harkin_32.jpg \n inflating: /data/training/Tom_Harkin_40.jpg \n inflating: /data/training/Tom_Harkin_41.jpg \n inflating: /data/training/Tom_Harkin_42.jpg \n inflating: /data/training/Tom_Osborne_20.jpg \n inflating: /data/training/Tom_Osborne_21.jpg \n inflating: /data/training/Tom_Osborne_22.jpg \n inflating: /data/training/Tom_Osborne_30.jpg \n inflating: /data/training/Tom_Osborne_31.jpg \n inflating: /data/training/Tom_Osborne_32.jpg \n inflating: /data/training/Tom_Osborne_50.jpg \n inflating: /data/training/Tom_Osborne_51.jpg \n inflating: /data/training/Tom_Osborne_52.jpg \n inflating: /data/training/Tom_Ridge_20.jpg \n inflating: /data/training/Tom_Ridge_21.jpg \n inflating: /data/training/Tom_Ridge_22.jpg \n inflating: /data/training/Tom_Ridge_30.jpg \n inflating: /data/training/Tom_Ridge_31.jpg \n inflating: /data/training/Tom_Ridge_32.jpg \n inflating: /data/training/Tom_Ridge_50.jpg \n inflating: /data/training/Tom_Ridge_51.jpg \n inflating: /data/training/Tom_Ridge_52.jpg \n inflating: /data/training/Tom_Sizemore_00.jpg \n inflating: /data/training/Tom_Sizemore_01.jpg \n inflating: /data/training/Tom_Sizemore_02.jpg \n inflating: /data/training/Tom_Sizemore_10.jpg \n inflating: /data/training/Tom_Sizemore_11.jpg \n inflating: /data/training/Tom_Sizemore_12.jpg \n inflating: /data/training/Tom_Sizemore_20.jpg \n inflating: /data/training/Tom_Sizemore_21.jpg \n inflating: /data/training/Tom_Sizemore_22.jpg \n inflating: /data/training/Valerie_Harper_00.jpg \n inflating: /data/training/Valerie_Harper_01.jpg \n inflating: /data/training/Valerie_Harper_02.jpg \n inflating: /data/training/Valerie_Harper_30.jpg \n inflating: /data/training/Valerie_Harper_31.jpg \n inflating: /data/training/Valerie_Harper_32.jpg \n inflating: /data/training/Valerie_Harper_40.jpg \n inflating: /data/training/Valerie_Harper_41.jpg \n inflating: /data/training/Valerie_Harper_42.jpg \n inflating: /data/training/Valerie_Harper_50.jpg \n inflating: /data/training/Valerie_Harper_51.jpg \n inflating: /data/training/Valerie_Harper_52.jpg \n inflating: /data/training/Vicente_Fox_10.jpg \n inflating: /data/training/Vicente_Fox_11.jpg \n inflating: /data/training/Vicente_Fox_12.jpg \n inflating: /data/training/Vicente_Fox_20.jpg \n inflating: /data/training/Vicente_Fox_21.jpg \n inflating: /data/training/Vicente_Fox_22.jpg \n inflating: /data/training/Vicente_Fox_30.jpg \n inflating: /data/training/Vicente_Fox_31.jpg \n inflating: /data/training/Vicente_Fox_32.jpg \n inflating: /data/training/Vojislav_Seselj_00.jpg \n inflating: /data/training/Vojislav_Seselj_01.jpg \n inflating: /data/training/Vojislav_Seselj_02.jpg \n inflating: /data/training/Vojislav_Seselj_20.jpg \n inflating: /data/training/Vojislav_Seselj_21.jpg \n inflating: /data/training/Vojislav_Seselj_22.jpg \n inflating: /data/training/Vojislav_Seselj_40.jpg \n inflating: /data/training/Vojislav_Seselj_41.jpg \n inflating: /data/training/Vojislav_Seselj_42.jpg \n inflating: /data/training/Vojislav_Seselj_50.jpg \n inflating: /data/training/Vojislav_Seselj_51.jpg \n inflating: /data/training/Vojislav_Seselj_52.jpg \n inflating: /data/training/Warren_Beatty_10.jpg \n inflating: /data/training/Warren_Beatty_11.jpg \n inflating: /data/training/Warren_Beatty_12.jpg \n inflating: /data/training/Warren_Beatty_30.jpg \n inflating: /data/training/Warren_Beatty_31.jpg \n inflating: /data/training/Warren_Beatty_32.jpg \n inflating: /data/training/Warren_Beatty_50.jpg \n inflating: /data/training/Warren_Beatty_51.jpg \n inflating: /data/training/Warren_Beatty_52.jpg \n inflating: /data/training/Warren_Buffett_00.jpg \n inflating: /data/training/Warren_Buffett_01.jpg \n inflating: /data/training/Warren_Buffett_02.jpg \n inflating: /data/training/Warren_Buffett_30.jpg \n inflating: /data/training/Warren_Buffett_31.jpg \n inflating: /data/training/Warren_Buffett_32.jpg \n inflating: /data/training/Warren_Buffett_40.jpg \n inflating: /data/training/Warren_Buffett_41.jpg \n inflating: /data/training/Warren_Buffett_42.jpg \n inflating: /data/training/Warren_Buffett_50.jpg \n inflating: /data/training/Warren_Buffett_51.jpg \n inflating: /data/training/Warren_Buffett_52.jpg \n inflating: /data/training/Wayne_Allard_00.jpg \n inflating: /data/training/Wayne_Allard_01.jpg \n inflating: /data/training/Wayne_Allard_02.jpg \n inflating: /data/training/Wayne_Allard_20.jpg \n inflating: /data/training/Wayne_Allard_21.jpg \n inflating: /data/training/Wayne_Allard_22.jpg \n inflating: /data/training/Wayne_Allard_50.jpg \n inflating: /data/training/Wayne_Allard_51.jpg \n inflating: /data/training/Wayne_Allard_52.jpg \n inflating: /data/training/Wayne_Gretzky_20.jpg \n inflating: /data/training/Wayne_Gretzky_21.jpg \n inflating: /data/training/Wayne_Gretzky_22.jpg \n inflating: /data/training/Wayne_Gretzky_30.jpg \n inflating: /data/training/Wayne_Gretzky_31.jpg \n inflating: /data/training/Wayne_Gretzky_32.jpg \n inflating: /data/training/Wayne_Gretzky_40.jpg \n inflating: /data/training/Wayne_Gretzky_41.jpg \n inflating: /data/training/Wayne_Gretzky_42.jpg \n inflating: /data/training/Wayne_Newton_10.jpg \n inflating: /data/training/Wayne_Newton_11.jpg \n inflating: /data/training/Wayne_Newton_12.jpg \n inflating: /data/training/Wayne_Newton_20.jpg \n inflating: /data/training/Wayne_Newton_21.jpg \n inflating: /data/training/Wayne_Newton_22.jpg \n inflating: /data/training/Wayne_Newton_40.jpg \n inflating: /data/training/Wayne_Newton_41.jpg \n inflating: /data/training/Wayne_Newton_42.jpg \n inflating: /data/training/Wes_Craven_00.jpg \n inflating: /data/training/Wes_Craven_01.jpg \n inflating: /data/training/Wes_Craven_02.jpg \n inflating: /data/training/Wes_Craven_20.jpg \n inflating: /data/training/Wes_Craven_21.jpg \n inflating: /data/training/Wes_Craven_22.jpg \n inflating: /data/training/Wes_Craven_30.jpg \n inflating: /data/training/Wes_Craven_31.jpg \n inflating: /data/training/Wes_Craven_32.jpg \n inflating: /data/training/Wes_Craven_50.jpg \n inflating: /data/training/Wes_Craven_51.jpg \n inflating: /data/training/Wes_Craven_52.jpg \n inflating: /data/training/Wesley_Clark_00.jpg \n inflating: /data/training/Wesley_Clark_01.jpg \n inflating: /data/training/Wesley_Clark_02.jpg \n inflating: /data/training/Wesley_Clark_20.jpg \n inflating: /data/training/Wesley_Clark_21.jpg \n inflating: /data/training/Wesley_Clark_22.jpg \n inflating: /data/training/Wesley_Clark_30.jpg \n inflating: /data/training/Wesley_Clark_31.jpg \n inflating: /data/training/Wesley_Clark_32.jpg \n inflating: /data/training/Wesley_Clark_40.jpg \n inflating: /data/training/Wesley_Clark_41.jpg \n inflating: /data/training/Wesley_Clark_42.jpg \n inflating: /data/training/Whoopi_Goldberg_00.jpg \n inflating: /data/training/Whoopi_Goldberg_01.jpg \n inflating: /data/training/Whoopi_Goldberg_02.jpg \n inflating: /data/training/Whoopi_Goldberg_40.jpg \n inflating: /data/training/Whoopi_Goldberg_41.jpg \n inflating: /data/training/Whoopi_Goldberg_42.jpg \n inflating: /data/training/Whoopi_Goldberg_50.jpg \n inflating: /data/training/Whoopi_Goldberg_51.jpg \n inflating: /data/training/Whoopi_Goldberg_52.jpg \n inflating: /data/training/William_Delahunt_00.jpg \n inflating: /data/training/William_Delahunt_01.jpg \n inflating: /data/training/William_Delahunt_02.jpg \n inflating: /data/training/William_Delahunt_10.jpg \n inflating: /data/training/William_Delahunt_11.jpg \n inflating: /data/training/William_Delahunt_12.jpg \n inflating: /data/training/William_Delahunt_20.jpg \n inflating: /data/training/William_Delahunt_21.jpg \n inflating: /data/training/William_Delahunt_22.jpg \n inflating: /data/training/William_Donaldson_00.jpg \n inflating: /data/training/William_Donaldson_01.jpg \n inflating: /data/training/William_Donaldson_02.jpg \n inflating: /data/training/William_Donaldson_10.jpg \n inflating: /data/training/William_Donaldson_11.jpg \n inflating: /data/training/William_Donaldson_12.jpg \n inflating: /data/training/William_Donaldson_50.jpg \n inflating: /data/training/William_Donaldson_51.jpg \n inflating: /data/training/William_Donaldson_52.jpg \n inflating: /data/training/William_McDonough_00.jpg \n inflating: /data/training/William_McDonough_01.jpg \n inflating: /data/training/William_McDonough_02.jpg \n inflating: /data/training/William_McDonough_20.jpg \n inflating: /data/training/William_McDonough_21.jpg \n inflating: /data/training/William_McDonough_22.jpg \n inflating: /data/training/William_McDonough_30.jpg \n inflating: /data/training/William_McDonough_31.jpg \n inflating: /data/training/William_McDonough_32.jpg \n inflating: /data/training/William_McDonough_40.jpg \n inflating: /data/training/William_McDonough_41.jpg \n inflating: /data/training/William_McDonough_42.jpg \n inflating: /data/training/Yang_Jianli_00.jpg \n inflating: /data/training/Yang_Jianli_01.jpg \n inflating: /data/training/Yang_Jianli_02.jpg \n inflating: /data/training/Yang_Jianli_10.jpg \n inflating: /data/training/Yang_Jianli_11.jpg \n inflating: /data/training/Yang_Jianli_12.jpg \n inflating: /data/training/Yang_Jianli_30.jpg \n inflating: /data/training/Yang_Jianli_31.jpg \n inflating: /data/training/Yang_Jianli_32.jpg \n inflating: /data/training/Yuri_Fedotov_20.jpg \n inflating: /data/training/Yuri_Fedotov_21.jpg \n inflating: /data/training/Yuri_Fedotov_22.jpg \n inflating: /data/training/Yuri_Fedotov_30.jpg \n inflating: /data/training/Yuri_Fedotov_31.jpg \n inflating: /data/training/Yuri_Fedotov_32.jpg \n inflating: /data/training/Yuri_Fedotov_40.jpg \n inflating: /data/training/Yuri_Fedotov_41.jpg \n inflating: /data/training/Yuri_Fedotov_42.jpg \n inflating: /data/training/Zhang_Ziyi_10.jpg \n inflating: /data/training/Zhang_Ziyi_11.jpg \n inflating: /data/training/Zhang_Ziyi_12.jpg \n inflating: /data/training/Zhang_Ziyi_20.jpg \n inflating: /data/training/Zhang_Ziyi_21.jpg \n inflating: /data/training/Zhang_Ziyi_22.jpg \n inflating: /data/training/Zhang_Ziyi_40.jpg \n inflating: /data/training/Zhang_Ziyi_41.jpg \n inflating: /data/training/Zhang_Ziyi_42.jpg \n inflating: /data/training/Zhong_Nanshan_00.jpg \n inflating: /data/training/Zhong_Nanshan_01.jpg \n inflating: /data/training/Zhong_Nanshan_02.jpg \n inflating: /data/training/Zhong_Nanshan_10.jpg \n inflating: /data/training/Zhong_Nanshan_11.jpg \n inflating: /data/training/Zhong_Nanshan_12.jpg \n inflating: /data/training/Zhong_Nanshan_50.jpg \n inflating: /data/training/Zhong_Nanshan_51.jpg \n inflating: /data/training/Zhong_Nanshan_52.jpg \n inflating: /data/training_frames_keypoints.csv \n"
],
[
"# import the required libraries\nimport glob\nimport os\nimport numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport matplotlib.image as mpimg\n\nimport cv2",
"_____no_output_____"
]
],
[
[
"Then, let's load in our training data and display some stats about that dat ato make sure it's been loaded in correctly!",
"_____no_output_____"
]
],
[
[
"key_pts_frame = pd.read_csv('/data/training_frames_keypoints.csv')\n\nn = 0\nimage_name = key_pts_frame.iloc[n, 0]\nkey_pts = key_pts_frame.iloc[n, 1:].as_matrix()\nkey_pts = key_pts.astype('float').reshape(-1, 2)\n\nprint('Image name: ', image_name)\nprint('Landmarks shape: ', key_pts.shape)\nprint('First 4 key pts: {}'.format(key_pts[:4]))",
"Image name: Luis_Fonsi_21.jpg\nLandmarks shape: (68, 2)\nFirst 4 key pts: [[ 45. 98.]\n [ 47. 106.]\n [ 49. 110.]\n [ 53. 119.]]\n"
],
[
"# print out some stats about the data\nprint('Number of images: ', key_pts_frame.shape[0])\n\nimg2 = cv2.imread('/data/training/Nicolas_Sarkozy_02.jpg')\nwidth, height, channels = img2.shape\nnpixel = img2.size\nprint(\"npixel\", npixel)\nprint(\"width {}, height {}, channels {}\".format(width, height, channels))\n",
"Number of images: 3462\nnpixel 53940\nwidth 155, height 116, channels 3\n"
]
],
[
[
"## Look at some images\n\nBelow, is a function `show_keypoints` that takes in an image and keypoints and displays them. As you look at this data, **note that these images are not all of the same size**, and neither are the faces! To eventually train a neural network on these images, we'll need to standardize their shape.",
"_____no_output_____"
]
],
[
[
"def show_keypoints(image, key_pts):\n \"\"\"Show image with keypoints\"\"\"\n plt.imshow(image)\n plt.scatter(key_pts[:, 0], key_pts[:, 1], s=20, marker='.', c='m')\n",
"_____no_output_____"
],
[
"# Display a few different types of images by changing the index n\n\n# select an image by index in our data frame\nn = 0\nimage_name = key_pts_frame.iloc[n, 0]\nkey_pts = key_pts_frame.iloc[n, 1:].as_matrix()\nkey_pts = key_pts.astype('float').reshape(-1, 2)\n\nplt.figure(figsize=(5, 5))\nshow_keypoints(mpimg.imread(os.path.join('/data/training/', image_name)), key_pts)\nplt.show()",
"_____no_output_____"
]
],
[
[
