path
stringlengths 13
17
| screenshot_names
sequencelengths 1
873
| code
stringlengths 0
40.4k
| cell_type
stringclasses 1
value |
|---|---|---|---|
129000049/cell_45 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LogisticRegression
import numpy as np
import pandas as pd
wdf = pd.read_csv('/kaggle/input/historical-weather-data-for-indian-cities/jaipur.csv', parse_dates=['date_time'], index_col='date_time')
train_X.shape
train_y.shape
model = LinearRegression()
model.fit(train_X, train_y)
prediction = model.predict(test_X)
mae = np.mean(np.absolute(prediction - test_y))
variance_score = model.score(test_X, test_y)
prediction = np.round(prediction, 2)
results = pd.DataFrame({'Actual': test_y, 'Prediction': prediction, 'Difference': test_y - prediction})
model = LogisticRegression()
model.fit(train_X, train_y)
prediction = model.predict(test_X)
results = pd.DataFrame({'Actual': test_y, 'Prediction': prediction, 'Difference': test_y - prediction})
print(results) | code |
129000049/cell_18 | [
"text_plain_output_1.png"
] | import pandas as pd
wdf = pd.read_csv('/kaggle/input/historical-weather-data-for-indian-cities/jaipur.csv', parse_dates=['date_time'], index_col='date_time')
wdf.columns
wdf.shape
wdf.isnull().any()
wdf_num = wdf.loc[:, ['mintempC', 'tempC', 'HeatIndexC', 'pressure']]
wdf_num.shape
wdf_num.columns
weth = wdf_num['2019':'2020']
weth.head() | code |
129000049/cell_8 | [
"text_plain_output_1.png"
] | import pandas as pd
wdf = pd.read_csv('/kaggle/input/historical-weather-data-for-indian-cities/jaipur.csv', parse_dates=['date_time'], index_col='date_time')
wdf.columns
wdf.shape | code |
129000049/cell_15 | [
"text_plain_output_1.png"
] | import pandas as pd
wdf = pd.read_csv('/kaggle/input/historical-weather-data-for-indian-cities/jaipur.csv', parse_dates=['date_time'], index_col='date_time')
wdf.columns
wdf.shape
wdf.isnull().any()
wdf_num = wdf.loc[:, ['mintempC', 'tempC', 'HeatIndexC', 'pressure']]
wdf_num.shape | code |
129000049/cell_38 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LogisticRegression
import numpy as np
import pandas as pd
wdf = pd.read_csv('/kaggle/input/historical-weather-data-for-indian-cities/jaipur.csv', parse_dates=['date_time'], index_col='date_time')
train_X.shape
train_y.shape
model = LinearRegression()
model.fit(train_X, train_y)
prediction = model.predict(test_X)
mae = np.mean(np.absolute(prediction - test_y))
variance_score = model.score(test_X, test_y)
prediction = np.round(prediction, 2)
results = pd.DataFrame({'Actual': test_y, 'Prediction': prediction, 'Difference': test_y - prediction})
model = LogisticRegression()
model.fit(train_X, train_y) | code |
129000049/cell_17 | [
"text_html_output_1.png"
] | import pandas as pd
wdf = pd.read_csv('/kaggle/input/historical-weather-data-for-indian-cities/jaipur.csv', parse_dates=['date_time'], index_col='date_time')
wdf.columns
wdf.shape
wdf.isnull().any()
wdf_num = wdf.loc[:, ['mintempC', 'tempC', 'HeatIndexC', 'pressure']]
wdf_num.shape
wdf_num.columns | code |
129000049/cell_31 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LinearRegression
from sklearn.linear_model import LinearRegression
train_X.shape
train_y.shape
model = LinearRegression()
model.fit(train_X, train_y) | code |
129000049/cell_22 | [
"text_html_output_1.png"
] | train_X.shape | code |
1008454/cell_6 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd
gTemp = pd.read_csv('../input/GlobalTemperatures.csv')
gTempCountry = pd.read_csv('../input/GlobalLandTemperaturesByCountry.csv')
gTempState = pd.read_csv('../input/GlobalLandTemperaturesByState.csv')
gTempMajorCity = pd.read_csv('../input/GlobalLandTemperaturesByMajorCity.csv')
gTempCity = pd.read_csv('../input/GlobalLandTemperaturesByCity.csv')
plt.figure(figsize=(12, 6))
gTemp.groupby(by='Year').mean()['LandAndOceanAverageTemperature'].dropna().plot() | code |
1008454/cell_3 | [
"text_html_output_1.png"
] | import pandas as pd
gTemp = pd.read_csv('../input/GlobalTemperatures.csv')
gTempCountry = pd.read_csv('../input/GlobalLandTemperaturesByCountry.csv')
gTempState = pd.read_csv('../input/GlobalLandTemperaturesByState.csv')
gTempMajorCity = pd.read_csv('../input/GlobalLandTemperaturesByMajorCity.csv')
gTempCity = pd.read_csv('../input/GlobalLandTemperaturesByCity.csv')
gTemp.head(5) | code |
1008454/cell_5 | [
"text_plain_output_1.png"
] | import pandas as pd
gTemp = pd.read_csv('../input/GlobalTemperatures.csv')
gTempCountry = pd.read_csv('../input/GlobalLandTemperaturesByCountry.csv')
gTempState = pd.read_csv('../input/GlobalLandTemperaturesByState.csv')
gTempMajorCity = pd.read_csv('../input/GlobalLandTemperaturesByMajorCity.csv')
gTempCity = pd.read_csv('../input/GlobalLandTemperaturesByCity.csv')
gTemp.info() | code |
330673/cell_13 | [
"text_plain_output_1.png"
] | from scipy.stats import chisquare
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/people.csv')
from scipy.stats import chisquare
chars = [i for i in df.columns.values if 'char_' in i]
flags = []
for feat in df[chars]:
group = df[chars].groupby(feat)
for otherfeat in df[chars].drop(feat, axis=1):
summary = group[otherfeat].count()
if chisquare(summary)[1] < 0.05:
flags.append(feat)
flags.append(otherfeat)
flags = set(flags)
dums = pd.get_dummies(df[chars])
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
scaledums = MinMaxScaler().fit_transform(dums)
pca = PCA(n_components=2)
featurecomponents = pca.fit_transform(scaledums)
import numpy as np
components = {}
index = 0
for feature in dums.columns.values:
components[feature] = [pca.components_[0][index]]
index += 1
sortedcomps = pca.components_[0]
sortedcomps.sort()
maxcap = sortedcomps[-3]
mincap = sortedcomps[2]
components = {i: x for i, x in components.items() if x >= maxcap or x <= mincap}
components = pd.DataFrame(components)
featurecomponents = pd.DataFrame(featurecomponents, columns=['Principle Component 1', 'Principle Component 2'])
df['Principle Component 1'] = featurecomponents['Principle Component 1']
featurecomponents['group_1'] = df['group_1']
groupslist = list(set(featurecomponents['group_1'].tolist()))
group = featurecomponents[featurecomponents['group_1'] == groupslist[0]]
group.plot(kind='scatter', x='Principle Component 1', y='Principle Component 2', figsize=(3, 3))
print('There are {} data points in this group.'.format(len(group.index))) | code |
330673/cell_9 | [
"text_plain_output_1.png"
] | from scipy.stats import chisquare
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/people.csv')
from scipy.stats import chisquare
chars = [i for i in df.columns.values if 'char_' in i]
flags = []
for feat in df[chars]:
group = df[chars].groupby(feat)
for otherfeat in df[chars].drop(feat, axis=1):
summary = group[otherfeat].count()
if chisquare(summary)[1] < 0.05:
flags.append(feat)
flags.append(otherfeat)
flags = set(flags)
dums = pd.get_dummies(df[chars])
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
scaledums = MinMaxScaler().fit_transform(dums)
pca = PCA(n_components=2)
featurecomponents = pca.fit_transform(scaledums)
import numpy as np
components = {}
index = 0
for feature in dums.columns.values:
components[feature] = [pca.components_[0][index]]
index += 1
sortedcomps = pca.components_[0]
sortedcomps.sort()
maxcap = sortedcomps[-3]
mincap = sortedcomps[2]
components = {i: x for i, x in components.items() if x >= maxcap or x <= mincap}
components = pd.DataFrame(components)
components.plot(kind='bar', figsize=(12, 4)) | code |
330673/cell_4 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from scipy.stats import chisquare
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/people.csv')
from scipy.stats import chisquare
chars = [i for i in df.columns.values if 'char_' in i]
flags = []
for feat in df[chars]:
group = df[chars].groupby(feat)
for otherfeat in df[chars].drop(feat, axis=1):
summary = group[otherfeat].count()
if chisquare(summary)[1] < 0.05:
flags.append(feat)
flags.append(otherfeat)
flags = set(flags)
print('It looks like {}% of the characteristics might be related to one another.'.format(len(flags) / len(chars) * 100)) | code |
330673/cell_6 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from scipy.stats import chisquare
