path
stringlengths 13
17
| screenshot_names
sequencelengths 1
873
| code
stringlengths 0
40.4k
| cell_type
stringclasses 1
value |
|---|---|---|---|
90156125/cell_16 | [
"text_plain_output_1.png"
] | import cudf as pd
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
df_train = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Training.csv')
match_2020 = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2020.csv')
previous_match = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2008-2019.csv')
players = pd.read_csv('../input/ipl-2020-player-performance/IPL 2020 Squads.csv', encoding='windows-1254')
previous_match.shape
previous_match.isnull().sum() | code |
90156125/cell_17 | [
"text_html_output_1.png"
] | import cudf as pd
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
df_train = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Training.csv')
match_2020 = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2020.csv')
previous_match = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2008-2019.csv')
players = pd.read_csv('../input/ipl-2020-player-performance/IPL 2020 Squads.csv', encoding='windows-1254')
previous_match.shape
previous_match.isnull().sum()
previous_match.dropna() | code |
90156125/cell_24 | [
"text_html_output_1.png"
] | import cudf as pd
import matplotlib.pyplot as plt
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
df_train = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Training.csv')
match_2020 = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2020.csv')
previous_match = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2008-2019.csv')
players = pd.read_csv('../input/ipl-2020-player-performance/IPL 2020 Squads.csv', encoding='windows-1254')
previous_match.shape
previous_match.isnull().sum()
previous_match.dropna()
previous_match.season.unique()
order = previous_match.city.value_counts().iloc[:10].index
plt.figure(figsize=(30, 10))
order = previous_match.winner.value_counts().iloc[:10].index
sns.countplot(x=previous_match['winner'], palette='rainbow', data=previous_match, order=order)
plt.show() | code |
90156125/cell_14 | [
"text_html_output_1.png"
] | import cudf as pd
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df_train = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Training.csv')
match_2020 = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2020.csv')
previous_match = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2008-2019.csv')
players = pd.read_csv('../input/ipl-2020-player-performance/IPL 2020 Squads.csv', encoding='windows-1254')
previous_match.shape | code |
90156125/cell_22 | [
"text_plain_output_1.png"
] | import cudf as pd
import matplotlib.pyplot as plt
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
df_train = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Training.csv')
match_2020 = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2020.csv')
previous_match = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2008-2019.csv')
players = pd.read_csv('../input/ipl-2020-player-performance/IPL 2020 Squads.csv', encoding='windows-1254')
previous_match.shape
previous_match.isnull().sum()
previous_match.dropna()
previous_match.season.unique()
plt.subplots(figsize=(15, 6))
order = previous_match.city.value_counts().iloc[:10].index
sns.countplot(x=previous_match['city'], data=previous_match, order=order)
plt.show() | code |
90156125/cell_10 | [
"text_html_output_1.png"
] | import cudf as pd
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df_train = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Training.csv')
match_2020 = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2020.csv')
previous_match = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2008-2019.csv')
players = pd.read_csv('../input/ipl-2020-player-performance/IPL 2020 Squads.csv', encoding='windows-1254')
df_train.isnull().sum()
df_train.describe().T
df_train.head() | code |
90156125/cell_5 | [
"image_output_1.png"
] | import cudf as pd
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df_train = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Training.csv')
match_2020 = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2020.csv')
previous_match = pd.read_csv('/kaggle/input/ipl-2020-player-performance/Matches IPL 2008-2019.csv')
players = pd.read_csv('../input/ipl-2020-player-performance/IPL 2020 Squads.csv', encoding='windows-1254')
df_train.head() | code |
32068954/cell_21 | [
"application_vnd.jupyter.stderr_output_1.png"
] | from pyspark.sql import SparkSession
import json
import os
spark = SparkSession.builder.appName('SimpleApp').getOrCreate()
sc = spark.sparkContext
def findArticlePath(article_sha_id):
"""
This function finds the full path given an article_sha_id
"""
ROOT_PATH = '/kaggle/input/CORD-19-research-challenge/biorxiv_medrxiv/biorxiv_medrxiv/'
FILE_SUFFIX = '.json'
article_path = ROOT_PATH + article_sha_id + FILE_SUFFIX
return article_path
def retrieveJson(article_sha_id):
"""
Given a 1-word string containing a JSON key, return the data for those keys.
Also return the location of those keys?
"""
article_file_path = findArticlePath(article_sha_id)
with open(article_file_path, 'r') as read_file:
json_string = json.load(read_file)
return json_string
def retrieveKey(json_input, key_input):
"""
Uses retrieveTopKey and retrieveSubKey to return key values from anywhere in the JSON data string.
"""
result = []
istopitem = False
if not isinstance(json_input, list):
json_string = json_input
json_input = []
json_input.append(json_string)
istopitem = True
if istopitem:
for key in json_string:
if key == key_input:
top_key_value = json_string[key]
result.append(top_key_value)
return result
for json_string in json_input:
if isinstance(json_string, dict):
for key in json_string:
top_key_value = json_string[key]
if isinstance(top_key_value, dict):
for sub_key in top_key_value:
sub_key_value = top_key_value[sub_key]
if sub_key == key_input:
result.append(sub_key_value)
else:
sub_result = retrieveKey(sub_key_value, key_input)
if len(sub_result) != 0:
result.append(sub_result)
elif isinstance(top_key_value, list):
for top_key_value_item in top_key_value:
if isinstance(top_key_value_item, dict):
for sub_key in top_key_value_item:
sub_key_value = top_key_value_item[sub_key]
if sub_key == key_input:
result.append(sub_key_value)
else:
sub_result = retrieveKey(sub_key_value, key_input)
if len(sub_result) != 0:
result.append(sub_result)
else:
result.append(top_key_value_item)
return result
def searchWordInTitle(word):
article_list = []
for files in os.listdir(path='/kaggle/input/CORD-19-research-challenge/biorxiv_medrxiv/biorxiv_medrxiv/'):
article_sha_id = files.split('.')[0]
article_json = retrieveJson(article_sha_id)
title = retrieveKey(article_json, 'title')
if isinstance(title, list):
for item in title:
if isinstance(item, list):
for it in item:
find_result = it.find(word)
if find_result >= 0:
if not article_sha_id in article_list:
article_list.append(article_sha_id)
else:
find_result = item.find(word)
if find_result >= 0:
if not article_sha_id in article_list:
article_list.append(article_sha_id)
else:
find_words = title.find(word)
if find_words >= 0:
if not article_sha_id in article_list:
article_list.append(article_sha_id)
return article_list
def abstract(data):
""" Abstract """
data_item = data[0]
row_entries = data_item[0]
for row in row_entries:
cite_span = row[0]
ref_span = row[1]
section = row[2]
text = row[3]
return (cite_span, ref_span, section, text)
def back_matter(data):
""" Back Matter """
data_item = data[0]
row_entries = data_item[0]
for row in row_entries:
cite_span = row[0]
ref_span = row[1]
section = row[2]
text = row[3]
return (cite_span, ref_span, section, text)
def bib_entries(data):
""" Bib Entries """
data_item = data[0]
row_entries = data_item[0]
for row in row_entries:
cite_span = row[0]
ref_span = row[1]
section = row[2]
text = row[3]
return (cite_span, ref_span, section, text)
def body_text(data):
""" Body Text """
data_item = data[0]
row_entries = data_item[0]
for row in row_entries:
cite_span = row[0]
ref_span = row[1]
section = row[2]
text = row[3]