"## Dataset class and Transformations\n\nTo prepare our data for training, we'll be using PyTorch's Dataset class. Much of this this code is a modified version of what can be found in the [PyTorch data loading tutorial](http://pytorch.org/tutorials/beginner/data_loading_tutorial.html).\n\n#### Dataset class\n\n``torch.utils.data.Dataset`` is an abstract class representing a\ndataset. This class will allow us to load batches of image/keypoint data, and uniformly apply transformations to our data, such as rescaling and normalizing images for training a neural network.\n\n\nYour custom dataset should inherit ``Dataset`` and override the following\nmethods:\n\n- ``__len__`` so that ``len(dataset)`` returns the size of the dataset.\n- ``__getitem__`` to support the indexing such that ``dataset[i]`` can\n be used to get the i-th sample of image/keypoint data.\n\nLet's create a dataset class for our face keypoints dataset. We will\nread the CSV file in ``__init__`` but leave the reading of images to\n``__getitem__``. This is memory efficient because all the images are not\nstored in the memory at once but read as required.\n\nA sample of our dataset will be a dictionary\n``{'image': image, 'keypoints': key_pts}``. Our dataset will take an\noptional argument ``transform`` so that any required processing can be\napplied on the sample. We will see the usefulness of ``transform`` in the\nnext section.\n",
"_____no_output_____"
]
],
[
[
"from torch.utils.data import Dataset, DataLoader\n\nclass FacialKeypointsDataset(Dataset):\n \"\"\"Face Landmarks dataset.\"\"\"\n\n def __init__(self, csv_file, root_dir, transform=None):\n \"\"\"\n Args:\n csv_file (string): Path to the csv file with annotations.\n root_dir (string): Directory with all the images.\n transform (callable, optional): Optional transform to be applied\n on a sample.\n \"\"\"\n self.key_pts_frame = pd.read_csv(csv_file)\n self.root_dir = root_dir\n self.transform = transform\n\n def __len__(self):\n return len(self.key_pts_frame)\n\n def __getitem__(self, idx):\n image_name = os.path.join(self.root_dir,\n self.key_pts_frame.iloc[idx, 0])\n \n image = mpimg.imread(image_name)\n \n # if image has an alpha color channel, get rid of it\n if(image.shape[2] == 4):\n image = image[:,:,0:3]\n \n key_pts = self.key_pts_frame.iloc[idx, 1:].as_matrix()\n key_pts = key_pts.astype('float').reshape(-1, 2)\n sample = {'image': image, 'keypoints': key_pts}\n\n if self.transform:\n sample = self.transform(sample)\n\n return sample",
"_____no_output_____"
]
],
[
[
"Now that we've defined this class, let's instantiate the dataset and display some images.",
"_____no_output_____"
]
],
[
[
"# Construct the dataset\nface_dataset = FacialKeypointsDataset(csv_file='/data/training_frames_keypoints.csv',\n root_dir='/data/training/')\n\n# print some stats about the dataset\nprint('Length of dataset: ', len(face_dataset))",
"Length of dataset: 3462\n"
],
[
"# Display a few of the images from the dataset\nnum_to_display = 3\n\nfor i in range(num_to_display):\n \n # define the size of images\n fig = plt.figure(figsize=(20,10))\n \n # randomly select a sample\n rand_i = np.random.randint(0, len(face_dataset))\n sample = face_dataset[rand_i]\n\n # print the shape of the image and keypoints\n print(i, sample['image'].shape, sample['keypoints'].shape)\n\n ax = plt.subplot(1, num_to_display, i + 1)\n ax.set_title('Sample #{}'.format(i))\n \n # Using the same display function, defined earlier\n show_keypoints(sample['image'], sample['keypoints'])\n",
"0 (360, 423, 3) (68, 2)\n1 (331, 287, 3) (68, 2)\n2 (315, 271, 3) (68, 2)\n"
]
],
[
[
"## Transforms\n\nNow, the images above are not of the same size, and neural networks often expect images that are standardized; a fixed size, with a normalized range for color ranges and coordinates, and (for PyTorch) converted from numpy lists and arrays to Tensors.\n\nTherefore, we will need to write some pre-processing code.\nLet's create four transforms:\n\n- ``Normalize``: to convert a color image to grayscale values with a range of [0,1] and normalize the keypoints to be in a range of about [-1, 1]\n- ``Rescale``: to rescale an image to a desired size.\n- ``RandomCrop``: to crop an image randomly.\n- ``ToTensor``: to convert numpy images to torch images.\n\n\nWe will write them as callable classes instead of simple functions so\nthat parameters of the transform need not be passed everytime it's\ncalled. For this, we just need to implement ``__call__`` method and \n(if we require parameters to be passed in), the ``__init__`` method. \nWe can then use a transform like this:\n\n tx = Transform(params)\n transformed_sample = tx(sample)\n\nObserve below how these transforms are generally applied to both the image and its keypoints.\n\n",
"_____no_output_____"
]
],
[
[
"import torch\nfrom torchvision import transforms, utils\n# tranforms\n\nclass Normalize(object):\n \"\"\"Convert a color image to grayscale and normalize the color range to [0,1].\"\"\" \n\n def __call__(self, sample):\n image, key_pts = sample['image'], sample['keypoints']\n \n image_copy = np.copy(image)\n key_pts_copy = np.copy(key_pts)\n\n # convert image to grayscale\n image_copy = cv2.cvtColor(image, cv2.COLOR_RGB2GRAY)\n \n # scale color range from [0, 255] to [0, 1]\n image_copy= image_copy/255.0\n \n # scale keypoints to be centered around 0 with a range of [-1, 1]\n # mean = 100, sqrt = 50, so, pts should be (pts - 100)/50\n key_pts_copy = (key_pts_copy - 100)/50.0\n\n\n return {'image': image_copy, 'keypoints': key_pts_copy}\n\n\nclass Rescale(object):\n \"\"\"Rescale the image in a sample to a given size.\n\n Args:\n output_size (tuple or int): Desired output size. If tuple, output is\n matched to output_size. If int, smaller of image edges is matched\n to output_size keeping aspect ratio the same.\n \"\"\"\n\n def __init__(self, output_size):\n assert isinstance(output_size, (int, tuple))\n self.output_size = output_size\n\n def __call__(self, sample):\n image, key_pts = sample['image'], sample['keypoints']\n\n h, w = image.shape[:2]\n if isinstance(self.output_size, int):\n if h > w:\n new_h, new_w = self.output_size * h / w, self.output_size\n else:\n new_h, new_w = self.output_size, self.output_size * w / h\n else:\n new_h, new_w = self.output_size\n\n new_h, new_w = int(new_h), int(new_w)\n\n img = cv2.resize(image, (new_w, new_h))\n \n # scale the pts, too\n key_pts = key_pts * [new_w / w, new_h / h]\n\n return {'image': img, 'keypoints': key_pts}\n\n\nclass RandomCrop(object):\n \"\"\"Crop randomly the image in a sample.\n\n Args:\n output_size (tuple or int): Desired output size. If int, square crop\n is made.\n \"\"\"\n\n def __init__(self, output_size):\n assert isinstance(output_size, (int, tuple))\n if isinstance(output_size, int):\n self.output_size = (output_size, output_size)\n else:\n assert len(output_size) == 2\n self.output_size = output_size\n\n def __call__(self, sample):\n image, key_pts = sample['image'], sample['keypoints']\n\n h, w = image.shape[:2]\n new_h, new_w = self.output_size\n\n top = np.random.randint(0, h - new_h)\n left = np.random.randint(0, w - new_w)\n\n image = image[top: top + new_h,\n left: left + new_w]\n\n key_pts = key_pts - [left, top]\n\n return {'image': image, 'keypoints': key_pts}\n\n\nclass ToTensor(object):\n \"\"\"Convert ndarrays in sample to Tensors.\"\"\"\n\n def __call__(self, sample):\n image, key_pts = sample['image'], sample['keypoints']\n \n # if image has no grayscale color channel, add one\n if(len(image.shape) == 2):\n # add that third color dim\n image = image.reshape(image.shape[0], image.shape[1], 1)\n \n # swap color axis because\n # numpy image: H x W x C\n # torch image: C X H X W\n image = image.transpose((2, 0, 1))\n \n return {'image': torch.from_numpy(image),\n 'keypoints': torch.from_numpy(key_pts)}",
"_____no_output_____"
]
],
[
[
"## Test out the transforms\n\nLet's test these transforms out to make sure they behave as expected. As you look at each transform, note that, in this case, **order does matter**. For example, you cannot crop a image using a value smaller than the original image (and the orginal images vary in size!), but, if you first rescale the original image, you can then crop it to any size smaller than the rescaled size.",
"_____no_output_____"
]
],
[
[
"# test out some of these transforms\nrescale = Rescale(100)\ncrop = RandomCrop(50)\ncomposed = transforms.Compose([Rescale(250),\n RandomCrop(224)])\n\n# apply the transforms to a sample image\ntest_num = 500\nsample = face_dataset[test_num]\n\nfig = plt.figure()\nfor i, tx in enumerate([rescale, crop, composed]):\n transformed_sample = tx(sample)\n\n ax = plt.subplot(1, 3, i + 1)\n plt.tight_layout()\n ax.set_title(type(tx).__name__)\n show_keypoints(transformed_sample['image'], transformed_sample['keypoints'])\n\nplt.show()",
"_____no_output_____"
]
],
[
[
"## Create the transformed dataset\n\nApply the transforms in order to get grayscale images of the same shape. Verify that your transform works by printing out the shape of the resulting data (printing out a few examples should show you a consistent tensor size).",
"_____no_output_____"
]
],
[
[
"# define the data tranform\n# order matters! i.e. rescaling should come before a smaller crop\ndata_transform = transforms.Compose([Rescale(250),\n RandomCrop(224),\n Normalize(),\n ToTensor()])\n\n# create the transformed dataset\ntransformed_dataset = FacialKeypointsDataset(csv_file='/data/training_frames_keypoints.csv',\n root_dir='/data/training/',\n transform=data_transform)\n",
"_____no_output_____"
],
[
"# print some stats about the transformed data\nprint('Number of images: ', len(transformed_dataset))\n\n# make sure the sample tensors are the expected size\nfor i in range(5):\n sample = transformed_dataset[i]\n print(i, sample['image'].size(), sample['keypoints'].size())\n",
"Number of images: 3462\n0 torch.Size([1, 224, 224]) torch.Size([68, 2])\n1 torch.Size([1, 224, 224]) torch.Size([68, 2])\n2 torch.Size([1, 224, 224]) torch.Size([68, 2])\n3 torch.Size([1, 224, 224]) torch.Size([68, 2])\n4 torch.Size([1, 224, 224]) torch.Size([68, 2])\n"
]
],
[
[
"## Data Iteration and Batching\n\nRight now, we are iterating over this data using a ``for`` loop, but we are missing out on a lot of PyTorch's dataset capabilities, specifically the abilities to:\n\n- Batch the data\n- Shuffle the data\n- Load the data in parallel using ``multiprocessing`` workers.\n\n``torch.utils.data.DataLoader`` is an iterator which provides all these\nfeatures, and we'll see this in use in the *next* notebook, Notebook 2, when we load data in batches to train a neural network!\n\n---\n\n",
"_____no_output_____"
],
[
"## Ready to Train!\n\nNow that you've seen how to load and transform our data, you're ready to build a neural network to train on this data.\n\nIn the next notebook, you'll be tasked with creating a CNN for facial keypoint detection.",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
]
] |
cb949a51a37efd87477d120622652c9682fe9756
| 15,261 |
ipynb
|
Jupyter Notebook
|
scripts/.ipynb_checkpoints/rf_on_ls_protein_compound-checkpoint.ipynb
|
raghvendra5688/Drug-Repurposing
|
5416b9806818f988ca84e1a0e31b1b74adf7e60b
|
[
"MIT"
] | 2 |
2020-08-10T02:44:29.000Z
|
2022-02-24T14:29:06.000Z
|
scripts/.ipynb_checkpoints/rf_on_ls_protein_compound-checkpoint.ipynb
|
raghvendra5688/Drug-Repurposing
|
5416b9806818f988ca84e1a0e31b1b74adf7e60b
|
[
"MIT"
] | 4 |
2020-10-16T07:28:41.000Z
|
2022-02-10T02:21:25.000Z
|
scripts/.ipynb_checkpoints/rf_on_ls_protein_compound-checkpoint.ipynb
|
raghvendra5688/Drug-Repurposing
|
5416b9806818f988ca84e1a0e31b1b74adf7e60b
|
[
"MIT"
] | 3 |
2020-08-13T15:07:28.000Z
|
2021-03-21T11:05:06.000Z
| 42.628492 | 157 | 0.549112 |
[
[
[
"import os\nimport pickle\nimport re\nimport pandas as pd\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport seaborn as sn\nimport numpy as np\nimport re \n\nimport xgboost as xgb\nimport shap\n\nfrom sklearn import ensemble\nfrom sklearn import dummy\nfrom sklearn import linear_model\nfrom sklearn import svm\nfrom sklearn import neural_network\nfrom sklearn import metrics\nfrom sklearn import preprocessing\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.utils.fixes import loguniform\nimport scipy\n\nfrom misc import save_model, load_model, regression_results, grid_search_cv",
"_____no_output_____"
],
[
"# Options of settings with different Xs and Ys \noptions = [\"../data/Train_Compound_Viral_interactions_for_Supervised_Learning_with_LS_LS.csv\",\n \"../data/Train_Compound_Viral_interactions_for_Supervised_Learning_with_MFP_LS.csv\",\n \"..\"] #(to be continued)\n\ndata_type_options = [\"LS_Compound_LS_Protein\",\n \"MFP_Compound_LS_Protein\",\n \"..\"\n ]\n\n# input option is also used to control the model parameters below\ninput_option = 0\n\nclassification_task = False\nclassification_th = 85\n\ndata_type=data_type_options[input_option]\nfilename = options[input_option]\n\nwith open(filename, \"rb\") as file:\n print(\"Loading \", filename)\n big_df = pd.read_csv(filename, header='infer', delimiter=\",\")\n total_length = len(big_df.columns)\n X = big_df.iloc[:,range(5,total_length)]\n Y = big_df[['pchembl_value']].to_numpy().flatten()\n meta_X = big_df.iloc[:,[0,1,2,3]]\n print(\"Lengths --> X = %d, Y = %d\" % (len(X), len(Y)))\n\nprint(X.columns)\nn_samples = len(X)\nindices = np.arange(n_samples)\n\nX_train = X\ny_train = Y\nprint(X_train[:10])\nprint(X_train.shape,y_train.shape)\nprint(X_train.columns)\nprint(big_df.isnull().sum().sum())",
