path
stringlengths 13
17
| screenshot_names
sequencelengths 1
873
| code
stringlengths 0
40.4k
| cell_type
stringclasses 1
value |
|---|---|---|---|
72111100/cell_11 | [
"text_html_output_1.png"
] | train = pd.read_csv('../input/car-price/train set.csv')
names = [x.split(' ')[0] for x in list(train['name'])]
train.insert(0, 'brand', names)
train = train.drop(['name', 'seller_type', 'owner', 'torque', 'fuel'], axis=1)
train.head(2) | code |
72111100/cell_18 | [
"image_output_1.png"
] | train = pd.read_csv('../input/car-price/train set.csv')
names = [x.split(' ')[0] for x in list(train['name'])]
train.insert(0, 'brand', names)
train = train.drop(['name', 'seller_type', 'owner', 'torque', 'fuel'], axis=1)
train['engine'] = [int(x.split(' ')[0]) for x in list(train['engine'])]
train['mileage'] = [float(x.split(' ')[0]) for x in list(train['mileage'])]
train['max_power'] = [float(x.split(' ')[0]) for x in list(train['max_power'])]
num_features = [x for x in train.columns if type(train[x][0]) is not str]
cat_features = [x for x in train.columns if x not in num_features]
train.head(3) | code |
72111100/cell_32 | [
"text_plain_output_1.png"
] | from math import sqrt
from sklearn.linear_model import LinearRegression as LR, Perceptron
from sklearn.model_selection import cross_val_score
from sklearn.preprocessing import StandardScaler
from sklearn.preprocessing import StandardScaler
from sklearn.svm import SVR
train = pd.read_csv('../input/car-price/train set.csv')
names = [x.split(' ')[0] for x in list(train['name'])]
train.insert(0, 'brand', names)
train = train.drop(['name', 'seller_type', 'owner', 'torque', 'fuel'], axis=1)
train['engine'] = [int(x.split(' ')[0]) for x in list(train['engine'])]
train['mileage'] = [float(x.split(' ')[0]) for x in list(train['mileage'])]
train['max_power'] = [float(x.split(' ')[0]) for x in list(train['max_power'])]
num_features = [x for x in train.columns if type(train[x][0]) is not str]
cat_features = [x for x in train.columns if x not in num_features]
X_train = train.drop('selling_price', axis=1).values[0:6850]
y_train = train['selling_price'].values[0:6850]
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
scaled_X_train = scaler.fit_transform(X_train)
from sklearn.model_selection import KFold
from sklearn.model_selection import cross_val_score
from sklearn.metrics import *
from math import sqrt
mcc = make_scorer(mean_absolute_error)
def evaluate_model(model):
model = model
import sklearn
scores = cross_val_score(model, scaled_X_train, y_train, scoring=mcc, cv=5, n_jobs=-1)
return scores.mean()
models = [KNR(), RNR(), LR(), RFR(n_estimators=300), Perceptron(), SVR(), MLPR()]
models_names = ['K_neighbors', 'radius_neighbors', 'linear_regression', 'random_forest_regressor', 'perceptron', 'SVR', 'MLP_Regression']
scores = list()
for clf, clf_name in zip(models, models_names):
k_mean = evaluate_model(clf)
scores.append(k_mean)
model = models[3]
model.fit(scaled_X_train, y_train)
train_prediction = model.predict(scaled_X_train)
train_pred = [int(x) for x in train_prediction.round()]
train_prediction = np.array(train_pred)
def transform(df):
brand = [x.split(' ')[0] for x in list(df['name'])]
df.insert(0, 'brand', brand)
df.drop(['name', 'seller_type', 'owner', 'torque', 'fuel'], axis=1, inplace=True)
df['engine'] = [int(x.split(' ')[0]) for x in list(df['engine'])]
df['mileage'] = [float(x.split(' ')[0]) for x in list(df['mileage'])]
df['max_power'] = [float(x.split(' ')[0]) for x in list(df['max_power'])]
df['transmission'] = [0 if x == 'Manual' else 1 for x in df['transmission']]
df['brand'] = [0 if x <= 1000000 else 1 if x <= 2000000 else 2 if x <= 4000000 else 3 for x in df['selling_price']]
X = df.drop('selling_price', axis=1).values
y = df['selling_price'].values
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
X = scaler.fit_transform(X)
return (X, y)
test_prediction = model.predict(X_test)
test_pred = [int(x) for x in test_prediction.round()]
test_prediction = np.array(test_pred)
print('Test R.M.S.E : ', sqrt(mean_squared_error(y_test, test_prediction))) | code |
72111100/cell_28 | [
"text_plain_output_1.png"
] | from math import sqrt
from sklearn.linear_model import LinearRegression as LR, Perceptron
from sklearn.model_selection import cross_val_score
from sklearn.preprocessing import StandardScaler
from sklearn.svm import SVR
train = pd.read_csv('../input/car-price/train set.csv')
names = [x.split(' ')[0] for x in list(train['name'])]
train.insert(0, 'brand', names)
train = train.drop(['name', 'seller_type', 'owner', 'torque', 'fuel'], axis=1)
train['engine'] = [int(x.split(' ')[0]) for x in list(train['engine'])]
train['mileage'] = [float(x.split(' ')[0]) for x in list(train['mileage'])]
train['max_power'] = [float(x.split(' ')[0]) for x in list(train['max_power'])]
num_features = [x for x in train.columns if type(train[x][0]) is not str]
cat_features = [x for x in train.columns if x not in num_features]
X_train = train.drop('selling_price', axis=1).values[0:6850]
y_train = train['selling_price'].values[0:6850]
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
scaled_X_train = scaler.fit_transform(X_train)
from sklearn.model_selection import KFold
from sklearn.model_selection import cross_val_score
from sklearn.metrics import *
from math import sqrt
mcc = make_scorer(mean_absolute_error)
def evaluate_model(model):
model = model
import sklearn
scores = cross_val_score(model, scaled_X_train, y_train, scoring=mcc, cv=5, n_jobs=-1)
return scores.mean()
models = [KNR(), RNR(), LR(), RFR(n_estimators=300), Perceptron(), SVR(), MLPR()]
models_names = ['K_neighbors', 'radius_neighbors', 'linear_regression', 'random_forest_regressor', 'perceptron', 'SVR', 'MLP_Regression']
scores = list()
for clf, clf_name in zip(models, models_names):
k_mean = evaluate_model(clf)
scores.append(k_mean)
model = models[3]
model.fit(scaled_X_train, y_train)
train_prediction = model.predict(scaled_X_train)
train_pred = [int(x) for x in train_prediction.round()]
train_prediction = np.array(train_pred)
print('Train R.M.S.E : ', sqrt(mean_squared_error(y_train, train_prediction))) | code |
72111100/cell_16 | [
"image_output_1.png"
] | train = pd.read_csv('../input/car-price/train set.csv')
names = [x.split(' ')[0] for x in list(train['name'])]
train.insert(0, 'brand', names)
train = train.drop(['name', 'seller_type', 'owner', 'torque', 'fuel'], axis=1)
train['engine'] = [int(x.split(' ')[0]) for x in list(train['engine'])]
train['mileage'] = [float(x.split(' ')[0]) for x in list(train['mileage'])]
train['max_power'] = [float(x.split(' ')[0]) for x in list(train['max_power'])]
num_features = [x for x in train.columns if type(train[x][0]) is not str]
cat_features = [x for x in train.columns if x not in num_features]
train.head(3) | code |
72111100/cell_3 | [
"text_plain_output_1.png"
] | train = pd.read_csv('../input/car-price/train set.csv')
train.head(2) | code |
72111100/cell_24 | [
"text_html_output_1.png"
] | from sklearn.linear_model import LinearRegression as LR, Perceptron
from sklearn.model_selection import cross_val_score
from sklearn.preprocessing import StandardScaler
from sklearn.svm import SVR
train = pd.read_csv('../input/car-price/train set.csv')
names = [x.split(' ')[0] for x in list(train['name'])]
train.insert(0, 'brand', names)
train = train.drop(['name', 'seller_type', 'owner', 'torque', 'fuel'], axis=1)
train['engine'] = [int(x.split(' ')[0]) for x in list(train['engine'])]
train['mileage'] = [float(x.split(' ')[0]) for x in list(train['mileage'])]
train['max_power'] = [float(x.split(' ')[0]) for x in list(train['max_power'])]
num_features = [x for x in train.columns if type(train[x][0]) is not str]
cat_features = [x for x in train.columns if x not in num_features]
X_train = train.drop('selling_price', axis=1).values[0:6850]
y_train = train['selling_price'].values[0:6850]
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
scaled_X_train = scaler.fit_transform(X_train)
from sklearn.model_selection import KFold
from sklearn.model_selection import cross_val_score
from sklearn.metrics import *
from math import sqrt
mcc = make_scorer(mean_absolute_error)
def evaluate_model(model):
model = model
import sklearn
scores = cross_val_score(model, scaled_X_train, y_train, scoring=mcc, cv=5, n_jobs=-1)
return scores.mean()
models = [KNR(), RNR(), LR(), RFR(n_estimators=300), Perceptron(), SVR(), MLPR()]
models_names = ['K_neighbors', 'radius_neighbors', 'linear_regression', 'random_forest_regressor', 'perceptron', 'SVR', 'MLP_Regression']
scores = list()
for clf, clf_name in zip(models, models_names):
k_mean = evaluate_model(clf)
print(f'score of {clf_name} : ', round(k_mean, 3))
scores.append(k_mean) | code |
72111100/cell_14 | [
"text_html_output_1.png"
] | train = pd.read_csv('../input/car-price/train set.csv')
names = [x.split(' ')[0] for x in list(train['name'])]
train.insert(0, 'brand', names)
train = train.drop(['name', 'seller_type', 'owner', 'torque', 'fuel'], axis=1)
train['engine'] = [int(x.split(' ')[0]) for x in list(train['engine'])]
train['mileage'] = [float(x.split(' ')[0]) for x in list(train['mileage'])]
train['max_power'] = [float(x.split(' ')[0]) for x in list(train['max_power'])]
num_features = [x for x in train.columns if type(train[x][0]) is not str]
cat_features = [x for x in train.columns if x not in num_features]
import seaborn as sns, matplotlib.pyplot as plt
import seaborn as sns, matplotlib.pyplot as plt
sns.barplot(x=train['transmission'], y=train['selling_price'])
plt.show() | code |
72111100/cell_5 | [
"text_plain_output_1.png"
] | train = pd.read_csv('../input/car-price/train set.csv')
test = pd.read_csv('../input/car-price/test set.csv')
test.head(2) | code |
74067093/cell_7 | [
"text_plain_output_5.png",
"text_plain_output_4.png",
"text_plain_output_6.png",
"text_plain_output_3.png",
"text_plain_output_2.png",
"text_plain_output_1.png"
] | from tensorflow.keras.applications import EfficientNetB0
from tensorflow.keras.layers import Input, GlobalAveragePooling2D, Dense, Dropout
from tensorflow.keras.layers.experimental import preprocessing
from tensorflow.keras.metrics import AUC
from tensorflow.keras.models import Model
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers.experimental import preprocessing
from tensorflow.keras.models import Sequential
img_augmentation = Sequential([preprocessing.CenterCrop(224, 224)], name='img_augmentation')
from tensorflow.keras.applications import EfficientNetB0
from tensorflow.keras.layers import Input, GlobalAveragePooling2D, Dense, Dropout
from tensorflow.keras.layers.experimental.preprocessing import CenterCrop
from tensorflow.keras.models import Model
from tensorflow.keras.metrics import AUC
auc = AUC()
inputs = Input(shape=(299, 299, 3))
x = img_augmentation(inputs)
x = EfficientNetB0(weights='imagenet', include_top=False)(x)
x = GlobalAveragePooling2D()(x)
x = Dense(512, activation='relu')(x)
x = Dropout(rate=0.25)(x)
outputs = Dense(1, activation='sigmoid')(x)
model = Model(inputs=inputs, outputs=outputs)
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=[auc])
model.summary() | code |
17138373/cell_4 | [
"image_output_1.png"
] | import pandas as pd
dfBlackFriday = pd.read_csv('../input/BlackFriday.csv', delimiter=',')
dfBlackFriday.head() | code |
17138373/cell_6 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import pandas as pd
dfBlackFriday = pd.read_csv('../input/BlackFriday.csv', delimiter=',')
dfBlackFriday.isnull().sum() | code |
17138373/cell_11 | [
"text_html_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
dfBlackFriday = pd.read_csv('../input/BlackFriday.csv', delimiter=',')
dfBlackFriday.isnull().sum()
produtosmaisComprados = dfBlackFriday['Product_ID'].value_counts().head(10)
produtosmaisComprados.plot(kind='bar', title='10 Produtos mais comprados')
plt.xlabel('Produtos')
plt.ylabel('Quantidade') | code |
17138373/cell_8 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
dfBlackFriday = pd.read_csv('../input/BlackFriday.csv', delimiter=',')
dfBlackFriday.isnull().sum()
sns.violinplot(dfBlackFriday['Age'].sort_values(), dfBlackFriday['Purchase'], data=dfBlackFriday)
plt.show() | code |
17138373/cell_15 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
dfBlackFriday = pd.read_csv('../input/BlackFriday.csv', delimiter=',')
dfBlackFriday.isnull().sum()
dfBlackFridayCons = dfBlackFriday.query('Purchase > 9000')
sns.violinplot(dfBlackFridayCons['Marital_Status'], dfBlackFridayCons['Occupation'], data=dfBlackFridayCons) | code |
17138373/cell_10 | [
"text_html_output_1.png"
] | import pandas as pd
dfBlackFriday = pd.read_csv('../input/BlackFriday.csv', delimiter=',')
dfBlackFriday.isnull().sum()
dfBlackFriday['Product_ID'].value_counts() | code |
17138373/cell_5 | [
"text_plain_output_1.png"
] | import pandas as pd
dfBlackFriday = pd.read_csv('../input/BlackFriday.csv', delimiter=',')
dfBlackFriday.describe() | code |
122252864/cell_13 | [
"text_html_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
death = pd.read_csv('/kaggle/input/character-deathscsv/character-deaths.csv')