return (cite_span, ref_span, section, text)
def metadata(data):
""" Metadata """
data_item = data[0]
row_entries = data_item[0]
num_entries = len(row_entries)
i = 0
for row in row_entries:
if i < num_entries - 1:
rowitems = row_entries[i][0]
affiliation = rowitems[0]
institution = affiliation[0]
laboratory = affiliation[1]
location = affiliation[2]
email = rowitems[1]
first = rowitems[2]
last = rowitems[3]
middle = rowitems[4]
suffix = rowitems[5]
else:
title = row
i += 1
return (affiliation, institution, laboratory, location, email, first, last, middle, suffix)
def paper_id(data):
""" Paper Id """
data_item = data[0]
paper_id = data_item[0]
return paper_id
def ref_entries(data):
""" Ref Entries """
data_item = data[0]
row_entries = data_item[0]
for row in row_entries:
latex = row[0]
text = row[1]
type_data = row[2]
return (latex, text, type_data)
column_functions = {0: abstract, 1: back_matter, 2: bib_entries, 3: body_text, 4: metadata, 5: paper_id, 6: ref_entries}
def returnTextualReferences(article_list):
article_list = []
for files in os.listdir(path='/kaggle/input/CORD-19-research-challenge/biorxiv_medrxiv/biorxiv_medrxiv/'):
article_sha_id = files.split('.')[0]
article_path = retrieveArticlePath(article_sha_id)
article_df = spark.read.json(article_path)
article_df.createOrReplaceTempView('article')
column_results = []
column_names = []
for column_name in df.schema.names:
strSQL = 'SELECT ' + column_name + ' from article'
column_result = spark.sql(strSQL)
column_results.append(column_result)
column_names.append(column_name)
i = 0
for column_result in column_results:
column_name = column_names[i]
column_schema = column_result.schema
column_rdd = column_result.rdd
result = column_result.collect()
func = column_functions.get(i)
item_result = func(result)
i += 1
vaccine_in_title = searchWordInTitle('vaccine')
returnTextualReferences(vaccine_in_title) | code |
32068954/cell_2 | [
"text_plain_output_1.png"
] | !pip install pyspark | code |
104129811/cell_9 | [
"text_html_output_1.png"
] | import pandas as pd
import plotly.graph_objects as go
train = pd.read_csv('../input/standup-targets/train.csv')
y0 = train.score.values
fig = go.Figure()
fig.add_trace(go.Box(y=y0, name='Train', marker_color='#1e90ff'))
fig.update_layout(title_text='Score stats (individual shot)')
fig.update_yaxes()
fig.update_xaxes(showticklabels=False)
fig.show() | code |
104129811/cell_4 | [
"text_html_output_2.png"
] | import pandas as pd
train = pd.read_csv('../input/standup-targets/train.csv')
train.head() | code |
104129811/cell_7 | [
"text_html_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/standup-targets/train.csv')
train | code |
104129811/cell_8 | [
"text_html_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/standup-targets/train.csv')
print(f'Number of bullet impacts : {train.shape[0]}')
print(f'Average number of impacts per target : {train.shape[0] / len(train.image_name.unique())}') | code |
104129811/cell_15 | [
"text_plain_output_1.png"
] | import pandas as pd
import plotly.graph_objects as go
train = pd.read_csv('../input/standup-targets/train.csv')
y0 = train.score.values
fig = go.Figure()
fig.add_trace(go.Box(y=y0, name='Train', marker_color='#1e90ff'))
fig.update_layout(title_text='Score stats (individual shot)')
fig.update_yaxes()
fig.update_xaxes(showticklabels=False)
colors = ['#1e90ff'] * 7
colors[0] = '#ff6347'
dict_ = dict(train.score.value_counts())
x = list(dict_.keys())
y = list(dict_.values())
fig = go.Figure(data=[go.Bar(x=x, y=y, marker_color=colors)])
fig.update_layout(title_text='Score distribution (individual shot)')
y0 = list(train.groupby(['image_name'])[['score']].mean().values.ravel())
fig = go.Figure()
fig.add_trace(go.Box(y=y0, name='Train', marker_color='#1e90ff'))
fig.update_layout(title_text='Score stats (per series)')
fig.update_yaxes()
fig.update_xaxes(showticklabels=False)
df_to_merge_on = pd.DataFrame(train.image_name.value_counts())
train = train.merge(df_to_merge_on, right_on=train.image_name.value_counts().index, left_on='image_name').drop(['image_name_x'], axis=1)
train = train.rename(columns={'image_name_y': 'nb_bullets'})
y0 = list(train.loc[train['nb_bullets'] == 1, :].groupby(['image_name'])[['score']].mean().values.ravel())
y1 = list(train.loc[train['nb_bullets'] == 2, :].groupby(['image_name'])[['score']].mean().values.ravel())
y2 = list(train.loc[train['nb_bullets'] == 3, :].groupby(['image_name'])[['score']].mean().values.ravel())
y3 = list(train.loc[train['nb_bullets'] == 4, :].groupby(['image_name'])[['score']].mean().values.ravel())
y4 = list(train.loc[train['nb_bullets'] == 5, :].groupby(['image_name'])[['score']].mean().values.ravel())
y5 = list(train.loc[train['nb_bullets'] == 6, :].groupby(['image_name'])[['score']].mean().values.ravel())
fig = go.Figure()
fig.add_trace(go.Box(y=y0, name='1 bullet series'))
fig.add_trace(go.Box(y=y1, name='2 bullet series', marker_color='blue'))
fig.add_trace(go.Box(y=y2, name='3 bullet series', marker_color='blue'))
fig.add_trace(go.Box(y=y3, name='4 bullet series', marker_color='blue'))
fig.add_trace(go.Box(y=y4, name='5 bullet series', marker_color='blue'))
fig.add_trace(go.Box(y=y5, name='6 bullet series', marker_color='blue'))
fig.update_layout(legend_title_text='Discourse effectiveness', title_text='Mean size of sentences in discourse (logscale)')
fig.update_xaxes(showticklabels=False)
fig.show() | code |
104129811/cell_10 | [
"text_html_output_1.png"
] | import pandas as pd
import plotly.graph_objects as go
train = pd.read_csv('../input/standup-targets/train.csv')
y0 = train.score.values
fig = go.Figure()
fig.add_trace(go.Box(y=y0, name='Train', marker_color='#1e90ff'))
fig.update_layout(title_text='Score stats (individual shot)')
fig.update_yaxes()
fig.update_xaxes(showticklabels=False)
colors = ['#1e90ff'] * 7
colors[0] = '#ff6347'
dict_ = dict(train.score.value_counts())
x = list(dict_.keys())
y = list(dict_.values())
fig = go.Figure(data=[go.Bar(x=x, y=y, marker_color=colors)])
fig.update_layout(title_text='Score distribution (individual shot)') | code |
104129811/cell_12 | [
"text_html_output_1.png"
] | import pandas as pd
import plotly.graph_objects as go
train = pd.read_csv('../input/standup-targets/train.csv')
y0 = train.score.values
fig = go.Figure()
fig.add_trace(go.Box(y=y0, name='Train', marker_color='#1e90ff'))
fig.update_layout(title_text='Score stats (individual shot)')
fig.update_yaxes()
fig.update_xaxes(showticklabels=False)
colors = ['#1e90ff'] * 7
colors[0] = '#ff6347'
dict_ = dict(train.score.value_counts())
x = list(dict_.keys())
y = list(dict_.values())
fig = go.Figure(data=[go.Bar(x=x, y=y, marker_color=colors)])
fig.update_layout(title_text='Score distribution (individual shot)')
y0 = list(train.groupby(['image_name'])[['score']].mean().values.ravel())
fig = go.Figure()
fig.add_trace(go.Box(y=y0, name='Train', marker_color='#1e90ff'))
fig.update_layout(title_text='Score stats (per series)')
fig.update_yaxes()
fig.update_xaxes(showticklabels=False)
fig.show() | code |
74052264/cell_4 | [
"text_plain_output_1.png"
] | import pandas as pd
df = pd.read_csv('../input/covidqa/community.csv')
df.head() | code |
74052264/cell_6 | [
"text_plain_output_4.png",
"text_plain_output_3.png",
"text_plain_output_2.png",
"text_plain_output_1.png"
] | from sklearn.feature_extraction.text import TfidfVectorizer
import numpy as np
import pandas as pd
df = pd.read_csv('../input/covidqa/community.csv')
vectorizer = TfidfVectorizer()
vectorizer.fit(np.concatenate((df.question, df.answer))) | code |
74052264/cell_11 | [
"text_html_output_1.png"
] | from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
import numpy as np
import pandas as pd
df = pd.read_csv('../input/covidqa/community.csv')
vectorizer = TfidfVectorizer()
vectorizer.fit(np.concatenate((df.question, df.answer)))
Question_vectors = vectorizer.transform(df.question)
name = input('Enter your name : ')
print(f'BOT : Hello {name}! How can I help you ?')
while True:
input_question = input(f'{name} : ')
if input_question.lower() == 'bye':
print(f'BOT : Bye! Have a nice day and maintain proper norms and regulations to stop the spread of COVID.')