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/people.csv')
from scipy.stats import chisquare
chars = [i for i in df.columns.values if 'char_' in i]
flags = []
for feat in df[chars]:
group = df[chars].groupby(feat)
for otherfeat in df[chars].drop(feat, axis=1):
summary = group[otherfeat].count()
if chisquare(summary)[1] < 0.05:
flags.append(feat)
flags.append(otherfeat)
flags = set(flags)
dums = pd.get_dummies(df[chars])
print('Before PCA the full size of the characteristics is {} features'.format(len(dums.columns.values))) | code |
330673/cell_2 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import pandas as pd
import pandas as pd
df = pd.read_csv('../input/people.csv')
print(df.head()) | code |
330673/cell_11 | [
"text_plain_output_1.png"
] | from scipy.stats import chisquare
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/people.csv')
from scipy.stats import chisquare
chars = [i for i in df.columns.values if 'char_' in i]
flags = []
for feat in df[chars]:
group = df[chars].groupby(feat)
for otherfeat in df[chars].drop(feat, axis=1):
summary = group[otherfeat].count()
if chisquare(summary)[1] < 0.05:
flags.append(feat)
flags.append(otherfeat)
flags = set(flags)
dums = pd.get_dummies(df[chars])
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
scaledums = MinMaxScaler().fit_transform(dums)
pca = PCA(n_components=2)
featurecomponents = pca.fit_transform(scaledums)
import numpy as np
components = {}
index = 0
for feature in dums.columns.values:
components[feature] = [pca.components_[0][index]]
index += 1
sortedcomps = pca.components_[0]
sortedcomps.sort()
maxcap = sortedcomps[-3]
mincap = sortedcomps[2]
components = {i: x for i, x in components.items() if x >= maxcap or x <= mincap}
components = pd.DataFrame(components)
featurecomponents = pd.DataFrame(featurecomponents, columns=['Principle Component 1', 'Principle Component 2'])
df['Principle Component 1'] = featurecomponents['Principle Component 1']
featurecomponents.plot(kind='scatter', x='Principle Component 1', y='Principle Component 2', figsize=(12, 12), s=1) | code |
330673/cell_7 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from scipy.stats import chisquare
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/people.csv')
from scipy.stats import chisquare
chars = [i for i in df.columns.values if 'char_' in i]
flags = []
for feat in df[chars]:
group = df[chars].groupby(feat)
for otherfeat in df[chars].drop(feat, axis=1):
summary = group[otherfeat].count()
if chisquare(summary)[1] < 0.05:
flags.append(feat)
flags.append(otherfeat)
flags = set(flags)
dums = pd.get_dummies(df[chars])
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
scaledums = MinMaxScaler().fit_transform(dums)
pca = PCA(n_components=2)
featurecomponents = pca.fit_transform(scaledums)
print(pca.explained_variance_ratio_) | code |
330673/cell_15 | [
"text_plain_output_2.png",
"application_vnd.jupyter.stderr_output_1.png"
] | from scipy.stats import chisquare
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/people.csv')
from scipy.stats import chisquare
chars = [i for i in df.columns.values if 'char_' in i]
flags = []
for feat in df[chars]:
group = df[chars].groupby(feat)
for otherfeat in df[chars].drop(feat, axis=1):
summary = group[otherfeat].count()
if chisquare(summary)[1] < 0.05:
flags.append(feat)
flags.append(otherfeat)
flags = set(flags)
dums = pd.get_dummies(df[chars])
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
scaledums = MinMaxScaler().fit_transform(dums)
pca = PCA(n_components=2)
featurecomponents = pca.fit_transform(scaledums)
import numpy as np
components = {}
index = 0
for feature in dums.columns.values:
components[feature] = [pca.components_[0][index]]
index += 1
sortedcomps = pca.components_[0]
sortedcomps.sort()
maxcap = sortedcomps[-3]
mincap = sortedcomps[2]
components = {i: x for i, x in components.items() if x >= maxcap or x <= mincap}
components = pd.DataFrame(components)
featurecomponents = pd.DataFrame(featurecomponents, columns=['Principle Component 1', 'Principle Component 2'])
df['Principle Component 1'] = featurecomponents['Principle Component 1']
featurecomponents['group_1'] = df['group_1']
groupslist = list(set(featurecomponents['group_1'].tolist()))
group = featurecomponents[featurecomponents['group_1'] == groupslist[0]]
group = featurecomponents[featurecomponents['group_1'] == groupslist[5]]
group = featurecomponents[featurecomponents['group_1'] == groupslist[6]]
group.plot(kind='scatter', x='Principle Component 1', y='Principle Component 2', figsize=(3, 3))
print('There are {} data points in this group.'.format(len(group.index))) | code |
330673/cell_17 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from scipy.stats import chisquare
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/people.csv')
from scipy.stats import chisquare
chars = [i for i in df.columns.values if 'char_' in i]
flags = []
for feat in df[chars]:
group = df[chars].groupby(feat)
for otherfeat in df[chars].drop(feat, axis=1):
summary = group[otherfeat].count()
if chisquare(summary)[1] < 0.05:
flags.append(feat)
flags.append(otherfeat)
flags = set(flags)
dums = pd.get_dummies(df[chars])
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
scaledums = MinMaxScaler().fit_transform(dums)
pca = PCA(n_components=2)
featurecomponents = pca.fit_transform(scaledums)
import numpy as np
components = {}
index = 0
for feature in dums.columns.values:
components[feature] = [pca.components_[0][index]]
index += 1
sortedcomps = pca.components_[0]
sortedcomps.sort()
maxcap = sortedcomps[-3]
mincap = sortedcomps[2]
components = {i: x for i, x in components.items() if x >= maxcap or x <= mincap}
components = pd.DataFrame(components)
cares = [i / 100 for i in range(75, 100, 5)]
for i in range(20, len(dums.columns.values)):
pca = PCA(n_components=i)
pca.fit(scaledums)
try:
if pca.explained_variance_ratio_.sum() > cares[0]:
print("To explain {0} of the variance you'll need {1} components".format(cares[0], i))
cares = cares[1:]
except:
break | code |
330673/cell_14 | [
"text_plain_output_1.png"
] | from scipy.stats import chisquare
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
import pandas as pd
import pandas as pd
df = pd.read_csv('../input/people.csv')
from scipy.stats import chisquare
chars = [i for i in df.columns.values if 'char_' in i]
flags = []
for feat in df[chars]:
group = df[chars].groupby(feat)
for otherfeat in df[chars].drop(feat, axis=1):
summary = group[otherfeat].count()
if chisquare(summary)[1] < 0.05:
flags.append(feat)
flags.append(otherfeat)
flags = set(flags)
dums = pd.get_dummies(df[chars])
from sklearn.decomposition import PCA
from sklearn.preprocessing import MinMaxScaler
scaledums = MinMaxScaler().fit_transform(dums)
pca = PCA(n_components=2)
featurecomponents = pca.fit_transform(scaledums)
import numpy as np
components = {}
index = 0
for feature in dums.columns.values:
components[feature] = [pca.components_[0][index]]
index += 1
sortedcomps = pca.components_[0]
sortedcomps.sort()
maxcap = sortedcomps[-3]
mincap = sortedcomps[2]
components = {i: x for i, x in components.items() if x >= maxcap or x <= mincap}
components = pd.DataFrame(components)
featurecomponents = pd.DataFrame(featurecomponents, columns=['Principle Component 1', 'Principle Component 2'])
df['Principle Component 1'] = featurecomponents['Principle Component 1']
featurecomponents['group_1'] = df['group_1']
groupslist = list(set(featurecomponents['group_1'].tolist()))
group = featurecomponents[featurecomponents['group_1'] == groupslist[0]]
group = featurecomponents[featurecomponents['group_1'] == groupslist[5]]
group.plot(kind='scatter', x='Principle Component 1', y='Principle Component 2', figsize=(3, 3))
print('There are {} data points in this group.'.format(len(group.index))) | code |
325017/cell_5 | [
"application_vnd.jupyter.stderr_output_1.png"
] | import pandas as pd
masterDF = pd.read_csv('../input/emails.csv')
messageList = masterDF['message'].tolist()
bodyList = []
for message in messageList:
firstSplit = message.split('X-FileName: ', 1)[1]
secondSplit = firstSplit.split('.')