death.shape | code |
122252864/cell_9 | [
"image_output_1.png"
] | import pandas as pd
import seaborn as sns
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.shape
battle.rename(columns={'attacker_1': 'primary_attacker'}, inplace=True)
battle.rename(columns={'defender_1': 'primary_defender'}, inplace=True)
sns.set(rc={'figure.figsize': (13, 5)})
sns.barplot(x='attacker_king', y='attacker_size', data=battle)
plt.show() | code |
122252864/cell_4 | [
"text_plain_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.shape | code |
122252864/cell_6 | [
"text_plain_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.shape
battle.rename(columns={'attacker_1': 'primary_attacker'}, inplace=True)
battle.rename(columns={'defender_1': 'primary_defender'}, inplace=True)
battle.head() | code |
122252864/cell_11 | [
"text_plain_output_1.png"
] | import pandas as pd
import seaborn as sns
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.shape
battle.rename(columns={'attacker_1': 'primary_attacker'}, inplace=True)
battle.rename(columns={'defender_1': 'primary_defender'}, inplace=True)
sns.set(rc={'figure.figsize': (13, 5)})
sns.set(rc={'figure.figsize': (13, 5)})
sns.countplot(x=battle['attacker_king'], hue=battle['battle_type'])
plt.show() | code |
122252864/cell_7 | [
"text_plain_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.shape
battle.rename(columns={'attacker_1': 'primary_attacker'}, inplace=True)
battle.rename(columns={'defender_1': 'primary_defender'}, inplace=True)
battle['attacker_king'].value_counts() | code |
122252864/cell_18 | [
"image_output_1.png"
] | import pandas as pd
import seaborn as sns
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.shape
battle.rename(columns={'attacker_1': 'primary_attacker'}, inplace=True)
battle.rename(columns={'defender_1': 'primary_defender'}, inplace=True)
sns.set(rc={'figure.figsize': (13, 5)})
sns.set(rc={'figure.figsize': (13, 5)})
death = pd.read_csv('/kaggle/input/character-deathscsv/character-deaths.csv')
death.shape
sns.set(rc={'figure.figsize': (30, 10)})
sns.countplot(data=death, x='Allegiances', width=0.8)
plt.show() | code |
122252864/cell_8 | [
"image_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.shape
battle.rename(columns={'attacker_1': 'primary_attacker'}, inplace=True)
battle.rename(columns={'defender_1': 'primary_defender'}, inplace=True)
battle['location'].value_counts() | code |
122252864/cell_15 | [
"text_plain_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
death = pd.read_csv('/kaggle/input/character-deathscsv/character-deaths.csv')
death.shape
death['Nobility'].value_counts() | code |
122252864/cell_16 | [
"application_vnd.jupyter.stderr_output_1.png",
"image_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
death = pd.read_csv('/kaggle/input/character-deathscsv/character-deaths.csv')
death.shape
death['Death Year'].value_counts() | code |
122252864/cell_3 | [
"text_html_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.head() | code |
122252864/cell_17 | [
"application_vnd.jupyter.stderr_output_1.png",
"image_output_1.png"
] | import pandas as pd
import seaborn as sns
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.shape
battle.rename(columns={'attacker_1': 'primary_attacker'}, inplace=True)
battle.rename(columns={'defender_1': 'primary_defender'}, inplace=True)
sns.set(rc={'figure.figsize': (13, 5)})
sns.set(rc={'figure.figsize': (13, 5)})
death = pd.read_csv('/kaggle/input/character-deathscsv/character-deaths.csv')
death.shape
sns.countplot(data=death, x='Death Year')
plt.show() | code |
122252864/cell_14 | [
"text_plain_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
death = pd.read_csv('/kaggle/input/character-deathscsv/character-deaths.csv')
death.shape
death['Gender'].value_counts() | code |
122252864/cell_10 | [
"text_html_output_1.png"
] | import pandas as pd
import seaborn as sns
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.shape
battle.rename(columns={'attacker_1': 'primary_attacker'}, inplace=True)
battle.rename(columns={'defender_1': 'primary_defender'}, inplace=True)
sns.set(rc={'figure.figsize': (13, 5)})
sns.set(rc={'figure.figsize': (13, 5)})
sns.barplot(x='defender_king', y='defender_size', data=battle)
plt.show() | code |
122252864/cell_12 | [
"text_html_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
death = pd.read_csv('/kaggle/input/character-deathscsv/character-deaths.csv')
death.head() | code |
122252864/cell_5 | [
"text_plain_output_1.png"
] | import pandas as pd
battle = pd.read_csv('/kaggle/input/games-of-thrones/battles.csv')
battle.shape
battle.rename(columns={'attacker_1': 'primary_attacker'}, inplace=True)
battle.head() | code |
129020918/cell_20 | [
"application_vnd.jupyter.stderr_output_1.png"
] | from PIL import Image
from matplotlib import patches, patheffects
from torch.utils.data import ConcatDataset
from torch.utils.data import Dataset
from torch.utils.data import Dataset, DataLoader
from torch.utils.data import Subset
from torchvision import datasets, transforms
from torchvision.transforms.functional import to_tensor
from tqdm.notebook import trange, tqdm
import math
import matplotlib.pyplot as plt
import numpy as np
import os
import torch as tc
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
import torchvision.models as models
import torchvision.models as models
import xml.etree.ElementTree as ET
imagenet_stats = ([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
def one_epoch(net, loss, dl, opt=None, metric=None):
if opt: net.train() # only affects some layers
else: net.eval()
L, M = [], []
dl_it = iter(dl)
for xb, yb in tqdm(dl_it, leave=False):
xb = xb.cuda()
if not isinstance(yb, list): yb = [yb] # this is new(!)
yb = [yb_.cuda() for yb_ in yb]
if opt:
y_ = net(xb)
l = loss(y_, yb)
opt.zero_grad()
l.backward()
opt.step()
else:
with tc.no_grad():
y_ = net(xb)
l = loss(y_, yb)
L.append(l.detach().cpu().numpy())
if isinstance(metric, list):
for m in metric:
M.append(m(tc.sigmoid(y_), yb[0]))
elif metric:
M.append(metric(tc.sigmoid(y_), yb[0]))
return L, M
def fit(net, tr_dl, val_dl, loss=nn.CrossEntropyLoss(), epochs=3, lr=3e-3, wd=1e-3, plot=True):
opt = optim.Adam(net.parameters(), lr=lr, weight_decay=wd)
Ltr_hist, Lval_hist = [], []
for epoch in trange(epochs):
Ltr, _ = one_epoch(net, loss, tr_dl, opt)
Lval, Aval = one_epoch(net, loss, val_dl, None, batch_iou)
Ltr_hist.append(np.mean(Ltr))
Lval_hist.append(np.mean(Lval))
#print(f'epoch: {epoch}\ttraining loss: {np.mean(Ltr):0.4f}\tvalidation loss: {np.mean(Lval):0.4f} \tvalidation accuracy: {mean(Aval):0.2f}')
print(f'epoch: {epoch}\ttraining loss: {np.mean(Ltr):0.4f}\tvalidation loss: {np.mean(Lval):0.4f}, overlap accuracy: {np.array(Aval).mean():0.2f}')
# plot the losses
if plot:
_,ax = plt.subplots(1,1,figsize=(16,4))
ax.plot(1+np.arange(len(Ltr_hist)),Ltr_hist)
ax.plot(1+np.arange(len(Lval_hist)),Lval_hist)
ax.grid('on')
ax.set_xlim(left=1, right=len(Ltr_hist))
ax.legend(['training loss', 'validation loss']);
return Ltr_hist, Lval_hist
def denorm(x, stats=imagenet_stats):
return x * tc.Tensor(stats[1])[:,None,None] + tc.Tensor(stats[0])[:,None,None]
transform = transforms.Compose([
transforms.RandomApply([transforms.ColorJitter(brightness=0.1, contrast=0.1, saturation=0.1, hue=0.1)], p=1),
transforms.ToTensor()])
def draw_rect(ax, xy, w, h):
patch = ax.add_patch(patches.Rectangle(xy, w, h, fill=False, edgecolor='yellow', lw=2))
patch.set_path_effects([patheffects.Stroke(linewidth=6, foreground='black'), patheffects.Normal()])
def _freeze(md, fr=True):
ch = list(md.children())
for c in ch: _freeze(c, fr)
if not ch and not isinstance(md, tc.nn.modules.batchnorm.BatchNorm2d): # not freezing the BatchNorm layers!
for p in md.parameters():
p.requires_grad = not fr
def freeze_to(md, ix=-1):
ch_all = list(md.children())
for ch in ch_all[:ix]: _freeze(ch, True)
def unfreeze_to(md, ix=-1):
ch_all = list(md.children())
for ch in ch_all[:ix]: _freeze(ch, False)
def calculate_iou(box_true, box_pred):
x1_tr, y1_tr, w_tr, h_tr = box_true
x1_pr, y1_pr, w_pr, h_pr = box_pred
intersection_x1 = max(x1_tr, x1_pr)
intersection_y1 = max(y1_tr, y1_pr)
intersection_x2 = min(x1_tr + w_tr, x1_pr + w_pr)
intersection_y2 = min(y1_tr + h_tr, y1_pr + h_pr)
intersection_w = max(0, intersection_x2 - intersection_x1)
intersection_h = max(0, intersection_y2 - intersection_y1)
intersection_area = intersection_w * intersection_h
area_true = w_tr * h_tr
area_pred = w_pr * h_pr
union_area = area_true + area_pred - intersection_area
iou = intersection_area / union_area
return iou
def batch_iou(box_true_batch, box_pred_batch):
iou_list = []
for box_true, box_pred in zip(box_true_batch, box_pred_batch):
iou = calculate_iou(box_true, box_pred)
iou_list.append(iou.item())
iou_array = np.array(iou_list).mean()
return iou_array
class Data2tensor(Dataset):
def __init__(self, data_dir, transforms=None):
self.data_dir = data_dir
self.transforms = transforms
self.img_paths = []
self.xml_paths = []
for filename in os.listdir(data_dir):
name, ext = os.path.splitext(filename)
if ext == '.jpg':
img_path = os.path.join(data_dir, filename)
xml_path = os.path.join(data_dir, name + '.xml')
if os.path.isfile(xml_path):
self.img_paths.append(img_path)
self.xml_paths.append(xml_path)
def __len__(self):
return len(self.img_paths)
def __getitem__(self, idx):
img_path = self.img_paths[idx]
xml_path = self.xml_paths[idx]
img = Image.open(img_path).convert('RGB')
w, h = img.size
tree = ET.parse(xml_path)
root = tree.getroot()
bboxes = []
for obj in root.findall('object'):
bbox = obj.find('bndbox')
xmin = int(bbox.find('xmin').text)
ymin = int(bbox.find('ymin').text)
xmax = int(bbox.find('xmax').text)
ymax = int(bbox.find('ymax').text)
y_bbox = np.array([xmin, ymin, xmax, ymax])
y_bbox = y_bbox / np.array([w, h, w, h])
y_bbox = [*y_bbox[:2], *y_bbox[2:] - y_bbox[:2]]
bboxes.append(y_bbox)
if self.transforms:
img = self.transforms(img)
else:
img = to_tensor(img)
bboxes = [tc.tensor(bbox, dtype=tc.float32) for bbox in bboxes]
return (img, bboxes)
def ds_train_val(train_dir, val_dir, p, transforms):
ds_full = Data2tensor(train_dir, transforms)
num_data = len(ds_full)
num_train = math.ceil(num_data * p)
train_indices = list(range(num_train))
ds_tr = Subset(ds_full, train_indices)
ds_val = Data2tensor(val_dir, None)
return (ds_tr, ds_val)
train_dir = '/kaggle/input/spot250/train'
val_dir = '/kaggle/input/spot250/valid'
ds_tr0, ds_val = ds_train_val(train_dir, val_dir, p=1, transforms=transform)
from torch.utils.data import ConcatDataset
transform1 = transforms.Compose([transforms.RandomApply([transforms.ColorJitter(brightness=0.1, contrast=0.1, saturation=0.1, hue=0.1)], p=1), transforms.ToTensor()])
transform2 = transforms.Compose([transforms.RandomApply([transforms.ColorJitter(brightness=0.4, contrast=0.5, saturation=0.2, hue=0.1)], p=1), transforms.ToTensor()])
ds_tr, _ = ds_train_val(train_dir, val_dir, p=1, transforms=None)
transf_ds1, _ = ds_train_val(train_dir, val_dir, p=0.4, transforms=transform1)
ds_tr1 = ConcatDataset([ds_tr, transf_ds1])
transf_ds2, _ = ds_train_val(train_dir, val_dir, p=0.4, transforms=transform2)
transformed_ds_tr = ConcatDataset([ds_tr1, transf_ds2])
bs = 500
train_and_transformed = DataLoader(transformed_ds_tr, batch_size=bs, shuffle=True, num_workers=0)
train_dl = DataLoader(ds_tr, batch_size=bs, shuffle=True, num_workers=0)
val_dl = DataLoader(ds_val, batch_size=2 * bs, shuffle=True, num_workers=0)
xt, yt = next(iter(train_and_transformed))
xb, yb = next(iter(train_dl))
x_v, y_v = next(iter(val_dl))
md_full = models.resnet34()
num_ftrs = md_full.fc.in_features
md_full.fc = nn.Linear(num_ftrs, 4)
def myloss(y, y_b, reduction='mean'):
inp_reg = y
tar_reg = y_b[0]
loss_reg = F.mse_loss(1.2 * tc.sigmoid(inp_reg) - 0.1, tar_reg, reduction=reduction)
if reduction == 'none':
loss_reg = loss_reg.mean(dim=-1)
return loss_reg
bs = 32
train_and_transformed = DataLoader(transformed_ds_tr, batch_size=bs, shuffle=True, num_workers=0)
train_dl = DataLoader(ds_tr0, batch_size=bs, shuffle=True, num_workers=0)
val_dl = DataLoader(ds_val, batch_size=bs, shuffle=True, num_workers=0)
fit(md_full.cuda(), train_dl, val_dl, loss=myloss, epochs=5, wd=0.001, lr=0.003) | code |
129020918/cell_11 | [
"image_output_1.png"
] | from PIL import Image
from matplotlib import patches, patheffects
from torch.utils.data import Dataset
from torch.utils.data import Dataset, DataLoader
from torchvision import datasets, transforms
from torchvision.transforms.functional import to_tensor
from tqdm.notebook import trange, tqdm
import matplotlib.pyplot as plt
import numpy as np
import os
import torch as tc
import torch.nn as nn
import torch.optim as optim
import xml.etree.ElementTree as ET
imagenet_stats = ([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
def one_epoch(net, loss, dl, opt=None, metric=None):
if opt: net.train() # only affects some layers
else: net.eval()
L, M = [], []
dl_it = iter(dl)
for xb, yb in tqdm(dl_it, leave=False):
xb = xb.cuda()
if not isinstance(yb, list): yb = [yb] # this is new(!)