break
input_question_vector = vectorizer.transform([input_question])
similarities = cosine_similarity(input_question_vector, Question_vectors)
closest = np.argmax(similarities, axis=1)
print(f'BOT : {df.answer.iloc[closest].values[0]}') | code |
2039183/cell_9 | [
"application_vnd.jupyter.stderr_output_1.png"
] | from gensim.models import Phrases
from gensim.models import word2vec
from gensim.models.phrases import Phraser
from nltk.corpus import stopwords
import logging
import pandas as pd
import pickle
import pickle
import re
import pandas as pd
import re
from nltk.corpus import stopwords
from gensim.models import word2vec
import pickle
import nltk.data
import os
tokenizer = nltk.data.load('tokenizers/punkt/english.pickle')
path = '../input/'
TRAIN_DATA_FILE = f'{path}train.csv'
TEST_DATA_FILE = f'{path}test.csv'
train = pd.read_csv(TRAIN_DATA_FILE, header=0)
test = pd.read_csv(TEST_DATA_FILE, header=0)
all_comments = train['comment_text'].fillna('_na_').tolist() + test['comment_text'].fillna('_na_').tolist()
with open('all_comments.csv', 'w+') as comments_file:
i = 0
for comment in all_comments:
comment = re.sub('[^a-zA-Z]', ' ', str(comment))
comments_file.write('%s\n' % comment)
class FileToComments(object):
def __init__(self, filename):
self.filename = filename
self.stop = set(nltk.corpus.stopwords.words('english'))
def __iter__(self):
def comment_to_wordlist(comment, remove_stopwords=True):
comment = str(comment)
words = comment.lower().split()
return words
for line in open(self.filename, 'r'):
tokenized_comment = comment_to_wordlist(line, tokenizer)
yield tokenized_comment
all_comments = FileToComments('all_comments.csv')
from gensim.models import Phrases
from gensim.models.phrases import Phraser
bigram = Phrases(all_comments, min_count=30, threshold=15)
bigram_phraser = Phraser(bigram)
all_tokens = [bigram_phraser[comment] for comment in all_comments]
stops = set(stopwords.words('english'))
clean_all_tokens = []
for token in all_tokens:
words = [w for w in token if not w in stops]
clean_all_tokens += [words]
import pickle
with open('tokenized_all_comments.pickle', 'wb') as filename:
pickle.dump(clean_all_tokens, filename, protocol=pickle.HIGHEST_PROTOCOL)
with open('data/tokenized_comments/tokenized_all_comments.pickle', 'rb') as filename:
all_comments = pickle.load(filename)
import logging
logging.basicConfig(format='%(asctime)s : %(levelname)s : %(message)s', level=logging.INFO)
num_features = 300
min_word_count = 20
num_workers = 16
context = 10
downsampling = 0.001
print('Training model...')
model = word2vec.Word2Vec(all_comments, workers=num_workers, size=num_features, min_count=min_word_count, window=context, sample=downsampling)
model.init_sims(replace=True)
model_name = 'models/%sfeatures_%sminwords_%scontext' % (num_features, min_word_count, context)
model.save(model_name) | code |
2039183/cell_6 | [
"text_plain_output_1.png"
] | from gensim.models import Phrases
from gensim.models.phrases import Phraser
from nltk.corpus import stopwords
import pandas as pd
import re
import pandas as pd
import re
from nltk.corpus import stopwords
from gensim.models import word2vec
import pickle
import nltk.data
import os
tokenizer = nltk.data.load('tokenizers/punkt/english.pickle')
path = '../input/'
TRAIN_DATA_FILE = f'{path}train.csv'
TEST_DATA_FILE = f'{path}test.csv'
train = pd.read_csv(TRAIN_DATA_FILE, header=0)
test = pd.read_csv(TEST_DATA_FILE, header=0)
all_comments = train['comment_text'].fillna('_na_').tolist() + test['comment_text'].fillna('_na_').tolist()
with open('all_comments.csv', 'w+') as comments_file:
i = 0
for comment in all_comments:
comment = re.sub('[^a-zA-Z]', ' ', str(comment))
comments_file.write('%s\n' % comment)
class FileToComments(object):
def __init__(self, filename):
self.filename = filename
self.stop = set(nltk.corpus.stopwords.words('english'))
def __iter__(self):
def comment_to_wordlist(comment, remove_stopwords=True):
comment = str(comment)
words = comment.lower().split()
return words
for line in open(self.filename, 'r'):
tokenized_comment = comment_to_wordlist(line, tokenizer)
yield tokenized_comment
all_comments = FileToComments('all_comments.csv')
from gensim.models import Phrases
from gensim.models.phrases import Phraser
bigram = Phrases(all_comments, min_count=30, threshold=15)
bigram_phraser = Phraser(bigram)
all_tokens = [bigram_phraser[comment] for comment in all_comments]
stops = set(stopwords.words('english'))
clean_all_tokens = []
for token in all_tokens:
words = [w for w in token if not w in stops]
clean_all_tokens += [words]
print('tokens cleaned') | code |
2039183/cell_2 | [
"text_plain_output_1.png"
] | import pandas as pd
import re
path = '../input/'
TRAIN_DATA_FILE = f'{path}train.csv'
TEST_DATA_FILE = f'{path}test.csv'
train = pd.read_csv(TRAIN_DATA_FILE, header=0)
test = pd.read_csv(TEST_DATA_FILE, header=0)
print('Read %d labeled train reviews and %d unlabelled test reviews' % (len(train), len(test)))
all_comments = train['comment_text'].fillna('_na_').tolist() + test['comment_text'].fillna('_na_').tolist()
with open('all_comments.csv', 'w+') as comments_file:
i = 0
for comment in all_comments:
comment = re.sub('[^a-zA-Z]', ' ', str(comment))
comments_file.write('%s\n' % comment) | code |
2039183/cell_7 | [
"text_plain_output_1.png"
] | from gensim.models import Phrases
from gensim.models.phrases import Phraser
from nltk.corpus import stopwords
import pandas as pd
import pickle
import pickle
import re
import pandas as pd
import re
from nltk.corpus import stopwords
from gensim.models import word2vec
import pickle
import nltk.data
import os
tokenizer = nltk.data.load('tokenizers/punkt/english.pickle')
path = '../input/'
TRAIN_DATA_FILE = f'{path}train.csv'
TEST_DATA_FILE = f'{path}test.csv'
train = pd.read_csv(TRAIN_DATA_FILE, header=0)
test = pd.read_csv(TEST_DATA_FILE, header=0)
all_comments = train['comment_text'].fillna('_na_').tolist() + test['comment_text'].fillna('_na_').tolist()
with open('all_comments.csv', 'w+') as comments_file:
i = 0
for comment in all_comments:
comment = re.sub('[^a-zA-Z]', ' ', str(comment))
comments_file.write('%s\n' % comment)
class FileToComments(object):
def __init__(self, filename):
self.filename = filename
self.stop = set(nltk.corpus.stopwords.words('english'))
def __iter__(self):
def comment_to_wordlist(comment, remove_stopwords=True):
comment = str(comment)
words = comment.lower().split()
return words
for line in open(self.filename, 'r'):
tokenized_comment = comment_to_wordlist(line, tokenizer)
yield tokenized_comment
all_comments = FileToComments('all_comments.csv')
from gensim.models import Phrases
from gensim.models.phrases import Phraser
bigram = Phrases(all_comments, min_count=30, threshold=15)
bigram_phraser = Phraser(bigram)
all_tokens = [bigram_phraser[comment] for comment in all_comments]
stops = set(stopwords.words('english'))
clean_all_tokens = []
for token in all_tokens:
words = [w for w in token if not w in stops]
clean_all_tokens += [words]
import pickle
with open('tokenized_all_comments.pickle', 'wb') as filename:
pickle.dump(clean_all_tokens, filename, protocol=pickle.HIGHEST_PROTOCOL)
print('files saved to tokenized_all_comments.pickle...') | code |
122259592/cell_4 | [
"application_vnd.jupyter.stderr_output_1.png"
] | import pandas as pd
sample_sub = pd.read_csv('sample_submission.csv')
test_data = pd.read_csv('test_dataset.csv', index_col=0)
train_data = pd.read_csv('train_dataset.csv', index_col=0)
sample_sub.head() | code |
48162572/cell_9 | [
"text_html_output_1.png"
] | import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
submission.head() | code |
48162572/cell_34 | [
"text_plain_output_1.png"
] | import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
ALL.columns | code |
48162572/cell_23 | [
"text_html_output_1.png"
] | import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape | code |
48162572/cell_30 | [
"image_output_1.png"
] | import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
ALL.head() | code |
48162572/cell_44 | [
"text_plain_output_1.png"
] | from sklearn.decomposition import PCA
import pandas as pd
import statistics
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
def standardization(l):
l_mean = statistics.mean(l)
l_stdev = statistics.stdev(l)