"Loading ../data/Train_Compound_Viral_interactions_for_Supervised_Learning_with_LS_LS.csv\nLengths --> X = 54175, Y = 54175\nIndex(['LS_0', 'LS_1', 'LS_2', 'LS_3', 'LS_4', 'LS_5', 'LS_6', 'LS_7', 'LS_8',\n 'LS_9',\n ...\n 'PLS_54', 'PLS_55', 'PLS_56', 'PLS_57', 'PLS_58', 'PLS_59', 'PLS_60',\n 'PLS_61', 'PLS_62', 'PLS_63'],\n dtype='object', length=320)\n LS_0 LS_1 LS_2 LS_3 LS_4 LS_5 LS_6 \\\n0 0.565035 -0.213065 -0.000764 0.015025 -0.031855 -0.012590 -0.000102 \n1 0.272218 -0.258328 0.000385 0.007196 -0.028770 0.062976 -0.000169 \n2 0.567573 0.057800 0.000016 0.012931 -0.012279 0.378251 -0.000176 \n3 0.801987 0.023184 0.000136 0.000772 -0.013787 -0.682254 -0.000193 \n4 0.833605 -0.166840 0.000225 0.007307 -0.028559 -0.172600 -0.000118 \n5 0.628016 -0.419587 -0.000024 0.001701 -0.006175 -0.149002 -0.000135 \n6 0.854284 0.167687 0.000309 0.019150 -0.011129 0.158235 -0.000310 \n7 0.898972 -0.057246 0.002465 0.004290 -0.010132 0.133053 -0.000133 \n8 0.445106 0.099132 0.000561 0.013880 -0.044212 0.032194 -0.000432 \n9 0.916135 0.124614 0.000845 0.014388 -0.029252 0.131530 -0.000099 \n\n LS_7 LS_8 LS_9 ... PLS_54 PLS_55 PLS_56 \\\n0 0.000173 0.043410 -0.000032 ... -7.175925 -1.589016 -1.450142 \n1 0.000112 -0.048768 -0.000148 ... 16.786903 0.683173 7.260680 \n2 0.003412 -0.551354 -0.000105 ... 16.786903 0.683173 7.260680 \n3 0.002080 -0.425750 -0.000040 ... 16.786903 0.683173 7.260680 \n4 0.000324 0.208194 -0.000021 ... 16.786903 0.683173 7.260680 \n5 0.000187 0.086830 -0.000137 ... 16.256847 6.623242 6.126185 \n6 0.000683 -0.230810 -0.000049 ... 16.786903 0.683173 7.260680 \n7 0.000369 0.154465 -0.000041 ... 16.786903 0.683173 7.260680 \n8 0.001020 -0.259994 -0.000025 ... 16.786903 0.683173 7.260680 \n9 0.001031 -0.423787 -0.000053 ... 16.786903 0.683173 7.260680 \n\n PLS_57 PLS_58 PLS_59 PLS_60 PLS_61 PLS_62 PLS_63 \n0 -97.836319 -12.442183 8.642500 14.662628 -5.112053 -4.386289 1.195183 \n1 -132.681641 9.667920 -0.224070 -0.795782 10.744227 -5.754991 7.939027 \n2 -132.681641 9.667920 -0.224070 -0.795782 10.744227 -5.754991 7.939027 \n3 -132.681641 9.667920 -0.224070 -0.795782 10.744227 -5.754991 7.939027 \n4 -132.681641 9.667920 -0.224070 -0.795782 10.744227 -5.754991 7.939027 \n5 -140.331253 -2.526256 -9.824449 -7.932061 -11.736423 -8.355380 -9.171484 \n6 -132.681641 9.667920 -0.224070 -0.795782 10.744227 -5.754991 7.939027 \n7 -132.681641 9.667920 -0.224070 -0.795782 10.744227 -5.754991 7.939027 \n8 -132.681641 9.667920 -0.224070 -0.795782 10.744227 -5.754991 7.939027 \n9 -132.681641 9.667920 -0.224070 -0.795782 10.744227 -5.754991 7.939027 \n\n[10 rows x 320 columns]\n(54175, 320) (54175,)\nIndex(['LS_0', 'LS_1', 'LS_2', 'LS_3', 'LS_4', 'LS_5', 'LS_6', 'LS_7', 'LS_8',\n 'LS_9',\n ...\n 'PLS_54', 'PLS_55', 'PLS_56', 'PLS_57', 'PLS_58', 'PLS_59', 'PLS_60',\n 'PLS_61', 'PLS_62', 'PLS_63'],\n dtype='object', length=320)\n0\n"
],
[
"def calculate_classification_metrics(labels, predictions):\n \n predictions = predictions.round()\n fpr, tpr, thresholds = metrics.roc_curve(labels, predictions)\n auc = metrics.auc(fpr, tpr)\n aupr = metrics.average_precision_score(labels,predictions)\n \n return metrics.accuracy_score(labels, predictions),\\\n metrics.f1_score(labels, predictions, average='binary'),\\\n auc,\\\n aupr\n\n\ndef calculate_regression_metrics(labels, predictions):\n return metrics.mean_absolute_error(labels, predictions),\\\n metrics.mean_squared_error(labels, predictions),\\\n metrics.r2_score(labels, predictions),\\\n scipy.stats.pearsonr(np.array(labels).flatten(),np.array(predictions.flatten()))[0],\\\n scipy.stats.spearmanr(np.array(labels).flatten(),np.array(predictions.flatten()))[0]\n\n\n",
"_____no_output_____"
],
[
"def supervised_learning_steps(method,scoring,data_type,task,model,params,X_train,y_train,n_iter):\n \n gs = grid_search_cv(model, params, X_train, y_train, scoring=scoring, n_iter = n_iter)\n\n y_pred = gs.predict(X_train)\n y_pred[y_pred < 0] = 0\n\n if task:\n results=calculate_classification_metrics(y_train, y_pred)\n print(\"Acc: %.3f, F1: %.3f, AUC: %.3f, AUPR: %.3f\" % (results[0], results[1], results[2], results[3]))\n else:\n results=calculate_regression_metrics(y_train,y_pred)\n print(\"MAE: %.3f, MSE: %.3f, R2: %.3f, Pearson R: %.3f, Spearman R: %.3f\" % (results[0], results[1], results[2], results[3], results[4]))\n \n print('Parameters')\n print('----------')\n for p,v in gs.best_estimator_.get_params().items():\n print(p, \":\", v)\n print('-' * 80)\n\n if task:\n save_model(gs, \"%s_models/%s_%s_classifier_gs.pk\" % (method,method,data_type))\n save_model(gs.best_estimator_, \"%s_models/%s_%s_classifier_best_estimator.pk\" %(method,method,data_type))\n else:\n save_model(gs, \"%s_models/%s_%s_regressor_gs.pk\" % (method,method,data_type))\n save_model(gs.best_estimator_, \"%s_models/%s_%s_regressor_best_estimator.pk\" %(method,method,data_type))\n \n return(gs)",
"_____no_output_____"
],
[
"if classification_task:\n model = ensemble.RandomForestRegressor(n_estimators=100, criterion='auc',\n max_depth=None, min_samples_split=2,\n min_samples_leaf=1, min_weight_fraction_leaf=0.0,\n max_features='auto', max_leaf_nodes=None,\n min_impurity_decrease=0.0, min_impurity_split=None,\n bootstrap=True, oob_score=False,\n n_jobs=-1, random_state=328, verbose=1,\n warm_start=False, ccp_alpha=0.0, max_samples=None)\n\nelse:\n model = ensemble.RandomForestRegressor(n_estimators=100, criterion='mse',\n max_depth=None, min_samples_split=2,\n min_samples_leaf=1, min_weight_fraction_leaf=0.0,\n max_features='auto', max_leaf_nodes=None,\n min_impurity_decrease=0.0, min_impurity_split=None,\n bootstrap=True, oob_score=False,\n n_jobs=-1, random_state=328, verbose=1,\n warm_start=False, ccp_alpha=0.0, max_samples=None)\n\n\n# Grid parameters\nparam_rf = {\"n_estimators\": scipy.stats.randint(20, 500),\n \"max_depth\": scipy.stats.randint(1, 9),\n \"min_samples_leaf\": scipy.stats.randint(1, 10),\n \"max_features\": scipy.stats.uniform.ppf([0.1,0.7])\n}\n\nn_iter=200\n\nif classification_task:\n rf_gs=supervised_learning_steps(\"rf\",\"roc_auc\",data_type,classification_task,model,param_rf,X_train,y_train,n_iter)\nelse:\n rf_gs=supervised_learning_steps(\"rf\",\"r2\",data_type,classification_task,model,param_rf,X_train,y_train,n_iter)\n\nrf_gs.cv_results_",
"_____no_output_____"
],
[
"rf_gs = load_model(\"rf_models/rf__LS_Drug_LS_Protein_regressor_gs.pk\")\nnp.max(rf_gs.cv_results_[\"mean_test_score\"])\n\nfile_list = [\"../data/Test_Compound_Viral_interactions_for_Supervised_Learning_with_LS_LS.csv\",\n \"../data/Test_Compound_Viral_interactions_for_Supervised_Learning_with_MFP_LS.csv\"]\n\nfilename = file_list[input_option]\nwith open(filename, \"rb\") as file:\n print(\"Loading \", filename)\n big_df = pd.read_csv(filename, header='infer', delimiter=\",\")\n total_length = len(big_df.columns)\n X = big_df.iloc[:,range(5,total_length)]\n Y = big_df[['pchembl_value']].to_numpy().flatten()\n meta_X = big_df.iloc[:,[0,1,2,3]]\n print(\"Lengths --> X = %d, Y = %d\" % (len(X), len(Y)))\n\nprint(X.columns)\nn_samples = len(X)\nindices = np.arange(n_samples)\n\nX_test = X\ny_test = Y\nrf_best = rf_gs.best_estimator_\ny_pred_rf=rf_best.predict(X_test)\nprint(calculate_regression_metrics(y_test,y_pred_rf))\n\n#Write the output in the results folder\nmeta_X[\"predictions\"]=y_pred_xgb\nmeta_X[\"labels\"]=y_test\nrev_output_df = meta_X.iloc[:,[0,2,4,5]].copy()\nrev_output_df.to_csv(\"../results/RF_\"+data_type_options[input_option]+\"supervised_test_predictions.csv\",index=False)",
"_____no_output_____"
],
[
"## load JS visualization code to notebook (Doesn't work for random forest)\n#shap.initjs()\n\n## explain the model's predictions using SHAP values\n#explainer = shap.TreeExplainer(xgb_gs.best_estimator_)\n#shap_values = explainer.shap_values(X_train)\n#shap.summary_plot(shap_values, X_train)",
"_____no_output_____"
],
[
"##Get results for SARS-COV-2\n#big_X_test = pd.read_csv(\"../data/COVID-19/sars_cov_2_additional_compound_viral_interactions_to_predict_with_LS_v2.csv\",header='infer',sep=\",\")\n#total_length = len(big_X_test.columns)\n#X_test = big_X_test.iloc[:,range(8,total_length)]\n#rf_best = load_model(\"../models/rf_models/rf__LS_Drug_LS_Protein_regressor_best_estimator.pk\")\n#y_pred = rf_best.predict(X_test)\n\n#meta_X_test = big_X_test.iloc[:,[0,2]].copy()\n#meta_X_test.loc[:,'predictions']=y_pred\n#meta_X_test.loc[:,'labels']=0\n#meta_X_test.to_csv(\"../results/RF_supervised_sars_cov2_additional_test_predictions.csv\",index=False)",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb94a9b3ca1066e49784b66999709d3b3647fff1
| 4,161 |
ipynb
|
Jupyter Notebook
|
bases/br_anatel_banda_larga_fixa/code/br_anatel_banda_larga_fixa_backhaul.ipynb
|
isadorabugarin/mais
|
77f1e23b56ba01a9a329206a8d30a57028349c36
|
[
"MIT"
] | 290 |
2020-10-14T17:18:21.000Z
|
2022-03-31T20:56:07.000Z
|
bases/br_anatel_banda_larga_fixa/code/br_anatel_banda_larga_fixa_backhaul.ipynb
|
isadorabugarin/mais
|
77f1e23b56ba01a9a329206a8d30a57028349c36
|
[
"MIT"
] | 756 |
2020-10-09T16:37:57.000Z
|
2022-03-31T18:28:18.000Z
|
bases/br_anatel_banda_larga_fixa/code/br_anatel_banda_larga_fixa_backhaul.ipynb
|
isadorabugarin/mais
|
77f1e23b56ba01a9a329206a8d30a57028349c36
|
[
"MIT"
] | 81 |
2020-10-15T18:21:42.000Z
|
2022-03-31T03:25:13.000Z
| 30.595588 | 163 | 0.522711 |
[
[
[
"from google.colab import drive\ndrive.mount('/content/drive', force_remount=True)",
"Mounted at /content/drive\n"
],
[
"import pandas as pd\nimport numpy as np",
"_____no_output_____"
],
[
"backhaul = pd.read_excel('/content/drive/MyDrive/Anatel Data/Backhaul 2018.xlsx')",
"_____no_output_____"
],
[
"#exclusão de observações da planilha original\nbackhaul.dropna(inplace=True)\n#exclusão de variáveis\nbackhaul.drop(['População', 'Município', 'PGO', 'Prestadora'], axis=1, inplace=True)",
"_____no_output_____"
],
[
"#renomeação das variaveis\nbackhaul.rename(columns={'Cód. IBGE':'id_municipio', 'UF':'sigla_uf', 'Situação':'situacao', 'Ano de atendimento':'ano_atendimento', \n 'Tecnologia de atendimento':'tecnologia', 'Concessionária':'concessionaria', 'Capacidade do Backhaul':'capacidade_backhaul', \n 'Capacidade Ocupada':'capacidadde_ocupada', 'Capacidade Disponível':'capacidade_disponivel'}, inplace=True)",
"_____no_output_____"
],
[
"#exclusão de astericos das bases pra melhor compatibilização dos dados\nbackhaul['ano_atendimento'] = backhaul['ano_atendimento'].apply(lambda x: str(x).replace('***',''))\nbackhaul['ano_atendimento'] = backhaul['ano_atendimento'].apply(lambda x: str(x).replace('**',''))\nbackhaul['tecnologia'] = backhaul['tecnologia'].apply(lambda x: str(x).replace('**','Não Informado'))\nbackhaul['tecnologia'] = backhaul['tecnologia'].apply(lambda x: str(x).replace('***',''))\nbackhaul['capacidade_backhaul'] = backhaul['capacidade_backhaul'].apply(lambda x: str(x).replace('**',''))\nbackhaul['capacidadde_ocupada'] = backhaul['capacidadde_ocupada'].apply(lambda x: str(x).replace('**',''))\nbackhaul['capacidade_disponivel'] = backhaul['capacidade_disponivel'].apply(lambda x: str(x).replace('**',''))",
"_____no_output_____"
],
[
"#mudança no formato da variavel\nbackhaul['id_municipio'] = backhaul['id_municipio'].astype('int64')",
"_____no_output_____"
],
[
"#exportando o arquivo\nbackhaul.to_csv('/content/drive/MyDrive/br_anatel/banda_larga_fixa/output/banda_larga_backhaul.csv', index=False)",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb94b03dad5c7619b7a886e600dd6389b400cf85
| 92,220 |
ipynb
|
Jupyter Notebook
|
2.4 Deep Taylor Decomposition (2).ipynb
|
1202kbs/Understanding-NN
|
110b2508071290ba1a82bcb26cab2e00104dee8e
|
[
"MIT"
] | 324 |
2018-01-18T02:11:16.000Z
|
2022-03-27T16:34:37.000Z
|
2.4 Deep Taylor Decomposition (2).ipynb
|
Jae-35/Understanding-NN
|
f9ddafc353f183cd0d055c64ec8a02a4178c729d
|
[
"MIT"
] | 5 |
2018-03-31T16:09:19.000Z
|
2018-11-14T05:40:55.000Z
|
2.4 Deep Taylor Decomposition (2).ipynb
|
Jae-35/Understanding-NN
|
f9ddafc353f183cd0d055c64ec8a02a4178c729d
|
[
"MIT"
] | 90 |
2018-01-08T03:59:09.000Z
|
2022-03-01T04:32:03.000Z
| 256.166667 | 82,070 | 0.910063 |
[
[
[
"# 2.4 Deep Taylor Decomposition Part 2.",
"_____no_output_____"
],
[
"## Tensorflow Walkthrough",
"_____no_output_____"
],
[
"### 1. Import Dependencies\n\nI made a custom `Taylor` class for Deep Taylor Decomposition. If you are interested in the details, check out `models_3_2.py` in the models directory.",
"_____no_output_____"
]
],
[
[
"import os\nimport re\n\nfrom tensorflow.examples.tutorials.mnist import input_data\nfrom tensorflow.python.ops import nn_ops, gen_nn_ops\nimport matplotlib.pyplot as plt\nimport tensorflow as tf\nimport numpy as np\n\nfrom models.models_2_4 import MNIST_CNN, Taylor\n\n%matplotlib inline\n\nmnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)\nimages = mnist.train.images\nlabels = mnist.train.labels\n\nlogdir = './tf_logs/2_4_DTD/'\nckptdir = logdir + 'model'\n\nif not os.path.exists(logdir):\n os.mkdir(logdir)",
"Extracting MNIST_data/train-images-idx3-ubyte.gz\nExtracting MNIST_data/train-labels-idx1-ubyte.gz\nExtracting MNIST_data/t10k-images-idx3-ubyte.gz\nExtracting MNIST_data/t10k-labels-idx1-ubyte.gz\n"
]
],
[
[
"### 2. Building Graph",
"_____no_output_____"
]
],
[
[
"with tf.name_scope('Classifier'):\n\n # Initialize neural network\n DNN = MNIST_CNN('CNN')\n\n # Setup training process\n X = tf.placeholder(tf.float32, [None, 784], name='X')\n Y = tf.placeholder(tf.float32, [None, 10], name='Y')\n\n activations, logits = DNN(X)\n \n tf.add_to_collection('DTD', X)\n \n for activation in activations:\n tf.add_to_collection('DTD', activation)\n\n cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=Y))\n\n optimizer = tf.train.AdamOptimizer().minimize(cost, var_list=DNN.vars)\n\n correct_prediction = tf.equal(tf.argmax(logits, 1), tf.argmax(Y, 1))\n accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n\ncost_summary = tf.summary.scalar('Cost', cost)\naccuray_summary = tf.summary.scalar('Accuracy', accuracy)\nsummary = tf.summary.merge_all()",