yb = [yb_.cuda() for yb_ in yb]
if opt:
y_ = net(xb)
l = loss(y_, yb)
opt.zero_grad()
l.backward()
opt.step()
else:
with tc.no_grad():
y_ = net(xb)
l = loss(y_, yb)
L.append(l.detach().cpu().numpy())
if isinstance(metric, list):
for m in metric:
M.append(m(tc.sigmoid(y_), yb[0]))
elif metric:
M.append(metric(tc.sigmoid(y_), yb[0]))
return L, M
def fit(net, tr_dl, val_dl, loss=nn.CrossEntropyLoss(), epochs=3, lr=3e-3, wd=1e-3, plot=True):
opt = optim.Adam(net.parameters(), lr=lr, weight_decay=wd)
Ltr_hist, Lval_hist = [], []
for epoch in trange(epochs):
Ltr, _ = one_epoch(net, loss, tr_dl, opt)
Lval, Aval = one_epoch(net, loss, val_dl, None, batch_iou)
Ltr_hist.append(np.mean(Ltr))
Lval_hist.append(np.mean(Lval))
#print(f'epoch: {epoch}\ttraining loss: {np.mean(Ltr):0.4f}\tvalidation loss: {np.mean(Lval):0.4f} \tvalidation accuracy: {mean(Aval):0.2f}')
print(f'epoch: {epoch}\ttraining loss: {np.mean(Ltr):0.4f}\tvalidation loss: {np.mean(Lval):0.4f}, overlap accuracy: {np.array(Aval).mean():0.2f}')
# plot the losses
if plot:
_,ax = plt.subplots(1,1,figsize=(16,4))
ax.plot(1+np.arange(len(Ltr_hist)),Ltr_hist)
ax.plot(1+np.arange(len(Lval_hist)),Lval_hist)
ax.grid('on')
ax.set_xlim(left=1, right=len(Ltr_hist))
ax.legend(['training loss', 'validation loss']);
return Ltr_hist, Lval_hist
def denorm(x, stats=imagenet_stats):
return x * tc.Tensor(stats[1])[:,None,None] + tc.Tensor(stats[0])[:,None,None]
transform = transforms.Compose([
transforms.RandomApply([transforms.ColorJitter(brightness=0.1, contrast=0.1, saturation=0.1, hue=0.1)], p=1),
transforms.ToTensor()])
def draw_rect(ax, xy, w, h):
patch = ax.add_patch(patches.Rectangle(xy, w, h, fill=False, edgecolor='yellow', lw=2))
patch.set_path_effects([patheffects.Stroke(linewidth=6, foreground='black'), patheffects.Normal()])
def _freeze(md, fr=True):
ch = list(md.children())
for c in ch: _freeze(c, fr)
if not ch and not isinstance(md, tc.nn.modules.batchnorm.BatchNorm2d): # not freezing the BatchNorm layers!
for p in md.parameters():
p.requires_grad = not fr
def freeze_to(md, ix=-1):
ch_all = list(md.children())
for ch in ch_all[:ix]: _freeze(ch, True)
def unfreeze_to(md, ix=-1):
ch_all = list(md.children())
for ch in ch_all[:ix]: _freeze(ch, False)
def calculate_iou(box_true, box_pred):
x1_tr, y1_tr, w_tr, h_tr = box_true
x1_pr, y1_pr, w_pr, h_pr = box_pred
intersection_x1 = max(x1_tr, x1_pr)
intersection_y1 = max(y1_tr, y1_pr)
intersection_x2 = min(x1_tr + w_tr, x1_pr + w_pr)
intersection_y2 = min(y1_tr + h_tr, y1_pr + h_pr)
intersection_w = max(0, intersection_x2 - intersection_x1)
intersection_h = max(0, intersection_y2 - intersection_y1)
intersection_area = intersection_w * intersection_h
area_true = w_tr * h_tr
area_pred = w_pr * h_pr
union_area = area_true + area_pred - intersection_area
iou = intersection_area / union_area
return iou
def batch_iou(box_true_batch, box_pred_batch):
iou_list = []
for box_true, box_pred in zip(box_true_batch, box_pred_batch):
iou = calculate_iou(box_true, box_pred)
iou_list.append(iou.item())
iou_array = np.array(iou_list).mean()
return iou_array
class Data2tensor(Dataset):
def __init__(self, data_dir, transforms=None):
self.data_dir = data_dir
self.transforms = transforms
self.img_paths = []
self.xml_paths = []
for filename in os.listdir(data_dir):
name, ext = os.path.splitext(filename)
if ext == '.jpg':
img_path = os.path.join(data_dir, filename)
xml_path = os.path.join(data_dir, name + '.xml')
if os.path.isfile(xml_path):
self.img_paths.append(img_path)
self.xml_paths.append(xml_path)
def __len__(self):
return len(self.img_paths)
def __getitem__(self, idx):
img_path = self.img_paths[idx]
xml_path = self.xml_paths[idx]
img = Image.open(img_path).convert('RGB')
w, h = img.size
tree = ET.parse(xml_path)
root = tree.getroot()
bboxes = []
for obj in root.findall('object'):
bbox = obj.find('bndbox')
xmin = int(bbox.find('xmin').text)
ymin = int(bbox.find('ymin').text)
xmax = int(bbox.find('xmax').text)
ymax = int(bbox.find('ymax').text)
y_bbox = np.array([xmin, ymin, xmax, ymax])
y_bbox = y_bbox / np.array([w, h, w, h])
y_bbox = [*y_bbox[:2], *y_bbox[2:] - y_bbox[:2]]
bboxes.append(y_bbox)
if self.transforms:
img = self.transforms(img)
else:
img = to_tensor(img)
bboxes = [tc.tensor(bbox, dtype=tc.float32) for bbox in bboxes]
return (img, bboxes)
def show_img_and_bbox(x, y, ax=None):
# plot the image:
if not ax: _, ax = plt.subplots(1, 1, figsize=(5, 5))
if len(x.shape) == 3:
H,W = x.shape[1:]
x = x.numpy().transpose(1, 2, 0)
ax.imshow(x)
ax.axis('off')
# showing bounding box
bbox = y
x, y, w, h = bbox[0], bbox[1], bbox[2] - bbox[0], bbox[3] - bbox[1]
draw_rect(ax, [x*W, y*H], x*W+ w*W, y*H+h*H)
x, y = ds_tr0[3]
show_img_and_bbox(x, y[0]) | code |
129020918/cell_15 | [
"text_plain_output_1.png"
] | from matplotlib import patches, patheffects
from torch.utils.data import ConcatDataset
from torch.utils.data import Dataset, DataLoader
from torch.utils.data import Subset
from torchvision import datasets, transforms
from tqdm.notebook import trange, tqdm
import math
import matplotlib.pyplot as plt
import numpy as np
import torch as tc
import torch.nn as nn
import torch.optim as optim
imagenet_stats = ([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
def one_epoch(net, loss, dl, opt=None, metric=None):
if opt: net.train() # only affects some layers
else: net.eval()
L, M = [], []
dl_it = iter(dl)
for xb, yb in tqdm(dl_it, leave=False):
xb = xb.cuda()
if not isinstance(yb, list): yb = [yb] # this is new(!)
yb = [yb_.cuda() for yb_ in yb]
if opt:
y_ = net(xb)
l = loss(y_, yb)
opt.zero_grad()
l.backward()
opt.step()
else:
with tc.no_grad():
y_ = net(xb)
l = loss(y_, yb)
L.append(l.detach().cpu().numpy())
if isinstance(metric, list):
for m in metric:
M.append(m(tc.sigmoid(y_), yb[0]))
elif metric:
M.append(metric(tc.sigmoid(y_), yb[0]))
return L, M
def fit(net, tr_dl, val_dl, loss=nn.CrossEntropyLoss(), epochs=3, lr=3e-3, wd=1e-3, plot=True):
opt = optim.Adam(net.parameters(), lr=lr, weight_decay=wd)
Ltr_hist, Lval_hist = [], []
for epoch in trange(epochs):
Ltr, _ = one_epoch(net, loss, tr_dl, opt)
Lval, Aval = one_epoch(net, loss, val_dl, None, batch_iou)
Ltr_hist.append(np.mean(Ltr))
Lval_hist.append(np.mean(Lval))
#print(f'epoch: {epoch}\ttraining loss: {np.mean(Ltr):0.4f}\tvalidation loss: {np.mean(Lval):0.4f} \tvalidation accuracy: {mean(Aval):0.2f}')
print(f'epoch: {epoch}\ttraining loss: {np.mean(Ltr):0.4f}\tvalidation loss: {np.mean(Lval):0.4f}, overlap accuracy: {np.array(Aval).mean():0.2f}')
# plot the losses
if plot:
_,ax = plt.subplots(1,1,figsize=(16,4))
ax.plot(1+np.arange(len(Ltr_hist)),Ltr_hist)
ax.plot(1+np.arange(len(Lval_hist)),Lval_hist)
ax.grid('on')
ax.set_xlim(left=1, right=len(Ltr_hist))
ax.legend(['training loss', 'validation loss']);
return Ltr_hist, Lval_hist
def denorm(x, stats=imagenet_stats):
return x * tc.Tensor(stats[1])[:,None,None] + tc.Tensor(stats[0])[:,None,None]
transform = transforms.Compose([
transforms.RandomApply([transforms.ColorJitter(brightness=0.1, contrast=0.1, saturation=0.1, hue=0.1)], p=1),
transforms.ToTensor()])
def draw_rect(ax, xy, w, h):
patch = ax.add_patch(patches.Rectangle(xy, w, h, fill=False, edgecolor='yellow', lw=2))
patch.set_path_effects([patheffects.Stroke(linewidth=6, foreground='black'), patheffects.Normal()])
def _freeze(md, fr=True):
ch = list(md.children())
for c in ch: _freeze(c, fr)
if not ch and not isinstance(md, tc.nn.modules.batchnorm.BatchNorm2d): # not freezing the BatchNorm layers!
for p in md.parameters():
p.requires_grad = not fr
def freeze_to(md, ix=-1):
ch_all = list(md.children())
for ch in ch_all[:ix]: _freeze(ch, True)
def unfreeze_to(md, ix=-1):
ch_all = list(md.children())
for ch in ch_all[:ix]: _freeze(ch, False)
def ds_train_val(train_dir, val_dir, p, transforms):
ds_full = Data2tensor(train_dir, transforms)
num_data = len(ds_full)
num_train = math.ceil(num_data * p)
train_indices = list(range(num_train))
ds_tr = Subset(ds_full, train_indices)
ds_val = Data2tensor(val_dir, None)
return (ds_tr, ds_val)
train_dir = '/kaggle/input/spot250/train'
val_dir = '/kaggle/input/spot250/valid'
ds_tr0, ds_val = ds_train_val(train_dir, val_dir, p=1, transforms=transform)
from torch.utils.data import ConcatDataset
transform1 = transforms.Compose([transforms.RandomApply([transforms.ColorJitter(brightness=0.1, contrast=0.1, saturation=0.1, hue=0.1)], p=1), transforms.ToTensor()])
transform2 = transforms.Compose([transforms.RandomApply([transforms.ColorJitter(brightness=0.4, contrast=0.5, saturation=0.2, hue=0.1)], p=1), transforms.ToTensor()])
ds_tr, _ = ds_train_val(train_dir, val_dir, p=1, transforms=None)
transf_ds1, _ = ds_train_val(train_dir, val_dir, p=0.4, transforms=transform1)
ds_tr1 = ConcatDataset([ds_tr, transf_ds1])
transf_ds2, _ = ds_train_val(train_dir, val_dir, p=0.4, transforms=transform2)
transformed_ds_tr = ConcatDataset([ds_tr1, transf_ds2])
bs = 500
train_and_transformed = DataLoader(transformed_ds_tr, batch_size=bs, shuffle=True, num_workers=0)
train_dl = DataLoader(ds_tr, batch_size=bs, shuffle=True, num_workers=0)
val_dl = DataLoader(ds_val, batch_size=2 * bs, shuffle=True, num_workers=0)
xt, yt = next(iter(train_and_transformed))
xb, yb = next(iter(train_dl))
x_v, y_v = next(iter(val_dl))
print(f'Train and transformed data: x shape = {xt.shape}, bbox shape = {yt[0].shape} \n')
print(f'Training data: image shape = {xb.shape}, bbox shape = {yb[0].shape} \n')
print(f'Validation data: image shape = {x_v.shape}, bbox shape = {y_v[0].shape}') | code |
90136510/cell_13 | [
"text_html_output_1.png"
] | from sqlalchemy import create_engine
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df = df.rename(columns={'Company Name': 'CompanyName', 'Job Title': 'JobTitle', 'Salaries Reported': 'SalariesReported'}, inplace=False)
from sqlalchemy import create_engine
engine = create_engine('sqlite://', echo=False)
df.to_sql('SalaryDataset', con=engine)
sql = '\n\nSelect * from SalaryDataset\nlimit 5\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\n\nSelect * from SalaryDataset\nwhere CompanyName='IBM'\n\n\n"
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nSelect distinct(jobtitle) as job_title from SalaryDataset\n\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nSelect \njobtitle,\ncount(salariesreported) as count_of_reports \nfrom SalaryDataset\ngroup by jobtitle\norder by count(salariesreported) desc\n\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\n\nSelect\nlocation,\nsum(case when JobTitle='Data Scientist' then 1 else 0 end) as Data_Scientist,\nsum(case when JobTitle='Data Analyst' then 1 else 0 end) as Data_Analyst,\nsum(case when JobTitle='Data Engineer' then 1 else 0 end) as Data_Engineer,\nsum(case when JobTitle='Machine Learning Engineer' then 1 else 0 end) as Machine_Learning_Engineer,\nsum(case when JobTitle='Junior Data Scientist' then 1 else 0 end) as Junior_Data_Scientist,\nsum(case when JobTitle='Senior Machine Learning Engineer' then 1 else 0 end) as Senior_Machine_Learning_Engineer,\nsum(case when JobTitle='Lead Data Scientist' then 1 else 0 end) as Lead_Data_Scientist,\nsum(case when JobTitle='Software Engineer- Machine learning' then 1 else 0 end) as Software_Engineer_Machine_learning,\nsum(case when JobTitle='Machine Learning Scientist' then 1 else 0 end) as Machine_Learning_Scientist,\nsum(case when JobTitle='Machine Learning Developer' then 1 else 0 end) as Machine_Learning_Developer,\nsum(case when JobTitle='Machine Learning Consultant' then 1 else 0 end) as Machine_Learning_Consultant,\nsum(case when JobTitle='Machine Learning Software Engineer' then 1 else 0 end) as Machine_Learning_Software_Engineer,\nsum(case when JobTitle='Machine Learning Engineer/Data Scientist' then 1 else 0 end) as Machine_Learning_Engineer_Data_Scientist,\nsum(case when JobTitle='Machine Learning Data Associate II' then 1 else 0 end) as Machine_Learning_Data_Associate_II,\nsum(case when JobTitle='Machine Learning Data Associate I' then 1 else 0 end) as Machine_Learning_Data_Associate_I,\nsum(case when JobTitle='Machine Learning Data Associate' then 1 else 0 end) as Machine_Learning_Data_Associate,\nsum(case when JobTitle='Machine Learning Data Analyst' then 1 else 0 end) as Machine_Learning_Data_Analyst,\nsum(case when JobTitle='Machine Learning Associate' then 1 else 0 end) as Machine_Learning_Associate,\nsum(case when JobTitle='Data Scientist - Trainee' then 1 else 0 end) as Data_Scientist_Trainee,\nsum(case when JobTitle='Data Science Manager' then 1 else 0 end) as Data_Science_Manager,\nsum(case when JobTitle='Data Science Lead' then 1 else 0 end) as Data_Science_Lead,\nsum(case when JobTitle='Data Science Associate' then 1 else 0 end) as Data_Science_Associate,\nsum(case when JobTitle='Associate Machine Learning Engineer' then 1 else 0 end) as Associate_Machine_Learning_Engineer,\nsum(case when JobTitle='Data Science Consultant' then 1 else 0 end) as Data_Science_Consultant,\nsum(case when JobTitle='Senior Data Scientist' then 1 else 0 end) as Senior_Data_Scientist\nfrom SalaryDataset\ngroup by 1\n\n\n"
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nselect *,\n\ncase \nwhen a.count_of_branches>3 then "big"\nwhen a.count_of_branches>1 and a.count_of_branches<=3 then "medium"\nwhen a.count_of_branches=1 then "small"\nelse \'no size\' end as size_of_the_company\n\nfrom\n(Select \ncompanyname,\ncount(distinct location) as count_of_branches\nfrom SalaryDataset\nwhere companyname !=\'None\'\ngroup by 1\norder by count(distinct location) desc\n) a\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\nselect distinct companyname\nfrom salarydataset\n'
df_sql = pd.read_sql_query(sql, con=engine)