ret = []
for x in l:
y = (x - l_mean) / l_stdev
ret.append(y)
return ret
ALL.columns
GENES = [col for col in train_features.columns if col.startswith('g-')]
CELLS = [col for col in train_features.columns if col.startswith('c-')]
ADD_CNT = 12
for x in GENES:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in CELLS:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in ALL:
if x in ALL.columns[-ADD_CNT:]:
ALL[x] = standardization(ALL[x])
ALL = ALL.drop(['sig_id'], axis=1)
from sklearn.decomposition import PCA
pca = PCA(150)
ALL = pca.fit_transform(ALL)
Columns = []
for i in range(150):
Columns.append('d' + str(i + 1))
ALL = pd.DataFrame(ALL, columns=Columns)
categorical_features = Columns
ALL.head() | code |
48162572/cell_6 | [
"text_plain_output_1.png"
] | import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
train_features.head() | code |
48162572/cell_39 | [
"text_html_output_1.png"
] | import pandas as pd
import statistics
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
def standardization(l):
l_mean = statistics.mean(l)
l_stdev = statistics.stdev(l)
ret = []
for x in l:
y = (x - l_mean) / l_stdev
ret.append(y)
return ret
ALL.columns
GENES = [col for col in train_features.columns if col.startswith('g-')]
CELLS = [col for col in train_features.columns if col.startswith('c-')]
ADD_CNT = 12
for x in GENES:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in CELLS:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in ALL:
if x in ALL.columns[-ADD_CNT:]:
ALL[x] = standardization(ALL[x])
ALL.describe() | code |
48162572/cell_41 | [
"image_output_1.png"
] | import pandas as pd
import statistics
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
def standardization(l):
l_mean = statistics.mean(l)
l_stdev = statistics.stdev(l)
ret = []
for x in l:
y = (x - l_mean) / l_stdev
ret.append(y)
return ret
ALL.columns
GENES = [col for col in train_features.columns if col.startswith('g-')]
CELLS = [col for col in train_features.columns if col.startswith('c-')]
ADD_CNT = 12
for x in GENES:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in CELLS:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in ALL:
if x in ALL.columns[-ADD_CNT:]:
ALL[x] = standardization(ALL[x])
ALL = ALL.drop(['sig_id'], axis=1)
ALL.head() | code |
48162572/cell_19 | [
"text_html_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
plt.figure(figsize=(8, 6))
plt.subplot(2, 2, 1)
plt.hist(test_features['g-0'])
plt.subplot(2, 2, 2)
plt.hist(test_features['c-0'])
plt.subplot(2, 2, 3)
plt.hist(test_features['g-178'])
plt.subplot(2, 2, 4)
plt.hist(test_features['c-32'])
plt.show() | code |
48162572/cell_52 | [
"text_html_output_1.png"
] | x_train.head() | code |
48162572/cell_7 | [
"text_html_output_1.png"
] | import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
train_targets.head() | code |
48162572/cell_45 | [
"text_plain_output_1.png"
] | from sklearn.decomposition import PCA
import pandas as pd
import statistics
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
def standardization(l):
l_mean = statistics.mean(l)
l_stdev = statistics.stdev(l)
ret = []
for x in l:
y = (x - l_mean) / l_stdev
ret.append(y)
return ret
ALL.columns
GENES = [col for col in train_features.columns if col.startswith('g-')]
CELLS = [col for col in train_features.columns if col.startswith('c-')]
ADD_CNT = 12
for x in GENES:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in CELLS:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in ALL:
if x in ALL.columns[-ADD_CNT:]:
ALL[x] = standardization(ALL[x])
ALL = ALL.drop(['sig_id'], axis=1)
from sklearn.decomposition import PCA
pca = PCA(150)
ALL = pca.fit_transform(ALL)
Columns = []
for i in range(150):
Columns.append('d' + str(i + 1))
ALL = pd.DataFrame(ALL, columns=Columns)
categorical_features = Columns
pca.explained_variance_ratio_ | code |
48162572/cell_49 | [
"text_html_output_1.png"
] | from sklearn.decomposition import PCA
import pandas as pd
import statistics
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
def standardization(l):
l_mean = statistics.mean(l)
l_stdev = statistics.stdev(l)
ret = []
for x in l:
y = (x - l_mean) / l_stdev
ret.append(y)
return ret
ALL.columns
GENES = [col for col in train_features.columns if col.startswith('g-')]
CELLS = [col for col in train_features.columns if col.startswith('c-')]
ADD_CNT = 12
for x in GENES:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in CELLS:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in ALL:
if x in ALL.columns[-ADD_CNT:]:
ALL[x] = standardization(ALL[x])
ALL = ALL.drop(['sig_id'], axis=1)
from sklearn.decomposition import PCA
pca = PCA(150)
ALL = pca.fit_transform(ALL)
Columns = []
for i in range(150):
Columns.append('d' + str(i + 1))
ALL = pd.DataFrame(ALL, columns=Columns)
categorical_features = Columns
ALL.head() | code |
48162572/cell_18 | [
"text_html_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
plt.figure(figsize=(8, 6))
plt.subplot(2, 2, 1)
plt.hist(train_features['g-0'])
plt.subplot(2, 2, 2)
plt.hist(train_features['c-0'])
plt.subplot(2, 2, 3)
plt.hist(train_features['g-178'])
plt.subplot(2, 2, 4)
plt.hist(train_features['c-32'])
plt.show() | code |
48162572/cell_8 | [
"text_html_output_1.png"
] | import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
test_features.head() | code |
48162572/cell_15 | [
"text_html_output_1.png"
] | import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
test_features.describe() | code |
48162572/cell_31 | [
"image_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
plt.figure(figsize=(8, 6))
plt.subplot(2, 2, 1)
plt.hist(ALL['g-0'])
plt.subplot(2, 2, 2)
plt.hist(ALL['c-0'])
plt.subplot(2, 2, 3)
plt.hist(ALL['g-178'])
plt.subplot(2, 2, 4)
plt.hist(ALL['c-32'])
plt.show() | code |
48162572/cell_46 | [
"text_html_output_1.png"
] | from sklearn.decomposition import PCA
import pandas as pd
import statistics
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
def standardization(l):
l_mean = statistics.mean(l)
l_stdev = statistics.stdev(l)
ret = []
for x in l:
y = (x - l_mean) / l_stdev
ret.append(y)
return ret
ALL.columns
GENES = [col for col in train_features.columns if col.startswith('g-')]
CELLS = [col for col in train_features.columns if col.startswith('c-')]
ADD_CNT = 12
for x in GENES:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in CELLS:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
for x in ALL:
if x in ALL.columns[-ADD_CNT:]:
ALL[x] = standardization(ALL[x])
ALL = ALL.drop(['sig_id'], axis=1)
from sklearn.decomposition import PCA
pca = PCA(150)
ALL = pca.fit_transform(ALL)
Columns = []
for i in range(150):
Columns.append('d' + str(i + 1))
ALL = pd.DataFrame(ALL, columns=Columns)
categorical_features = Columns
pca.explained_variance_ratio_
pca.explained_variance_ratio_.cumsum() | code |
48162572/cell_24 | [
"text_html_output_1.png"
] | import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
ALL.head() | code |
48162572/cell_14 | [
"text_html_output_1.png"
] | import pandas as pd
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
train_features.describe() | code |
48162572/cell_37 | [
"text_html_output_1.png"
] | import pandas as pd
import statistics
train_features = pd.read_csv('/kaggle/input/lish-moa/train_features.csv')
train_targets = pd.read_csv('/kaggle/input/lish-moa/train_targets_scored.csv')
test_features = pd.read_csv('/kaggle/input/lish-moa/test_features.csv')
submission = pd.read_csv('/kaggle/input/lish-moa/sample_submission.csv')
ALL = pd.concat([train_features, test_features])
ALL.shape
def standardization(l):
l_mean = statistics.mean(l)
l_stdev = statistics.stdev(l)
ret = []
for x in l:
y = (x - l_mean) / l_stdev
ret.append(y)
return ret
ALL.columns
GENES = [col for col in train_features.columns if col.startswith('g-')]
CELLS = [col for col in train_features.columns if col.startswith('c-')]
ADD_CNT = 12
for x in GENES:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1
print('OK')
for x in CELLS:
ALL[x + 'High'] = (ALL[x] ** 2) ** 0.5
MAX = ALL[x + 'High'].max()
for y in ALL[x + 'High']:
if y <= MAX / 2:
y = 0
else:
y = y - MAX / 2
ALL[x + 'High'] = ALL[x + 'High'] * (abs(ALL[x]) / ALL[x])
ADD_CNT += 1 | code |
88092902/cell_9 | [
"text_html_output_4.png",
"text_html_output_2.png",