"_____no_output_____"
]
],
[
[
"### 3. Training Network\n\nThis is the step where the DNN is trained to classify the 10 digits of the MNIST images. Summaries are written into the logdir and you can visualize the statistics using tensorboard by typing this command: `tensorboard --lodir=./tf_logs`",
"_____no_output_____"
]
],
[
[
"sess = tf.InteractiveSession()\nsess.run(tf.global_variables_initializer())\n\nsaver = tf.train.Saver()\nfile_writer = tf.summary.FileWriter(logdir, tf.get_default_graph())\n\n# Hyper parameters\ntraining_epochs = 15\nbatch_size = 100\n\nfor epoch in range(training_epochs):\n total_batch = int(mnist.train.num_examples / batch_size)\n avg_cost = 0\n avg_acc = 0\n \n for i in range(total_batch):\n batch_xs, batch_ys = mnist.train.next_batch(batch_size)\n _, c, a, summary_str = sess.run([optimizer, cost, accuracy, summary], feed_dict={X: batch_xs, Y: batch_ys})\n avg_cost += c / total_batch\n avg_acc += a / total_batch\n \n file_writer.add_summary(summary_str, epoch * total_batch + i)\n\n print('Epoch:', '%04d' % (epoch + 1), 'cost =', '{:.9f}'.format(avg_cost), 'accuracy =', '{:.9f}'.format(avg_acc))\n \n saver.save(sess, ckptdir)\n\nprint('Accuracy:', sess.run(accuracy, feed_dict={X: mnist.test.images, Y: mnist.test.labels}))\n\nsess.close()",
"Epoch: 0001 cost = 0.155111813 accuracy = 0.951527280\nEpoch: 0002 cost = 0.044594308 accuracy = 0.985636373\nEpoch: 0003 cost = 0.029799474 accuracy = 0.990345463\nEpoch: 0004 cost = 0.021161627 accuracy = 0.992818188\nEpoch: 0005 cost = 0.017527232 accuracy = 0.994145460\nEpoch: 0006 cost = 0.014087625 accuracy = 0.995490913\nEpoch: 0007 cost = 0.013123321 accuracy = 0.995472732\nEpoch: 0008 cost = 0.009770884 accuracy = 0.996800003\nEpoch: 0009 cost = 0.009048406 accuracy = 0.996981821\nEpoch: 0010 cost = 0.009553186 accuracy = 0.996981821\nEpoch: 0011 cost = 0.007268998 accuracy = 0.997509093\nEpoch: 0012 cost = 0.007693169 accuracy = 0.997618184\nEpoch: 0013 cost = 0.005034720 accuracy = 0.998527274\nEpoch: 0014 cost = 0.006306725 accuracy = 0.997745457\nEpoch: 0015 cost = 0.004300387 accuracy = 0.998581820\nAccuracy: 0.9917\n"
]
],
[
[
"### 4. Restoring Subgraph\n\nHere we first rebuild the DNN graph from metagraph, restore DNN parameters from the checkpoint and then gather the necessary weight and biases for Deep Taylor Decomposition using the `tf.get_collection()` function.",
"_____no_output_____"
]
],
[
[
"tf.reset_default_graph()\n\nsess = tf.InteractiveSession()\n\nnew_saver = tf.train.import_meta_graph(ckptdir + '.meta')\nnew_saver.restore(sess, tf.train.latest_checkpoint(logdir))\n\nweights = tf.get_collection(tf.GraphKeys.TRAINABLE_VARIABLES, scope='CNN')\nactivations = tf.get_collection('DTD')\nX = activations[0]",
"INFO:tensorflow:Restoring parameters from ./tf_logs/3_2_DTD/model\n"
]
],
[
[
"### 5. Attaching Subgraph for Calculating Relevance Scores",
"_____no_output_____"
]
],
[
[
"conv_ksize = [1, 3, 3, 1]\npool_ksize = [1, 2, 2, 1]\nconv_strides = [1, 1, 1, 1]\npool_strides = [1, 2, 2, 1]\n\nweights.reverse()\nactivations.reverse()\n\ntaylor = Taylor(activations, weights, conv_ksize, pool_ksize, conv_strides, pool_strides, 'Taylor')\n\nRs = []\nfor i in range(10):\n Rs.append(taylor(i))",
"_____no_output_____"
]
],
[
[
"### 6. Calculating Relevance Scores $R(x_i)$",
"_____no_output_____"
]
],
[
[
"sample_imgs = []\nfor i in range(10):\n sample_imgs.append(images[np.argmax(labels, axis=1) == i][3])\n\nimgs = []\nfor i in range(10):\n imgs.append(sess.run(Rs[i], feed_dict={X: sample_imgs[i][None,:]}))",
"_____no_output_____"
]
],
[
[
"### 7. Displaying Images\n\nThe relevance scores are visualized as heat maps. You can see which features/data points influenced the DNN most its decision making.",
"_____no_output_____"
]
],
[
[
"plt.figure(figsize=(15,15))\nfor i in range(5):\n plt.subplot(5, 2, 2 * i + 1)\n plt.imshow(np.reshape(imgs[2 * i], [28, 28]), cmap='hot_r')\n plt.title('Digit: {}'.format(2 * i))\n plt.colorbar()\n \n plt.subplot(5, 2, 2 * i + 2)\n plt.imshow(np.reshape(imgs[2 * i + 1], [28, 28]), cmap='hot_r')\n plt.title('Digit: {}'.format(2 * i + 1))\n plt.colorbar()\n\nplt.tight_layout()",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
]
] |
cb94c19bcf821864b2bd33a966408e8a26e8fd7a
| 276,240 |
ipynb
|
Jupyter Notebook
|
_posts/DS/Deep-Learning/Deep-Learning-Coursera/2. Improving Deep Neural Networks- Hyperparameter tuning- Regularization and Optimization/Initialization.ipynb
|
wansook0316/wansook.github.io
|
9a5c2cb84d7a9d64f7dfc7e9d4110d4e5af67efb
|
[
"MIT"
] | 136 |
2018-04-02T11:08:06.000Z
|
2022-02-27T21:31:17.000Z
|
_posts/DS/Deep-Learning/Deep-Learning-Coursera/2. Improving Deep Neural Networks- Hyperparameter tuning- Regularization and Optimization/Initialization.ipynb
|
wansook0316/wansook.github.io
|
9a5c2cb84d7a9d64f7dfc7e9d4110d4e5af67efb
|
[
"MIT"
] | 1 |
2019-01-20T06:47:19.000Z
|
2019-01-20T06:47:19.000Z
|
_posts/DS/Deep-Learning/Deep-Learning-Coursera/2. Improving Deep Neural Networks- Hyperparameter tuning- Regularization and Optimization/Initialization.ipynb
|
wansook0316/wansook.github.io
|
9a5c2cb84d7a9d64f7dfc7e9d4110d4e5af67efb
|
[
"MIT"
] | 201 |
2018-04-19T22:06:50.000Z
|
2022-03-13T16:21:58.000Z
| 254.364641 | 57,656 | 0.899895 |
[
[
[
"# Initialization\n\nWelcome to the first assignment of \"Improving Deep Neural Networks\". \n\nTraining your neural network requires specifying an initial value of the weights. A well chosen initialization method will help learning. \n\nIf you completed the previous course of this specialization, you probably followed our instructions for weight initialization, and it has worked out so far. But how do you choose the initialization for a new neural network? In this notebook, you will see how different initializations lead to different results. \n\nA well chosen initialization can:\n- Speed up the convergence of gradient descent\n- Increase the odds of gradient descent converging to a lower training (and generalization) error \n\nTo get started, run the following cell to load the packages and the planar dataset you will try to classify.",
"_____no_output_____"
]
],
[
[
"import numpy as np\nimport matplotlib.pyplot as plt\nimport sklearn\nimport sklearn.datasets\nfrom init_utils import sigmoid, relu, compute_loss, forward_propagation, backward_propagation\nfrom init_utils import update_parameters, predict, load_dataset, plot_decision_boundary, predict_dec\n\n%matplotlib inline\nplt.rcParams['figure.figsize'] = (7.0, 4.0) # set default size of plots\nplt.rcParams['image.interpolation'] = 'nearest'\nplt.rcParams['image.cmap'] = 'gray'\n\n# load image dataset: blue/red dots in circles\ntrain_X, train_Y, test_X, test_Y = load_dataset()",
"_____no_output_____"
]
],
[
[
"You would like a classifier to separate the blue dots from the red dots.",
"_____no_output_____"
],
[
"## 1 - Neural Network model ",
"_____no_output_____"
],
[
"You will use a 3-layer neural network (already implemented for you). Here are the initialization methods you will experiment with: \n- *Zeros initialization* -- setting `initialization = \"zeros\"` in the input argument.\n- *Random initialization* -- setting `initialization = \"random\"` in the input argument. This initializes the weights to large random values. \n- *He initialization* -- setting `initialization = \"he\"` in the input argument. This initializes the weights to random values scaled according to a paper by He et al., 2015. \n\n**Instructions**: Please quickly read over the code below, and run it. In the next part you will implement the three initialization methods that this `model()` calls.",
"_____no_output_____"
]
],
[
[
"def model(X, Y, learning_rate = 0.01, num_iterations = 15000, print_cost = True, initialization = \"he\"):\n \"\"\"\n Implements a three-layer neural network: LINEAR->RELU->LINEAR->RELU->LINEAR->SIGMOID.\n \n Arguments:\n X -- input data, of shape (2, number of examples)\n Y -- true \"label\" vector (containing 0 for red dots; 1 for blue dots), of shape (1, number of examples)\n learning_rate -- learning rate for gradient descent \n num_iterations -- number of iterations to run gradient descent\n print_cost -- if True, print the cost every 1000 iterations\n initialization -- flag to choose which initialization to use (\"zeros\",\"random\" or \"he\")\n \n Returns:\n parameters -- parameters learnt by the model\n \"\"\"\n \n grads = {}\n costs = [] # to keep track of the loss\n m = X.shape[1] # number of examples\n layers_dims = [X.shape[0], 10, 5, 1]\n \n # Initialize parameters dictionary.\n if initialization == \"zeros\":\n parameters = initialize_parameters_zeros(layers_dims)\n elif initialization == \"random\":\n parameters = initialize_parameters_random(layers_dims)\n elif initialization == \"he\":\n parameters = initialize_parameters_he(layers_dims)\n\n # Loop (gradient descent)\n\n for i in range(0, num_iterations):\n\n # Forward propagation: LINEAR -> RELU -> LINEAR -> RELU -> LINEAR -> SIGMOID.\n a3, cache = forward_propagation(X, parameters)\n \n # Loss\n cost = compute_loss(a3, Y)\n\n # Backward propagation.\n grads = backward_propagation(X, Y, cache)\n \n # Update parameters.\n parameters = update_parameters(parameters, grads, learning_rate)\n \n # Print the loss every 1000 iterations\n if print_cost and i % 1000 == 0:\n print(\"Cost after iteration {}: {}\".format(i, cost))\n costs.append(cost)\n \n # plot the loss\n plt.plot(costs)\n plt.ylabel('cost')\n plt.xlabel('iterations (per hundreds)')\n plt.title(\"Learning rate =\" + str(learning_rate))\n plt.show()\n \n return parameters",
"_____no_output_____"
]
],
[
[
"## 2 - Zero initialization\n\nThere are two types of parameters to initialize in a neural network:\n- the weight matrices $(W^{[1]}, W^{[2]}, W^{[3]}, ..., W^{[L-1]}, W^{[L]})$\n- the bias vectors $(b^{[1]}, b^{[2]}, b^{[3]}, ..., b^{[L-1]}, b^{[L]})$\n\n**Exercise**: Implement the following function to initialize all parameters to zeros. You'll see later that this does not work well since it fails to \"break symmetry\", but lets try it anyway and see what happens. Use np.zeros((..,..)) with the correct shapes.",
"_____no_output_____"
]
],
[
[
"# GRADED FUNCTION: initialize_parameters_zeros \n\ndef initialize_parameters_zeros(layers_dims):\n \"\"\"\n Arguments:\n layer_dims -- python array (list) containing the size of each layer.\n \n Returns:\n parameters -- python dictionary containing your parameters \"W1\", \"b1\", ..., \"WL\", \"bL\":\n W1 -- weight matrix of shape (layers_dims[1], layers_dims[0])\n b1 -- bias vector of shape (layers_dims[1], 1)\n ...\n WL -- weight matrix of shape (layers_dims[L], layers_dims[L-1])\n bL -- bias vector of shape (layers_dims[L], 1)\n \"\"\"\n \n parameters = {}\n L = len(layers_dims) # number of layers in the network\n \n for l in range(1, L):\n ### START CODE HERE ### (≈ 2 lines of code)\n parameters['W' + str(l)] = np.zeros((layers_dims[l], layers_dims[l-1]))\n parameters['b' + str(l)] = np.zeros((layers_dims[l],1))\n ### END CODE HERE ###\n return parameters",
"_____no_output_____"
],
[
"parameters = initialize_parameters_zeros([3,2,1])\nprint(\"W1 = \" + str(parameters[\"W1\"]))\nprint(\"b1 = \" + str(parameters[\"b1\"]))\nprint(\"W2 = \" + str(parameters[\"W2\"]))\nprint(\"b2 = \" + str(parameters[\"b2\"]))",
"W1 = [[ 0. 0. 0.]\n [ 0. 0. 0.]]\nb1 = [[ 0.]\n [ 0.]]\nW2 = [[ 0. 0.]]\nb2 = [[ 0.]]\n"
]
],
[
[
"**Expected Output**:\n\n<table> \n <tr>\n <td>\n **W1**\n </td>\n <td>\n [[ 0. 0. 0.]\n [ 0. 0. 0.]]\n </td>\n </tr>\n <tr>\n <td>\n **b1**\n </td>\n <td>\n [[ 0.]\n [ 0.]]\n </td>\n </tr>\n <tr>\n <td>\n **W2**\n </td>\n <td>\n [[ 0. 0.]]\n </td>\n </tr>\n <tr>\n <td>\n **b2**\n </td>\n <td>\n [[ 0.]]\n </td>\n </tr>\n\n</table> ",
"_____no_output_____"
],
[
"Run the following code to train your model on 15,000 iterations using zeros initialization.",
"_____no_output_____"
]
],
[
[
"parameters = model(train_X, train_Y, initialization = \"zeros\")\nprint (\"On the train set:\")\npredictions_train = predict(train_X, train_Y, parameters)\nprint (\"On the test set:\")\npredictions_test = predict(test_X, test_Y, parameters)",
"Cost after iteration 0: 0.6931471805599453\nCost after iteration 1000: 0.6931471805599453\nCost after iteration 2000: 0.6931471805599453\nCost after iteration 3000: 0.6931471805599453\nCost after iteration 4000: 0.6931471805599453\nCost after iteration 5000: 0.6931471805599453\nCost after iteration 6000: 0.6931471805599453\nCost after iteration 7000: 0.6931471805599453\nCost after iteration 8000: 0.6931471805599453\nCost after iteration 9000: 0.6931471805599453\nCost after iteration 10000: 0.6931471805599455\nCost after iteration 11000: 0.6931471805599453\nCost after iteration 12000: 0.6931471805599453\nCost after iteration 13000: 0.6931471805599453\nCost after iteration 14000: 0.6931471805599453\n"
]
],
[
[
"The performance is really bad, and the cost does not really decrease, and the algorithm performs no better than random guessing. Why? Lets look at the details of the predictions and the decision boundary:",
"_____no_output_____"
]
],
[
[
"print (\"predictions_train = \" + str(predictions_train))\nprint (\"predictions_test = \" + str(predictions_test))",
"predictions_train = [[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0]]\npredictions_test = [[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]]\n"
],
[