df_sql.head(100)
df_sql.to_csv('company.csv', index=False)
print('run') | code |
90136510/cell_9 | [
"text_html_output_1.png"
] | from sqlalchemy import create_engine
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df = df.rename(columns={'Company Name': 'CompanyName', 'Job Title': 'JobTitle', 'Salaries Reported': 'SalariesReported'}, inplace=False)
from sqlalchemy import create_engine
engine = create_engine('sqlite://', echo=False)
df.to_sql('SalaryDataset', con=engine)
sql = '\n\nSelect * from SalaryDataset\nlimit 5\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\n\nSelect * from SalaryDataset\nwhere CompanyName='IBM'\n\n\n"
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\nSelect distinct(jobtitle) as job_title from SalaryDataset\n'
df_sql = pd.read_sql_query(sql, con=engine)
df_sql.head() | code |
90136510/cell_4 | [
"text_html_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df = df.rename(columns={'Company Name': 'CompanyName', 'Job Title': 'JobTitle', 'Salaries Reported': 'SalariesReported'}, inplace=False)
df.head() | code |
90136510/cell_11 | [
"text_html_output_1.png"
] | from sqlalchemy import create_engine
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df = df.rename(columns={'Company Name': 'CompanyName', 'Job Title': 'JobTitle', 'Salaries Reported': 'SalariesReported'}, inplace=False)
from sqlalchemy import create_engine
engine = create_engine('sqlite://', echo=False)
df.to_sql('SalaryDataset', con=engine)
sql = '\n\nSelect * from SalaryDataset\nlimit 5\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\n\nSelect * from SalaryDataset\nwhere CompanyName='IBM'\n\n\n"
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nSelect distinct(jobtitle) as job_title from SalaryDataset\n\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nSelect \njobtitle,\ncount(salariesreported) as count_of_reports \nfrom SalaryDataset\ngroup by jobtitle\norder by count(salariesreported) desc\n\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\nSelect\nlocation,\nsum(case when JobTitle='Data Scientist' then 1 else 0 end) as Data_Scientist,\nsum(case when JobTitle='Data Analyst' then 1 else 0 end) as Data_Analyst,\nsum(case when JobTitle='Data Engineer' then 1 else 0 end) as Data_Engineer,\nsum(case when JobTitle='Machine Learning Engineer' then 1 else 0 end) as Machine_Learning_Engineer,\nsum(case when JobTitle='Junior Data Scientist' then 1 else 0 end) as Junior_Data_Scientist,\nsum(case when JobTitle='Senior Machine Learning Engineer' then 1 else 0 end) as Senior_Machine_Learning_Engineer,\nsum(case when JobTitle='Lead Data Scientist' then 1 else 0 end) as Lead_Data_Scientist,\nsum(case when JobTitle='Software Engineer- Machine learning' then 1 else 0 end) as Software_Engineer_Machine_learning,\nsum(case when JobTitle='Machine Learning Scientist' then 1 else 0 end) as Machine_Learning_Scientist,\nsum(case when JobTitle='Machine Learning Developer' then 1 else 0 end) as Machine_Learning_Developer,\nsum(case when JobTitle='Machine Learning Consultant' then 1 else 0 end) as Machine_Learning_Consultant,\nsum(case when JobTitle='Machine Learning Software Engineer' then 1 else 0 end) as Machine_Learning_Software_Engineer,\nsum(case when JobTitle='Machine Learning Engineer/Data Scientist' then 1 else 0 end) as Machine_Learning_Engineer_Data_Scientist,\nsum(case when JobTitle='Machine Learning Data Associate II' then 1 else 0 end) as Machine_Learning_Data_Associate_II,\nsum(case when JobTitle='Machine Learning Data Associate I' then 1 else 0 end) as Machine_Learning_Data_Associate_I,\nsum(case when JobTitle='Machine Learning Data Associate' then 1 else 0 end) as Machine_Learning_Data_Associate,\nsum(case when JobTitle='Machine Learning Data Analyst' then 1 else 0 end) as Machine_Learning_Data_Analyst,\nsum(case when JobTitle='Machine Learning Associate' then 1 else 0 end) as Machine_Learning_Associate,\nsum(case when JobTitle='Data Scientist - Trainee' then 1 else 0 end) as Data_Scientist_Trainee,\nsum(case when JobTitle='Data Science Manager' then 1 else 0 end) as Data_Science_Manager,\nsum(case when JobTitle='Data Science Lead' then 1 else 0 end) as Data_Science_Lead,\nsum(case when JobTitle='Data Science Associate' then 1 else 0 end) as Data_Science_Associate,\nsum(case when JobTitle='Associate Machine Learning Engineer' then 1 else 0 end) as Associate_Machine_Learning_Engineer,\nsum(case when JobTitle='Data Science Consultant' then 1 else 0 end) as Data_Science_Consultant,\nsum(case when JobTitle='Senior Data Scientist' then 1 else 0 end) as Senior_Data_Scientist\nfrom SalaryDataset\ngroup by 1\n"
df_sql = pd.read_sql_query(sql, con=engine)
df_sql.head(100) | code |
90136510/cell_1 | [
"text_plain_output_1.png"
] | import os
import numpy as np
import pandas as pd
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename)) | code |
90136510/cell_7 | [
"text_plain_output_1.png"
] | from sqlalchemy import create_engine
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df = df.rename(columns={'Company Name': 'CompanyName', 'Job Title': 'JobTitle', 'Salaries Reported': 'SalariesReported'}, inplace=False)
from sqlalchemy import create_engine
engine = create_engine('sqlite://', echo=False)
df.to_sql('SalaryDataset', con=engine)
sql = '\nSelect * from SalaryDataset\nlimit 5\n'
df_sql = pd.read_sql_query(sql, con=engine)
df_sql.head() | code |
90136510/cell_8 | [
"text_html_output_1.png"
] | from sqlalchemy import create_engine
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df = df.rename(columns={'Company Name': 'CompanyName', 'Job Title': 'JobTitle', 'Salaries Reported': 'SalariesReported'}, inplace=False)
from sqlalchemy import create_engine
engine = create_engine('sqlite://', echo=False)
df.to_sql('SalaryDataset', con=engine)
sql = '\n\nSelect * from SalaryDataset\nlimit 5\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\nSelect * from SalaryDataset\nwhere CompanyName='IBM'\n"
df_sql = pd.read_sql_query(sql, con=engine)
df_sql.head() | code |
90136510/cell_15 | [
"text_html_output_1.png"
] | from sqlalchemy import create_engine
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df = df.rename(columns={'Company Name': 'CompanyName', 'Job Title': 'JobTitle', 'Salaries Reported': 'SalariesReported'}, inplace=False)
from sqlalchemy import create_engine
engine = create_engine('sqlite://', echo=False)
df.to_sql('SalaryDataset', con=engine)
sql = '\n\nSelect * from SalaryDataset\nlimit 5\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\n\nSelect * from SalaryDataset\nwhere CompanyName='IBM'\n\n\n"
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nSelect distinct(jobtitle) as job_title from SalaryDataset\n\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nSelect \njobtitle,\ncount(salariesreported) as count_of_reports \nfrom SalaryDataset\ngroup by jobtitle\norder by count(salariesreported) desc\n\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\n\nSelect\nlocation,\nsum(case when JobTitle='Data Scientist' then 1 else 0 end) as Data_Scientist,\nsum(case when JobTitle='Data Analyst' then 1 else 0 end) as Data_Analyst,\nsum(case when JobTitle='Data Engineer' then 1 else 0 end) as Data_Engineer,\nsum(case when JobTitle='Machine Learning Engineer' then 1 else 0 end) as Machine_Learning_Engineer,\nsum(case when JobTitle='Junior Data Scientist' then 1 else 0 end) as Junior_Data_Scientist,\nsum(case when JobTitle='Senior Machine Learning Engineer' then 1 else 0 end) as Senior_Machine_Learning_Engineer,\nsum(case when JobTitle='Lead Data Scientist' then 1 else 0 end) as Lead_Data_Scientist,\nsum(case when JobTitle='Software Engineer- Machine learning' then 1 else 0 end) as Software_Engineer_Machine_learning,\nsum(case when JobTitle='Machine Learning Scientist' then 1 else 0 end) as Machine_Learning_Scientist,\nsum(case when JobTitle='Machine Learning Developer' then 1 else 0 end) as Machine_Learning_Developer,\nsum(case when JobTitle='Machine Learning Consultant' then 1 else 0 end) as Machine_Learning_Consultant,\nsum(case when JobTitle='Machine Learning Software Engineer' then 1 else 0 end) as Machine_Learning_Software_Engineer,\nsum(case when JobTitle='Machine Learning Engineer/Data Scientist' then 1 else 0 end) as Machine_Learning_Engineer_Data_Scientist,\nsum(case when JobTitle='Machine Learning Data Associate II' then 1 else 0 end) as Machine_Learning_Data_Associate_II,\nsum(case when JobTitle='Machine Learning Data Associate I' then 1 else 0 end) as Machine_Learning_Data_Associate_I,\nsum(case when JobTitle='Machine Learning Data Associate' then 1 else 0 end) as Machine_Learning_Data_Associate,\nsum(case when JobTitle='Machine Learning Data Analyst' then 1 else 0 end) as Machine_Learning_Data_Analyst,\nsum(case when JobTitle='Machine Learning Associate' then 1 else 0 end) as Machine_Learning_Associate,\nsum(case when JobTitle='Data Scientist - Trainee' then 1 else 0 end) as Data_Scientist_Trainee,\nsum(case when JobTitle='Data Science Manager' then 1 else 0 end) as Data_Science_Manager,\nsum(case when JobTitle='Data Science Lead' then 1 else 0 end) as Data_Science_Lead,\nsum(case when JobTitle='Data Science Associate' then 1 else 0 end) as Data_Science_Associate,\nsum(case when JobTitle='Associate Machine Learning Engineer' then 1 else 0 end) as Associate_Machine_Learning_Engineer,\nsum(case when JobTitle='Data Science Consultant' then 1 else 0 end) as Data_Science_Consultant,\nsum(case when JobTitle='Senior Data Scientist' then 1 else 0 end) as Senior_Data_Scientist\nfrom SalaryDataset\ngroup by 1\n\n\n"
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nselect *,\n\ncase \nwhen a.count_of_branches>3 then "big"\nwhen a.count_of_branches>1 and a.count_of_branches<=3 then "medium"\nwhen a.count_of_branches=1 then "small"\nelse \'no size\' end as size_of_the_company\n\nfrom\n(Select \ncompanyname,\ncount(distinct location) as count_of_branches\nfrom SalaryDataset\nwhere companyname !=\'None\'\ngroup by 1\norder by count(distinct location) desc\n) a\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nselect distinct companyname\nfrom salarydataset\n'
df_sql = pd.read_sql_query(sql, con=engine)
df_sql.to_csv('company.csv', index=False)
sql = "\nselect \na.location,\nsum(case when a.companyname='Mu Sigma' then 1 else 0 end) as Mu_Sigma,\nsum(case when a.companyname='IBM' then 1 else 0 end) as IBM,\nsum(case when a.companyname='Tata Consultancy Services' then 1 else 0 end) as Tata_Consultancy_Services,\nsum(case when a.companyname='Impact Analytics' then 1 else 0 end) as Impact_Analytics,\nsum(case when a.companyname='Accenture' then 1 else 0 end) as Accenture,\nsum(case when a.companyname='Infosys' then 1 else 0 end) as Infosys,\nsum(case when a.companyname='Capgemini' then 1 else 0 end) as Capgemini,\nsum(case when a.companyname='Cognizant Technology Solutions' then 1 else 0 end) as Cognizant_Technology_Solutions\nfrom\n(Select \ncompanyname,\nlocation,\ncount(distinct location) as count_of_branches\nfrom SalaryDataset\nwhere companyname!='None'\ngroup by 1,2\norder by count(distinct location) desc) a\ngroup by 1\n"
df_sql = pd.read_sql_query(sql, con=engine)
df_sql.head(100) | code |
90136510/cell_3 | [
"text_html_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df.head() | code |
90136510/cell_10 | [
"text_html_output_1.png"
] | from sqlalchemy import create_engine
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df = df.rename(columns={'Company Name': 'CompanyName', 'Job Title': 'JobTitle', 'Salaries Reported': 'SalariesReported'}, inplace=False)
from sqlalchemy import create_engine
engine = create_engine('sqlite://', echo=False)
df.to_sql('SalaryDataset', con=engine)
sql = '\n\nSelect * from SalaryDataset\nlimit 5\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\n\nSelect * from SalaryDataset\nwhere CompanyName='IBM'\n\n\n"
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nSelect distinct(jobtitle) as job_title from SalaryDataset\n\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\nSelect \njobtitle,\ncount(salariesreported) as count_of_reports \nfrom SalaryDataset\ngroup by jobtitle\norder by count(salariesreported) desc\n'
df_sql = pd.read_sql_query(sql, con=engine)
df_sql.head() | code |
90136510/cell_12 | [
"text_plain_output_1.png"
] | from sqlalchemy import create_engine
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df = df.rename(columns={'Company Name': 'CompanyName', 'Job Title': 'JobTitle', 'Salaries Reported': 'SalariesReported'}, inplace=False)
from sqlalchemy import create_engine
engine = create_engine('sqlite://', echo=False)
df.to_sql('SalaryDataset', con=engine)
sql = '\n\nSelect * from SalaryDataset\nlimit 5\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\n\nSelect * from SalaryDataset\nwhere CompanyName='IBM'\n\n\n"
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nSelect distinct(jobtitle) as job_title from SalaryDataset\n\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\n\nSelect \njobtitle,\ncount(salariesreported) as count_of_reports \nfrom SalaryDataset\ngroup by jobtitle\norder by count(salariesreported) desc\n\n\n\n'
df_sql = pd.read_sql_query(sql, con=engine)