"text_html_output_1.png",
"text_plain_output_1.png",
"text_html_output_3.png"
] | from cuml.svm import SVR
import matplotlib.pyplot as plt
import pandas as pd
import random
data_types_dict = {'time_id': 'int16', 'investment_id': 'int16', 'target': 'float32'}
features = [f'f_{i}' for i in range(300)]
for f in features:
data_types_dict[f] = 'float32'
train = pd.read_csv('../input/ubiquant-market-prediction/train.csv', usecols=data_types_dict.keys(), dtype=data_types_dict)
N_DEVIDE_DATA = 20
n_row = len(train)
idx_list = list(range(n_row))
random.shuffle(idx_list)
models = []
for i in range(N_DEVIDE_DATA - 1):
start_idx = int(n_row / N_DEVIDE_DATA) * i
end_idx = int(n_row / N_DEVIDE_DATA) * (i + 1)
devide_idx_list = idx_list[start_idx:end_idx]
tr = train.iloc[devide_idx_list]
X = tr[features].to_numpy()
y = tr['target'].to_numpy()
model = SVR(C=5.0, kernel='rbf', epsilon=0.1)
model.fit(X, y)
r2 = model.score(X, y)
models.append(model)
start_idx = int(n_row / N_DEVIDE_DATA) * 19
devide_idx_list = idx_list[start_idx:]
test = train.iloc[devide_idx_list]
X_test = test[features].to_numpy()
y_test = test['target'].to_numpy()
pre_y = 0
for model in models:
pre_y += model.predict(X_test)
pre_y /= len(models)
plt.scatter(y_test, pre_y) | code |
88092902/cell_6 | [
"text_plain_output_1.png"
] | from cuml.svm import SVR
import pandas as pd
import random
data_types_dict = {'time_id': 'int16', 'investment_id': 'int16', 'target': 'float32'}
features = [f'f_{i}' for i in range(300)]
for f in features:
data_types_dict[f] = 'float32'
train = pd.read_csv('../input/ubiquant-market-prediction/train.csv', usecols=data_types_dict.keys(), dtype=data_types_dict)
N_DEVIDE_DATA = 20
n_row = len(train)
idx_list = list(range(n_row))
random.shuffle(idx_list)
models = []
for i in range(N_DEVIDE_DATA - 1):
start_idx = int(n_row / N_DEVIDE_DATA) * i
end_idx = int(n_row / N_DEVIDE_DATA) * (i + 1)
devide_idx_list = idx_list[start_idx:end_idx]
tr = train.iloc[devide_idx_list]
X = tr[features].to_numpy()
y = tr['target'].to_numpy()
model = SVR(C=5.0, kernel='rbf', epsilon=0.1)
model.fit(X, y)
r2 = model.score(X, y)
print(i, 'R^2:', r2)
models.append(model) | code |
88092902/cell_16 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from cuml.svm import SVR
import pandas as pd
import pickle
import random
import ubiquant
data_types_dict = {'time_id': 'int16', 'investment_id': 'int16', 'target': 'float32'}
features = [f'f_{i}' for i in range(300)]
for f in features:
data_types_dict[f] = 'float32'
train = pd.read_csv('../input/ubiquant-market-prediction/train.csv', usecols=data_types_dict.keys(), dtype=data_types_dict)
N_DEVIDE_DATA = 20
n_row = len(train)
idx_list = list(range(n_row))
random.shuffle(idx_list)
models = []
for i in range(N_DEVIDE_DATA - 1):
start_idx = int(n_row / N_DEVIDE_DATA) * i
end_idx = int(n_row / N_DEVIDE_DATA) * (i + 1)
devide_idx_list = idx_list[start_idx:end_idx]
tr = train.iloc[devide_idx_list]
X = tr[features].to_numpy()
y = tr['target'].to_numpy()
model = SVR(C=5.0, kernel='rbf', epsilon=0.1)
model.fit(X, y)
r2 = model.score(X, y)
models.append(model)
start_idx = int(n_row / N_DEVIDE_DATA) * 19
devide_idx_list = idx_list[start_idx:]
test = train.iloc[devide_idx_list]
X_test = test[features].to_numpy()
y_test = test['target'].to_numpy()
pre_y = 0
for model in models:
pre_y += model.predict(X_test)
pre_y /= len(models)
for i, model in enumerate(models):
filename = 'model_svr_{}.sav'.format(i)
pickle.dump(model, open(filename, 'wb'))
models = []
for i in range(19):
filename = 'model_svr_{}.sav'.format(i)
loaded_model = pickle.load(open(filename, 'rb'))
models.append(loaded_model)
import ubiquant
env = ubiquant.make_env()
iter_test = env.iter_test()
for test_df, sample_prediction_df in iter_test:
test_x = test_df[features].to_numpy()
for loaded_model in models:
sample_prediction_df['target'] += loaded_model.predict(test_x)
sample_prediction_df['target'] /= len(models) + 1
env.predict(sample_prediction_df)
display(sample_prediction_df) | code |
88092902/cell_10 | [
"text_plain_output_1.png"
] | from cuml.svm import SVR
import numpy as np
import pandas as pd
import random
data_types_dict = {'time_id': 'int16', 'investment_id': 'int16', 'target': 'float32'}
features = [f'f_{i}' for i in range(300)]
for f in features:
data_types_dict[f] = 'float32'
train = pd.read_csv('../input/ubiquant-market-prediction/train.csv', usecols=data_types_dict.keys(), dtype=data_types_dict)
N_DEVIDE_DATA = 20
n_row = len(train)
idx_list = list(range(n_row))
random.shuffle(idx_list)
models = []
for i in range(N_DEVIDE_DATA - 1):
start_idx = int(n_row / N_DEVIDE_DATA) * i
end_idx = int(n_row / N_DEVIDE_DATA) * (i + 1)
devide_idx_list = idx_list[start_idx:end_idx]
tr = train.iloc[devide_idx_list]
X = tr[features].to_numpy()
y = tr['target'].to_numpy()
model = SVR(C=5.0, kernel='rbf', epsilon=0.1)
model.fit(X, y)
r2 = model.score(X, y)
models.append(model)
start_idx = int(n_row / N_DEVIDE_DATA) * 19
devide_idx_list = idx_list[start_idx:]
test = train.iloc[devide_idx_list]
X_test = test[features].to_numpy()
y_test = test['target'].to_numpy()
pre_y = 0
for model in models:
pre_y += model.predict(X_test)
pre_y /= len(models)
np.corrcoef(y_test.tolist(), pre_y.tolist()) | code |
128005771/cell_2 | [
"application_vnd.jupyter.stderr_output_1.png"
] | import pandas as pd
df = pd.read_csv('diabetes_prediction_dataset.csv')
df.head() | code |
32064989/cell_21 | [
"text_html_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
sum(model['errors0']) | code |
32064989/cell_9 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import matplotlib.pyplot as plt #for visualizing
import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
plt.scatter(x='YearsExperience', y='Salary', data=Salary) | code |
32064989/cell_25 | [
"text_plain_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
sum(model['errors2']) | code |
32064989/cell_34 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LinearRegression
lr_model = LinearRegression()
lr_model.fit(x_train, y_train)
lr_model.intercept_ | code |
32064989/cell_23 | [
"text_plain_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
sum(model['errors1']) | code |
32064989/cell_30 | [
"text_plain_output_1.png"
] | print(x_train.shape, y_train.shape) | code |
32064989/cell_33 | [
"text_html_output_1.png"
] | from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LinearRegression
lr_model = LinearRegression()
lr_model.fit(x_train, y_train) | code |
32064989/cell_20 | [
"text_html_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
model | code |
32064989/cell_40 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt #for visualizing
import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
fig,ax = plt.subplots()
ax.scatter(x='YearsExperience',y='Salary',data=Salary)
ax.add_line(plt.Line2D(model['YearsExperience'],model.predicted0,color='red'))
ax.add_line(plt.Line2D(model['YearsExperience'],model.predicted1,color='Black'))
ax.add_line(plt.Line2D(model['YearsExperience'],model.predicted2,color='Green'))
x = Salary.loc[:, ['YearsExperience']]
y = Salary.loc[:, ['Salary']]
fig, ax = plt.subplots()
ax.scatter(x='YearsExperience', y='Salary', data=Salary)
ax.add_line(plt.Line2D(model['YearsExperience'], model.predicted3, color='red'))
ax.add_line(plt.Line2D(model['YearsExperience'], model.predicted0, color='Green'))
ax.add_line(plt.Line2D(model['YearsExperience'], model.predicted2, color='Black')) | code |
32064989/cell_39 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt #for visualizing
import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
fig,ax = plt.subplots()
ax.scatter(x='YearsExperience',y='Salary',data=Salary)
ax.add_line(plt.Line2D(model['YearsExperience'],model.predicted0,color='red'))
ax.add_line(plt.Line2D(model['YearsExperience'],model.predicted1,color='Black'))
ax.add_line(plt.Line2D(model['YearsExperience'],model.predicted2,color='Green'))
model | code |
32064989/cell_26 | [
"text_plain_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
sum(model['errors2'] ** 2) | code |
32064989/cell_2 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename)) | code |
32064989/cell_7 | [
"text_html_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
Salary.info()