"plt.title(\"Model with Zeros initialization\")\naxes = plt.gca()\naxes.set_xlim([-1.5,1.5])\naxes.set_ylim([-1.5,1.5])\nplot_decision_boundary(lambda x: predict_dec(parameters, x.T), train_X, train_Y)",
"_____no_output_____"
]
],
[
[
"The model is predicting 0 for every example. \n\nIn general, initializing all the weights to zero results in the network failing to break symmetry. This means that every neuron in each layer will learn the same thing, and you might as well be training a neural network with $n^{[l]}=1$ for every layer, and the network is no more powerful than a linear classifier such as logistic regression. ",
"_____no_output_____"
],
[
"<font color='blue'>\n**What you should remember**:\n- The weights $W^{[l]}$ should be initialized randomly to break symmetry. \n- It is however okay to initialize the biases $b^{[l]}$ to zeros. Symmetry is still broken so long as $W^{[l]}$ is initialized randomly. \n",
"_____no_output_____"
],
[
"## 3 - Random initialization\n\nTo break symmetry, lets intialize the weights randomly. Following random initialization, each neuron can then proceed to learn a different function of its inputs. In this exercise, you will see what happens if the weights are intialized randomly, but to very large values. \n\n**Exercise**: Implement the following function to initialize your weights to large random values (scaled by \\*10) and your biases to zeros. Use `np.random.randn(..,..) * 10` for weights and `np.zeros((.., ..))` for biases. We are using a fixed `np.random.seed(..)` to make sure your \"random\" weights match ours, so don't worry if running several times your code gives you always the same initial values for the parameters. ",
"_____no_output_____"
]
],
[
[
"# GRADED FUNCTION: initialize_parameters_random\n\ndef initialize_parameters_random(layers_dims):\n \"\"\"\n Arguments:\n layer_dims -- python array (list) containing the size of each layer.\n \n Returns:\n parameters -- python dictionary containing your parameters \"W1\", \"b1\", ..., \"WL\", \"bL\":\n W1 -- weight matrix of shape (layers_dims[1], layers_dims[0])\n b1 -- bias vector of shape (layers_dims[1], 1)\n ...\n WL -- weight matrix of shape (layers_dims[L], layers_dims[L-1])\n bL -- bias vector of shape (layers_dims[L], 1)\n \"\"\"\n \n np.random.seed(3) # This seed makes sure your \"random\" numbers will be the as ours\n parameters = {}\n L = len(layers_dims) # integer representing the number of layers\n \n for l in range(1, L):\n ### START CODE HERE ### (≈ 2 lines of code)\n parameters['W' + str(l)] = np.random.randn(layers_dims[l], layers_dims[l-1]) * 10\n parameters['b' + str(l)] = np.zeros((layers_dims[l], 1))\n ### END CODE HERE ###\n\n return parameters",
"_____no_output_____"
],
[
"parameters = initialize_parameters_random([3, 2, 1])\nprint(\"W1 = \" + str(parameters[\"W1\"]))\nprint(\"b1 = \" + str(parameters[\"b1\"]))\nprint(\"W2 = \" + str(parameters[\"W2\"]))\nprint(\"b2 = \" + str(parameters[\"b2\"]))",
"W1 = [[ 17.88628473 4.36509851 0.96497468]\n [-18.63492703 -2.77388203 -3.54758979]]\nb1 = [[ 0.]\n [ 0.]]\nW2 = [[-0.82741481 -6.27000677]]\nb2 = [[ 0.]]\n"
]
],
[
[
"**Expected Output**:\n\n<table> \n <tr>\n <td>\n **W1**\n </td>\n <td>\n [[ 17.88628473 4.36509851 0.96497468]\n [-18.63492703 -2.77388203 -3.54758979]]\n </td>\n </tr>\n <tr>\n <td>\n **b1**\n </td>\n <td>\n [[ 0.]\n [ 0.]]\n </td>\n </tr>\n <tr>\n <td>\n **W2**\n </td>\n <td>\n [[-0.82741481 -6.27000677]]\n </td>\n </tr>\n <tr>\n <td>\n **b2**\n </td>\n <td>\n [[ 0.]]\n </td>\n </tr>\n\n</table> ",
"_____no_output_____"
],
[
"Run the following code to train your model on 15,000 iterations using random initialization.",
"_____no_output_____"
]
],
[
[
"parameters = model(train_X, train_Y, initialization = \"random\")\nprint (\"On the train set:\")\npredictions_train = predict(train_X, train_Y, parameters)\nprint (\"On the test set:\")\npredictions_test = predict(test_X, test_Y, parameters)",
"/home/jovyan/work/week5/Initialization/init_utils.py:145: RuntimeWarning: divide by zero encountered in log\n logprobs = np.multiply(-np.log(a3),Y) + np.multiply(-np.log(1 - a3), 1 - Y)\n/home/jovyan/work/week5/Initialization/init_utils.py:145: RuntimeWarning: invalid value encountered in multiply\n logprobs = np.multiply(-np.log(a3),Y) + np.multiply(-np.log(1 - a3), 1 - Y)\n"
]
],
[
[
"If you see \"inf\" as the cost after the iteration 0, this is because of numerical roundoff; a more numerically sophisticated implementation would fix this. But this isn't worth worrying about for our purposes. \n\nAnyway, it looks like you have broken symmetry, and this gives better results. than before. The model is no longer outputting all 0s. ",
"_____no_output_____"
]
],
[
[
"print (predictions_train)\nprint (predictions_test)",
"[[1 0 1 1 0 0 1 1 1 1 1 0 1 0 0 1 0 1 1 0 0 0 1 0 1 1 1 1 1 1 0 1 1 0 0 1 1\n 1 1 1 1 1 1 0 1 1 1 1 0 1 0 1 1 1 1 0 0 1 1 1 1 0 1 1 0 1 0 1 1 1 1 0 0 0\n 0 0 1 0 1 0 1 1 1 0 0 1 1 1 1 1 1 0 0 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 0 1\n 1 0 0 1 0 0 1 1 0 1 1 1 0 1 0 0 1 0 1 1 1 1 1 1 1 0 1 1 0 0 1 1 0 0 0 1 0\n 1 0 1 0 1 1 1 0 0 1 1 1 1 0 1 1 0 1 0 1 1 0 1 0 1 1 1 1 0 1 1 1 1 0 1 0 1\n 0 1 1 1 1 0 1 1 0 1 1 0 1 1 0 1 0 1 1 1 0 1 1 1 0 1 0 1 0 0 1 0 1 1 0 1 1\n 0 1 1 0 1 1 1 0 1 1 1 1 0 1 0 0 1 1 0 1 1 1 0 0 0 1 1 0 1 1 1 1 0 1 1 0 1\n 1 1 0 0 1 0 0 0 1 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 1 1 1 1 1 1 1 0 0 0 1\n 1 1 1 0]]\n[[1 1 1 1 0 1 0 1 1 0 1 1 1 0 0 0 0 1 0 1 0 0 1 0 1 0 1 1 1 1 1 0 0 0 0 1 0\n 1 1 0 0 1 1 1 1 1 0 1 1 1 0 1 0 1 1 0 1 0 1 0 1 1 1 1 1 1 1 1 1 0 1 0 1 1\n 1 1 1 0 1 0 0 1 0 0 0 1 1 0 1 1 0 0 0 1 1 0 1 1 0 0]]\n"
],
[
"plt.title(\"Model with large random initialization\")\naxes = plt.gca()\naxes.set_xlim([-1.5,1.5])\naxes.set_ylim([-1.5,1.5])\nplot_decision_boundary(lambda x: predict_dec(parameters, x.T), train_X, train_Y)",
"_____no_output_____"
]
],
[
[
"**Observations**:\n- The cost starts very high. This is because with large random-valued weights, the last activation (sigmoid) outputs results that are very close to 0 or 1 for some examples, and when it gets that example wrong it incurs a very high loss for that example. Indeed, when $\\log(a^{[3]}) = \\log(0)$, the loss goes to infinity.\n- Poor initialization can lead to vanishing/exploding gradients, which also slows down the optimization algorithm. \n- If you train this network longer you will see better results, but initializing with overly large random numbers slows down the optimization.\n\n<font color='blue'>\n**In summary**:\n- Initializing weights to very large random values does not work well. \n- Hopefully intializing with small random values does better. The important question is: how small should be these random values be? Lets find out in the next part! ",
"_____no_output_____"
],
[
"## 4 - He initialization\n\nFinally, try \"He Initialization\"; this is named for the first author of He et al., 2015. (If you have heard of \"Xavier initialization\", this is similar except Xavier initialization uses a scaling factor for the weights $W^{[l]}$ of `sqrt(1./layers_dims[l-1])` where He initialization would use `sqrt(2./layers_dims[l-1])`.)\n\n**Exercise**: Implement the following function to initialize your parameters with He initialization.\n\n**Hint**: This function is similar to the previous `initialize_parameters_random(...)`. The only difference is that instead of multiplying `np.random.randn(..,..)` by 10, you will multiply it by $\\sqrt{\\frac{2}{\\text{dimension of the previous layer}}}$, which is what He initialization recommends for layers with a ReLU activation. ",
"_____no_output_____"
]
],
[
[
"# GRADED FUNCTION: initialize_parameters_he\n\ndef initialize_parameters_he(layers_dims):\n \"\"\"\n Arguments:\n layer_dims -- python array (list) containing the size of each layer.\n \n Returns:\n parameters -- python dictionary containing your parameters \"W1\", \"b1\", ..., \"WL\", \"bL\":\n W1 -- weight matrix of shape (layers_dims[1], layers_dims[0])\n b1 -- bias vector of shape (layers_dims[1], 1)\n ...\n WL -- weight matrix of shape (layers_dims[L], layers_dims[L-1])\n bL -- bias vector of shape (layers_dims[L], 1)\n \"\"\"\n \n np.random.seed(3)\n parameters = {}\n L = len(layers_dims) - 1 # integer representing the number of layers\n \n for l in range(1, L + 1):\n ### START CODE HERE ### (≈ 2 lines of code)\n parameters['W' + str(l)] = np.random.randn(layers_dims[l], layers_dims[l-1]) * np.sqrt(2./layers_dims[l-1])\n parameters['b' + str(l)] = np.zeros((layers_dims[l], 1))\n ### END CODE HERE ###\n \n return parameters",
"_____no_output_____"
],
[
"parameters = initialize_parameters_he([2, 4, 1])\nprint(\"W1 = \" + str(parameters[\"W1\"]))\nprint(\"b1 = \" + str(parameters[\"b1\"]))\nprint(\"W2 = \" + str(parameters[\"W2\"]))\nprint(\"b2 = \" + str(parameters[\"b2\"]))",
"W1 = [[ 1.78862847 0.43650985]\n [ 0.09649747 -1.8634927 ]\n [-0.2773882 -0.35475898]\n [-0.08274148 -0.62700068]]\nb1 = [[ 0.]\n [ 0.]\n [ 0.]\n [ 0.]]\nW2 = [[-0.03098412 -0.33744411 -0.92904268 0.62552248]]\nb2 = [[ 0.]]\n"
]
],
[
[
"**Expected Output**:\n\n<table> \n <tr>\n <td>\n **W1**\n </td>\n <td>\n [[ 1.78862847 0.43650985]\n [ 0.09649747 -1.8634927 ]\n [-0.2773882 -0.35475898]\n [-0.08274148 -0.62700068]]\n </td>\n </tr>\n <tr>\n <td>\n **b1**\n </td>\n <td>\n [[ 0.]\n [ 0.]\n [ 0.]\n [ 0.]]\n </td>\n </tr>\n <tr>\n <td>\n **W2**\n </td>\n <td>\n [[-0.03098412 -0.33744411 -0.92904268 0.62552248]]\n </td>\n </tr>\n <tr>\n <td>\n **b2**\n </td>\n <td>\n [[ 0.]]\n </td>\n </tr>\n\n</table> ",
"_____no_output_____"
],
[
"Run the following code to train your model on 15,000 iterations using He initialization.",
"_____no_output_____"
]
],
[
[
"parameters = model(train_X, train_Y, initialization = \"he\")\nprint (\"On the train set:\")\npredictions_train = predict(train_X, train_Y, parameters)\nprint (\"On the test set:\")\npredictions_test = predict(test_X, test_Y, parameters)",
"Cost after iteration 0: 0.8830537463419761\nCost after iteration 1000: 0.6879825919728063\nCost after iteration 2000: 0.6751286264523371\nCost after iteration 3000: 0.6526117768893807\nCost after iteration 4000: 0.6082958970572938\nCost after iteration 5000: 0.5304944491717495\nCost after iteration 6000: 0.4138645817071794\nCost after iteration 7000: 0.3117803464844441\nCost after iteration 8000: 0.23696215330322562\nCost after iteration 9000: 0.18597287209206836\nCost after iteration 10000: 0.1501555628037182\nCost after iteration 11000: 0.12325079292273548\nCost after iteration 12000: 0.09917746546525937\nCost after iteration 13000: 0.0845705595402428\nCost after iteration 14000: 0.07357895962677366\n"
],
[
"plt.title(\"Model with He initialization\")\naxes = plt.gca()\naxes.set_xlim([-1.5,1.5])\naxes.set_ylim([-1.5,1.5])\nplot_decision_boundary(lambda x: predict_dec(parameters, x.T), train_X, train_Y)",
"_____no_output_____"
]
],
[
[
"**Observations**:\n- The model with He initialization separates the blue and the red dots very well in a small number of iterations.\n",
"_____no_output_____"
],
[
"## 5 - Conclusions",
"_____no_output_____"
],
[
"You have seen three different types of initializations. For the same number of iterations and same hyperparameters the comparison is:\n\n<table> \n <tr>\n <td>\n **Model**\n </td>\n <td>\n **Train accuracy**\n </td>\n <td>\n **Problem/Comment**\n </td>\n\n </tr>\n <td>\n 3-layer NN with zeros initialization\n </td>\n <td>\n 50%\n </td>\n <td>\n fails to break symmetry\n </td>\n <tr>\n <td>\n 3-layer NN with large random initialization\n </td>\n <td>\n 83%\n </td>\n <td>\n too large weights \n </td>\n </tr>\n <tr>\n <td>\n 3-layer NN with He initialization\n </td>\n <td>\n 99%\n </td>\n <td>\n recommended method\n </td>\n </tr>\n</table> ",
"_____no_output_____"
],
[
"<font color='blue'>\n**What you should remember from this notebook**:\n- Different initializations lead to different results\n- Random initialization is used to break symmetry and make sure different hidden units can learn different things\n- Don't intialize to values that are too large\n- He initialization works well for networks with ReLU activations. ",
"_____no_output_____"
]
],
[
[
"!tar cvfz notebook.tar.gz *",
"Initialization-Copy1.ipynb\nInitialization.ipynb\n__pycache__/\n__pycache__/init_utils.cpython-36.pyc\ninit_utils.py\n"
],
[
"!tar cvfz notebook.tar.gz *",
"Initialization-Copy1.ipynb\nInitialization.ipynb\n__pycache__/\n__pycache__/init_utils.cpython-36.pyc\ninit_utils.py\nnotebook.tar.gz\ntar: notebook.tar.gz: file changed as we read it\n"
],
[
"!tar cvfz notebook.tar.gz *",
"Initialization-Copy1.ipynb\nInitialization.ipynb\n__pycache__/\n__pycache__/init_utils.cpython-36.pyc\ninit_utils.py\nnotebook.tar.gz\ntar: notebook.tar.gz: file changed as we read it\n"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code"
]
] |
cb94cb5fbcc88412e05039fc1d29688c669d64c3
| 10,213 |
ipynb
|
Jupyter Notebook
|
notepads/day-3.ipynb
|
ellewoot/advent-of-code-2020
|
54822324e0f802720fb4f3ee20208f7c93fcfac5
|
[
"MIT"
] | null | null | null |
notepads/day-3.ipynb
|
ellewoot/advent-of-code-2020
|
54822324e0f802720fb4f3ee20208f7c93fcfac5
|
[
"MIT"
] | null | null | null |
notepads/day-3.ipynb
|
ellewoot/advent-of-code-2020
|
54822324e0f802720fb4f3ee20208f7c93fcfac5
|
[
"MIT"
] | null | null | null | 25.987277 | 898 | 0.49672 |
[
[
[
"# Day 3 - Advent of Code 2020 (https://adventofcode.com/2020/day/3)\n## Data Preparation\n\nAs with the previous days, I will be starting with getting my input into a good format before even taken a reat pass at the problem...",