if len(secondSplit) > 1:
secondSplit = secondSplit[1]
body = ''.join(secondSplit)[4:]
bodyList.append(body) | code |
16135671/cell_9 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
def test_plot(x):
pass
test_plot(x_train_re) | code |
16135671/cell_6 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
def test_plot(x):
pass
test_plot(x_test) | code |
16135671/cell_1 | [
"application_vnd.jupyter.stderr_output_2.png",
"text_plain_output_1.png"
] | !pip install tensorflow-gpu
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import tensorflow as tf
import keras
from keras.models import Model
from keras.layers import Input, Dense, Conv2D, MaxPooling2D, UpSampling2D, BatchNormalization, Activation
from keras.datasets import cifar10 | code |
16135671/cell_3 | [
"image_output_1.png"
] | from keras.datasets import cifar10
def load_images():
(x_train, _), (x_test, _) = cifar10.load_data()
return (x_train, x_test)
x_train, x_test = load_images() | code |
16135671/cell_14 | [
"image_output_1.png"
] | from keras.datasets import cifar10
from keras.layers import Input, Dense, Conv2D, MaxPooling2D, UpSampling2D, BatchNormalization, Activation
from keras.models import Model
import keras
import matplotlib.pyplot as plt
def load_images():
(x_train, _), (x_test, _) = cifar10.load_data()
return (x_train, x_test)
def test_plot(x):
pass
def normalize(x_train, x_test):
x_train = keras.utils.normalize(x_train)
x_test = keras.utils.normalize(x_test)
return (x_train, x_test)
input_shape = x_train.shape[1:]
receptive_field = (3, 3)
pooling_field = (2, 2)
def CONVautoencoder(x_train_re, x_test_re, epochs=200):
input_img = Input(shape=(32, 32, 3))
x = Conv2D(64, (3, 3), padding='same')(input_img)
x = BatchNormalization()(x)
x = Activation('relu')(x)
x = MaxPooling2D((2, 2), padding='same')(x)
x = Conv2D(32, (3, 3), padding='same')(x)
x = BatchNormalization()(x)
x = Activation('relu')(x)
x = MaxPooling2D((2, 2), padding='same')(x)
x = Conv2D(16, (3, 3), padding='same')(x)
x = BatchNormalization()(x)
x = Activation('relu')(x)
encoded = MaxPooling2D((2, 2), padding='same')(x)
x = Conv2D(16, (3, 3), padding='same')(encoded)
x = BatchNormalization()(x)
x = Activation('relu')(x)
x = UpSampling2D((2, 2))(x)
x = Conv2D(32, (3, 3), padding='same')(x)
x = BatchNormalization()(x)
x = Activation('relu')(x)
x = UpSampling2D((2, 2))(x)
x = Conv2D(64, (3, 3), padding='same')(x)
x = BatchNormalization()(x)
x = Activation('relu')(x)
x = UpSampling2D((2, 2))(x)
x = Conv2D(3, (3, 3), padding='same')(x)
x = BatchNormalization()(x)
decoded = Activation('sigmoid')(x)
autoencoder = Model(input_img, decoded)
encoder = Model(input_img, encoded)
autoencoder.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
autoencoder.fit(x_train_re, x_train_re, epochs=epochs, batch_size=250, validation_data=(x_test_re, x_test_re))
encoded_imgs = encoder.predict(x_test_re)
predicted = autoencoder.predict(x_test_re)
return (encoded_imgs, predicted)
return (encoded_imgs, predicted)
def plotting(x_test,encoded_imgs,predicted):
plt.figure(figsize=(40, 4))
for i in range(10):
# display original images
ax = plt.subplot(3, 20, i + 1)
plt.imshow(x_test[i].reshape(32, 32,3))
#plt.gray()
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)
# display reconstructed images
ax = plt.subplot(3, 20, 2*20 +i+ 1)
plt.imshow(predicted[i].reshape(32, 32,3))
#plt.gray()
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)
def main():
x_train, x_test = load_images()
x_train, x_test = normalize(x_train, x_test)
encoded_imgs, predicted = CONVautoencoder(x_train_re, x_test_re)
main() | code |
16135671/cell_10 | [
"text_plain_output_1.png"
] | import keras
def normalize(x_train, x_test):
x_train = keras.utils.normalize(x_train)
x_test = keras.utils.normalize(x_test)
return (x_train, x_test)
print(x_train_re.shape[1:])
print(x_test_re.shape)
input_shape = x_train.shape[1:]
receptive_field = (3, 3)
pooling_field = (2, 2) | code |
16135671/cell_5 | [
"image_output_1.png"
] | import matplotlib.pyplot as plt
def test_plot(x):
pass
test_plot(x_train) | code |
122264859/cell_9 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from sklearn.model_selection import train_test_split
from tensorflow import keras
from tensorflow.keras.optimizers import RMSprop
from tensorflow.keras.preprocessing.image import ImageDataGenerator
import numpy as np
import pandas as pd
import tensorflow as tf
import tensorflow as tf
mnist = keras.datasets.mnist
(train_images_mnist, train_labels_mnist), (test_images_mnist, test_labels_mnist) = mnist.load_data()
train_images_mnist = np.reshape(train_images_mnist, (train_images_mnist.shape[0], 28, 28, 1))
test_images_mnist = np.reshape(test_images_mnist, (test_images_mnist.shape[0], 28, 28, 1))
az_path = '/kaggle/input/az-handwritten-alphabets-in-csv-format/A_Z Handwritten Data.csv'
AZ_data = pd.read_csv(az_path, header=None)
AZ_labels = AZ_data.values[:, 0]
AZ_images = AZ_data.values[:, 1:]
AZ_images = np.reshape(AZ_images, (AZ_images.shape[0], 28, 28, 1))
from sklearn.model_selection import train_test_split
test_size = float(len(test_labels_mnist)) / len(train_labels_mnist)
train_images_AZ, test_images_AZ, train_labels_AZ, test_labels_AZ = train_test_split(AZ_images, AZ_labels, test_size=test_size)
train_labels_mnist = train_labels_mnist + max(AZ_labels) + 1
test_labels_mnist = test_labels_mnist + max(AZ_labels) + 1
train_images = np.concatenate((train_images_AZ, train_images_mnist), axis=0)
train_labels = np.concatenate((train_labels_AZ, train_labels_mnist))
test_images = np.concatenate((test_images_AZ, test_images_mnist), axis=0)
test_labels = np.concatenate((test_labels_AZ, test_labels_mnist))
import tensorflow as tf
from tensorflow.keras.optimizers import RMSprop
model = tf.keras.models.Sequential([tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Conv2D(32, (3, 3), activation='relu'), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Flatten(), tf.keras.layers.Dense(512, activation='relu'), tf.keras.layers.Dense(len(np.unique(train_labels)), activation='softmax')])
model.compile(optimizer=RMSprop(learning_rate=0.0001), loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.summary()
from tensorflow.keras.preprocessing.image import ImageDataGenerator
train_datagen = ImageDataGenerator(rescale=1.0 / 255, rotation_range=15, width_shift_range=0.1, height_shift_range=0.1, shear_range=0.1, zoom_range=0.2, horizontal_flip=False, fill_mode='nearest')
test_datagen = ImageDataGenerator(rescale=1.0 / 255)
train_generator = train_datagen.flow(train_images, train_labels, batch_size=50, shuffle=True)
validation_generator = test_datagen.flow(test_images, test_labels, batch_size=50, shuffle=True)
history = model.fit(train_generator, steps_per_epoch=750, epochs=100, validation_data=validation_generator, validation_steps=75, verbose=2)
model.save('/kaggle/working/model_v2')
model.evaluate(validation_generator) | code |
122264859/cell_4 | [
"image_output_1.png"
] | from sklearn.model_selection import train_test_split
from tensorflow import keras
import numpy as np
import pandas as pd
mnist = keras.datasets.mnist
(train_images_mnist, train_labels_mnist), (test_images_mnist, test_labels_mnist) = mnist.load_data()
train_images_mnist = np.reshape(train_images_mnist, (train_images_mnist.shape[0], 28, 28, 1))
test_images_mnist = np.reshape(test_images_mnist, (test_images_mnist.shape[0], 28, 28, 1))
az_path = '/kaggle/input/az-handwritten-alphabets-in-csv-format/A_Z Handwritten Data.csv'
AZ_data = pd.read_csv(az_path, header=None)
AZ_labels = AZ_data.values[:, 0]
AZ_images = AZ_data.values[:, 1:]
AZ_images = np.reshape(AZ_images, (AZ_images.shape[0], 28, 28, 1))
from sklearn.model_selection import train_test_split
test_size = float(len(test_labels_mnist)) / len(train_labels_mnist)
print(f'test set size: {test_size}')
train_images_AZ, test_images_AZ, train_labels_AZ, test_labels_AZ = train_test_split(AZ_images, AZ_labels, test_size=test_size)
train_labels_mnist = train_labels_mnist + max(AZ_labels) + 1
test_labels_mnist = test_labels_mnist + max(AZ_labels) + 1
train_images = np.concatenate((train_images_AZ, train_images_mnist), axis=0)
train_labels = np.concatenate((train_labels_AZ, train_labels_mnist))
test_images = np.concatenate((test_images_AZ, test_images_mnist), axis=0)
test_labels = np.concatenate((test_labels_AZ, test_labels_mnist))
print('Data ready') | code |
122264859/cell_19 | [
"image_output_1.png"
] | from imutils.contours import sort_contours
from matplotlib import cm
from sklearn.model_selection import train_test_split
from tensorflow import keras
from tensorflow.keras.models import load_model
from tensorflow.keras.optimizers import RMSprop
from tensorflow.keras.preprocessing.image import ImageDataGenerator
import cv2
import imutils
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import tensorflow as tf
import tensorflow as tf
import tensorflow as tf
mnist = keras.datasets.mnist
(train_images_mnist, train_labels_mnist), (test_images_mnist, test_labels_mnist) = mnist.load_data()
train_images_mnist = np.reshape(train_images_mnist, (train_images_mnist.shape[0], 28, 28, 1))