Salary.describe() | code |
32064989/cell_32 | [
"text_plain_output_1.png"
] | print(x_test.shape, y_test.shape) | code |
32064989/cell_15 | [
"text_html_output_1.png",
"text_plain_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
model | code |
32064989/cell_35 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LinearRegression
lr_model = LinearRegression()
lr_model.fit(x_train, y_train)
lr_model.intercept_
lr_model.coef_ | code |
32064989/cell_31 | [
"text_plain_output_1.png",
"image_output_1.png"
] | x_train | code |
32064989/cell_24 | [
"text_plain_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
sum(model['errors1'] ** 2) | code |
32064989/cell_22 | [
"text_html_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
sum(model['errors0'] ** 2) | code |
32064989/cell_27 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt #for visualizing
import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
fig, ax = plt.subplots()
ax.scatter(x='YearsExperience', y='Salary', data=Salary)
ax.add_line(plt.Line2D(model['YearsExperience'], model.predicted0, color='red'))
ax.add_line(plt.Line2D(model['YearsExperience'], model.predicted1, color='Black'))
ax.add_line(plt.Line2D(model['YearsExperience'], model.predicted2, color='Green')) | code |
32064989/cell_37 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LinearRegression
lr_model = LinearRegression()
lr_model.fit(x_train, y_train)
lr_model.intercept_
lr_model.coef_
lr_model.score(x_train, y_train)
lr_model.score(x_test, y_test) | code |
32064989/cell_12 | [
"text_plain_output_1.png"
] | import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
pd.options.display.float_format = '{:20,.2f}'.format
import os
Salary = pd.read_csv('../input/salary/Salary.csv')
model = pd.read_csv('../input/salary/Salary.csv')
model | code |
32064989/cell_36 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LinearRegression
lr_model = LinearRegression()
lr_model.fit(x_train, y_train)
lr_model.intercept_
lr_model.coef_
lr_model.score(x_train, y_train) | code |
130025335/cell_25 | [
"image_output_1.png"
] | from tensorflow.keras.utils import load_img
import cv2
import hashlib
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df)
RESIZED_IMAGE_PATH = '../input/resized-plant2021/img_sz_384/'
train_df.labels.value_counts()
label_counts = train_df['labels'].value_counts()
colormap = plt.cm.get_cmap('tab10')
num_labels = len(label_counts)
colors = [colormap(i) for i in range(num_labels)]
plt.xticks(rotation=45)
for i, count in enumerate(label_counts):
plt.text(i, count + 50, str(count), ha='center')
total_count = label_counts.sum()
plt.text(0.5, 1.08, f'Total Count: {total_count}', transform=plt.gca().transAxes, ha='center', fontsize=12, fontweight='bold')
max_count = max(label_counts)
y_limit = max_count + max_count * 0.1
plt.ylim(top=y_limit)
initial_length = len(train_df)
hashes = {}
duplicates = []
originals = []
for index, row in train_df.iterrows():
filename = row['image']
with open(os.path.join(RESIZED_IMAGE_PATH, filename), 'rb') as f:
hash = hashlib.md5(f.read()).hexdigest()
if hash in hashes:
duplicates.append(filename)
originals.append(hashes[hash])
train_df.drop(index, inplace=True)
else:
hashes[hash] = filename
# List of image filenames
image_filenames = [ originals[0], duplicates[0],
originals[1], duplicates[1],
originals[2], duplicates[2]]
# Create a figure with a 3x2 grid of subplots
fig, axes = plt.subplots(3, 2, figsize=(10, 8))
# Iterate over the image filenames and corresponding subplots
for i, (image_filename, ax) in enumerate(zip(image_filenames, axes.flatten())):
# Load the image
image = plt.imread(format_tpu_path(image_filename))
# Show the image in the subplot
ax.imshow(image)
ax.axis('off')
# Set subplot title
if i % 2 == 0:
title = 'Original'
else:
title = 'Duplicate'
ax.set_title(title)
# Adjust the layout of subplots to avoid overlapping
plt.tight_layout()
# Display the figure
plt.show()
train_df.labels.value_counts()
label_counts = train_df['labels'].value_counts()
colormap = plt.cm.get_cmap('tab10')
num_labels = len(label_counts)
colors = [colormap(i) for i in range(num_labels)]
plt.figure(figsize=(10, 6))
plt.bar(label_counts.index, label_counts.values, color=colors)
plt.xlabel('Labels')
plt.ylabel('Count')
plt.title('Label Distribution')
plt.xticks(rotation=45)
for i, count in enumerate(label_counts):
plt.text(i, count + 50, str(count), ha='center')
total_count = label_counts.sum()
plt.axhline(total_count, color='black', linestyle='--', alpha=0.5)
plt.text(0.5, 1.08, f'Total Count: {total_count}', transform=plt.gca().transAxes, ha='center', fontsize=12, fontweight='bold')
max_count = max(label_counts)
y_limit = max_count + max_count * 0.1
plt.ylim(top=y_limit)
plt.show() | code |
130025335/cell_30 | [
"text_html_output_1.png"
] | from sklearn.preprocessing import MultiLabelBinarizer
from tensorflow.keras.utils import load_img
import cv2
import hashlib
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df)
RESIZED_IMAGE_PATH = '../input/resized-plant2021/img_sz_384/'
train_df.labels.value_counts()
initial_length = len(train_df)
hashes = {}
duplicates = []
originals = []
for index, row in train_df.iterrows():
filename = row['image']
with open(os.path.join(RESIZED_IMAGE_PATH, filename), 'rb') as f:
hash = hashlib.md5(f.read()).hexdigest()
if hash in hashes:
duplicates.append(filename)
originals.append(hashes[hash])
train_df.drop(index, inplace=True)
else:
hashes[hash] = filename
train_df.labels.value_counts()
labels = train_df['labels'].tolist()
unique_labels = set()
for label in labels:
unique_labels.update(label.split())
common_labels = [label[0] for label in pd.Series(labels).str.split(expand=True).stack().value_counts()[:6].items()]
mlb = MultiLabelBinarizer(classes=common_labels)
label_matrix = mlb.fit_transform(train_df['labels'].str.split())
label_df = pd.DataFrame(label_matrix, columns=common_labels)
train_df.reset_index(drop=True, inplace=True)
label_df.reset_index(drop=True, inplace=True)
new_df = pd.concat([train_df, label_df], axis=1)
train_df = new_df.drop('labels', axis=1)
train_df.head() | code |
130025335/cell_20 | [
"text_plain_output_1.png"
] | from tensorflow.keras.utils import load_img
import cv2
import hashlib
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df)
RESIZED_IMAGE_PATH = '../input/resized-plant2021/img_sz_384/'
train_df.labels.value_counts()
initial_length = len(train_df)
hashes = {}
duplicates = []
originals = []
for index, row in train_df.iterrows():
filename = row['image']
with open(os.path.join(RESIZED_IMAGE_PATH, filename), 'rb') as f:
hash = hashlib.md5(f.read()).hexdigest()
if hash in hashes:
duplicates.append(filename)
originals.append(hashes[hash])
train_df.drop(index, inplace=True)
else:
hashes[hash] = filename
print(f'Number of duplicates found: {initial_length - len(hashes)}') | code |
130025335/cell_29 | [
"image_output_1.png"
] | from tensorflow.keras.utils import load_img
import cv2
import hashlib
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df)
RESIZED_IMAGE_PATH = '../input/resized-plant2021/img_sz_384/'
train_df.labels.value_counts()
initial_length = len(train_df)
hashes = {}
duplicates = []
originals = []
for index, row in train_df.iterrows():
filename = row['image']
with open(os.path.join(RESIZED_IMAGE_PATH, filename), 'rb') as f:
hash = hashlib.md5(f.read()).hexdigest()
if hash in hashes:
duplicates.append(filename)
originals.append(hashes[hash])
train_df.drop(index, inplace=True)
else:
hashes[hash] = filename
train_df.labels.value_counts()
train_df | code |
130025335/cell_11 | [
"text_plain_output_1.png"
] | from tensorflow.keras.utils import load_img
import cv2
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df) | code |
130025335/cell_1 | [
"text_plain_output_2.png",
"application_vnd.jupyter.stderr_output_1.png"
] | import numpy as np
import pandas as pd
import random
from tensorflow.keras.utils import load_img
import matplotlib.pyplot as plt
import glob as gb
from kaggle_datasets import KaggleDatasets
!pip install -q efficientnet
import efficientnet.tfkeras as efn
import tensorflow as tf
from sklearn.utils import shuffle
from sklearn.model_selection import train_test_split
from tensorflow.keras.models import Sequential
from tensorflow.keras.applications import EfficientNetB7