"_____no_output_____"
]
],
[
[
"real_run = False\nfile_name = \"day3-input.txt\" if real_run else \"day3-test.txt\"\n\n# create a list from the file, removing any '\\n' characters\ndata = [line.rstrip('\\n') for line in open(file_name)]\n\n# print data to check it's what we want it to be\nprint(data)",
"['..##.......', '#...#...#..', '.#....#..#.', '..#.#...#.#', '.#...##..#.', '..#.##.....', '.#.#.#....#', '.#........#', '#.##...#...', '#...##....#', '.#..#...#.#']\n"
]
],
[
[
"## Part One\n\nEach line of our data is a layer in a toboggan slope... It repeats inifinitely out to either side and '#' are trees and '.' are open spaces.\n\nWe have a route given to us (down 1, right 3) and we want to return the amount of trees we are given. If we start from (0,0), we will go down hitting (1,3) (2,6) (3,9) etc.",
"_____no_output_____"
]
],
[
[
"length = len(data)\n\nfor row in range(length):\n col = row * 3\n \n item = data[row][col]\n \n print(item)",
".\n.\n#\n.\n"
]
],
[
[
"Since the length of each row is less than 3x the length we will get a string index out of range exception. Since we are starting at a 0 index and we know the length of each row, we can use modulus.",
"_____no_output_____"
]
],
[
[
"length = len(data)\nrow_length = len(data[0])\n\nfor row in range(length):\n col = row * 3 % row_length\n \n item = data[row][col]\n \n print(item)",
".\n.\n#\n.\n#\n#\n.\n#\n#\n#\n#\n"
]
],
[
[
"Now we make sure to keep track of the count as it increases.",
"_____no_output_____"
]
],
[
[
"length = len(data)\nrow_length = len(data[0])\ntree_count = 0\n\nfor row in range(length):\n col = (row * 3) % row_length\n \n item = data[row][col]\n \n if item == '#':\n tree_count += 1\n\nprint(tree_count)",
"7\n"
]
],
[
[
"## Part Two\n\nNow, we do the same for a group of:\n* Right 1, down 1.\n* Right 3, down 1. (This is the slope we already checked.)\n* Right 5, down 1.\n* Right 7, down 1.\n* Right 1, down 2.\n\nSo we can generalise and pass in a parameter for what the column position. We can address 4/5 requirements by just making this one change!",
"_____no_output_____"
]
],
[
[
"def traverse_path(path_data, right):\n length = len(path_data)\n row_length = len(path_data[0])\n tree_count = 0\n\n for row in range(length):\n col = (row * right) % row_length\n\n item = data[row][col]\n\n if item == '#':\n tree_count += 1\n\n return tree_count",
"_____no_output_____"
],
[
"# Test with the one we already know...\nr3_d1 = traverse_path(data, 3)\nprint(r3_d1)",
"7\n"
]
],
[
[
"And to figure the \"down\" motion of right 1, down 2 we can make some adjustments to the row number and divide the length so we don't go down beyond the rows. ",
"_____no_output_____"
]
],
[
[
"import math",
"_____no_output_____"
],
[
"def traverse_path(path_data, right, down=1):\n # Use ceil as to make sure we get the final rows! (due to range taking us up to strictly less than the total)\n length = math.ceil(len(path_data) / down)\n row_length = len(path_data[0])\n tree_count = 0\n\n for row in range(length):\n col = (row * right) % row_length\n row_num = row * down\n \n item = data[row_num][col]\n\n if item == '#':\n tree_count += 1\n\n return tree_count",
"_____no_output_____"
],
[
"# Test with the one we already know...\nr3_d1 = traverse_path(data, 3)\nprint(r3_d1)\n\n# and do all the others:\nr1_d1 = traverse_path(data, 1)\nr5_d1 = traverse_path(data, 5)\nr7_d1 = traverse_path(data, 7)\nr1_d2 = traverse_path(data, 1, 2)\n\nproduct = r3_d1 * r1_d1 * r5_d1 * r7_d1 * r1_d2\n\nprint(product)",
"7\n336\n"
]
],
[
[
"#### There we have it...\n\nBut perhaps we can go one step further, reduce that repeated code and give the instructions as a dict...",
"_____no_output_____"
]
],
[
[
"instructions = [\n {'right':1, 'down':1}, \n {'right':3, 'down':1}, \n {'right':5, 'down':1}, \n {'right':7, 'down':1}, \n {'right':1, 'down':2}\n]",
"_____no_output_____"
],
[
"def prod_trav_paths(instrs, path_data):\n results = []\n \n for instr in instrs:\n res = traverse_path(path_data, instr['right'], instr['down'])\n results.append(res)\n \n return math.prod(results)",
"_____no_output_____"
],
[
"prod_trav_paths(instructions, data)",
"_____no_output_____"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
]
] |
cb94d16e3fdf40c40cc1ac5bbefdf5c8f850ecfe
| 58,207 |
ipynb
|
Jupyter Notebook
|
inactive/02-python-jupyter/02-pandas/Chapter 7 - Cleaning up messy data.ipynb
|
allixender/geopython-ut-2018
|
33ffec0fc787708515e3ad458e94843afbfc4c0c
|
[
"MIT"
] | 1 |
2019-12-01T05:05:02.000Z
|
2019-12-01T05:05:02.000Z
|
02-pandas/Chapter 7 - Cleaning up messy data.ipynb
|
allixender/python-jupyter-pandas-geo-tutorials
|
e6807cc90921582971a9a9c8cca4aea50e851194
|
[
"BSD-3-Clause"
] | null | null | null |
02-pandas/Chapter 7 - Cleaning up messy data.ipynb
|
allixender/python-jupyter-pandas-geo-tutorials
|
e6807cc90921582971a9a9c8cca4aea50e851194
|
[
"BSD-3-Clause"
] | 1 |
2019-09-19T12:47:50.000Z
|
2019-09-19T12:47:50.000Z
| 41.725448 | 921 | 0.426804 |
[
[
[
"# The usual preamble\n%matplotlib inline\n\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport numpy as np\n\n# Make the graphs a bit prettier, and bigger\nplt.style.use('ggplot')\nplt.rcParams['figure.figsize'] = (15, 5)\nplt.rcParams['font.family'] = 'sans-serif'\n\n# This is necessary to show lots of columns in pandas 0.12. \n# Not necessary in pandas 0.13.\npd.set_option('display.width', 5000) \npd.set_option('display.max_columns', 60)",
"_____no_output_____"
]
],
[
[
"One of the main problems with messy data is: how do you know if it's messy or not?\n\nWe're going to use the NYC 311 service request dataset again here, since it's big and a bit unwieldy.",
"_____no_output_____"
]
],
[
[
"requests = pd.read_csv('data/311-service-requests.csv', low_memory=False)",
"_____no_output_____"
]
],
[
[
"# 7.1 How do we know if it's messy? ",
"_____no_output_____"
],
[
"We're going to look at a few columns here. I know already that there are some problems with the zip code, so let's look at that first.\n \nTo get a sense for whether a column has problems, I usually use `.unique()` to look at all its values. If it's a numeric column, I'll instead plot a histogram to get a sense of the distribution.\n\nWhen we look at the unique values in \"Incident Zip\", it quickly becomes clear that this is a mess.\n\nSome of the problems:\n\n* Some have been parsed as strings, and some as floats\n* There are `nan`s \n* Some of the zip codes are `29616-0759` or `83`\n* There are some N/A values that pandas didn't recognize, like 'N/A' and 'NO CLUE'\n\nWhat we can do:\n\n* Normalize 'N/A' and 'NO CLUE' into regular nan values\n* Look at what's up with the 83, and decide what to do\n* Make everything strings",
"_____no_output_____"
]
],
[
[
"requests['Incident Zip'].unique()",
"_____no_output_____"
]
],
[
[
"# 7.2 Fixing the nan values and string/float confusion",
"_____no_output_____"
],
[
"We can pass a `na_values` option to `pd.read_csv` to clean this up a little bit. We can also specify that the type of Incident Zip is a string, not a float.",
"_____no_output_____"
]
],
[
[
"na_values = ['NO CLUE', 'N/A', '0']\nrequests = pd.read_csv('data/311-service-requests.csv', na_values=na_values, dtype={'Incident Zip': str})",
"_____no_output_____"
],
[
"requests['Incident Zip'].unique()",
"_____no_output_____"
]
],
[
[
"# 7.3 What's up with the dashes?",
"_____no_output_____"
]
],
[
[
"rows_with_dashes = requests['Incident Zip'].str.contains('-').fillna(False)\nlen(requests[rows_with_dashes])",
"_____no_output_____"
],
[
"requests[rows_with_dashes]",
"_____no_output_____"
]
],
[
[
"I thought these were missing data and originally deleted them like this:\n\n`requests['Incident Zip'][rows_with_dashes] = np.nan`\n\nBut then my friend Dave pointed out that 9-digit zip codes are normal. Let's look at all the zip codes with more than 5 digits, make sure they're okay, and then truncate them.",
"_____no_output_____"
]
],
[
[
"long_zip_codes = requests['Incident Zip'].str.len() > 5\nrequests['Incident Zip'][long_zip_codes].unique()",
"_____no_output_____"
]
],
[
[
"Those all look okay to truncate to me.",
"_____no_output_____"
]
],
[
[
"requests['Incident Zip'] = requests['Incident Zip'].str.slice(0, 5)",
"_____no_output_____"
]
],
[
[
"Done.",
"_____no_output_____"
],
[
"Earlier I thought 00083 was a broken zip code, but turns out Central Park's zip code 00083! Shows what I know. I'm still concerned about the 00000 zip codes, though: let's look at that. ",
"_____no_output_____"
]
],
[
[
"requests[requests['Incident Zip'] == '00000']",
"_____no_output_____"
]
],
[
[
"This looks bad to me. Let's set these to nan.",
"_____no_output_____"
]
],
[
[
"zero_zips = requests['Incident Zip'] == '00000'\nrequests.loc[zero_zips, 'Incident Zip'] = np.nan",
"_____no_output_____"
]
],
[
[
"Great. Let's see where we are now:",
"_____no_output_____"
]
],
[
[
"unique_zips = requests['Incident Zip'].unique()\nunique_zips",
"_____no_output_____"
]
],
[
[
"Amazing! This is much cleaner. There's something a bit weird here, though -- I looked up 77056 on Google maps, and that's in Texas.\n\nLet's take a closer look:",
"_____no_output_____"
]
],
[
[
"zips = requests['Incident Zip']\n# Let's say the zips starting with '0' and '1' are okay, for now. (this isn't actually true -- 13221 is in Syracuse, and why?)\nis_close = zips.str.startswith('0') | zips.str.startswith('1')\n# There are a bunch of NaNs, but we're not interested in them right now, so we'll say they're False\nis_far = ~(is_close) & zips.notnull()",
"_____no_output_____"
],
[
"zips[is_far]",
"_____no_output_____"
],
[
"requests[is_far][['Incident Zip', 'Descriptor', 'City']].sort_values(by='Incident Zip')",
"_____no_output_____"
]
],
[
[
"Okay, there really are requests coming from LA and Houston! Good to know. Filtering by zip code is probably a bad way to handle this -- we should really be looking at the city instead.",
"_____no_output_____"
]
],
[
[
"requests['City'].str.upper().value_counts()",
"_____no_output_____"
]
],
[
[
"It looks like these are legitimate complaints, so we'll just leave them alone.",
"_____no_output_____"
],
[
"# 7.4 Putting it together",
"_____no_output_____"
],
[
"Here's what we ended up doing to clean up our zip codes, all together:",
"_____no_output_____"
]
],
[
[
"na_values = ['NO CLUE', 'N/A', '0']\nrequests = pd.read_csv('data/311-service-requests.csv', \n na_values=na_values, \n dtype={'Incident Zip': str})",
"_____no_output_____"
],
[
"def fix_zip_codes(zips):\n # Truncate everything to length 5 \n zips = zips.str.slice(0, 5)\n \n # Set 00000 zip codes to nan\n zero_zips = zips == '00000'\n zips[zero_zips] = np.nan\n \n return zips",
"_____no_output_____"
],
[
"requests['Incident Zip'] = fix_zip_codes(requests['Incident Zip'])",
"_____no_output_____"
],
[
"requests['Incident Zip'].unique()",
"_____no_output_____"
]
],
[
[
"<style>\n @font-face {\n font-family: \"Computer Modern\";\n src: url('http://mirrors.ctan.org/fonts/cm-unicode/fonts/otf/cmunss.otf');\n }\n div.cell{\n width:800px;\n margin-left:16% !important;\n margin-right:auto;\n }\n h1 {\n font-family: Helvetica, serif;\n }\n h4{\n margin-top:12px;\n margin-bottom: 3px;\n }\n div.text_cell_render{\n font-family: Computer Modern, \"Helvetica Neue\", Arial, Helvetica, Geneva, sans-serif;\n line-height: 145%;\n font-size: 130%;\n width:800px;\n margin-left:auto;\n margin-right:auto;\n }\n .CodeMirror{\n font-family: \"Source Code Pro\", source-code-pro,Consolas, monospace;\n }\n .text_cell_render h5 {\n font-weight: 300;\n font-size: 22pt;\n color: #4057A1;\n font-style: italic;\n margin-bottom: .5em;\n margin-top: 0.5em;\n display: block;\n }\n \n .warning{\n color: rgb( 240, 20, 20 )\n } ",
"_____no_output_____"
]
]
] |
[
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown"
] |
[
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
]
] |
cb94d76475912465b239cd8799925bdc6a2e7e55
| 6,195 |
ipynb
|
Jupyter Notebook
|
experimental/pytables_h5py/features_lessspikes.ipynb
|
klusta-team/kwiklib
|
617a6ceff55957728c3dc94109b64e4c427429c2
|
[
"BSD-3-Clause"
] | 7 |
2015-01-20T13:55:51.000Z
|
2018-02-06T09:31:21.000Z
|
experimental/pytables_h5py/features_lessspikes.ipynb
|
klusta-team/kwiklib
|
617a6ceff55957728c3dc94109b64e4c427429c2
|
[
"BSD-3-Clause"
] | 6 |
2015-01-08T18:13:53.000Z
|
2016-06-22T09:53:53.000Z
|
experimental/pytables_h5py/features_lessspikes.ipynb
|
klusta-team/kwiklib
|
617a6ceff55957728c3dc94109b64e4c427429c2
|
[
"BSD-3-Clause"
] | 8 |
2015-01-22T22:57:19.000Z
|
2020-03-19T11:43:56.000Z
| 26.587983 | 98 | 0.444391 |
[
[
[
"empty"
]
]
] |
[
"empty"
] |
[
[
"empty"
]
] |
cb94d793abd17af24b801f3e01b1af22970472af
| 80,776 |
ipynb
|
Jupyter Notebook
|
setup/new-dataset-for-e2e-ml.ipynb
|
alar0330/amazon-sagemaker-build-train-deploy
|
b476c5ba5b3bd55a99709e7788079763fa498682
|
[
"Apache-2.0"
] | 80 |
2020-09-16T13:29:43.000Z
|
2022-03-28T10:13:52.000Z
|
setup/new-dataset-for-e2e-ml.ipynb
|
alar0330/amazon-sagemaker-build-train-deploy
|
b476c5ba5b3bd55a99709e7788079763fa498682
|
[
"Apache-2.0"
] | 1 |
2021-07-05T08:07:51.000Z
|
2021-07-05T08:07:51.000Z
|
setup/new-dataset-for-e2e-ml.ipynb
|
alar0330/amazon-sagemaker-build-train-deploy
|
b476c5ba5b3bd55a99709e7788079763fa498682
|
[
"Apache-2.0"
] | 17 |
2020-12-06T12:40:01.000Z
|
2022-03-21T07:23:51.000Z
| 64.107937 | 10,020 | 0.6624 |
[
[
[