sql = "\n\nSelect\nlocation,\nsum(case when JobTitle='Data Scientist' then 1 else 0 end) as Data_Scientist,\nsum(case when JobTitle='Data Analyst' then 1 else 0 end) as Data_Analyst,\nsum(case when JobTitle='Data Engineer' then 1 else 0 end) as Data_Engineer,\nsum(case when JobTitle='Machine Learning Engineer' then 1 else 0 end) as Machine_Learning_Engineer,\nsum(case when JobTitle='Junior Data Scientist' then 1 else 0 end) as Junior_Data_Scientist,\nsum(case when JobTitle='Senior Machine Learning Engineer' then 1 else 0 end) as Senior_Machine_Learning_Engineer,\nsum(case when JobTitle='Lead Data Scientist' then 1 else 0 end) as Lead_Data_Scientist,\nsum(case when JobTitle='Software Engineer- Machine learning' then 1 else 0 end) as Software_Engineer_Machine_learning,\nsum(case when JobTitle='Machine Learning Scientist' then 1 else 0 end) as Machine_Learning_Scientist,\nsum(case when JobTitle='Machine Learning Developer' then 1 else 0 end) as Machine_Learning_Developer,\nsum(case when JobTitle='Machine Learning Consultant' then 1 else 0 end) as Machine_Learning_Consultant,\nsum(case when JobTitle='Machine Learning Software Engineer' then 1 else 0 end) as Machine_Learning_Software_Engineer,\nsum(case when JobTitle='Machine Learning Engineer/Data Scientist' then 1 else 0 end) as Machine_Learning_Engineer_Data_Scientist,\nsum(case when JobTitle='Machine Learning Data Associate II' then 1 else 0 end) as Machine_Learning_Data_Associate_II,\nsum(case when JobTitle='Machine Learning Data Associate I' then 1 else 0 end) as Machine_Learning_Data_Associate_I,\nsum(case when JobTitle='Machine Learning Data Associate' then 1 else 0 end) as Machine_Learning_Data_Associate,\nsum(case when JobTitle='Machine Learning Data Analyst' then 1 else 0 end) as Machine_Learning_Data_Analyst,\nsum(case when JobTitle='Machine Learning Associate' then 1 else 0 end) as Machine_Learning_Associate,\nsum(case when JobTitle='Data Scientist - Trainee' then 1 else 0 end) as Data_Scientist_Trainee,\nsum(case when JobTitle='Data Science Manager' then 1 else 0 end) as Data_Science_Manager,\nsum(case when JobTitle='Data Science Lead' then 1 else 0 end) as Data_Science_Lead,\nsum(case when JobTitle='Data Science Associate' then 1 else 0 end) as Data_Science_Associate,\nsum(case when JobTitle='Associate Machine Learning Engineer' then 1 else 0 end) as Associate_Machine_Learning_Engineer,\nsum(case when JobTitle='Data Science Consultant' then 1 else 0 end) as Data_Science_Consultant,\nsum(case when JobTitle='Senior Data Scientist' then 1 else 0 end) as Senior_Data_Scientist\nfrom SalaryDataset\ngroup by 1\n\n\n"
df_sql = pd.read_sql_query(sql, con=engine)
sql = '\nselect *,\ncase \nwhen a.count_of_branches>3 then "big"\nwhen a.count_of_branches>1 and a.count_of_branches<=3 then "medium"\nwhen a.count_of_branches=1 then "small"\nelse \'no size\' end as size_of_the_company\nfrom\n(Select \ncompanyname,\ncount(distinct location) as count_of_branches\nfrom SalaryDataset\nwhere companyname !=\'None\'\ngroup by 1\norder by count(distinct location) desc\n) a\n'
df_sql = pd.read_sql_query(sql, con=engine)
df_sql.head(100) | code |
90136510/cell_5 | [
"text_html_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas as pd ## import liberaries pandas
import pandas as pd
df = pd.read_csv('/kaggle/input/analytics-industry-salaries-2022-india/Salary Dataset.csv')
df = df.rename(columns={'Company Name': 'CompanyName', 'Job Title': 'JobTitle', 'Salaries Reported': 'SalariesReported'}, inplace=False)
print('The columns of the dataset are:- ', df.columns) | code |
90144122/cell_13 | [
"text_plain_output_1.png",
"image_output_2.png",
"image_output_1.png"
] | from sklearn.model_selection import train_test_split
from tensorflow.keras.layers import Conv2D, MaxPooling2D
from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten
from tensorflow.keras.models import Sequential
import numpy as np
import pandas as pd
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
X = np.array([e[0] for e in data]).astype('float32')
y = np.array([e[1] for e in data])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
X_train = np.resize(X_train, (len(X_train), 50, 50, 1))
X_test = np.resize(X_test, (len(X_test), 50, 50, 1))
X_train /= 255.0
X_test /= 255.0
model1 = Sequential()
model1.add(Conv2D(64, kernel_size=(3, 3), input_shape=X_train[0].shape))
model1.add(Activation('relu'))
model1.add(MaxPooling2D(pool_size=(2, 2)))
model1.add(Dropout(0.5))
model1.add(Conv2D(64, kernel_size=(3, 3)))
model1.add(Activation('relu'))
model1.add(MaxPooling2D(pool_size=(2, 2)))
model1.add(Flatten())
model1.add(Dense(64))
model1.add(Activation('relu'))
model1.add(Dropout(0.3))
model1.add(Dense(1))
model1.add(Activation('sigmoid'))
model1.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history1 = model1.fit(X_train, y_train, batch_size=32, epochs=10, validation_split=0.2)
history_df1 = pd.DataFrame(history1.history)
model2 = Sequential()
model2.add(Conv2D(128, kernel_size=(3, 3), input_shape=X_train[0].shape))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Flatten())
model2.add(Dense(128))
model2.add(Activation('relu'))
model2.add(Dropout(0.3))
model2.add(Dense(1))
model2.add(Activation('sigmoid'))
model2.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history2 = model2.fit(X_train, y_train, batch_size=64, epochs=40, validation_split=0.2)
history_df2 = pd.DataFrame(history2.history)
model3 = Sequential()
model3.add(Conv2D(256, kernel_size=(3, 2), input_shape=X_train[0].shape))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Dropout(0.3))
model3.add(Conv2D(256, kernel_size=(2, 3)))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Dropout(0.5))
model3.add(Conv2D(256, kernel_size=(2, 2)))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Dropout(0.5))
model3.add(Conv2D(256, kernel_size=(2, 2)))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Flatten())
model3.add(Dense(128))
model3.add(Activation('relu'))
model3.add(Dropout(0.5))
model3.add(Dense(1))
model3.add(Activation('sigmoid'))
model3.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history3 = model3.fit(X_train, y_train, batch_size=256, epochs=50, validation_split=0.2)
history_df3 = pd.DataFrame(history3.history)
history_df3.loc[:, ['loss', 'val_loss']].plot(title='Loss')
history_df3.loc[:, ['accuracy', 'val_accuracy']].plot(title='Accuracy') | code |
90144122/cell_9 | [
"text_plain_output_1.png"
] | from sklearn.model_selection import train_test_split
from tensorflow.keras.layers import Conv2D, MaxPooling2D
from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten
from tensorflow.keras.models import Sequential
import numpy as np
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
X = np.array([e[0] for e in data]).astype('float32')
y = np.array([e[1] for e in data])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
X_train = np.resize(X_train, (len(X_train), 50, 50, 1))
X_test = np.resize(X_test, (len(X_test), 50, 50, 1))
X_train /= 255.0
X_test /= 255.0
model2 = Sequential()
model2.add(Conv2D(128, kernel_size=(3, 3), input_shape=X_train[0].shape))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Flatten())
model2.add(Dense(128))
model2.add(Activation('relu'))
model2.add(Dropout(0.3))
model2.add(Dense(1))
model2.add(Activation('sigmoid'))
model2.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history2 = model2.fit(X_train, y_train, batch_size=64, epochs=40, validation_split=0.2) | code |
90144122/cell_6 | [
"text_plain_output_1.png"
] | from sklearn.model_selection import train_test_split
from tensorflow.keras.layers import Conv2D, MaxPooling2D
from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten
from tensorflow.keras.models import Sequential
import numpy as np
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
X = np.array([e[0] for e in data]).astype('float32')
y = np.array([e[1] for e in data])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
X_train = np.resize(X_train, (len(X_train), 50, 50, 1))
X_test = np.resize(X_test, (len(X_test), 50, 50, 1))
X_train /= 255.0
X_test /= 255.0
model1 = Sequential()
model1.add(Conv2D(64, kernel_size=(3, 3), input_shape=X_train[0].shape))
model1.add(Activation('relu'))
model1.add(MaxPooling2D(pool_size=(2, 2)))
model1.add(Dropout(0.5))
model1.add(Conv2D(64, kernel_size=(3, 3)))
model1.add(Activation('relu'))
model1.add(MaxPooling2D(pool_size=(2, 2)))
model1.add(Flatten())
model1.add(Dense(64))
model1.add(Activation('relu'))
model1.add(Dropout(0.3))
model1.add(Dense(1))
model1.add(Activation('sigmoid'))
model1.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history1 = model1.fit(X_train, y_train, batch_size=32, epochs=10, validation_split=0.2) | code |
90144122/cell_11 | [
"text_plain_output_1.png"
] | from sklearn.metrics import classification_report
from sklearn.metrics import confusion_matrix
from sklearn.model_selection import train_test_split
from tensorflow.keras.layers import Conv2D, MaxPooling2D
from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten
from tensorflow.keras.models import Sequential
import numpy as np
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
X = np.array([e[0] for e in data]).astype('float32')
y = np.array([e[1] for e in data])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
X_train = np.resize(X_train, (len(X_train), 50, 50, 1))
X_test = np.resize(X_test, (len(X_test), 50, 50, 1))
X_train /= 255.0
X_test /= 255.0
model2 = Sequential()
model2.add(Conv2D(128, kernel_size=(3, 3), input_shape=X_train[0].shape))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Flatten())
model2.add(Dense(128))
model2.add(Activation('relu'))
model2.add(Dropout(0.3))
model2.add(Dense(1))
model2.add(Activation('sigmoid'))
model2.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history2 = model2.fit(X_train, y_train, batch_size=64, epochs=40, validation_split=0.2)
y_pred2 = model2.predict(X_test) >= 0.5
print('Confusion matrix:\n', confusion_matrix(y_test, y_pred2))
print('Classification report:\n', classification_report(y_test, y_pred2)) | code |
90144122/cell_7 | [
"text_plain_output_1.png"
] | from sklearn.model_selection import train_test_split
from tensorflow.keras.layers import Conv2D, MaxPooling2D
from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten
from tensorflow.keras.models import Sequential
import numpy as np
import pandas as pd
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
X = np.array([e[0] for e in data]).astype('float32')
y = np.array([e[1] for e in data])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
X_train = np.resize(X_train, (len(X_train), 50, 50, 1))
X_test = np.resize(X_test, (len(X_test), 50, 50, 1))
X_train /= 255.0
X_test /= 255.0
model1 = Sequential()
model1.add(Conv2D(64, kernel_size=(3, 3), input_shape=X_train[0].shape))
model1.add(Activation('relu'))
model1.add(MaxPooling2D(pool_size=(2, 2)))
model1.add(Dropout(0.5))
model1.add(Conv2D(64, kernel_size=(3, 3)))
model1.add(Activation('relu'))
model1.add(MaxPooling2D(pool_size=(2, 2)))
model1.add(Flatten())
model1.add(Dense(64))
model1.add(Activation('relu'))
model1.add(Dropout(0.3))
model1.add(Dense(1))
model1.add(Activation('sigmoid'))
model1.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history1 = model1.fit(X_train, y_train, batch_size=32, epochs=10, validation_split=0.2)
history_df1 = pd.DataFrame(history1.history)
history_df1.loc[:, ['loss', 'val_loss']].plot(title='Loss')
history_df1.loc[:, ['accuracy', 'val_accuracy']].plot(title='Accuracy') | code |
90144122/cell_8 | [
"text_plain_output_1.png",
"image_output_2.png",
"image_output_1.png"
] | from sklearn.metrics import classification_report
from sklearn.metrics import confusion_matrix
from sklearn.model_selection import train_test_split
from tensorflow.keras.layers import Conv2D, MaxPooling2D
from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten
from tensorflow.keras.models import Sequential
import numpy as np
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
X = np.array([e[0] for e in data]).astype('float32')
y = np.array([e[1] for e in data])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
X_train = np.resize(X_train, (len(X_train), 50, 50, 1))
X_test = np.resize(X_test, (len(X_test), 50, 50, 1))
X_train /= 255.0
X_test /= 255.0
model1 = Sequential()
model1.add(Conv2D(64, kernel_size=(3, 3), input_shape=X_train[0].shape))
model1.add(Activation('relu'))
model1.add(MaxPooling2D(pool_size=(2, 2)))
model1.add(Dropout(0.5))
model1.add(Conv2D(64, kernel_size=(3, 3)))
model1.add(Activation('relu'))
model1.add(MaxPooling2D(pool_size=(2, 2)))
model1.add(Flatten())
model1.add(Dense(64))
model1.add(Activation('relu'))
model1.add(Dropout(0.3))
model1.add(Dense(1))
model1.add(Activation('sigmoid'))
model1.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history1 = model1.fit(X_train, y_train, batch_size=32, epochs=10, validation_split=0.2)
y_pred1 = model1.predict(X_test) >= 0.5
print('Confusion matrix:\n', confusion_matrix(y_test, y_pred1))
print('Classification report:\n', classification_report(y_test, y_pred1)) | code |
90144122/cell_3 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
fig, axes = plt.subplots(3, 3, figsize=(15, 15))
index = 75
for i in range(3):
for j in range(3):
axes[i, j].imshow(data[index][0], cmap='gray')
index += 1
plt.show() | code |
90144122/cell_14 | [
"text_plain_output_1.png"
] | from sklearn.metrics import classification_report
from sklearn.metrics import confusion_matrix
from sklearn.model_selection import train_test_split
from tensorflow.keras.layers import Conv2D, MaxPooling2D
from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten
from tensorflow.keras.models import Sequential
import numpy as np
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
X = np.array([e[0] for e in data]).astype('float32')
y = np.array([e[1] for e in data])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
X_train = np.resize(X_train, (len(X_train), 50, 50, 1))
X_test = np.resize(X_test, (len(X_test), 50, 50, 1))
X_train /= 255.0
X_test /= 255.0
model3 = Sequential()
model3.add(Conv2D(256, kernel_size=(3, 2), input_shape=X_train[0].shape))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Dropout(0.3))