from tensorflow.keras.applications import EfficientNetB4
from tensorflow.keras.applications import EfficientNetB0
from tensorflow.keras.callbacks import EarlyStopping, ReduceLROnPlateau, TensorBoard, ModelCheckpoint
from tensorflow.keras.utils import plot_model
from IPython.display import SVG, Image
import cv2
from sklearn.preprocessing import MultiLabelBinarizer
import os
import hashlib
from PIL import Image
import albumentations as A | code |
130025335/cell_18 | [
"text_plain_output_1.png"
] | from tensorflow.keras.utils import load_img
import cv2
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df)
train_df.labels.value_counts()
label_counts = train_df['labels'].value_counts()
colormap = plt.cm.get_cmap('tab10')
num_labels = len(label_counts)
colors = [colormap(i) for i in range(num_labels)]
plt.figure(figsize=(10, 6))
plt.bar(label_counts.index, label_counts.values, color=colors)
plt.xlabel('Labels')
plt.ylabel('Count')
plt.title('Label Distribution')
plt.xticks(rotation=45)
for i, count in enumerate(label_counts):
plt.text(i, count + 50, str(count), ha='center')
total_count = label_counts.sum()
plt.axhline(total_count, color='black', linestyle='--', alpha=0.5)
plt.text(0.5, 1.08, f'Total Count: {total_count}', transform=plt.gca().transAxes, ha='center', fontsize=12, fontweight='bold')
max_count = max(label_counts)
y_limit = max_count + max_count * 0.1
plt.ylim(top=y_limit)
plt.show() | code |
130025335/cell_28 | [
"text_plain_output_1.png"
] | from tensorflow.keras.utils import load_img
import cv2
import hashlib
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df)
RESIZED_IMAGE_PATH = '../input/resized-plant2021/img_sz_384/'
train_df.labels.value_counts()
initial_length = len(train_df)
hashes = {}
duplicates = []
originals = []
for index, row in train_df.iterrows():
filename = row['image']
with open(os.path.join(RESIZED_IMAGE_PATH, filename), 'rb') as f:
hash = hashlib.md5(f.read()).hexdigest()
if hash in hashes:
duplicates.append(filename)
originals.append(hashes[hash])
train_df.drop(index, inplace=True)
else:
hashes[hash] = filename
train_df.labels.value_counts()
labels = train_df['labels'].tolist()
unique_labels = set()
for label in labels:
unique_labels.update(label.split())
print(unique_labels, 'suma:', len(unique_labels)) | code |
130025335/cell_3 | [
"text_plain_output_1.png"
] | import tensorflow as tf
gpus = tf.config.list_physical_devices('GPU')
print(gpus)
if len(gpus) == 1:
strategy = tf.distribute.OneDeviceStrategy(device='/gpu:0')
else:
strategy = tf.distribute.MirroredStrategy() | code |
130025335/cell_17 | [
"text_plain_output_1.png"
] | from tensorflow.keras.utils import load_img
import cv2
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df)
train_df.labels.value_counts() | code |
130025335/cell_31 | [
"text_html_output_1.png"
] | from sklearn.preprocessing import MultiLabelBinarizer
from tensorflow.keras.utils import load_img
import cv2
import hashlib
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df)
RESIZED_IMAGE_PATH = '../input/resized-plant2021/img_sz_384/'
train_df.labels.value_counts()
initial_length = len(train_df)
hashes = {}
duplicates = []
originals = []
for index, row in train_df.iterrows():
filename = row['image']
with open(os.path.join(RESIZED_IMAGE_PATH, filename), 'rb') as f:
hash = hashlib.md5(f.read()).hexdigest()
if hash in hashes:
duplicates.append(filename)
originals.append(hashes[hash])
train_df.drop(index, inplace=True)
else:
hashes[hash] = filename
train_df.labels.value_counts()
labels = train_df['labels'].tolist()
unique_labels = set()
for label in labels:
unique_labels.update(label.split())
common_labels = [label[0] for label in pd.Series(labels).str.split(expand=True).stack().value_counts()[:6].items()]
mlb = MultiLabelBinarizer(classes=common_labels)
label_matrix = mlb.fit_transform(train_df['labels'].str.split())
label_df = pd.DataFrame(label_matrix, columns=common_labels)
train_df.reset_index(drop=True, inplace=True)
label_df.reset_index(drop=True, inplace=True)
new_df = pd.concat([train_df, label_df], axis=1)
train_df = new_df.drop('labels', axis=1)
train_df | code |
130025335/cell_24 | [
"text_plain_output_1.png"
] | from tensorflow.keras.utils import load_img
import cv2
import hashlib
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df)
RESIZED_IMAGE_PATH = '../input/resized-plant2021/img_sz_384/'
train_df.labels.value_counts()
initial_length = len(train_df)
hashes = {}
duplicates = []
originals = []
for index, row in train_df.iterrows():
filename = row['image']
with open(os.path.join(RESIZED_IMAGE_PATH, filename), 'rb') as f:
hash = hashlib.md5(f.read()).hexdigest()
if hash in hashes:
duplicates.append(filename)
originals.append(hashes[hash])
train_df.drop(index, inplace=True)
else:
hashes[hash] = filename
train_df.labels.value_counts() | code |
130025335/cell_22 | [
"image_output_1.png"
] | from tensorflow.keras.utils import load_img
import cv2
import hashlib
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import random
def load_image(filename):
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + filename)
def load_random_image(filenames):
sample = random.choice(filenames)
image = load_img('../input/plant-pathology-2021-fgvc8/train_images/' + sample)
def load_augmented_random_image(filenames):
sample = random.choice(filenames)
image = load_img('/kaggle/working/' + sample)
def load_image_for_augmentation(image_path):
image = cv2.imread(image_path)
if image.shape[-1] == 1:
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB)
image = np.array(image)
return image
def format_path_gcs(st):
return GCS_DS_PATH + '/train_images/' + st
def format_resized_image_path_gcs(st):
return GCS_DS_PATH + '/img_sz_384/' + st
def format_tpu_path(st):
return '/kaggle/input/resized-plant2021' + '/img_sz_384/' + st
# Defining mem usage reduction function to try to make dataframes lighter
def reduce_mem_usage(df, verbose=True):
numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
start_mem = df.memory_usage().sum() / 1024**2
for col in df.columns:
col_type = df[col].dtypes
if col_type in numerics:
c_min = df[col].min()
c_max = df[col].max()
if str(col_type)[:3] == 'int':
if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
df[col] = df[col].astype(np.int8)
elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
df[col] = df[col].astype(np.int16)
elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
df[col] = df[col].astype(np.int32)
elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
df[col] = df[col].astype(np.int64)
else:
if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
df[col] = df[col].astype(np.float16)
elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
df[col] = df[col].astype(np.float32)
else:
df[col] = df[col].astype(np.float64)
end_mem = df.memory_usage().sum() / 1024**2
print('Memory usage after optimization is: {:.2f} MB'.format(end_mem))
print('Decreased by {:.1f}%'.format(100 * (start_mem - end_mem) / start_mem))
return df
train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')
IMAGE_PATH = '../input/plant-pathology-2021-fgvc8/test-images/'
train_df = reduce_mem_usage(train_df)
RESIZED_IMAGE_PATH = '../input/resized-plant2021/img_sz_384/'
train_df.labels.value_counts()
label_counts = train_df['labels'].value_counts()
colormap = plt.cm.get_cmap('tab10')
num_labels = len(label_counts)
colors = [colormap(i) for i in range(num_labels)]
plt.xticks(rotation=45)
for i, count in enumerate(label_counts):
plt.text(i, count + 50, str(count), ha='center')
total_count = label_counts.sum()
plt.text(0.5, 1.08, f'Total Count: {total_count}', transform=plt.gca().transAxes, ha='center', fontsize=12, fontweight='bold')
max_count = max(label_counts)
y_limit = max_count + max_count * 0.1
plt.ylim(top=y_limit)
initial_length = len(train_df)
hashes = {}
duplicates = []
originals = []
for index, row in train_df.iterrows():
filename = row['image']
with open(os.path.join(RESIZED_IMAGE_PATH, filename), 'rb') as f:
hash = hashlib.md5(f.read()).hexdigest()
if hash in hashes:
duplicates.append(filename)
originals.append(hashes[hash])
train_df.drop(index, inplace=True)
else:
hashes[hash] = filename
image_filenames = [originals[0], duplicates[0], originals[1], duplicates[1], originals[2], duplicates[2]]