"import boto3\nimport sagemaker\nimport time\nimport pandas as pd\nimport numpy as np\n\nrole = sagemaker.get_execution_role()\nregion = boto3.Session().region_name\nsagemaker_session = sagemaker.Session()\nbucket_name = sagemaker_session.default_bucket()\nprefix = 'endtoendmlsm'\n\nprint(region)\nprint(role)\nprint(bucket_name)",
"eu-west-1\narn:aws:iam::041631420165:role/service-role/AmazonSageMaker-ExecutionRole-20180507T143636\nsagemaker-eu-west-1-041631420165\n"
],
[
"s3 = boto3.resource('s3')\n\nfile_key = 'data/raw/windturbine_raw_data.csv'\ncopy_source = {\n 'Bucket': 'gianpo-public',\n 'Key': 'endtoendml/{0}'.format(file_key)\n}\n\ns3.Bucket(bucket_name).copy(copy_source, '{0}/'.format(prefix) + file_key)\n#sagemaker_session.upload_data('/home/ec2-user/SageMaker/windturbine_raw_data_2.csv', bucket=bucket_name, key_prefix='endtoendmlsm/data/raw')\n",
"_____no_output_____"
],
[
"sagemaker_session.download_data(path='.', bucket=bucket_name, key_prefix='endtoendmlsm/data/raw/windturbine_raw_data.csv', extra_args=None)",
"_____no_output_____"
],
[
"original_ds = pd.read_csv('./windturbine_raw_data.csv', names = ['turbine_id','turbine_type','wind_speed','rpm_blade','oil_temperature','oil_level','temperature','humidity','vibrations_frequency','pressure','wind_direction','breakdown'])\noriginal_ds.head()",
"_____no_output_____"
],
[
"original_ds.describe(include='all')",
"_____no_output_____"
],
[
"%matplotlib inline\nimport matplotlib.pyplot as plt",
"_____no_output_____"
],
[
"plt.scatter(original_ds.wind_speed, original_ds.rpm_blade)",
"_____no_output_____"
],
[
"plt.hist(original_ds.wind_speed, bins=70)",
"_____no_output_____"
],
[
"#wind_speed: gaussian, mean=50, std=30\nws = abs(np.random.normal(50, 30, 1000000)).astype(int)\n#temperature: gaussian, mean=20, std=18\ntemp = abs(np.random.normal(20, 18, 1000000)).astype(int)\n#humidity: gaussian, mean=50, std=5\nhum = abs(np.random.normal(50, 5, 1000000)).astype(int)\n#pressure: gaussian, mean=40, std=25\npress = abs(np.random.normal(40, 25, 1000000)).astype(int)\n#oil_level: uniform, min=5, max=35\noil_lev = np.random.uniform(5,35,1000000).astype(int)\n#rpm_blade: alpha*wind_speed + error\nalpha = 5\nrpm_blade = abs(alpha*ws + np.random.normal(0,30,1000000)).astype(int)\n#vibration_freq: beta*rpm_blade + gamma*pressure + error\nbeta = 3.5\ngamma = 2\nvibration_freq = abs(beta*rpm_blade + gamma*press + np.random.normal(0,50,1000000)).astype(int)\n#oil_temp: delta*temp + error\n#delta = 4.5\n#oil_temperature = abs(delta*temp + np.random.normal(0,50,1000000)).astype(int)\n#breakdown: k1*rpm_blade + k2*vibration_freq + k3*oil_temp + error",
"_____no_output_____"
],
[
"new_dataset = pd.DataFrame()",
"_____no_output_____"
],
[
"new_dataset['turbine_id'] = original_ds['turbine_id']\nnew_dataset['turbine_type'] = original_ds['turbine_type']\nnew_dataset['wind_direction'] = original_ds['wind_direction']\nnew_dataset['wind_speed'] = ws\nnew_dataset['temperature'] = temp\nnew_dataset['humidity'] = hum\nnew_dataset['pressure'] = press\nnew_dataset['oil_level'] = oil_lev\nnew_dataset['rpm_blade'] = rpm_blade\nnew_dataset['vibrations_frequency'] = vibration_freq\nnew_dataset['oil_temperature'] = original_ds['oil_temperature']",
"_____no_output_____"
],
[
"new_dataset.describe()",
"_____no_output_____"
],
[
"plt.scatter(new_dataset['wind_speed'][:10000], new_dataset['rpm_blade'][:10000])",
"_____no_output_____"
],
[
"plt.hist(new_dataset['rpm_blade'][:10000])",
"_____no_output_____"
],
[
"from scipy.special import expit",
"_____no_output_____"
],
[
"k1=0.0003\nk2=0.0005\nk3=0.0033\nbreakdown = k1*rpm_blade + k2*vibration_freq + k3*oil_lev + np.random.normal(0,0.1,1000000)\n\nnew_dataset['breakdown_num'] = breakdown\nnew_dataset.loc[new_dataset['breakdown_num'] <= 0.9, 'breakdown'] = 'no' \nnew_dataset.loc[new_dataset['breakdown_num'] > 0.9, 'breakdown'] = 'yes' ",
"_____no_output_____"
],
[
"new_dataset.describe(include='all')",
"_____no_output_____"
],
[
"plt.scatter(new_dataset['breakdown'][:10000], new_dataset['rpm_blade'][:10000])",
"_____no_output_____"
],
[
"plt.scatter(new_dataset['breakdown'][:10000], new_dataset['vibrations_frequency'][:10000])",
"_____no_output_____"
],
[
"final_dataset = new_dataset\nfinal_dataset = final_dataset.drop(columns=['breakdown_num'])\nfinal_dataset.to_csv('windturbine_raw_data.csv', index=False, columns = ['turbine_id','turbine_type','wind_speed','rpm_blade','oil_temperature','oil_level','temperature','humidity','vibrations_frequency','pressure','wind_direction','breakdown'])",
"_____no_output_____"
],
[
"sagemaker_session.upload_data('windturbine_raw_data.csv', bucket=bucket_name, key_prefix='endtoendmlsm/data/raw')",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb94d8429200cf47d19711238cb0f725d57f5523
| 421,927 |
ipynb
|
Jupyter Notebook
|
nbs/dl1/00_notebook_tutorial-Copy1.ipynb
|
sunbee/nbs-dl1
|
fda7711c6008bca97b71bf812b1fe50991760a9a
|
[
"Apache-2.0"
] | null | null | null |
nbs/dl1/00_notebook_tutorial-Copy1.ipynb
|
sunbee/nbs-dl1
|
fda7711c6008bca97b71bf812b1fe50991760a9a
|
[
"Apache-2.0"
] | null | null | null |
nbs/dl1/00_notebook_tutorial-Copy1.ipynb
|
sunbee/nbs-dl1
|
fda7711c6008bca97b71bf812b1fe50991760a9a
|
[
"Apache-2.0"
] | null | null | null | 503.49284 | 387,208 | 0.944554 |
[
[
[
"**Important note:** You should always work on a duplicate of the course notebook. On the page you used to open this, tick the box next to the name of the notebook and click duplicate to easily create a new version of this notebook.\n\nYou will get errors each time you try to update your course repository if you don't do this, and your changes will end up being erased by the original course version.",
"_____no_output_____"
],
[
"# Welcome to Jupyter Notebooks!",
"_____no_output_____"
],
[
"If you want to learn how to use this tool you've come to the right place. This article will teach you all you need to know to use Jupyter Notebooks effectively. You only need to go through Section 1 to learn the basics and you can go into Section 2 if you want to further increase your productivity.",
"_____no_output_____"
],
[
"You might be reading this tutorial in a web page (maybe Github or the course's webpage). We strongly suggest to read this tutorial in a (yes, you guessed it) Jupyter Notebook. This way you will be able to actually *try* the different commands we will introduce here.",
"_____no_output_____"
],
[
"## Section 1: Need to Know",
"_____no_output_____"
],
[
"### Introduction",
"_____no_output_____"
],
[
"Let's build up from the basics, what is a Jupyter Notebook? Well, you are reading one. It is a document made of cells. You can write like I am writing now (markdown cells) or you can perform calculations in Python (code cells) and run them like this:",
"_____no_output_____"
]
],
[
[
"1+1",
"_____no_output_____"
]
],
[
[
"Cool huh? This combination of prose and code makes Jupyter Notebook ideal for experimentation: we can see the rationale for each experiment, the code and the results in one comprehensive document. In fast.ai, each lesson is documented in a notebook and you can later use that notebook to experiment yourself. \n\nOther renowned institutions in academy and industry use Jupyter Notebook: Google, Microsoft, IBM, Bloomberg, Berkeley and NASA among others. Even Nobel-winning economists [use Jupyter Notebooks](https://paulromer.net/jupyter-mathematica-and-the-future-of-the-research-paper/) for their experiments and some suggest that Jupyter Notebooks will be the [new format for research papers](https://www.theatlantic.com/science/archive/2018/04/the-scientific-paper-is-obsolete/556676/).",
"_____no_output_____"
],
[
"### Writing",
"_____no_output_____"
],
[
"A type of cell in which you can write like this is called _Markdown_. [_Markdown_](https://en.wikipedia.org/wiki/Markdown) is a very popular markup language. To specify that a cell is _Markdown_ you need to click in the drop-down menu in the toolbar and select _Markdown_.",
"_____no_output_____"
],
[
"Click on the the '+' button on the left and select _Markdown_ from the toolbar.",
"_____no_output_____"
],
[
"My first markdown cell",
"_____no_output_____"
],
[
"Now you can type your first _Markdown_ cell. Write 'My first markdown cell' and press run.",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"You should see something like this:",
"_____no_output_____"
],
[
"My first markdown cell",
"_____no_output_____"
],
[
"Now try making your first _Code_ cell: follow the same steps as before but don't change the cell type (when you add a cell its default type is _Code_). Type something like 3/2. You should see '1.5' as output.",
"_____no_output_____"
]
],
[
[
"3/2",
"_____no_output_____"
],
[
"3/2",
"_____no_output_____"
]
],
[
[
"### Modes",
"_____no_output_____"
],
[
"If you made a mistake in your *Markdown* cell and you have already ran it, you will notice that you cannot edit it just by clicking on it. This is because you are in **Command Mode**. Jupyter Notebooks have two distinct modes:\n\n1. **Edit Mode**: Allows you to edit a cell's content.\n\n2. **Command Mode**: Allows you to edit the notebook as a whole and use keyboard shortcuts but not edit a cell's content. \n\nYou can toggle between these two by either pressing <kbd>ESC</kbd> and <kbd>Enter</kbd> or clicking outside a cell or inside it (you need to double click if its a Markdown cell). You can always know which mode you're on since the current cell has a green border if in **Edit Mode** and a blue border in **Command Mode**. Try it!",
"_____no_output_____"
],
[
"### Other Important Considerations",
"_____no_output_____"
],
[
"1. Your notebook is autosaved every 120 seconds. If you want to manually save it you can just press the save button on the upper left corner or press <kbd>s</kbd> in **Command Mode**.",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"2. To know if your kernel is computing or not you can check the dot in your upper right corner. If the dot is full, it means that the kernel is working. If not, it is idle. You can place the mouse on it and see the state of the kernel be displayed.",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"3. There are a couple of shortcuts you must know about which we use **all** the time (always in **Command Mode**). These are:\n\n<kbd>Shift</kbd>+<kbd>Enter</kbd>: Runs the code or markdown on a cell\n\n<kbd>Up Arrow</kbd>+<kbd>Down Arrow</kbd>: Toggle across cells\n\n<kbd>b</kbd>: Create new cell\n\n<kbd>0</kbd>+<kbd>0</kbd>: Reset Kernel\n\nYou can find more shortcuts in the Shortcuts section below.",
"_____no_output_____"
],
[
"4. You may need to use a terminal in a Jupyter Notebook environment (for example to git pull on a repository). That is very easy to do, just press 'New' in your Home directory and 'Terminal'. Don't know how to use the Terminal? We made a tutorial for that as well. You can find it [here](https://course.fast.ai/terminal_tutorial.html).",
"_____no_output_____"
],
[
"",
"_____no_output_____"
],
[
"That's it. This is all you need to know to use Jupyter Notebooks. That said, we have more tips and tricks below ↓↓↓",
"_____no_output_____"
],
[
"## Section 2: Going deeper",
"_____no_output_____"
],
[
"### Markdown formatting",
"_____no_output_____"
],
[
"#### Italics, Bold, Strikethrough, Inline, Blockquotes and Links",
"_____no_output_____"
],
[
"The five most important concepts to format your code appropriately when using markdown are:\n \n1. *Italics*: Surround your text with '\\_' or '\\*'\n2. **Bold**: Surround your text with '\\__' or '\\**'\n3. `inline`: Surround your text with '\\`'\n4. > blockquote: Place '\\>' before your text.\n5. [Links](https://course.fast.ai/): Surround the text you want to link with '\\[\\]' and place the link adjacent to the text, surrounded with '()'\n",
"_____no_output_____"
],
[
"#### Headings",
"_____no_output_____"
],
[
"Notice that including a hashtag before the text in a markdown cell makes the text a heading. The number of hashtags you include will determine the priority of the header ('#' is level one, '##' is level two, '###' is level three and '####' is level four). We will add three new cells with the '+' button on the left to see how every level of heading looks.",
"_____no_output_____"
],
[
"Double click on some headings and find out what level they are!",
"_____no_output_____"
],
[
"#### Lists",
"_____no_output_____"
],
[
"There are three types of lists in markdown.",
"_____no_output_____"
],
[
"Ordered list:\n\n1. Step 1\n 2. Step 1B\n3. Step 3",
"_____no_output_____"
],
[
"Unordered list\n\n* learning rate\n* cycle length\n* weight decay",
"_____no_output_____"
],
[
"Task list\n\n- [x] Learn Jupyter Notebooks\n - [x] Writing\n - [x] Modes\n - [x] Other Considerations\n- [ ] Change the world",
"_____no_output_____"
],
[
"Double click on each to see how they are built! ",
"_____no_output_____"
],
[
"### Code Capabilities",
"_____no_output_____"
],
[
"**Code** cells are different than **Markdown** cells in that they have an output cell. This means that we can _keep_ the results of our code within the notebook and share them. Let's say we want to show a graph that explains the result of an experiment. We can just run the necessary cells and save the notebook. The output will be there when we open it again! Try it out by running the next four cells.",
"_____no_output_____"
]
],
[
[
"# Import necessary libraries\nfrom fastai.vision import * \nimport matplotlib.pyplot as plt",