model3.add(Conv2D(256, kernel_size=(2, 3)))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Dropout(0.5))
model3.add(Conv2D(256, kernel_size=(2, 2)))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Dropout(0.5))
model3.add(Conv2D(256, kernel_size=(2, 2)))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Flatten())
model3.add(Dense(128))
model3.add(Activation('relu'))
model3.add(Dropout(0.5))
model3.add(Dense(1))
model3.add(Activation('sigmoid'))
model3.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history3 = model3.fit(X_train, y_train, batch_size=256, epochs=50, validation_split=0.2)
y_pred3 = model3.predict(X_test) >= 0.5
print('Confusion matrix:\n', confusion_matrix(y_test, y_pred3))
print('Classification report:\n', classification_report(y_test, y_pred3)) | code |
90144122/cell_10 | [
"image_output_1.png"
] | from sklearn.model_selection import train_test_split
from tensorflow.keras.layers import Conv2D, MaxPooling2D
from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten
from tensorflow.keras.models import Sequential
import numpy as np
import pandas as pd
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
X = np.array([e[0] for e in data]).astype('float32')
y = np.array([e[1] for e in data])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
X_train = np.resize(X_train, (len(X_train), 50, 50, 1))
X_test = np.resize(X_test, (len(X_test), 50, 50, 1))
X_train /= 255.0
X_test /= 255.0
model1 = Sequential()
model1.add(Conv2D(64, kernel_size=(3, 3), input_shape=X_train[0].shape))
model1.add(Activation('relu'))
model1.add(MaxPooling2D(pool_size=(2, 2)))
model1.add(Dropout(0.5))
model1.add(Conv2D(64, kernel_size=(3, 3)))
model1.add(Activation('relu'))
model1.add(MaxPooling2D(pool_size=(2, 2)))
model1.add(Flatten())
model1.add(Dense(64))
model1.add(Activation('relu'))
model1.add(Dropout(0.3))
model1.add(Dense(1))
model1.add(Activation('sigmoid'))
model1.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history1 = model1.fit(X_train, y_train, batch_size=32, epochs=10, validation_split=0.2)
history_df1 = pd.DataFrame(history1.history)
model2 = Sequential()
model2.add(Conv2D(128, kernel_size=(3, 3), input_shape=X_train[0].shape))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Dropout(0.5))
model2.add(Conv2D(128, kernel_size=(3, 3)))
model2.add(Activation('relu'))
model2.add(MaxPooling2D(pool_size=(2, 2)))
model2.add(Flatten())
model2.add(Dense(128))
model2.add(Activation('relu'))
model2.add(Dropout(0.3))
model2.add(Dense(1))
model2.add(Activation('sigmoid'))
model2.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history2 = model2.fit(X_train, y_train, batch_size=64, epochs=40, validation_split=0.2)
history_df2 = pd.DataFrame(history2.history)
history_df2.loc[:, ['loss', 'val_loss']].plot(title='Loss')
history_df2.loc[:, ['accuracy', 'val_accuracy']].plot(title='Accuracy') | code |
90144122/cell_12 | [
"text_plain_output_4.png",
"application_vnd.jupyter.stderr_output_3.png",
"text_plain_output_2.png",
"application_vnd.jupyter.stderr_output_1.png"
] | from sklearn.model_selection import train_test_split
from tensorflow.keras.layers import Conv2D, MaxPooling2D
from tensorflow.keras.layers import Dense, Dropout, Activation, Flatten
from tensorflow.keras.models import Sequential
import numpy as np
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
X = np.array([e[0] for e in data]).astype('float32')
y = np.array([e[1] for e in data])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
X_train = np.resize(X_train, (len(X_train), 50, 50, 1))
X_test = np.resize(X_test, (len(X_test), 50, 50, 1))
X_train /= 255.0
X_test /= 255.0
model3 = Sequential()
model3.add(Conv2D(256, kernel_size=(3, 2), input_shape=X_train[0].shape))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Dropout(0.3))
model3.add(Conv2D(256, kernel_size=(2, 3)))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Dropout(0.5))
model3.add(Conv2D(256, kernel_size=(2, 2)))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Dropout(0.5))
model3.add(Conv2D(256, kernel_size=(2, 2)))
model3.add(Activation('relu'))
model3.add(MaxPooling2D(pool_size=(2, 2)))
model3.add(Flatten())
model3.add(Dense(128))
model3.add(Activation('relu'))
model3.add(Dropout(0.5))
model3.add(Dense(1))
model3.add(Activation('sigmoid'))
model3.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history3 = model3.fit(X_train, y_train, batch_size=256, epochs=50, validation_split=0.2) | code |
90144122/cell_5 | [
"text_plain_output_1.png",
"image_output_2.png",
"image_output_1.png"
] | from sklearn.model_selection import train_test_split
import numpy as np
import pickle
data = pickle.load(open('../input/dog-and-cat/DOGnCAT50x50.pickle', 'rb'))
X = np.array([e[0] for e in data]).astype('float32')
y = np.array([e[1] for e in data])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
X_train = np.resize(X_train, (len(X_train), 50, 50, 1))
X_test = np.resize(X_test, (len(X_test), 50, 50, 1))
X_train /= 255.0
X_test /= 255.0
print('X_train shape: ', X_train.shape)
print('y_train shape: ', y_train.shape)
print('X_test shape: ', X_test.shape)
print('y_test shape: ', y_test.shape) | code |
90102125/cell_8 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd
df = pd.read_csv('../input/adult-census-income/adult.csv')
df.head() | code |
90102125/cell_24 | [
"text_plain_output_1.png"
] | from sklearn import metrics
from sklearn.model_selection import GridSearchCV
from sklearn.model_selection import train_test_split, cross_val_score
from sklearn.neural_network import MLPClassifier
from sklearn.preprocessing import MinMaxScaler
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/adult-census-income/adult.csv')
low = '<=50K'
y = df['income'].apply(lambda x: 0 if x == low else 1)
X = df.drop(['income'], axis=1)
X = pd.get_dummies(X)
from sklearn.model_selection import train_test_split, cross_val_score
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=1)
from sklearn.preprocessing import MinMaxScaler
scaler = MinMaxScaler()
scaler.fit(X_train)
X_train = scaler.transform(X_train)
X_test = scaler.transform(X_test)
from sklearn.neural_network import MLPClassifier
clf = MLPClassifier(random_state=1, max_iter=3000)
parameter_grid = {'hidden_layer_sizes': [3, (3, 3)]}
from sklearn.model_selection import GridSearchCV
gs = GridSearchCV(clf, parameter_grid, cv=5)
gs = gs.fit(X_train, y_train)
clf_best = gs.best_estimator_
clf_best = clf_best.fit(X_train, y_train)
from sklearn import metrics
y_pred = clf_best.predict(X_test)
print(metrics.accuracy_score(y_test, y_pred)) | code |
90102125/cell_14 | [
"text_html_output_1.png"
] | import pandas as pd
import pandas as pd
df = pd.read_csv('../input/adult-census-income/adult.csv')
X = df.drop(['income'], axis=1)
X = pd.get_dummies(X)
X.head(5) | code |
90102125/cell_22 | [
"text_html_output_1.png"
] | from sklearn.model_selection import GridSearchCV
from sklearn.model_selection import train_test_split, cross_val_score
from sklearn.neural_network import MLPClassifier
from sklearn.preprocessing import MinMaxScaler
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/adult-census-income/adult.csv')
low = '<=50K'
y = df['income'].apply(lambda x: 0 if x == low else 1)
X = df.drop(['income'], axis=1)
X = pd.get_dummies(X)
from sklearn.model_selection import train_test_split, cross_val_score
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=1)
from sklearn.preprocessing import MinMaxScaler
scaler = MinMaxScaler()
scaler.fit(X_train)
X_train = scaler.transform(X_train)
X_test = scaler.transform(X_test)
from sklearn.neural_network import MLPClassifier
clf = MLPClassifier(random_state=1, max_iter=3000)
parameter_grid = {'hidden_layer_sizes': [3, (3, 3)]}
from sklearn.model_selection import GridSearchCV
gs = GridSearchCV(clf, parameter_grid, cv=5)
gs = gs.fit(X_train, y_train)
clf_best = gs.best_estimator_
print('best model:', clf_best.get_params())
clf_best = clf_best.fit(X_train, y_train) | code |
90102125/cell_10 | [
"text_html_output_1.png"
] | import pandas as pd
import pandas as pd
df = pd.read_csv('../input/adult-census-income/adult.csv')
low = '<=50K'
y = df['income'].apply(lambda x: 0 if x == low else 1)
y.head(5) | code |
90102125/cell_12 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd
df = pd.read_csv('../input/adult-census-income/adult.csv')
X = df.drop(['income'], axis=1)
X.head(5) | code |
34137153/cell_4 | [
"text_html_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train = pd.read_csv('/kaggle/input/house-prices-advanced-regression-techniques/train.csv')
test = pd.read_csv('/kaggle/input/house-prices-advanced-regression-techniques/test.csv')
miss_values = train.isna().sum().sort_values(ascending=False).head(22).reset_index()
miss_values.rename(columns={'index': 'x', 0: 'y'}, inplace=True)
miss_values | code |
34137153/cell_1 | [
"text_plain_output_1.png"
] | import os
import numpy as np
import pandas as pd
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename)) | code |
34137153/cell_3 | [
"text_html_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train = pd.read_csv('/kaggle/input/house-prices-advanced-regression-techniques/train.csv')
test = pd.read_csv('/kaggle/input/house-prices-advanced-regression-techniques/test.csv')
train.describe() | code |
34137153/cell_5 | [
"image_output_1.png"
] | import matplotlib.pyplot as plt
import seaborn as sns
plt.rcParams['xtick.labelsize'] = 10
plt.rcParams['ytick.labelsize'] = 10
plt.figure(figsize=(4, 7))
color = sns.dark_palette('deeppink', reverse=True, n_colors=18)
ax = sns.barplot(x='y', y='x', data=miss_values, palette=color, orient='h')
plt.xticks(rotation=90)
plt.title('Columns Missing Values')
sns.despine()
plt.show() | code |
74052153/cell_9 | [
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
print(train.columns) | code |
74052153/cell_25 | [
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
train.dtypes.unique()
list(train.select_dtypes(include=['int64']).columns)
list(train.select_dtypes(include=['float64']).columns)
list(train.select_dtypes(include=['O']).columns)
train.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
test.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
train.isnull().mean().round(4).mul(100).sort_values(ascending=False)
train = train.loc[:, train.isnull().mean() < 0.4]
test = test.loc[:, test.isnull().mean() < 0.4]
train.shape
print(train.isnull().sum().sort_values(ascending=False)) | code |
74052153/cell_40 | [
"text_plain_output_1.png"
] | from sklearn.compose import ColumnTransformer
from sklearn.impute import SimpleImputer
from sklearn.impute import SimpleImputer
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import OneHotEncoder
import numpy as np
import pandas as pd
import xgboost as xgb
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
train.dtypes.unique()
list(train.select_dtypes(include=['int64']).columns)
list(train.select_dtypes(include=['float64']).columns)
list(train.select_dtypes(include=['O']).columns)
train.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
test.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
train.isnull().mean().round(4).mul(100).sort_values(ascending=False)
train = train.loc[:, train.isnull().mean() < 0.4]
test = test.loc[:, test.isnull().mean() < 0.4]
train.shape
from sklearn.impute import SimpleImputer
numerical_transformer = SimpleImputer(missing_values=np.NaN, strategy='mean')
categorical_transformer = Pipeline(steps=[('imputer', SimpleImputer(strategy='most_frequent')), ('onehot', OneHotEncoder(handle_unknown='ignore'))])
categorical_cols = [cname for cname in X_train.columns if X_train[cname].nunique() < 10 and X_train[cname].dtype == 'object']
numerical_cols = [cname for cname in X_train.columns if X_train[cname].dtype in ['int64', 'float64']]
my_cols = categorical_cols + numerical_cols
X_train = X_train[my_cols].copy()
X_test = X_test[my_cols].copy()
test = test[my_cols].copy()
preprocessor = ColumnTransformer(transformers=[('num', numerical_transformer, numerical_cols), ('cat', categorical_transformer, categorical_cols)])
X_train = preprocessor.fit_transform(X_train)
X_test = preprocessor.fit_transform(X_test)
test = preprocessor.fit_transform(test)
dtrain = xgb.DMatrix(X_train, label=y_train)
dtest = xgb.DMatrix(X_test, label=y_test)
params = {'max_depth': 6, 'min_child_weight': 1, 'eta': 0.3, 'subsample': 1, 'colsample_bytree': 1, 'objective': 'reg:linear'}
num_boost_round = 999
model = xgb.train(params, dtrain, num_boost_round=num_boost_round, evals=[(dtest, 'Test')], early_stopping_rounds=10)
cv_results = xgb.cv(params, dtrain, num_boost_round=num_boost_round, seed=42, nfold=5, metrics={'mae'}, early_stopping_rounds=10)
cv_results | code |
74052153/cell_39 | [
"text_plain_output_1.png"
] | from sklearn.compose import ColumnTransformer
from sklearn.impute import SimpleImputer
from sklearn.impute import SimpleImputer
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import OneHotEncoder
import numpy as np
import pandas as pd
import xgboost as xgb
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
train.dtypes.unique()
list(train.select_dtypes(include=['int64']).columns)
list(train.select_dtypes(include=['float64']).columns)
list(train.select_dtypes(include=['O']).columns)
train.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
test.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
train.isnull().mean().round(4).mul(100).sort_values(ascending=False)
train = train.loc[:, train.isnull().mean() < 0.4]
test = test.loc[:, test.isnull().mean() < 0.4]
train.shape
from sklearn.impute import SimpleImputer