test_images_mnist = np.reshape(test_images_mnist, (test_images_mnist.shape[0], 28, 28, 1))
az_path = '/kaggle/input/az-handwritten-alphabets-in-csv-format/A_Z Handwritten Data.csv'
AZ_data = pd.read_csv(az_path, header=None)
AZ_labels = AZ_data.values[:, 0]
AZ_images = AZ_data.values[:, 1:]
AZ_images = np.reshape(AZ_images, (AZ_images.shape[0], 28, 28, 1))
from sklearn.model_selection import train_test_split
test_size = float(len(test_labels_mnist)) / len(train_labels_mnist)
train_images_AZ, test_images_AZ, train_labels_AZ, test_labels_AZ = train_test_split(AZ_images, AZ_labels, test_size=test_size)
train_labels_mnist = train_labels_mnist + max(AZ_labels) + 1
test_labels_mnist = test_labels_mnist + max(AZ_labels) + 1
train_images = np.concatenate((train_images_AZ, train_images_mnist), axis=0)
train_labels = np.concatenate((train_labels_AZ, train_labels_mnist))
test_images = np.concatenate((test_images_AZ, test_images_mnist), axis=0)
test_labels = np.concatenate((test_labels_AZ, test_labels_mnist))
import tensorflow as tf
from tensorflow.keras.optimizers import RMSprop
model = tf.keras.models.Sequential([tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Conv2D(32, (3, 3), activation='relu'), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Flatten(), tf.keras.layers.Dense(512, activation='relu'), tf.keras.layers.Dense(len(np.unique(train_labels)), activation='softmax')])
model.compile(optimizer=RMSprop(learning_rate=0.0001), loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.summary()
from tensorflow.keras.preprocessing.image import ImageDataGenerator
train_datagen = ImageDataGenerator(rescale=1.0 / 255, rotation_range=15, width_shift_range=0.1, height_shift_range=0.1, shear_range=0.1, zoom_range=0.2, horizontal_flip=False, fill_mode='nearest')
test_datagen = ImageDataGenerator(rescale=1.0 / 255)
train_generator = train_datagen.flow(train_images, train_labels, batch_size=50, shuffle=True)
validation_generator = test_datagen.flow(test_images, test_labels, batch_size=50, shuffle=True)
history = model.fit(train_generator, steps_per_epoch=750, epochs=100, validation_data=validation_generator, validation_steps=75, verbose=2)
model.save('/kaggle/working/model_v2')
model.evaluate(validation_generator)
model_path = '/kaggle/working/model_v2'
model = load_model(model_path)
image_path = '/kaggle/input/tester2/pja 4.jpg'
image = cv2.imread(image_path)
# perform edge detection, find contours in the edge map, and sort the
# resulting contours from left-to-right
edged = cv2.Canny(blurred, 30, 250) #low_threshold, high_threshold
cnts = cv2.findContours(edged.copy(), cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE)
cnts = imutils.grab_contours(cnts)
cnts = sort_contours(cnts, method="left-to-right")[0]
figure = plt.figure(figsize=(7,7))
plt.axis('off');
plt.imshow(edged,cmap=cm.binary_r);
chars = []
for c in cnts:
x, y, w, h = cv2.boundingRect(c)
roi = cropped[y:y + h, x:x + w]
thresh = cv2.threshold(roi, 0, 255, cv2.THRESH_BINARY_INV | cv2.THRESH_OTSU)[1]
tH, tW = thresh.shape
if tW > tH:
thresh = imutils.resize(thresh, width=28)
else:
thresh = imutils.resize(thresh, height=28)
tH, tW = thresh.shape
dX = int(max(0, 28 - tW) / 2.0)
dY = int(max(0, 28 - tH) / 2.0)
padded = cv2.copyMakeBorder(thresh, top=dY, bottom=dY, left=dX, right=dX, borderType=cv2.BORDER_CONSTANT, value=(0, 0, 0))
padded = cv2.resize(padded, (28, 28))
padded = padded.astype('float32') / 255.0
padded = np.expand_dims(padded, axis=-1)
chars.append((padded, (x, y, w, h)))
# plot isolated characters
n_cols = 10
n_rows = int(np.floor(len(chars)/ n_cols)+1)
fig = plt.figure(figsize=(1.5*n_cols,1.5*n_rows))
for i,char in enumerate(chars):
ax = plt.subplot(n_rows,n_cols,i+1)
ax.imshow(char[0][:,:,0],cmap=cm.binary,aspect='auto')
#plt.axis('off')
plt.tight_layout()
boxes = [b[1] for b in chars]
chars = np.array([c[0] for c in chars], dtype='float32')
preds = model.predict(chars)
labelNames = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789'
image = cv2.imread(image_path)
cropped = image[120:, :]
for pred, (x, y, w, h) in zip(preds, boxes):
i = np.argmax(pred)
prob = pred[i]
label = labelNames[i]
label_text = f'{label},{prob * 100:.1f}%'
cv2.rectangle(cropped, (x, y), (x + w, y + h), (0, 255, 0), 2)
cv2.putText(cropped, label_text, (x - 10, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 1)
plt.figure(figsize=(15, 10))
plt.imshow(cropped) | code |
122264859/cell_18 | [
"text_plain_output_1.png"
] | from imutils.contours import sort_contours
from matplotlib import cm
from sklearn.model_selection import train_test_split
from tensorflow import keras
from tensorflow.keras.models import load_model
from tensorflow.keras.optimizers import RMSprop
from tensorflow.keras.preprocessing.image import ImageDataGenerator
import cv2
import imutils
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import tensorflow as tf
import tensorflow as tf
import tensorflow as tf
mnist = keras.datasets.mnist
(train_images_mnist, train_labels_mnist), (test_images_mnist, test_labels_mnist) = mnist.load_data()
train_images_mnist = np.reshape(train_images_mnist, (train_images_mnist.shape[0], 28, 28, 1))
test_images_mnist = np.reshape(test_images_mnist, (test_images_mnist.shape[0], 28, 28, 1))
az_path = '/kaggle/input/az-handwritten-alphabets-in-csv-format/A_Z Handwritten Data.csv'
AZ_data = pd.read_csv(az_path, header=None)
AZ_labels = AZ_data.values[:, 0]
AZ_images = AZ_data.values[:, 1:]
AZ_images = np.reshape(AZ_images, (AZ_images.shape[0], 28, 28, 1))
from sklearn.model_selection import train_test_split
test_size = float(len(test_labels_mnist)) / len(train_labels_mnist)
train_images_AZ, test_images_AZ, train_labels_AZ, test_labels_AZ = train_test_split(AZ_images, AZ_labels, test_size=test_size)
train_labels_mnist = train_labels_mnist + max(AZ_labels) + 1
test_labels_mnist = test_labels_mnist + max(AZ_labels) + 1
train_images = np.concatenate((train_images_AZ, train_images_mnist), axis=0)
train_labels = np.concatenate((train_labels_AZ, train_labels_mnist))
test_images = np.concatenate((test_images_AZ, test_images_mnist), axis=0)
test_labels = np.concatenate((test_labels_AZ, test_labels_mnist))
import tensorflow as tf
from tensorflow.keras.optimizers import RMSprop
model = tf.keras.models.Sequential([tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Conv2D(32, (3, 3), activation='relu'), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Flatten(), tf.keras.layers.Dense(512, activation='relu'), tf.keras.layers.Dense(len(np.unique(train_labels)), activation='softmax')])
model.compile(optimizer=RMSprop(learning_rate=0.0001), loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.summary()
from tensorflow.keras.preprocessing.image import ImageDataGenerator
train_datagen = ImageDataGenerator(rescale=1.0 / 255, rotation_range=15, width_shift_range=0.1, height_shift_range=0.1, shear_range=0.1, zoom_range=0.2, horizontal_flip=False, fill_mode='nearest')
test_datagen = ImageDataGenerator(rescale=1.0 / 255)
train_generator = train_datagen.flow(train_images, train_labels, batch_size=50, shuffle=True)
validation_generator = test_datagen.flow(test_images, test_labels, batch_size=50, shuffle=True)
history = model.fit(train_generator, steps_per_epoch=750, epochs=100, validation_data=validation_generator, validation_steps=75, verbose=2)
model.save('/kaggle/working/model_v2')
model.evaluate(validation_generator)
model_path = '/kaggle/working/model_v2'
model = load_model(model_path)
image_path = '/kaggle/input/tester2/pja 4.jpg'
image = cv2.imread(image_path)
# perform edge detection, find contours in the edge map, and sort the
# resulting contours from left-to-right
edged = cv2.Canny(blurred, 30, 250) #low_threshold, high_threshold
cnts = cv2.findContours(edged.copy(), cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE)
cnts = imutils.grab_contours(cnts)
cnts = sort_contours(cnts, method="left-to-right")[0]
figure = plt.figure(figsize=(7,7))
plt.axis('off');
plt.imshow(edged,cmap=cm.binary_r);
chars = []
for c in cnts:
x, y, w, h = cv2.boundingRect(c)
roi = cropped[y:y + h, x:x + w]
thresh = cv2.threshold(roi, 0, 255, cv2.THRESH_BINARY_INV | cv2.THRESH_OTSU)[1]
tH, tW = thresh.shape
if tW > tH:
thresh = imutils.resize(thresh, width=28)
else:
thresh = imutils.resize(thresh, height=28)
tH, tW = thresh.shape
dX = int(max(0, 28 - tW) / 2.0)
dY = int(max(0, 28 - tH) / 2.0)
padded = cv2.copyMakeBorder(thresh, top=dY, bottom=dY, left=dX, right=dX, borderType=cv2.BORDER_CONSTANT, value=(0, 0, 0))
padded = cv2.resize(padded, (28, 28))
padded = padded.astype('float32') / 255.0
padded = np.expand_dims(padded, axis=-1)
chars.append((padded, (x, y, w, h)))
# plot isolated characters
n_cols = 10
n_rows = int(np.floor(len(chars)/ n_cols)+1)
fig = plt.figure(figsize=(1.5*n_cols,1.5*n_rows))
for i,char in enumerate(chars):
ax = plt.subplot(n_rows,n_cols,i+1)
ax.imshow(char[0][:,:,0],cmap=cm.binary,aspect='auto')
#plt.axis('off')
plt.tight_layout()
boxes = [b[1] for b in chars]
chars = np.array([c[0] for c in chars], dtype='float32')
preds = model.predict(chars)
labelNames = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789' | code |