"_____no_output_____"
],
[
"from PIL import Image",
"_____no_output_____"
],
[
"a = 1\nb = a + 1\nc = b + a + 1\nd = c + b + a + 1\na, b, c ,d",
"_____no_output_____"
],
[
"plt.plot([a,b,c,d])\nplt.show()",
"_____no_output_____"
]
],
[
[
"We can also print images while experimenting. I am watching you.",
"_____no_output_____"
]
],
[
[
"Image.open('images/notebook_tutorial/cat_example.jpg')",
"_____no_output_____"
]
],
[
[
"### Running the app locally",
"_____no_output_____"
],
[
"You may be running Jupyter Notebook from an interactive coding environment like Gradient, Sagemaker or Salamander. You can also run a Jupyter Notebook server from your local computer. What's more, if you have installed Anaconda you don't even need to install Jupyter (if not, just `pip install jupyter`).\n\nYou just need to run `jupyter notebook` in your terminal. Remember to run it from a folder that contains all the folders/files you will want to access. You will be able to open, view and edit files located within the directory in which you run this command but not files in parent directories.\n\nIf a browser tab does not open automatically once you run the command, you should CTRL+CLICK the link starting with 'https://localhost:' and this will open a new tab in your default browser.",
"_____no_output_____"
],
[
"### Creating a notebook",
"_____no_output_____"
],
[
"Click on 'New' in the upper right corner and 'Python 3' in the drop-down list (we are going to use a [Python kernel](https://github.com/ipython/ipython) for all our experiments).\n\n\n\nNote: You will sometimes hear people talking about the Notebook 'kernel'. The 'kernel' is just the Python engine that performs the computations for you. ",
"_____no_output_____"
],
[
"### Shortcuts and tricks",
"_____no_output_____"
],
[
"#### Command Mode Shortcuts",
"_____no_output_____"
],
[
"There are a couple of useful keyboard shortcuts in `Command Mode` that you can leverage to make Jupyter Notebook faster to use. Remember that to switch back and forth between `Command Mode` and `Edit Mode` with <kbd>Esc</kbd> and <kbd>Enter</kbd>.",
"_____no_output_____"
],
[
"<kbd>m</kbd>: Convert cell to Markdown",
"_____no_output_____"
],
[
"<kbd>y</kbd>: Convert cell to Code",
"_____no_output_____"
],
[
"<kbd>D</kbd>+<kbd>D</kbd>: Delete the cell(if it's not the only cell) or delete the content of the cell and reset cell to Code(if only one cell left)",
"_____no_output_____"
],
[
"<kbd>o</kbd>: Toggle between hide or show output",
"_____no_output_____"
],
[
"<kbd>Shift</kbd>+<kbd>Arrow up/Arrow down</kbd>: Selects multiple cells. Once you have selected them you can operate on them like a batch (run, copy, paste etc).",
"_____no_output_____"
],
[
"<kbd>Shift</kbd>+<kbd>M</kbd>: Merge selected cells.",
"_____no_output_____"
],
[
"<kbd>Shift</kbd>+<kbd>Tab</kbd>: [press these two buttons at the same time, once] Tells you which parameters to pass on a function\n\n<kbd>Shift</kbd>+<kbd>Tab</kbd>: [press these two buttons at the same time, three times] Gives additional information on the method",
"_____no_output_____"
],
[
"#### Cell Tricks",
"_____no_output_____"
]
],
[
[
"from fastai import*\nfrom fastai.vision import *",
"_____no_output_____"
]
],
[
[
"There are also some tricks that you can code into a cell.",
"_____no_output_____"
],
[
"`?function-name`: Shows the definition and docstring for that function",
"_____no_output_____"
]
],
[
[
"?ImageDataBunch",
"_____no_output_____"
]
],
[
[
"`??function-name`: Shows the source code for that function",
"_____no_output_____"
]
],
[
[
"??ImageDataBunch",
"_____no_output_____"
]
],
[
[
"`doc(function-name)`: Shows the definition, docstring **and links to the documentation** of the function\n(only works with fastai library imported)",
"_____no_output_____"
]
],
[
[
"doc(ImageDataBunch)",
"_____no_output_____"
]
],
[
[
"#### Line Magics",
"_____no_output_____"
],
[
"Line magics are functions that you can run on cells and take as an argument the rest of the line from where they are called. You call them by placing a '%' sign before the command. The most useful ones are:",
"_____no_output_____"
],
[
"`%matplotlib inline`: This command ensures that all matplotlib plots will be plotted in the output cell within the notebook and will be kept in the notebook when saved.",
"_____no_output_____"
],
[
"`%reload_ext autoreload`, `%autoreload 2`: Reload all modules before executing a new line. If a module is edited, it is not necessary to rerun the import commands, the modules will be reloaded automatically.",
"_____no_output_____"
],
[
"These three commands are always called together at the beginning of every notebook.",
"_____no_output_____"
]
],
[
[
"%matplotlib inline\n%reload_ext autoreload\n%autoreload 2",
"_____no_output_____"
]
],
[
[
"`%timeit`: Runs a line a ten thousand times and displays the average time it took to run it.",
"_____no_output_____"
]
],
[
[
"%timeit [i+1 for i in range(1000)]",
"41.7 µs ± 872 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n"
]
],
[
[
"`%debug`: Allows to inspect a function which is showing an error using the [Python debugger](https://docs.python.org/3/library/pdb.html).",
"_____no_output_____"
]
],
[
[
"for i in range(1000):\n a = i+1\n b = 'string'\n c = b+1",
"_____no_output_____"
],
[
"%debug",
"> \u001b[0;32m<ipython-input-17-8d78ff778454>\u001b[0m(4)\u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m 1 \u001b[0;31m\u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1000\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m 2 \u001b[0;31m \u001b[0ma\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m 3 \u001b[0;31m \u001b[0mb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m'string'\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m----> 4 \u001b[0;31m \u001b[0mc\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\nipdb> \nipdb> \nipdb> c=b+sr(1)\n"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code",
"code",
"code",
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown",
"markdown",
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code"
]
] |
cb94fe2e849d80a8e78bc3b8eb33e841671de1a3
| 2,731 |
ipynb
|
Jupyter Notebook
|
round/jim.ipynb
|
tagordon/round
|
65e6329087f007e763893dd5103073390c9cbeb6
|
[
"MIT"
] | null | null | null |
round/jim.ipynb
|
tagordon/round
|
65e6329087f007e763893dd5103073390c9cbeb6
|
[
"MIT"
] | null | null | null |
round/jim.ipynb
|
tagordon/round
|
65e6329087f007e763893dd5103073390c9cbeb6
|
[
"MIT"
] | null | null | null | 23.144068 | 83 | 0.540095 |
[
[
[
"from round import lc\nimport matplotlib.pyplot as plt\nimport matplotlib.gridspec as gridspec\nimport sys\nimport os\nimport numpy as np\nimport matplotlib.pyplot as pl",
"_____no_output_____"
],
[
"# point to your own light curves\nfolder = '/Users/tgordon/research/tess_adap/lcs'\nfiles = os.listdir(folder)\nfitsfile = folder + \"/\" + files[0]\nlight_curve = lc.LightCurve.TESS_adap(fitsfile)",
"_____no_output_____"
],
[
"# plot the light curve before any detrending or clipping \nfig = pl.figure()\nlight_curve.plot_raw(fig.gca())",
"_____no_output_____"
],
[
"# third order polynomial subtracted, outliers clipped\nlight_curve.normalize()\nfig = pl.figure()\nlight_curve.plot(fig.gca())",
"_____no_output_____"
],
[
"# computes autocorrelation, optimizes GP, and runs mcmc \nlight_curve.compute(mcmc=True, mcmc_draws=500, tune=500, \n target_accept=0.9, prior_sig=3.0, \n with_SHOTerm=False, cores=2)",
"_____no_output_____"
],
[
"# corner plot of the mcmc along with a summary of the output \nlight_curve.plot_corner(smooth=True, \n truths=light_curve.mcmc_summary[\"mean\"].values, \n truth_color=\"#f55649\");",
"_____no_output_____"
],
[
"# summary of the mcmc in a pandas dataframe\nlight_curve.mcmc_summary",
"_____no_output_____"
]
]
] |
[
"code"
] |
[
[
"code",
"code",
"code",
"code",
"code",
"code",
"code"
]
] |
cb950580edc8f68fa8155185594f0a5cd1b4266e
| 8,828 |
ipynb
|
Jupyter Notebook
|
bbuzz/.ipynb_checkpoints/1_create_feature_groups-checkpoint.ipynb
|
javierdlrm/events
|
d950c8cdcf24181c2e748e51ea1b44fce6cfd30a
|
[
"Apache-2.0"
] | 1 |
2021-05-27T23:01:24.000Z
|
2021-05-27T23:01:24.000Z
|
bbuzz/1_create_feature_groups.ipynb
|
javierdlrm/events
|
d950c8cdcf24181c2e748e51ea1b44fce6cfd30a
|
[
"Apache-2.0"
] | null | null | null |
bbuzz/1_create_feature_groups.ipynb
|
javierdlrm/events
|
d950c8cdcf24181c2e748e51ea1b44fce6cfd30a
|
[
"Apache-2.0"
] | null | null | null | 31.194346 | 536 | 0.521296 |
[
[
[
"---\ntitle: \"Create empty feature groups for Online Feature Store\"\ndate: 2021-04-25\ntype: technical_note\ndraft: false\n---",
"_____no_output_____"
]
],
[
[
"import json\nfrom pyspark.sql.types import StructField, StructType, StringType, DoubleType, TimestampType, LongType, IntegerType",
"Starting Spark application\n"
]
],
[
[
"# Create empty feature groups \nIn this demo example we are expecting to recieve data from Kafka topic, read using spark streaming, do streamig aggregations and ingest aggregated data to feature groups. Thus we will create empy feature groups where we will ingest streaming data.",
"_____no_output_____"
],
[
"### Define schema for feature groups",
"_____no_output_____"
]
],
[
[
"card_schema = StructType([StructField('tid', StringType(), True),\n StructField('datetime', StringType(), True),\n StructField('cc_num', LongType(), True),\n StructField('amount', DoubleType(), True)])\n\nschema_10m = StructType([StructField('cc_num', LongType(), True),\n StructField('num_trans_per_10m', LongType(), True),\n StructField('avg_amt_per_10m', DoubleType(), True),\n StructField('stdev_amt_per_10m', DoubleType(), True)])\n\nschema_1h = StructType([StructField('cc_num', LongType(), True),\n StructField('num_trans_per_1h', LongType(), True),\n StructField('avg_amt_per_1h', DoubleType(), True),\n StructField('stdev_amt_per_1h', DoubleType(), True)])\n\nschema_12h = StructType([StructField('cc_num', LongType(), True),\n StructField('num_trans_per_12h', LongType(), True),\n StructField('avg_amt_per_12h', DoubleType(), True),\n StructField('stdev_amt_per_12h', DoubleType(), True)])",
"_____no_output_____"
]
],
[
[
"### Create empty spark dataframes ",
"_____no_output_____"
]
],
[
[
"empty_card_df = sqlContext.createDataFrame(sc.emptyRDD(), card_schema)\nempty_10m_agg_df = sqlContext.createDataFrame(sc.emptyRDD(), schema_10m)\nempty_1h_agg_df = sqlContext.createDataFrame(sc.emptyRDD(), schema_1h)\nempty_12h_agg_df = sqlContext.createDataFrame(sc.emptyRDD(), schema_12h)",
"_____no_output_____"
]
],
[
[
"### Establish a connection with your Hopsworks feature store.",
"_____no_output_____"
]
],
[
[
"import hsfs\nconnection = hsfs.connection()\n# get a reference to the feature store, you can access also shared feature stores by providing the feature store name\nfs = connection.get_feature_store()",
"Connected. Call `.close()` to terminate connection gracefully."
]
],
[
[
"### Create feature group metadata objects and save empty spark dataframes to materialise them in hopsworks feature store.\n\nNow We will create each feature group and enable online feature store. Since we are plannig to use these feature groups durring online model serving primary key(s) are required to retrieve feature vector from online feature store. ",
"_____no_output_____"
]
],
[
[
"card_transactions = fs.create_feature_group(\"card_transactions\", \n version = 1,\n online_enabled=False, \n statistics_config=False, \n primary_key=[\"tid\"])\n\ncard_transactions.save(empty_card_df)",
"<hsfs.feature_group.FeatureGroup object at 0x7f39a574b1d0>"
],
[
"card_transactions_10m_agg = fs.create_feature_group(\"card_transactions_10m_agg\", \n version = 1,\n online_enabled=True, \n statistics_config=False, \n primary_key=[\"cc_num\"])\n\ncard_transactions_10m_agg.save(empty_10m_agg_df)",
"<hsfs.feature_group.FeatureGroup object at 0x7f39a52e4fd0>"
],
[
"card_transactions_1h_agg = fs.create_feature_group(\"card_transactions_1h_agg\", \n version = 1,\n online_enabled=True, \n statistics_config=False, \n primary_key=[\"cc_num\"])\n\ncard_transactions_1h_agg.save(empty_1h_agg_df)",
"<hsfs.feature_group.FeatureGroup object at 0x7f39a52d7190>"
],
[
"card_transactions_12h_agg = fs.create_feature_group(\"card_transactions_12h_agg\", \n version = 1,\n online_enabled=True, \n statistics_config=False, \n primary_key=[\"cc_num\"])\n\ncard_transactions_12h_agg.save(empty_12h_agg_df)",
"<hsfs.feature_group.FeatureGroup object at 0x7f39a5768050>"
]
]
] |
[
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code",
"markdown",
"code"
] |
[
[
"markdown"
],
[
"code"
],
[
"markdown",
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code"
],
[
"markdown"
],
[
"code",
"code",
"code",
"code"
]
] |
cb9509789d5f1031b6eb124c30327ce7112d3bd6
| 131,467 |
ipynb
|
Jupyter Notebook
|
polymath/resources/draft-polymath.ipynb
|
ajay-k-nair/Superficial
|
9bd33a052d87d4340aace92f75b40329e4daf844
|
[
"MIT"
] | 1 |
2020-01-27T16:30:01.000Z
|
2020-01-27T16:30:01.000Z
|
polymath/resources/draft-polymath.ipynb
|
ajay-k-nair/Superficial
|
9bd33a052d87d4340aace92f75b40329e4daf844
|
[
"MIT"
] | 37 |
2017-12-21T05:43:47.000Z
|
2021-08-10T05:06:35.000Z
|
polymath/resources/draft-polymath.ipynb
|
ajay-k-nair/Superficial
|
9bd33a052d87d4340aace92f75b40329e4daf844
|
[
"MIT"
] | 18 |
2016-08-09T10:31:23.000Z
|
2021-02-02T06:55:46.000Z
| 144.152412 | 1,410 | 0.619836 |
[
[
[
"empty"
]
]
] |
[
"empty"
] |
[
[
"empty"
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.