numerical_transformer = SimpleImputer(missing_values=np.NaN, strategy='mean')
categorical_transformer = Pipeline(steps=[('imputer', SimpleImputer(strategy='most_frequent')), ('onehot', OneHotEncoder(handle_unknown='ignore'))])
categorical_cols = [cname for cname in X_train.columns if X_train[cname].nunique() < 10 and X_train[cname].dtype == 'object']
numerical_cols = [cname for cname in X_train.columns if X_train[cname].dtype in ['int64', 'float64']]
my_cols = categorical_cols + numerical_cols
X_train = X_train[my_cols].copy()
X_test = X_test[my_cols].copy()
test = test[my_cols].copy()
preprocessor = ColumnTransformer(transformers=[('num', numerical_transformer, numerical_cols), ('cat', categorical_transformer, categorical_cols)])
X_train = preprocessor.fit_transform(X_train)
X_test = preprocessor.fit_transform(X_test)
test = preprocessor.fit_transform(test)
dtrain = xgb.DMatrix(X_train, label=y_train)
dtest = xgb.DMatrix(X_test, label=y_test)
params = {'max_depth': 6, 'min_child_weight': 1, 'eta': 0.3, 'subsample': 1, 'colsample_bytree': 1, 'objective': 'reg:linear'}
num_boost_round = 999
model = xgb.train(params, dtrain, num_boost_round=num_boost_round, evals=[(dtest, 'Test')], early_stopping_rounds=10)
print('Best MAE: {:.2f} with {} rounds'.format(model.best_score, model.best_iteration + 1)) | code |
74052153/cell_41 | [
"text_plain_output_1.png"
] | from sklearn.compose import ColumnTransformer
from sklearn.impute import SimpleImputer
from sklearn.impute import SimpleImputer
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import OneHotEncoder
import numpy as np
import pandas as pd
import xgboost as xgb
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
train.dtypes.unique()
list(train.select_dtypes(include=['int64']).columns)
list(train.select_dtypes(include=['float64']).columns)
list(train.select_dtypes(include=['O']).columns)
train.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
test.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
train.isnull().mean().round(4).mul(100).sort_values(ascending=False)
train = train.loc[:, train.isnull().mean() < 0.4]
test = test.loc[:, test.isnull().mean() < 0.4]
train.shape
from sklearn.impute import SimpleImputer
numerical_transformer = SimpleImputer(missing_values=np.NaN, strategy='mean')
categorical_transformer = Pipeline(steps=[('imputer', SimpleImputer(strategy='most_frequent')), ('onehot', OneHotEncoder(handle_unknown='ignore'))])
categorical_cols = [cname for cname in X_train.columns if X_train[cname].nunique() < 10 and X_train[cname].dtype == 'object']
numerical_cols = [cname for cname in X_train.columns if X_train[cname].dtype in ['int64', 'float64']]
my_cols = categorical_cols + numerical_cols
X_train = X_train[my_cols].copy()
X_test = X_test[my_cols].copy()
test = test[my_cols].copy()
preprocessor = ColumnTransformer(transformers=[('num', numerical_transformer, numerical_cols), ('cat', categorical_transformer, categorical_cols)])
X_train = preprocessor.fit_transform(X_train)
X_test = preprocessor.fit_transform(X_test)
test = preprocessor.fit_transform(test)
dtrain = xgb.DMatrix(X_train, label=y_train)
dtest = xgb.DMatrix(X_test, label=y_test)
params = {'max_depth': 6, 'min_child_weight': 1, 'eta': 0.3, 'subsample': 1, 'colsample_bytree': 1, 'objective': 'reg:linear'}
num_boost_round = 999
model = xgb.train(params, dtrain, num_boost_round=num_boost_round, evals=[(dtest, 'Test')], early_stopping_rounds=10)
cv_results = xgb.cv(params, dtrain, num_boost_round=num_boost_round, seed=42, nfold=5, metrics={'mae'}, early_stopping_rounds=10)
cv_results
cv_results['test-mae-mean'].min() | code |
74052153/cell_11 | [
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
print(train.info) | code |
74052153/cell_7 | [
"text_html_output_1.png",
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.head() | code |
74052153/cell_18 | [
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
train.dtypes.unique()
list(train.select_dtypes(include=['int64']).columns)
list(train.select_dtypes(include=['float64']).columns)
list(train.select_dtypes(include=['O']).columns) | code |
74052153/cell_8 | [
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
print('The train dataset have the shape', train.shape)
print('The test dataset have the shape', test.shape) | code |
74052153/cell_16 | [
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
train.dtypes.unique()
list(train.select_dtypes(include=['int64']).columns)
list(train.select_dtypes(include=['float64']).columns) | code |
74052153/cell_35 | [
"text_plain_output_1.png"
] | from sklearn.impute import SimpleImputer
from sklearn.impute import SimpleImputer
from sklearn.metrics import mean_absolute_error
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import OneHotEncoder
import numpy as np
from sklearn.impute import SimpleImputer
numerical_transformer = SimpleImputer(missing_values=np.NaN, strategy='mean')
categorical_transformer = Pipeline(steps=[('imputer', SimpleImputer(strategy='most_frequent')), ('onehot', OneHotEncoder(handle_unknown='ignore'))])
mean_train = np.mean(y_train)
baseline_predictions = np.ones(y_test.shape) * mean_train
mae_baseline = mean_absolute_error(y_test, baseline_predictions)
print('Baseline MAE is {:.2f}'.format(mae_baseline)) | code |
74052153/cell_24 | [
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
train.dtypes.unique()
list(train.select_dtypes(include=['int64']).columns)
list(train.select_dtypes(include=['float64']).columns)
list(train.select_dtypes(include=['O']).columns)
train.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
test.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
train.isnull().mean().round(4).mul(100).sort_values(ascending=False)
train = train.loc[:, train.isnull().mean() < 0.4]
test = test.loc[:, test.isnull().mean() < 0.4]
train.shape | code |
74052153/cell_14 | [
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
train.dtypes.unique()
list(train.select_dtypes(include=['int64']).columns) | code |
74052153/cell_22 | [
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
train.dtypes.unique()
list(train.select_dtypes(include=['int64']).columns)
list(train.select_dtypes(include=['float64']).columns)
list(train.select_dtypes(include=['O']).columns)
train.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
test.drop(['Id', 'YrSold', 'MoSold'], inplace=True, axis=1)
train.isnull().mean().round(4).mul(100).sort_values(ascending=False) | code |
74052153/cell_10 | [
"text_html_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe | code |
74052153/cell_12 | [
"text_plain_output_1.png"
] | import pandas as pd
train = pd.read_csv('../input/home-data-for-ml-course/train.csv')
test = pd.read_csv('../input/home-data-for-ml-course/test.csv')
id = test['Id']
train.describe
train.dtypes.unique() | code |
49124211/cell_9 | [
"text_plain_output_1.png"
] | import cv2
import numpy as np
import os
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/drive-and-act/iccv_activities_3s/activities_3s/kinect_color/tasklevel.chunks_90.split_0.train.csv')
root_path = '../input/drive-and-act/kinect_color/kinect_color/'
sample_rate = 5
for j in range(1):
file_names = []
labels = []
length = []
for i in range(df.shape[0]):
if i > 0:
if root_path + df.iloc[i - 1, 1] + '.mp4' != root_path + df.iloc[i, 1] + '.mp4':
path_video = root_path + df.iloc[i, 1] + '.mp4'
cap = cv2.VideoCapture(path_video)
count_frame = 0
else:
path_video = root_path + df.iloc[i, 1] + '.mp4'
cap = cv2.VideoCapture(path_video)
count_frame = 0
frame_start = df.iloc[i, 3]
frame_end = df.iloc[i, 4]
label = df.iloc[i, 5]
length.append(frame_end - frame_start)
if frame_end - frame_start > 39:
root_save = '/kaggle/working/Trainsplit_0/' + df.iloc[i, 1] + '/' + df.iloc[i, 1].split('/')[-1] + '-' + str(frame_start) + '-' + str(frame_end - frame_start)
file_names.append(root_save)
labels.append(label)
lenn = 0
while True:
ret, frame = cap.read()
if count_frame >= frame_start and count_frame <= frame_end and (ret == True) and ((count_frame - frame_start) % sample_rate == 0) and (lenn < 9):
lenn += 1
frame = cv2.resize(frame, (256, 256))
if not os.path.exists(root_save):
os.makedirs(root_save)
save_path = df.iloc[i, 1].split('/')[-1] + str(count_frame) + '.jpg'
final_path = root_save + '/' + save_path
cv2.imwrite(final_path, frame)
count_frame += 1
if count_frame >= frame_end or ret == False:
break
np.save('/kaggle/working/Trainsplit_0_file_names.npy', file_names)
np.save('/kaggle/working/Trainsplit_0_labels.npy', labels)
len(labels) | code |
49124211/cell_4 | [
"text_html_output_1.png"
] | ! nvidia-smi | code |
49124211/cell_6 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd
df = pd.read_csv('../input/drive-and-act/iccv_activities_3s/activities_3s/kinect_color/tasklevel.chunks_90.split_0.train.csv')
df.head() | code |
49124211/cell_2 | [
"text_plain_output_1.png"
] | import psutil
import psutil
def get_size(bytes, suffix='B'):
factor = 1024
for unit in ['', 'K', 'M', 'G', 'T', 'P']:
if bytes < factor:
return f'{bytes:.2f}{unit}{suffix}'
bytes /= factor
print('=' * 40, 'Memory Information', '=' * 40)
svmem = psutil.virtual_memory()
print(f'Total: {get_size(svmem.total)}')
print(f'Available: {get_size(svmem.available)}')
print(f'Used: {get_size(svmem.used)}')
print(f'Percentage: {svmem.percent}%') | code |
49124211/cell_8 | [
"text_plain_output_1.png"
] | import cv2
import numpy as np
import os
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/drive-and-act/iccv_activities_3s/activities_3s/kinect_color/tasklevel.chunks_90.split_0.train.csv')
root_path = '../input/drive-and-act/kinect_color/kinect_color/'
sample_rate = 5
for j in range(1):
file_names = []
labels = []
length = []
for i in range(df.shape[0]):
if i > 0:
if root_path + df.iloc[i - 1, 1] + '.mp4' != root_path + df.iloc[i, 1] + '.mp4':
path_video = root_path + df.iloc[i, 1] + '.mp4'
cap = cv2.VideoCapture(path_video)
count_frame = 0
else:
path_video = root_path + df.iloc[i, 1] + '.mp4'
cap = cv2.VideoCapture(path_video)
count_frame = 0
frame_start = df.iloc[i, 3]
frame_end = df.iloc[i, 4]
label = df.iloc[i, 5]
length.append(frame_end - frame_start)
if frame_end - frame_start > 39:
root_save = '/kaggle/working/Trainsplit_0/' + df.iloc[i, 1] + '/' + df.iloc[i, 1].split('/')[-1] + '-' + str(frame_start) + '-' + str(frame_end - frame_start)
file_names.append(root_save)
labels.append(label)
lenn = 0
while True:
ret, frame = cap.read()
if count_frame >= frame_start and count_frame <= frame_end and (ret == True) and ((count_frame - frame_start) % sample_rate == 0) and (lenn < 9):
lenn += 1
frame = cv2.resize(frame, (256, 256))
if not os.path.exists(root_save):
os.makedirs(root_save)
save_path = df.iloc[i, 1].split('/')[-1] + str(count_frame) + '.jpg'
final_path = root_save + '/' + save_path
cv2.imwrite(final_path, frame)
count_frame += 1
if count_frame >= frame_end or ret == False:
break
np.save('/kaggle/working/Trainsplit_0_file_names.npy', file_names)
np.save('/kaggle/working/Trainsplit_0_labels.npy', labels)
np.unique(length, return_counts=True) | code |
49124211/cell_10 | [
"text_plain_output_1.png"
] | import cv2
import numpy as np
import os
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/drive-and-act/iccv_activities_3s/activities_3s/kinect_color/tasklevel.chunks_90.split_0.train.csv')
root_path = '../input/drive-and-act/kinect_color/kinect_color/'
sample_rate = 5
for j in range(1):
file_names = []
labels = []
length = []
for i in range(df.shape[0]):
if i > 0:
if root_path + df.iloc[i - 1, 1] + '.mp4' != root_path + df.iloc[i, 1] + '.mp4':
path_video = root_path + df.iloc[i, 1] + '.mp4'
cap = cv2.VideoCapture(path_video)
count_frame = 0
else:
path_video = root_path + df.iloc[i, 1] + '.mp4'
cap = cv2.VideoCapture(path_video)
count_frame = 0
frame_start = df.iloc[i, 3]
frame_end = df.iloc[i, 4]
label = df.iloc[i, 5]
length.append(frame_end - frame_start)
if frame_end - frame_start > 39:
root_save = '/kaggle/working/Trainsplit_0/' + df.iloc[i, 1] + '/' + df.iloc[i, 1].split('/')[-1] + '-' + str(frame_start) + '-' + str(frame_end - frame_start)
file_names.append(root_save)
labels.append(label)
lenn = 0
while True:
ret, frame = cap.read()
if count_frame >= frame_start and count_frame <= frame_end and (ret == True) and ((count_frame - frame_start) % sample_rate == 0) and (lenn < 9):
lenn += 1
frame = cv2.resize(frame, (256, 256))
if not os.path.exists(root_save):
os.makedirs(root_save)
save_path = df.iloc[i, 1].split('/')[-1] + str(count_frame) + '.jpg'
final_path = root_save + '/' + save_path
cv2.imwrite(final_path, frame)
count_frame += 1
if count_frame >= frame_end or ret == False:
break
np.save('/kaggle/working/Trainsplit_0_file_names.npy', file_names)
np.save('/kaggle/working/Trainsplit_0_labels.npy', labels)
len(file_names) | code |
104119293/cell_4 | [
"application_vnd.jupyter.stderr_output_1.png"
] | import matplotlib.pyplot as plt
import numpy as np
import numpy as np
import numpy.linalg
from sklearn import linear_model
from sklearn.linear_model import LinearRegression
from sklearn.preprocessing import PolynomialFeatures
import matplotlib.pyplot as plt
def theta_m2_m1(n2=1, n1=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = (a / mu_1) ** n1 * ((a + b + c) / (mu_1 + mu_2)) ** n2
return p
def theta_m1_m2(n2=1, n1=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = (b / mu_2) ** n2 * ((a + b + c) / (mu_1 + mu_2)) ** n1
return p
def theta_m2_0m1(n2=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = mu_1 / mu_2 * ((a + b + c) / (mu_1 + mu_2)) ** n2
return p
def theta_m1_0m2(n1=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = mu_1 / mu_2 * ((a + b + c) / (mu_1 + mu_2)) ** n1
return p