122264859/cell_8 | [
"text_plain_output_1.png"
] | from sklearn.model_selection import train_test_split
from tensorflow import keras
from tensorflow.keras.optimizers import RMSprop
from tensorflow.keras.preprocessing.image import ImageDataGenerator
import numpy as np
import pandas as pd
import tensorflow as tf
import tensorflow as tf
mnist = keras.datasets.mnist
(train_images_mnist, train_labels_mnist), (test_images_mnist, test_labels_mnist) = mnist.load_data()
train_images_mnist = np.reshape(train_images_mnist, (train_images_mnist.shape[0], 28, 28, 1))
test_images_mnist = np.reshape(test_images_mnist, (test_images_mnist.shape[0], 28, 28, 1))
az_path = '/kaggle/input/az-handwritten-alphabets-in-csv-format/A_Z Handwritten Data.csv'
AZ_data = pd.read_csv(az_path, header=None)
AZ_labels = AZ_data.values[:, 0]
AZ_images = AZ_data.values[:, 1:]
AZ_images = np.reshape(AZ_images, (AZ_images.shape[0], 28, 28, 1))
from sklearn.model_selection import train_test_split
test_size = float(len(test_labels_mnist)) / len(train_labels_mnist)
train_images_AZ, test_images_AZ, train_labels_AZ, test_labels_AZ = train_test_split(AZ_images, AZ_labels, test_size=test_size)
train_labels_mnist = train_labels_mnist + max(AZ_labels) + 1
test_labels_mnist = test_labels_mnist + max(AZ_labels) + 1
train_images = np.concatenate((train_images_AZ, train_images_mnist), axis=0)
train_labels = np.concatenate((train_labels_AZ, train_labels_mnist))
test_images = np.concatenate((test_images_AZ, test_images_mnist), axis=0)
test_labels = np.concatenate((test_labels_AZ, test_labels_mnist))
import tensorflow as tf
from tensorflow.keras.optimizers import RMSprop
model = tf.keras.models.Sequential([tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Conv2D(32, (3, 3), activation='relu'), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Flatten(), tf.keras.layers.Dense(512, activation='relu'), tf.keras.layers.Dense(len(np.unique(train_labels)), activation='softmax')])
model.compile(optimizer=RMSprop(learning_rate=0.0001), loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.summary()
from tensorflow.keras.preprocessing.image import ImageDataGenerator
train_datagen = ImageDataGenerator(rescale=1.0 / 255, rotation_range=15, width_shift_range=0.1, height_shift_range=0.1, shear_range=0.1, zoom_range=0.2, horizontal_flip=False, fill_mode='nearest')
test_datagen = ImageDataGenerator(rescale=1.0 / 255)
train_generator = train_datagen.flow(train_images, train_labels, batch_size=50, shuffle=True)
validation_generator = test_datagen.flow(test_images, test_labels, batch_size=50, shuffle=True)
history = model.fit(train_generator, steps_per_epoch=750, epochs=100, validation_data=validation_generator, validation_steps=75, verbose=2)
model.save('/kaggle/working/model_v2') | code |
122264859/cell_15 | [
"text_plain_output_2.png",
"text_plain_output_1.png"
] | from imutils.contours import sort_contours
from matplotlib import cm
import cv2
import imutils
import matplotlib.pyplot as plt
image_path = '/kaggle/input/tester2/pja 4.jpg'
image = cv2.imread(image_path)
edged = cv2.Canny(blurred, 30, 250)
cnts = cv2.findContours(edged.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
cnts = imutils.grab_contours(cnts)
cnts = sort_contours(cnts, method='left-to-right')[0]
figure = plt.figure(figsize=(7, 7))
plt.axis('off')
plt.imshow(edged, cmap=cm.binary_r) | code |
122264859/cell_17 | [
"text_plain_output_1.png"
] | from imutils.contours import sort_contours
from matplotlib import cm
from sklearn.model_selection import train_test_split
from tensorflow import keras
from tensorflow.keras.optimizers import RMSprop
import cv2
import imutils
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import tensorflow as tf
import tensorflow as tf
import tensorflow as tf
mnist = keras.datasets.mnist
(train_images_mnist, train_labels_mnist), (test_images_mnist, test_labels_mnist) = mnist.load_data()
train_images_mnist = np.reshape(train_images_mnist, (train_images_mnist.shape[0], 28, 28, 1))
test_images_mnist = np.reshape(test_images_mnist, (test_images_mnist.shape[0], 28, 28, 1))
az_path = '/kaggle/input/az-handwritten-alphabets-in-csv-format/A_Z Handwritten Data.csv'
AZ_data = pd.read_csv(az_path, header=None)
AZ_labels = AZ_data.values[:, 0]
AZ_images = AZ_data.values[:, 1:]
AZ_images = np.reshape(AZ_images, (AZ_images.shape[0], 28, 28, 1))
from sklearn.model_selection import train_test_split
test_size = float(len(test_labels_mnist)) / len(train_labels_mnist)
train_images_AZ, test_images_AZ, train_labels_AZ, test_labels_AZ = train_test_split(AZ_images, AZ_labels, test_size=test_size)
train_labels_mnist = train_labels_mnist + max(AZ_labels) + 1
test_labels_mnist = test_labels_mnist + max(AZ_labels) + 1
train_images = np.concatenate((train_images_AZ, train_images_mnist), axis=0)
train_labels = np.concatenate((train_labels_AZ, train_labels_mnist))
test_images = np.concatenate((test_images_AZ, test_images_mnist), axis=0)
test_labels = np.concatenate((test_labels_AZ, test_labels_mnist))
import tensorflow as tf
from tensorflow.keras.optimizers import RMSprop
model = tf.keras.models.Sequential([tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Conv2D(32, (3, 3), activation='relu'), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Flatten(), tf.keras.layers.Dense(512, activation='relu'), tf.keras.layers.Dense(len(np.unique(train_labels)), activation='softmax')])
model.compile(optimizer=RMSprop(learning_rate=0.0001), loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.summary()
image_path = '/kaggle/input/tester2/pja 4.jpg'
image = cv2.imread(image_path)
# perform edge detection, find contours in the edge map, and sort the
# resulting contours from left-to-right
edged = cv2.Canny(blurred, 30, 250) #low_threshold, high_threshold
cnts = cv2.findContours(edged.copy(), cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE)
cnts = imutils.grab_contours(cnts)
cnts = sort_contours(cnts, method="left-to-right")[0]
figure = plt.figure(figsize=(7,7))
plt.axis('off');
plt.imshow(edged,cmap=cm.binary_r);
chars = []
for c in cnts:
x, y, w, h = cv2.boundingRect(c)
roi = cropped[y:y + h, x:x + w]
thresh = cv2.threshold(roi, 0, 255, cv2.THRESH_BINARY_INV | cv2.THRESH_OTSU)[1]
tH, tW = thresh.shape
if tW > tH:
thresh = imutils.resize(thresh, width=28)
else:
thresh = imutils.resize(thresh, height=28)
tH, tW = thresh.shape
dX = int(max(0, 28 - tW) / 2.0)
dY = int(max(0, 28 - tH) / 2.0)
padded = cv2.copyMakeBorder(thresh, top=dY, bottom=dY, left=dX, right=dX, borderType=cv2.BORDER_CONSTANT, value=(0, 0, 0))
padded = cv2.resize(padded, (28, 28))
padded = padded.astype('float32') / 255.0
padded = np.expand_dims(padded, axis=-1)
chars.append((padded, (x, y, w, h)))
n_cols = 10
n_rows = int(np.floor(len(chars) / n_cols) + 1)
fig = plt.figure(figsize=(1.5 * n_cols, 1.5 * n_rows))
for i, char in enumerate(chars):
ax = plt.subplot(n_rows, n_cols, i + 1)
ax.imshow(char[0][:, :, 0], cmap=cm.binary, aspect='auto')
plt.tight_layout() | code |
122264859/cell_14 | [
"text_plain_output_1.png"
] | gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
cropped = gray[120:, :]
blurred = cv2.GaussianBlur(cropped, (5, 5), 0)
from matplotlib import cm
fig = plt.figure(figsize=(16, 4))
ax = plt.subplot(1, 4, 1)
ax.imshow(image)
ax.set_title('original image')
ax = plt.subplot(1, 4, 2)
ax.imshow(gray, cmap=cm.binary_r)
ax.set_axis_off()
ax.set_title('grayscale image')
ax = plt.subplot(1, 4, 3)
ax.imshow(cropped, cmap=cm.binary_r)
ax.set_axis_off()
ax.set_title('cropped image')
ax = plt.subplot(1, 4, 4)
ax.imshow(blurred, cmap=cm.binary_r)
ax.set_axis_off()
ax.set_title('blurred image') | code |
122264859/cell_10 | [
"text_plain_output_1.png"
] | pip install imutils | code |
122264859/cell_12 | [
"text_plain_output_1.png"
] | from tensorflow.keras.models import load_model
model_path = '/kaggle/working/model_v2'
print('Loading NN model...')
model = load_model(model_path)
print('Done') | code |
122264859/cell_5 | [
"image_output_1.png"
] | from sklearn.model_selection import train_test_split
from tensorflow import keras
from tensorflow.keras.optimizers import RMSprop
import numpy as np
import pandas as pd
import tensorflow as tf
import tensorflow as tf
mnist = keras.datasets.mnist
(train_images_mnist, train_labels_mnist), (test_images_mnist, test_labels_mnist) = mnist.load_data()
train_images_mnist = np.reshape(train_images_mnist, (train_images_mnist.shape[0], 28, 28, 1))
test_images_mnist = np.reshape(test_images_mnist, (test_images_mnist.shape[0], 28, 28, 1))
az_path = '/kaggle/input/az-handwritten-alphabets-in-csv-format/A_Z Handwritten Data.csv'
AZ_data = pd.read_csv(az_path, header=None)
AZ_labels = AZ_data.values[:, 0]
AZ_images = AZ_data.values[:, 1:]
AZ_images = np.reshape(AZ_images, (AZ_images.shape[0], 28, 28, 1))
from sklearn.model_selection import train_test_split
test_size = float(len(test_labels_mnist)) / len(train_labels_mnist)
train_images_AZ, test_images_AZ, train_labels_AZ, test_labels_AZ = train_test_split(AZ_images, AZ_labels, test_size=test_size)
train_labels_mnist = train_labels_mnist + max(AZ_labels) + 1