fig, axes = plt.subplots(3, 2, figsize=(10, 8))
for i, (image_filename, ax) in enumerate(zip(image_filenames, axes.flatten())):
image = plt.imread(format_tpu_path(image_filename))
ax.imshow(image)
ax.axis('off')
if i % 2 == 0:
title = 'Original'
else:
title = 'Duplicate'
ax.set_title(title)
plt.tight_layout()
plt.show() | code |
130025335/cell_5 | [
"image_output_1.png"
] | import tensorflow as tf
gpus = tf.config.list_physical_devices('GPU')
if len(gpus) == 1:
strategy = tf.distribute.OneDeviceStrategy(device='/gpu:0')
else:
strategy = tf.distribute.MirroredStrategy()
tf.config.optimizer.set_experimental_options({'auto_mixed_precision': True})
print('Mixed precision enabled') | code |
18159957/cell_13 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
labels_house = ['yes', 'no', 'unknown']
sizes_house = [2175, 1839, 105]
colors_house = ['#ff6666', '#ffcc99', '#ffb3e6']
labels_loan = ['yes', 'no', 'unknown']
sizes_loan = [665, 3349, 105]
colors_loan = ['#c2c2f0', '#ffb3e6', '#66b3ff']
labels_contact = ['cellular', 'telephone']
sizes_contact = [2652, 1467]
colors_contact = ['#ff9999', '#ffcc99']
labels_default = ['no', 'unknown', 'yes']
sizes_default = [3523, 454, 142]
colors_default = ['#99ff99', '#66b3ff', '#ff6666']
plt.rcParams.update({'font.size': 15})
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
sns.catplot(x='target', hue='loan', kind='count', data=df)
plt.title('result by personal loan status')
plt.xlabel('results(y) of current campaign')
plt.show() | code |
18159957/cell_9 | [
"image_output_1.png"
] | import pandas as pd
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
print(df.describe()) | code |
18159957/cell_6 | [
"image_output_1.png"
] | import pandas as pd
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
print(df.dtypes) | code |
18159957/cell_11 | [
"text_plain_output_1.png"
] | import pandas as pd
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
print(df['age'].value_counts())
print(df['job'].value_counts())
print(df['marital'].value_counts())
print(df['education'].value_counts())
print(df['default'].value_counts())
print(df['housing'].value_counts())
print(df['loan'].value_counts())
print(df['contact'].value_counts())
print(df['month'].value_counts())
print(df['day_of_week'].value_counts())
print(df['duration'].value_counts())
print(df['campaign'].value_counts())
print(df['pdays'].value_counts())
print(df['previous'].value_counts())
print(df['poutcome'].value_counts())
print(df['emp.var.rate'].value_counts())
print(df['cons.price.idx'].value_counts())
print(df['cons.conf.idx'].value_counts())
print(df['euribor3m'].value_counts())
print(df['nr.employed'].value_counts())
print(df['target'].value_counts()) | code |
18159957/cell_19 | [
"image_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
labels_house = ['yes', 'no', 'unknown']
sizes_house = [2175, 1839, 105]
colors_house = ['#ff6666', '#ffcc99', '#ffb3e6']
labels_loan = ['yes', 'no', 'unknown']
sizes_loan = [665, 3349, 105]
colors_loan = ['#c2c2f0', '#ffb3e6', '#66b3ff']
labels_contact = ['cellular', 'telephone']
sizes_contact = [2652, 1467]
colors_contact = ['#ff9999', '#ffcc99']
labels_default = ['no', 'unknown', 'yes']
sizes_default = [3523, 454, 142]
colors_default = ['#99ff99', '#66b3ff', '#ff6666']
plt.rcParams.update({'font.size': 15})
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
ldf = df[df.default == 'yes']
#result(y) vs previous campaign outcome(poutcome) vs duration
g = sns.catplot(x="duration", y="target", row = "poutcome",
kind="box", orient="h", height=2.5, aspect=5,
data=df)
sns.catplot(x='target', hue='job', kind='count', data=df)
plt.title('Result by job type of clients')
plt.xlabel('results(y) of current campaign')
plt.show() | code |
18159957/cell_7 | [
"image_output_1.png"
] | import pandas as pd
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
print(df.head()) | code |
18159957/cell_18 | [
"image_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
labels_house = ['yes', 'no', 'unknown']
sizes_house = [2175, 1839, 105]
colors_house = ['#ff6666', '#ffcc99', '#ffb3e6']
labels_loan = ['yes', 'no', 'unknown']
sizes_loan = [665, 3349, 105]
colors_loan = ['#c2c2f0', '#ffb3e6', '#66b3ff']
labels_contact = ['cellular', 'telephone']
sizes_contact = [2652, 1467]
colors_contact = ['#ff9999', '#ffcc99']
labels_default = ['no', 'unknown', 'yes']
sizes_default = [3523, 454, 142]
colors_default = ['#99ff99', '#66b3ff', '#ff6666']
plt.rcParams.update({'font.size': 15})
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
ldf = df[df.default == 'yes']
#result(y) vs previous campaign outcome(poutcome) vs duration
g = sns.catplot(x="duration", y="target", row = "poutcome",
kind="box", orient="h", height=2.5, aspect=5,
data=df)
sns.catplot(x='target', kind='count', data=df)
plt.title('results (target count) of current campaign')
plt.xlabel('results(target) of current campaign')
plt.show() | code |
18159957/cell_8 | [
"image_output_1.png"
] | import pandas as pd
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
print(df.shape) | code |
18159957/cell_15 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
labels_house = ['yes', 'no', 'unknown']
sizes_house = [2175, 1839, 105]
colors_house = ['#ff6666', '#ffcc99', '#ffb3e6']
labels_loan = ['yes', 'no', 'unknown']
sizes_loan = [665, 3349, 105]
colors_loan = ['#c2c2f0', '#ffb3e6', '#66b3ff']
labels_contact = ['cellular', 'telephone']
sizes_contact = [2652, 1467]
colors_contact = ['#ff9999', '#ffcc99']
labels_default = ['no', 'unknown', 'yes']
sizes_default = [3523, 454, 142]
colors_default = ['#99ff99', '#66b3ff', '#ff6666']
plt.rcParams.update({'font.size': 15})
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
sns.catplot(x='target', hue='default', kind='count', data=df)
plt.title('result by default status')
plt.xlabel('results(y) of current campaign')
plt.show() | code |
18159957/cell_16 | [
"image_output_4.png",
"image_output_3.png",
"image_output_2.png",
"image_output_1.png"
] | import pandas as pd
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
ldf = df[df.default == 'yes']
print(ldf) | code |
18159957/cell_17 | [
"image_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
labels_house = ['yes', 'no', 'unknown']
sizes_house = [2175, 1839, 105]
colors_house = ['#ff6666', '#ffcc99', '#ffb3e6']
labels_loan = ['yes', 'no', 'unknown']
sizes_loan = [665, 3349, 105]
colors_loan = ['#c2c2f0', '#ffb3e6', '#66b3ff']
labels_contact = ['cellular', 'telephone']
sizes_contact = [2652, 1467]
colors_contact = ['#ff9999', '#ffcc99']
labels_default = ['no', 'unknown', 'yes']
sizes_default = [3523, 454, 142]
colors_default = ['#99ff99', '#66b3ff', '#ff6666']
plt.rcParams.update({'font.size': 15})
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
ldf = df[df.default == 'yes']
g = sns.catplot(x='duration', y='target', row='poutcome', kind='box', orient='h', height=2.5, aspect=5, data=df) | code |
18159957/cell_14 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
labels_house = ['yes', 'no', 'unknown']
sizes_house = [2175, 1839, 105]
colors_house = ['#ff6666', '#ffcc99', '#ffb3e6']
labels_loan = ['yes', 'no', 'unknown']
sizes_loan = [665, 3349, 105]
colors_loan = ['#c2c2f0', '#ffb3e6', '#66b3ff']
labels_contact = ['cellular', 'telephone']
sizes_contact = [2652, 1467]
colors_contact = ['#ff9999', '#ffcc99']
labels_default = ['no', 'unknown', 'yes']
sizes_default = [3523, 454, 142]
colors_default = ['#99ff99', '#66b3ff', '#ff6666']
plt.rcParams.update({'font.size': 15})
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
centre_circle = plt.Circle((0, 0), 0.5, color='black', fc='white', linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
sns.catplot(x='target', hue='housing', kind='count', data=df)
plt.title('result by housing loan status')
plt.show() | code |
18159957/cell_10 | [
"text_plain_output_1.png"
] | import pandas as pd
df = pd.read_csv('../input/bank-additional-full.csv', sep=';', decimal='.', header=0, names=['age', 'job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration', 'campaign', 'pdays', 'previous', 'poutcome', 'emp.var.rate', 'cons.price.idx', 'cons.conf.idx', 'euribor3m', 'nr.employed', 'target'])
print(df.isnull().sum()) | code |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.