def theta_m2(n2=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = (mu_1 + mu_2) / (a + c) * (b / mu_2) ** n2
return p
def theta_m1(n1=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = (mu_1 + mu_2) / (b + c) * (a / mu_1) ** n1
return p
def theta_0s(a=1, b=1, c=1, mu_1=1, mu_2=1):
A = np.array([[0, a + b + c + mu_2, -(b + c / 2)], [a + b + c + mu_1, 0, -(a + c / 2)], [-mu_1, -mu_2, a + b + c]])
b = np.array([[(mu_1 + mu_2) * b / (a + c) + 2 * mu_1 ** 2 / mu_2], [(mu_1 + mu_2) * a / (b + c) + 2 * mu_1], [0]])
A_inv = np.linalg.inv(A)
ps = np.matmul(A_inv, b)
return ps
def calc_visschers(a=1, b=1, c=1, mu_1=1, mu_2=1, k=100):
p_list = []
t0 = theta_0s(a, b, c, mu_1, mu_2)[2][0]
p_list.append(t0)
t1busy = theta_0s(a, b, c, mu_1, mu_2)[0][0]
p_list.append(t1busy)
t2busy = theta_0s(a, b, c, mu_1, mu_2)[1][0]
p_list.append(t2busy)
tbothbusy = theta_m2_0m1(n2=0, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2) + theta_m1_0m2(n1=0, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(tbothbusy)
for n in range(1, k):
for m in range(1, k):
t1 = theta_m2_m1(n2=m, n1=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
t2 = theta_m1_m2(n2=m, n1=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(t1)
p_list.append(t2)
t3 = theta_m2_0m1(n2=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(t3)
t4 = theta_m1_0m2(n1=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(t4)
t5 = theta_m2(n2=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(t5)
t6 = theta_m1(n1=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(t6)
x = np.array(p_list)
x = x / x.sum()
probs_array = x[0:4]
return x
rlist = []
for k in range(100, 3000, 10):
r = calc_visschers(a=4.75, b=4.75, c=0, mu_1=10, mu_2=10, k=k)
rlist.append(r[0])
r = np.linspace(0, 1)
min_p = np.argmin(avg_wait)
min_val = r[min_p]
print(min_val)
def line(m, x, b):
return m * x + b
for i in range(min_p - 10, min_p):
x1 = r[i]
y1 = avg_wait[i]
c = 0
for j in range(min_p + 1, min_p + 10):
x2 = r[j]
y2 = avg_wait[j]
m = (y1 - y2) / (x1 - x2)
b = (x1 * y2 - x2 * y1) / (x1 - x2)
lines = []
true_func = []
for p in range(i + 1, j):
c += 1
x = r[p]
y_line = line(m=m, x=x, b=b)
lines.append(y_line)
y_true = avg_wait[p]
true_func.append(y_true)
if y_true > y_line:
print('non-convex between x1=' + str(x1) + ' and x2=' + str(x2) + ' at x=' + str(x) + ' with line value ' + str(y_line) + ' and average wait value ' + str(y_true))
if c % 10 == 0:
plt.plot(r[i + 1:j], lines)
plt.plot(r, avg_wait) | code |
104119293/cell_2 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import matplotlib.pyplot as plt
import numpy as np
import numpy as np
import numpy.linalg
from sklearn import linear_model
from sklearn.linear_model import LinearRegression
from sklearn.preprocessing import PolynomialFeatures
import matplotlib.pyplot as plt
def theta_m2_m1(n2=1, n1=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = (a / mu_1) ** n1 * ((a + b + c) / (mu_1 + mu_2)) ** n2
return p
def theta_m1_m2(n2=1, n1=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = (b / mu_2) ** n2 * ((a + b + c) / (mu_1 + mu_2)) ** n1
return p
def theta_m2_0m1(n2=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = mu_1 / mu_2 * ((a + b + c) / (mu_1 + mu_2)) ** n2
return p
def theta_m1_0m2(n1=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = mu_1 / mu_2 * ((a + b + c) / (mu_1 + mu_2)) ** n1
return p
def theta_m2(n2=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = (mu_1 + mu_2) / (a + c) * (b / mu_2) ** n2
return p
def theta_m1(n1=1, a=1, b=1, c=1, mu_1=1, mu_2=1):
p = (mu_1 + mu_2) / (b + c) * (a / mu_1) ** n1
return p
def theta_0s(a=1, b=1, c=1, mu_1=1, mu_2=1):
A = np.array([[0, a + b + c + mu_2, -(b + c / 2)], [a + b + c + mu_1, 0, -(a + c / 2)], [-mu_1, -mu_2, a + b + c]])
b = np.array([[(mu_1 + mu_2) * b / (a + c) + 2 * mu_1 ** 2 / mu_2], [(mu_1 + mu_2) * a / (b + c) + 2 * mu_1], [0]])
A_inv = np.linalg.inv(A)
ps = np.matmul(A_inv, b)
return ps
def calc_visschers(a=1, b=1, c=1, mu_1=1, mu_2=1, k=100):
p_list = []
t0 = theta_0s(a, b, c, mu_1, mu_2)[2][0]
p_list.append(t0)
t1busy = theta_0s(a, b, c, mu_1, mu_2)[0][0]
p_list.append(t1busy)
t2busy = theta_0s(a, b, c, mu_1, mu_2)[1][0]
p_list.append(t2busy)
tbothbusy = theta_m2_0m1(n2=0, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2) + theta_m1_0m2(n1=0, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(tbothbusy)
for n in range(1, k):
for m in range(1, k):
t1 = theta_m2_m1(n2=m, n1=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
t2 = theta_m1_m2(n2=m, n1=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(t1)
p_list.append(t2)
t3 = theta_m2_0m1(n2=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(t3)
t4 = theta_m1_0m2(n1=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(t4)
t5 = theta_m2(n2=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(t5)
t6 = theta_m1(n1=n, a=a, b=b, c=c, mu_1=mu_1, mu_2=mu_2)
p_list.append(t6)
x = np.array(p_list)
x = x / x.sum()
probs_array = x[0:4]
return x
rlist = []
for k in range(100, 3000, 10):
r = calc_visschers(a=4.75, b=4.75, c=0, mu_1=10, mu_2=10, k=k)
rlist.append(r[0])
plt.plot(range(100, 3000, 10), rlist) | code |
73075873/cell_21 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import matplotlib.pyplot as plt
import matplotlib.pyplot as plt
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
import seaborn as sns
data = pd.read_csv('/kaggle/input/loan-prediction-based-on-customer-behavior/Training Data.csv')
import matplotlib.pyplot as plt
total = list(data.Risk_Flag.value_counts())
Flag0 = total[0]
Flag1 = total[1]
import seaborn as sns
g=sns.catplot(x='STATE', data=data, height=12, aspect=1.5, kind='count', palette='deep')
g.set_xticklabels(rotation=60)
import seaborn as sns
import matplotlib.pyplot as plt
plt.xticks(rotation=60)
data['Age_group'] = pd.qcut(data.Age, 5)
g = sns.FacetGrid(data=data, row='House_Ownership', col='Married/Single', height=5, aspect=1.5)
g.map_dataframe(sns.barplot, x='Age_group', y='Risk_Flag', ci=None)
g.set_xticklabels(rotation=60)
data.corr().Risk_Flag.drop(['Risk_Flag', 'Id']).plot.bar() | code |
73075873/cell_4 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
data = pd.read_csv('/kaggle/input/loan-prediction-based-on-customer-behavior/Training Data.csv')
data.head() | code |
73075873/cell_23 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import matplotlib.pyplot as plt
import matplotlib.pyplot as plt
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
import seaborn as sns
data = pd.read_csv('/kaggle/input/loan-prediction-based-on-customer-behavior/Training Data.csv')
import matplotlib.pyplot as plt
total = list(data.Risk_Flag.value_counts())
Flag0 = total[0]
Flag1 = total[1]
import seaborn as sns
g=sns.catplot(x='STATE', data=data, height=12, aspect=1.5, kind='count', palette='deep')
g.set_xticklabels(rotation=60)
import seaborn as sns
import matplotlib.pyplot as plt
plt.xticks(rotation=60)
data['Age_group'] = pd.qcut(data.Age, 5)
g = sns.FacetGrid(data=data, row='House_Ownership', col='Married/Single', height=5, aspect=1.5)
g.map_dataframe(sns.barplot, x='Age_group', y='Risk_Flag', ci=None)
g.set_xticklabels(rotation=60)
dummies = pd.get_dummies(data[['STATE', 'Profession']])
dummies.drop(dummies.columns[[0, -1]], axis=1, inplace=True)
features = ['Income', 'Age', 'Experience', 'CURRENT_JOB_YRS', 'CURRENT_HOUSE_YRS', 'Married/Single', 'House_Ownership', 'Car_Ownership']
X = pd.concat([data[features], dummies], axis=1)
y = data['Risk_Flag']
X.head() | code |
73075873/cell_30 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from imblearn.ensemble import BalancedRandomForestClassifier
from imblearn.over_sampling import ADASYN
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import f1_score, accuracy_score, roc_auc_score, plot_roc_curve, plot_confusion_matrix
import matplotlib.pyplot as plt
import matplotlib.pyplot as plt
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
import seaborn as sns
data = pd.read_csv('/kaggle/input/loan-prediction-based-on-customer-behavior/Training Data.csv')
import matplotlib.pyplot as plt
total = list(data.Risk_Flag.value_counts())
Flag0 = total[0]
Flag1 = total[1]
import seaborn as sns
g=sns.catplot(x='STATE', data=data, height=12, aspect=1.5, kind='count', palette='deep')
g.set_xticklabels(rotation=60)
import seaborn as sns
import matplotlib.pyplot as plt
plt.xticks(rotation=60)
from imblearn.ensemble import BalancedRandomForestClassifier
from sklearn.metrics import f1_score, accuracy_score, roc_auc_score, plot_roc_curve, plot_confusion_matrix
brf=BalancedRandomForestClassifier().fit(X_train, y_train)
fig, (ax1, ax2) = plt.subplots(1, 2, figsize = (16,6))
plt.title('asfafasf')
ax1.set_title('Confusion matrix (Balanced RF)')
ax2.set_title('ROC curve (Balanced RF)')
ax2.plot([0,1], [0,1], 'g--', alpha=0.25)
plot_confusion_matrix(brf, X_test, y_test, cmap=plt.cm.Blues, normalize='true', ax=ax1)
plot_roc_curve(brf, X_test, y_test, ax=ax2)
y_pred = brf.predict(X_test)
acc_brf=accuracy_score(y_test, y_pred)
f1_brf=f1_score(y_test, y_pred)
roc_brf=roc_auc_score(y_test, y_pred)
print('Roc_Auc score: %.3f' %roc_brf)
from imblearn.over_sampling import ADASYN
ada = ADASYN(random_state=42)
X_ada, y_ada = ada.fit_resample(X_train, y_train)
from sklearn.ensemble import RandomForestClassifier
rf_ada = RandomForestClassifier().fit(X_ada, y_ada)
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(16, 6))
ax1.set_title('Confusion matrix (RF and ADASYN)')
ax2.set_title('ROC curve (RF and ADASYN)')
ax2.plot([0, 1], [0, 1], 'g--', alpha=0.25)
plot_confusion_matrix(rf_ada, X_test, y_test, cmap=plt.cm.Blues, normalize='true', ax=ax1)
plot_roc_curve(rf_ada, X_test, y_test, ax=ax2)
y_pred = rf_ada.predict(X_test)
acc_ada = accuracy_score(y_test, y_pred)
f1_ada = f1_score(y_test, y_pred)
roc_ada = roc_auc_score(y_test, y_pred)
print('Roc_Auc score: %.3f' % roc_ada) | code |
73075873/cell_33 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from imblearn.combine import SMOTETomek
from imblearn.ensemble import BalancedRandomForestClassifier
from imblearn.over_sampling import ADASYN
from imblearn.under_sampling import TomekLinks
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import f1_score, accuracy_score, roc_auc_score, plot_roc_curve, plot_confusion_matrix
import matplotlib.pyplot as plt
import matplotlib.pyplot as plt
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
import seaborn as sns
data = pd.read_csv('/kaggle/input/loan-prediction-based-on-customer-behavior/Training Data.csv')
import matplotlib.pyplot as plt
total = list(data.Risk_Flag.value_counts())
Flag0 = total[0]
Flag1 = total[1]
import seaborn as sns
g=sns.catplot(x='STATE', data=data, height=12, aspect=1.5, kind='count', palette='deep')
g.set_xticklabels(rotation=60)
import seaborn as sns
import matplotlib.pyplot as plt
plt.xticks(rotation=60)
from imblearn.ensemble import BalancedRandomForestClassifier
from sklearn.metrics import f1_score, accuracy_score, roc_auc_score, plot_roc_curve, plot_confusion_matrix
brf=BalancedRandomForestClassifier().fit(X_train, y_train)
fig, (ax1, ax2) = plt.subplots(1, 2, figsize = (16,6))
plt.title('asfafasf')
ax1.set_title('Confusion matrix (Balanced RF)')
ax2.set_title('ROC curve (Balanced RF)')
ax2.plot([0,1], [0,1], 'g--', alpha=0.25)
plot_confusion_matrix(brf, X_test, y_test, cmap=plt.cm.Blues, normalize='true', ax=ax1)
plot_roc_curve(brf, X_test, y_test, ax=ax2)
y_pred = brf.predict(X_test)
acc_brf=accuracy_score(y_test, y_pred)
f1_brf=f1_score(y_test, y_pred)
roc_brf=roc_auc_score(y_test, y_pred)
print('Roc_Auc score: %.3f' %roc_brf)
from imblearn.over_sampling import ADASYN
ada = ADASYN(random_state=42)
X_ada, y_ada = ada.fit_resample(X_train, y_train)
from sklearn.ensemble import RandomForestClassifier
rf_ada=RandomForestClassifier().fit(X_ada, y_ada)
fig, (ax1, ax2) = plt.subplots(1, 2, figsize = (16,6))
ax1.set_title('Confusion matrix (RF and ADASYN)')
ax2.set_title('ROC curve (RF and ADASYN)')
ax2.plot([0,1], [0,1], 'g--', alpha=0.25)
plot_confusion_matrix(rf_ada,X_test, y_test, cmap=plt.cm.Blues, normalize='true', ax=ax1)
plot_roc_curve(rf_ada, X_test, y_test, ax=ax2)
y_pred = rf_ada.predict(X_test)
acc_ada=accuracy_score(y_test, y_pred)
f1_ada=f1_score(y_test, y_pred)
roc_ada=roc_auc_score(y_test, y_pred)
print('Roc_Auc score: %.3f' %roc_ada)
from imblearn.combine import SMOTETomek
from imblearn.under_sampling import TomekLinks
smt = SMOTETomek(tomek=TomekLinks(sampling_strategy='majority'))
X_smt, y_smt = smt.fit_resample(X_train, y_train)
rf_smt = RandomForestClassifier().fit(X_smt, y_smt)
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(16, 6))
ax1.set_title('Confusion matrix (RF and SMOTETomek)')
ax2.set_title('ROC curve (RF and SMOTETomek)')
ax2.plot([0, 1], [0, 1], 'g--', alpha=0.25)
plot_confusion_matrix(rf_smt, X_test, y_test, cmap=plt.cm.Blues, normalize='true', ax=ax1)
plot_roc_curve(rf_smt, X_test, y_test, ax=ax2)
y_pred = rf_smt.predict(X_test)
acc_smt = accuracy_score(y_test, y_pred)
f1_smt = f1_score(y_test, y_pred)
roc_smt = roc_auc_score(y_test, y_pred)
print('Roc_Auc score: %.3f' % roc_smt) | code |
73075873/cell_29 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from imblearn.over_sampling import ADASYN
from imblearn.over_sampling import ADASYN
print('Initial size:', X_train.shape)
ada = ADASYN(random_state=42)
X_ada, y_ada = ada.fit_resample(X_train, y_train)
print('Resampled size:', X_ada.shape) | code |
73075873/cell_2 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import os
import numpy as np
import pandas as pd
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename)) | code |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.