test_labels_mnist = test_labels_mnist + max(AZ_labels) + 1
train_images = np.concatenate((train_images_AZ, train_images_mnist), axis=0)
train_labels = np.concatenate((train_labels_AZ, train_labels_mnist))
test_images = np.concatenate((test_images_AZ, test_images_mnist), axis=0)
test_labels = np.concatenate((test_labels_AZ, test_labels_mnist))
import tensorflow as tf
from tensorflow.keras.optimizers import RMSprop
model = tf.keras.models.Sequential([tf.keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Conv2D(32, (3, 3), activation='relu'), tf.keras.layers.MaxPooling2D(2, 2), tf.keras.layers.Flatten(), tf.keras.layers.Dense(512, activation='relu'), tf.keras.layers.Dense(len(np.unique(train_labels)), activation='softmax')])
model.compile(optimizer=RMSprop(learning_rate=0.0001), loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.summary() | code |
16115621/cell_9 | [
"image_output_1.png"
] | import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train = pd.read_csv('../input/train.csv')
test = pd.read_csv('../input/test.csv')
train.reset_index(drop=True, inplace=True)
train['SalePrice'] = np.log1p(train['SalePrice'])
train['SalePrice'].hist(bins=50)
y = train['SalePrice'].reset_index(drop=True) | code |
16115621/cell_6 | [
"text_html_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train = pd.read_csv('../input/train.csv')
test = pd.read_csv('../input/test.csv')
test.describe() | code |
16115621/cell_11 | [
"text_html_output_1.png"
] | import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train = pd.read_csv('../input/train.csv')
test = pd.read_csv('../input/test.csv')
train.reset_index(drop=True, inplace=True)
train['SalePrice'] = np.log1p(train['SalePrice'])
y = train['SalePrice'].reset_index(drop=True)
train = train.drop(['Id', 'SalePrice'], axis=1)
test = test.drop(['Id'], axis=1)
x = pd.concat([train, test]).reset_index(drop=True)
x.info() | code |
16115621/cell_7 | [
"application_vnd.jupyter.stderr_output_1.png",
"image_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train = pd.read_csv('../input/train.csv')
test = pd.read_csv('../input/test.csv')
train['SalePrice'].hist(bins=50) | code |
16115621/cell_15 | [
"text_html_output_1.png"
] | import matplotlib.pyplot as plt # visualization
import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns # visualization
train = pd.read_csv('../input/train.csv')
test = pd.read_csv('../input/test.csv')
train.reset_index(drop=True, inplace=True)
train['SalePrice'] = np.log1p(train['SalePrice'])
y = train['SalePrice'].reset_index(drop=True)
train = train.drop(['Id', 'SalePrice'], axis=1)
test = test.drop(['Id'], axis=1)
x = pd.concat([train, test]).reset_index(drop=True)
x['MSSubClass'] = x['MSSubClass'].apply(str)
x['YrSold'] = x['YrSold'].astype(str)
x['MoSold'] = x['MoSold'].astype(str)
x['Functional'] = x['Functional'].fillna('Typ')
x['Electrical'] = x['Electrical'].fillna('SBrkr')
x['KitchenQual'] = x['KitchenQual'].fillna('TA')
x['Exterior1st'] = x['Exterior1st'].fillna(x['Exterior1st'].mode()[0])
x['Exterior2nd'] = x['Exterior2nd'].fillna(x['Exterior2nd'].mode()[0])
x['SaleType'] = x['SaleType'].fillna(x['SaleType'].mode()[0])
for col in ('GarageYrBlt', 'GarageArea', 'GarageCars'):
x[col] = x[col].fillna(0)
for col in ['GarageType', 'GarageFinish', 'GarageQual', 'GarageCond']:
x[col] = x[col].fillna('None')
for col in ('BsmtFinSF1', 'BsmtFinSF2', 'BsmtUnfSF', 'TotalBsmtSF', 'BsmtFullBath', 'BsmtHalfBath'):
x[col] = x[col].fillna(0)
for col in ('BsmtQual', 'BsmtCond', 'BsmtExposure', 'BsmtFinType1', 'BsmtFinType2'):
x[col] = x[col].fillna('None')
objects = []
for i in x.columns:
if x[i].dtype == object:
objects.append(i)
x.update(x[objects].fillna('None'))
numeric_dtypes = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
numerics = []
for i in x.columns:
if x[i].dtype in numeric_dtypes:
numerics.append(i)
x.update(x[numerics].fillna(0))
x['total_sf'] = x['TotalBsmtSF'] + x['BsmtFinSF1'] + x['BsmtFinSF2'] + x['1stFlrSF'] + x['2ndFlrSF']
x['total_bathrooms'] = x['FullBath'] + 0.5 * x['HalfBath'] + x['BsmtFullBath'] + 0.5 * x['BsmtHalfBath']
x['total_porch_sf'] = x['OpenPorchSF'] + x['3SsnPorch'] + x['EnclosedPorch'] + x['ScreenPorch'] + x['WoodDeckSF']
x['hasPool'] = x['PoolArea'].apply(lambda x: 1 if x > 0 else 0)
x['has2ndFloor'] = x['2ndFlrSF'].apply(lambda x: 1 if x > 0 else 0)
x['hasGarage'] = x['GarageArea'].apply(lambda x: 1 if x > 0 else 0)
x['hasBasement'] = x['TotalBsmtSF'].apply(lambda x: 1 if x > 0 else 0)
x['hasFireplace'] = x['Fireplaces'].apply(lambda x: 1 if x > 0 else 0)
column_print = ['LotFrontage', 'LotArea', 'OverallQual', 'OverallQual']
f, axes = plt.subplots(2, 2, figsize=(7, 7), sharex=True)
for i in x.columns:
if i in column_print:
sns.distplot(x[i])
numeric_plot.append(i) | code |
16115621/cell_14 | [
"image_output_1.png"
] | import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train = pd.read_csv('../input/train.csv')
test = pd.read_csv('../input/test.csv')
train.reset_index(drop=True, inplace=True)
train['SalePrice'] = np.log1p(train['SalePrice'])
y = train['SalePrice'].reset_index(drop=True)
train = train.drop(['Id', 'SalePrice'], axis=1)
test = test.drop(['Id'], axis=1)
x = pd.concat([train, test]).reset_index(drop=True)
x['MSSubClass'] = x['MSSubClass'].apply(str)
x['YrSold'] = x['YrSold'].astype(str)
x['MoSold'] = x['MoSold'].astype(str)
x['Functional'] = x['Functional'].fillna('Typ')
x['Electrical'] = x['Electrical'].fillna('SBrkr')
x['KitchenQual'] = x['KitchenQual'].fillna('TA')
x['Exterior1st'] = x['Exterior1st'].fillna(x['Exterior1st'].mode()[0])
x['Exterior2nd'] = x['Exterior2nd'].fillna(x['Exterior2nd'].mode()[0])
x['SaleType'] = x['SaleType'].fillna(x['SaleType'].mode()[0])
for col in ('GarageYrBlt', 'GarageArea', 'GarageCars'):
x[col] = x[col].fillna(0)
for col in ['GarageType', 'GarageFinish', 'GarageQual', 'GarageCond']:
x[col] = x[col].fillna('None')
for col in ('BsmtFinSF1', 'BsmtFinSF2', 'BsmtUnfSF', 'TotalBsmtSF', 'BsmtFullBath', 'BsmtHalfBath'):
x[col] = x[col].fillna(0)
for col in ('BsmtQual', 'BsmtCond', 'BsmtExposure', 'BsmtFinType1', 'BsmtFinType2'):
x[col] = x[col].fillna('None')
objects = []
for i in x.columns:
if x[i].dtype == object:
objects.append(i)
x.update(x[objects].fillna('None'))
numeric_dtypes = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
numerics = []
for i in x.columns:
if x[i].dtype in numeric_dtypes:
numerics.append(i)
x.update(x[numerics].fillna(0))
x['total_sf'] = x['TotalBsmtSF'] + x['BsmtFinSF1'] + x['BsmtFinSF2'] + x['1stFlrSF'] + x['2ndFlrSF']
x['total_bathrooms'] = x['FullBath'] + 0.5 * x['HalfBath'] + x['BsmtFullBath'] + 0.5 * x['BsmtHalfBath']
x['total_porch_sf'] = x['OpenPorchSF'] + x['3SsnPorch'] + x['EnclosedPorch'] + x['ScreenPorch'] + x['WoodDeckSF']
x['hasPool'] = x['PoolArea'].apply(lambda x: 1 if x > 0 else 0)
x['has2ndFloor'] = x['2ndFlrSF'].apply(lambda x: 1 if x > 0 else 0)
x['hasGarage'] = x['GarageArea'].apply(lambda x: 1 if x > 0 else 0)
x['hasBasement'] = x['TotalBsmtSF'].apply(lambda x: 1 if x > 0 else 0)
x['hasFireplace'] = x['Fireplaces'].apply(lambda x: 1 if x > 0 else 0)
x.describe() | code |
16115621/cell_10 | [
"text_html_output_1.png"
] | import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train = pd.read_csv('../input/train.csv')
test = pd.read_csv('../input/test.csv')
train.reset_index(drop=True, inplace=True)
train['SalePrice'] = np.log1p(train['SalePrice'])
y = train['SalePrice'].reset_index(drop=True)
train = train.drop(['Id', 'SalePrice'], axis=1)
test = test.drop(['Id'], axis=1)
x = pd.concat([train, test]).reset_index(drop=True)
x.describe() | code |
16115621/cell_12 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train = pd.read_csv('../input/train.csv')
test = pd.read_csv('../input/test.csv')
train.reset_index(drop=True, inplace=True)
train['SalePrice'] = np.log1p(train['SalePrice'])
y = train['SalePrice'].reset_index(drop=True)
train = train.drop(['Id', 'SalePrice'], axis=1)
test = test.drop(['Id'], axis=1)
x = pd.concat([train, test]).reset_index(drop=True)
x['MSSubClass'] = x['MSSubClass'].apply(str)
x['YrSold'] = x['YrSold'].astype(str)
x['MoSold'] = x['MoSold'].astype(str)
x['Functional'] = x['Functional'].fillna('Typ')
x['Electrical'] = x['Electrical'].fillna('SBrkr')
x['KitchenQual'] = x['KitchenQual'].fillna('TA')
x['Exterior1st'] = x['Exterior1st'].fillna(x['Exterior1st'].mode()[0])
x['Exterior2nd'] = x['Exterior2nd'].fillna(x['Exterior2nd'].mode()[0])
x['SaleType'] = x['SaleType'].fillna(x['SaleType'].mode()[0])
for col in ('GarageYrBlt', 'GarageArea', 'GarageCars'):
x[col] = x[col].fillna(0)
for col in ['GarageType', 'GarageFinish', 'GarageQual', 'GarageCond']:
x[col] = x[col].fillna('None')
for col in ('BsmtFinSF1', 'BsmtFinSF2', 'BsmtUnfSF', 'TotalBsmtSF', 'BsmtFullBath', 'BsmtHalfBath'):
x[col] = x[col].fillna(0)
for col in ('BsmtQual', 'BsmtCond', 'BsmtExposure', 'BsmtFinType1', 'BsmtFinType2'):
x[col] = x[col].fillna('None')
objects = []
for i in x.columns:
if x[i].dtype == object:
objects.append(i)
x.update(x[objects].fillna('None'))
numeric_dtypes = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
numerics = []
for i in x.columns:
if x[i].dtype in numeric_dtypes:
numerics.append(i)
x.update(x[numerics].fillna(0))
x.info() | code |
16115621/cell_5 | [
"text_plain_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train = pd.read_csv('../input/train.csv')
test = pd.read_csv('../input/test.csv')
train.describe() | code |
72097728/cell_9 | [
"text_plain_output_1.png"
] | from sklearn.compose import ColumnTransformer
from sklearn.ensemble import RandomForestRegressor
from sklearn.impute import SimpleImputer
from sklearn.metrics import mean_squared_error
from sklearn.model_selection import train_test_split
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import OneHotEncoder
import pandas as pd
import pandas as pd
from sklearn.model_selection import train_test_split
df = pd.read_csv('../input/30-days-of-ml/train.csv')
test_df = pd.read_csv('../input/30-days-of-ml/test.csv')
sample_df = pd.read_csv('../input/30-days-of-ml/sample_submission.csv')
y = df.target
X = df.drop(['target'], axis=1)
X_train_full, X_valid_full, y_train, y_valid = train_test_split(X, y, train_size=0.8, test_size=0.2, random_state=0)
categorical_cols = [cname for cname in X_train_full.columns if X_train_full[cname].nunique() < 5 and X_train_full[cname].dtype == 'object']
numerical_cols = [cname for cname in X_train_full.columns if X_train_full[cname].dtype in ['int64', 'float64']]
my_cols = categorical_cols + numerical_cols
X_train = X_train_full[my_cols].copy()
X_valid = X_valid_full[my_cols].copy()
from sklearn.compose import ColumnTransformer
from sklearn.pipeline import Pipeline
from sklearn.impute import SimpleImputer
from sklearn.preprocessing import OneHotEncoder
numerical_transformer = SimpleImputer(strategy='constant')
categorical_transformer = Pipeline(steps=[('imputer', SimpleImputer(strategy='most_frequent')), ('onehot', OneHotEncoder(handle_unknown='ignore'))])
preprocessor = ColumnTransformer(transformers=[('num', numerical_transformer, numerical_cols), ('cat', categorical_transformer, categorical_cols)])
from sklearn.ensemble import RandomForestRegressor
model = RandomForestRegressor(n_estimators=100, random_state=0)
from sklearn.metrics import mean_absolute_error
my_pipeline = Pipeline(steps=[('preprocessor', preprocessor), ('model', model)])
my_pipeline.fit(X_train, y_train)
preds = my_pipeline.predict(X_valid)
from sklearn.metrics import mean_squared_error
score = mean_squared_error(y_valid, preds)
print('MSE:', score) | code |
72097728/cell_3 | [
"text_plain_output_1.png"
] | from sklearn.model_selection import train_test_split
import pandas as pd
import pandas as pd
from sklearn.model_selection import train_test_split
df = pd.read_csv('../input/30-days-of-ml/train.csv')
test_df = pd.read_csv('../input/30-days-of-ml/test.csv')
sample_df = pd.read_csv('../input/30-days-of-ml/sample_submission.csv')
y = df.target
X = df.drop(['target'], axis=1)
X_train_full, X_valid_full, y_train, y_valid = train_test_split(X, y, train_size=0.8, test_size=0.2, random_state=0)
categorical_cols = [cname for cname in X_train_full.columns if X_train_full[cname].nunique() < 5 and X_train_full[cname].dtype == 'object']
numerical_cols = [cname for cname in X_train_full.columns if X_train_full[cname].dtype in ['int64', 'float64']]
my_cols = categorical_cols + numerical_cols
X_train = X_train_full[my_cols].copy()
X_valid = X_valid_full[my_cols].copy()
[cname for cname in X_train_full.columns if X_train_full[cname].nunique() < 5 and X_train_full[cname].dtype == 'object'] | code |
2003139/cell_9 | [
"text_plain_output_1.png"
] | from PIL import Image
from keras import optimizers, losses, activations, models
from keras.layers import Convolution2D, Dense, Input, Flatten, Dropout, MaxPooling2D, BatchNormalization, GlobalMaxPool2D, Concatenate
from random import shuffle
import numpy as np # linear algebra
import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
list_paths = []
for subdir, dirs, files in os.walk('../input'):
for file in files:
filepath = subdir + os.sep + file
list_paths.append(filepath)
list_train = [filepath for filepath in list_paths if 'train/' in filepath]
shuffle(list_train)
list_test = [filepath for filepath in list_paths if 'test/' in filepath]
list_train = list_train
list_test = list_test
index = [os.path.basename(filepath) for filepath in list_test]
list_classes = list(set([os.path.dirname(filepath).split(os.sep)[-1] for filepath in list_paths if 'train' in filepath]))
def get_class_from_path(filepath):
return os.path.dirname(filepath).split(os.sep)[-1]
def read_and_resize(filepath):
im_array = np.array(Image.open(filepath), dtype='uint8')
pil_im = Image.fromarray(im_array)
new_array = np.array(pil_im.resize((256, 256)))
return new_array / 255
def label_transform(labels):
labels = pd.get_dummies(pd.Series(labels))
label_index = labels.columns.values
return (labels, label_index)
from keras.callbacks import ModelCheckpoint, LearningRateScheduler, EarlyStopping, ReduceLROnPlateau, TensorBoard
from keras import optimizers, losses, activations, models
from keras.layers import Convolution2D, Dense, Input, Flatten, Dropout, MaxPooling2D, BatchNormalization, GlobalMaxPool2D, Concatenate
input_shape = (256, 256, 3)
nclass = len(label_index)
def get_model():
nclass = len(label_index)
inp = Input(shape=input_shape)
norm_inp = BatchNormalization()(inp)
img_1 = Convolution2D(16, kernel_size=3, activation=activations.relu, padding='same')(norm_inp)
img_1 = Convolution2D(16, kernel_size=3, activation=activations.relu, padding='same')(img_1)
img_1 = MaxPooling2D(pool_size=(3, 3))(img_1)
img_1 = Dropout(rate=0.2)(img_1)
img_1 = Convolution2D(32, kernel_size=3, activation=activations.relu, padding='same')(img_1)
img_1 = Convolution2D(32, kernel_size=3, activation=activations.relu, padding='same')(img_1)
img_1 = MaxPooling2D(pool_size=(3, 3))(img_1)
img_1 = Dropout(rate=0.2)(img_1)
img_1 = Convolution2D(64, kernel_size=2, activation=activations.relu, padding='same')(img_1)
img_1 = Convolution2D(20, kernel_size=2, activation=activations.relu, padding='same')(img_1)
img_1 = GlobalMaxPool2D()(img_1)
img_1 = Dropout(rate=0.2)(img_1)
dense_1 = Dense(20, activation=activations.relu)(img_1)
dense_1 = Dense(nclass, activation=activations.softmax)(dense_1)
model = models.Model(inputs=inp, outputs=dense_1)
opt = optimizers.Adam()
model.compile(optimizer=opt, loss=losses.categorical_crossentropy, metrics=['acc'])
model.summary()
return model | code |
2003139/cell_1 | [
"text_plain_output_1.png"
] | from subprocess import check_output
import numpy as np
import pandas as pd
import os
from PIL import Image
from skimage.transform import resize
from random import shuffle
from subprocess import check_output
print(check_output(['ls', '../input']).decode('utf8')) | code |
2003139/cell_10 | [
"application_vnd.jupyter.stderr_output_1.png"
] | from PIL import Image
from keras import optimizers, losses, activations, models
from keras.callbacks import ModelCheckpoint, LearningRateScheduler, EarlyStopping, ReduceLROnPlateau, TensorBoard
from keras.layers import Convolution2D, Dense, Input, Flatten, Dropout, MaxPooling2D, BatchNormalization, GlobalMaxPool2D, Concatenate
from random import shuffle
import numpy as np # linear algebra
import os
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
list_paths = []
for subdir, dirs, files in os.walk('../input'):
for file in files:
filepath = subdir + os.sep + file
list_paths.append(filepath)
list_train = [filepath for filepath in list_paths if 'train/' in filepath]
shuffle(list_train)
list_test = [filepath for filepath in list_paths if 'test/' in filepath]
list_train = list_train
list_test = list_test
index = [os.path.basename(filepath) for filepath in list_test]
list_classes = list(set([os.path.dirname(filepath).split(os.sep)[-1] for filepath in list_paths if 'train' in filepath]))
def get_class_from_path(filepath):
return os.path.dirname(filepath).split(os.sep)[-1]
def read_and_resize(filepath):
im_array = np.array(Image.open(filepath), dtype='uint8')
pil_im = Image.fromarray(im_array)
new_array = np.array(pil_im.resize((256, 256)))
return new_array / 255
def label_transform(labels):
labels = pd.get_dummies(pd.Series(labels))
label_index = labels.columns.values
return (labels, label_index)
X_train = np.array([read_and_resize(filepath) for filepath in list_train])
X_test = np.array([read_and_resize(filepath) for filepath in list_test])
labels = [get_class_from_path(filepath) for filepath in list_train]
y, label_index = label_transform(labels)
y = np.array(y)
from keras.callbacks import ModelCheckpoint, LearningRateScheduler, EarlyStopping, ReduceLROnPlateau, TensorBoard
from keras import optimizers, losses, activations, models
from keras.layers import Convolution2D, Dense, Input, Flatten, Dropout, MaxPooling2D, BatchNormalization, GlobalMaxPool2D, Concatenate
input_shape = (256, 256, 3)
nclass = len(label_index)
def get_model():
nclass = len(label_index)
inp = Input(shape=input_shape)
norm_inp = BatchNormalization()(inp)
img_1 = Convolution2D(16, kernel_size=3, activation=activations.relu, padding='same')(norm_inp)
img_1 = Convolution2D(16, kernel_size=3, activation=activations.relu, padding='same')(img_1)
img_1 = MaxPooling2D(pool_size=(3, 3))(img_1)
img_1 = Dropout(rate=0.2)(img_1)
img_1 = Convolution2D(32, kernel_size=3, activation=activations.relu, padding='same')(img_1)
img_1 = Convolution2D(32, kernel_size=3, activation=activations.relu, padding='same')(img_1)
img_1 = MaxPooling2D(pool_size=(3, 3))(img_1)
img_1 = Dropout(rate=0.2)(img_1)
img_1 = Convolution2D(64, kernel_size=2, activation=activations.relu, padding='same')(img_1)
img_1 = Convolution2D(20, kernel_size=2, activation=activations.relu, padding='same')(img_1)
img_1 = GlobalMaxPool2D()(img_1)
img_1 = Dropout(rate=0.2)(img_1)
dense_1 = Dense(20, activation=activations.relu)(img_1)
dense_1 = Dense(nclass, activation=activations.softmax)(dense_1)
model = models.Model(inputs=inp, outputs=dense_1)
opt = optimizers.Adam()
model.compile(optimizer=opt, loss=losses.categorical_crossentropy, metrics=['acc'])
model.summary()
return model
model = get_model()
file_path = 'weights.best.hdf5'
checkpoint = ModelCheckpoint(file_path, monitor='val_acc', verbose=1, save_best_only=True, mode='max')
early = EarlyStopping(monitor='val_acc', mode='max', patience=1)
callbacks_list = [checkpoint, early]
history = model.fit(X_train, y, validation_split=0.1, epochs=3, shuffle=True, verbose=2, callbacks=callbacks_list)
model.load_weights(file_path) | code |
105185803/cell_1 | [
"text_plain_output_1.png"
] | import pandas as pd
import plotly.express as px
import os
!pip install kaleido | code |
105185803/cell_5 | [
"text_html_output_2.png"
] | import pandas as pd
import plotly.express as px
stream = open('../input/bbk-kunpeng/perf_test_result.txt', 'r')
stream = stream.read()
ds = pd.DataFrame(columns=['N', 'NRHS', 'data_type', 'gflops', 'uplo'])
datatype = 'single'
N = 0
NRHS = 0
uplo = 'U'
gflops = 0.0
cnt = 0
for line in stream.split('\n'):
if line == 'test SINGLE':
datatype = 'single'
elif line == 'test DOUBLE':
datatype = 'double'
elif line == 'test COMPLEX':
datatype = 'complex'
elif line == 'test COMPLEX16':
datatype = 'complex16'
elif line.startswith('N='):
splited = line.split(',')
N = int(splited[0].split('=')[1])
NRHS = int(splited[1].split('=')[1])
uplo = splited[2].split('=')[1]
elif line.startswith('gflops'):
gflops = float(line.split('=')[1])
ds.loc[cnt] = [N, NRHS, datatype, gflops, uplo]
cnt += 1
ds.to_csv('performance test.csv')
from _plotly_utils.basevalidators import textwrap
fig = px.line(ds, x='N', y='gflops', color='data_type', facet_col_wrap=2, title='Performance of BBK on 32 core')
fig.show() | code |
121149832/cell_4 | [
"application_vnd.jupyter.stderr_output_766.png",
"application_vnd.jupyter.stderr_output_116.png",
"application_vnd.jupyter.stderr_output_74.png",
"application_vnd.jupyter.stderr_output_268.png",
"application_vnd.jupyter.stderr_output_362.png",
"text_plain_output_743.png",
"text_plain_output_673.png",
"text_plain_output_445.png",
"text_plain_output_201.png",
"text_plain_output_261.png",
"text_plain_output_775.png",
"text_plain_output_819.png",
"text_plain_output_565.png",
"application_vnd.jupyter.stderr_output_566.png",
"application_vnd.jupyter.stderr_output_578.png",
"application_vnd.jupyter.stderr_output_516.png",
"application_vnd.jupyter.stderr_output_672.png",
"text_plain_output_521.png",
"text_plain_output_769.png",
"text_plain_output_205.png",
"application_vnd.jupyter.stderr_output_732.png",
"application_vnd.jupyter.stderr_output_222.png",
"application_vnd.jupyter.stderr_output_626.png",
"application_vnd.jupyter.stderr_output_96.png",
"text_plain_output_693.png",
"application_vnd.jupyter.stderr_output_642.png",
"application_vnd.jupyter.stderr_output_640.png",
"text_plain_output_511.png",
"application_vnd.jupyter.stderr_output_836.png",
"text_plain_output_271.png",
"application_vnd.jupyter.stderr_output_808.png",
"text_plain_output_475.png",
"application_vnd.jupyter.stderr_output_296.png",
"text_plain_output_455.png",
"text_plain_output_223.png",
"application_vnd.jupyter.stderr_output_110.png",
"text_plain_output_715.png",
"text_plain_output_579.png",
"text_plain_output_793.png",
"text_plain_output_629.png",
"text_plain_output_287.png",
"application_vnd.jupyter.stderr_output_112.png",
"text_plain_output_181.png",
"text_plain_output_137.png",
"application_vnd.jupyter.stderr_output_400.png",
"application_vnd.jupyter.stderr_output_212.png",
"application_vnd.jupyter.stderr_output_768.png",
"application_vnd.jupyter.stderr_output_700.png",
"application_vnd.jupyter.stderr_output_458.png",
"application_vnd.jupyter.stderr_output_634.png",
"application_vnd.jupyter.stderr_output_740.png",
"text_plain_output_139.png",
"application_vnd.jupyter.stderr_output_420.png",
"text_plain_output_813.png",
"text_plain_output_35.png",
"text_plain_output_697.png",
"text_plain_output_501.png",
"text_plain_output_593.png",
"application_vnd.jupyter.stderr_output_24.png",
"application_vnd.jupyter.stderr_output_354.png",
"text_plain_output_685.png",
"application_vnd.jupyter.stderr_output_16.png",
"application_vnd.jupyter.stderr_output_274.png",
"application_vnd.jupyter.stderr_output_610.png",
"application_vnd.jupyter.stderr_output_632.png",
"application_vnd.jupyter.stderr_output_368.png",
"text_plain_output_449.png",
"text_plain_output_117.png",
"application_vnd.jupyter.stderr_output_760.png",
"application_vnd.jupyter.stderr_output_474.png",
"application_vnd.jupyter.stderr_output_258.png",
"text_plain_output_367.png",
"application_vnd.jupyter.stderr_output_758.png",
"application_vnd.jupyter.stderr_output_668.png",
"application_vnd.jupyter.stderr_output_622.png",
"text_plain_output_395.png",
"application_vnd.jupyter.stderr_output_286.png",
"application_vnd.jupyter.stderr_output_426.png",
"text_plain_output_617.png",
"application_vnd.jupyter.stderr_output_152.png",
"application_vnd.jupyter.stderr_output_156.png",
"text_plain_output_307.png",
"application_vnd.jupyter.stderr_output_522.png",
"application_vnd.jupyter.stderr_output_804.png",
"application_vnd.jupyter.stderr_output_710.png",
"application_vnd.jupyter.stderr_output_684.png",
"application_vnd.jupyter.stderr_output_70.png",
"application_vnd.jupyter.stderr_output_310.png",
"text_plain_output_799.png",
"application_vnd.jupyter.stderr_output_554.png",
"text_plain_output_833.png",
"application_vnd.jupyter.stderr_output_204.png",
"text_plain_output_399.png",
"application_vnd.jupyter.stderr_output_284.png",
"text_plain_output_671.png",
"application_vnd.jupyter.stderr_output_124.png",
"application_vnd.jupyter.stderr_output_786.png",
"text_plain_output_195.png",
"application_vnd.jupyter.stderr_output_498.png",
"application_vnd.jupyter.stderr_output_838.png",
"text_plain_output_471.png",
"text_plain_output_219.png",
"application_vnd.jupyter.stderr_output_52.png",
"text_plain_output_485.png",
"text_plain_output_237.png",
"text_plain_output_43.png",
"application_vnd.jupyter.stderr_output_172.png",
"text_plain_output_187.png",
"text_plain_output_309.png",
"application_vnd.jupyter.stderr_output_512.png",
"text_plain_output_143.png",
"application_vnd.jupyter.stderr_output_348.png",
"text_plain_output_37.png",
"application_vnd.jupyter.stderr_output_32.png",
"application_vnd.jupyter.stderr_output_246.png",
"application_vnd.jupyter.stderr_output_704.png",
"application_vnd.jupyter.stderr_output_502.png",
"application_vnd.jupyter.stderr_output_722.png",
"application_vnd.jupyter.stderr_output_176.png",
"application_vnd.jupyter.stderr_output_356.png",
"text_plain_output_477.png",
"application_vnd.jupyter.stderr_output_802.png",
"text_plain_output_627.png",
"application_vnd.jupyter.stderr_output_506.png",
"text_plain_output_613.png",
"text_plain_output_147.png",
"text_plain_output_443.png",
"text_plain_output_327.png",
"application_vnd.jupyter.stderr_output_346.png",
"text_plain_output_79.png",
"text_plain_output_331.png",
"application_vnd.jupyter.stderr_output_382.png",
"text_plain_output_809.png",
"application_vnd.jupyter.stderr_output_170.png",
"application_vnd.jupyter.stderr_output_132.png",
"text_plain_output_5.png",
"text_plain_output_75.png",
"application_vnd.jupyter.stderr_output_692.png",
"application_vnd.jupyter.stderr_output_540.png",
"application_vnd.jupyter.stderr_output_48.png",
"application_vnd.jupyter.stderr_output_236.png",
"application_vnd.jupyter.stderr_output_418.png",
"application_vnd.jupyter.stderr_output_636.png",
"text_plain_output_167.png",
"application_vnd.jupyter.stderr_output_550.png",
"text_plain_output_213.png",
"text_plain_output_73.png",
"text_plain_output_687.png",
"application_vnd.jupyter.stderr_output_378.png",
"application_vnd.jupyter.stderr_output_764.png",
"application_vnd.jupyter.stderr_output_432.png",
"text_plain_output_321.png",
"application_vnd.jupyter.stderr_output_2.png",
"application_vnd.jupyter.stderr_output_472.png",
"application_vnd.jupyter.stderr_output_774.png",
"application_vnd.jupyter.stderr_output_776.png",
"application_vnd.jupyter.stderr_output_814.png",
"text_plain_output_115.png",
"application_vnd.jupyter.stderr_output_504.png",
"text_plain_output_407.png",
"application_vnd.jupyter.stderr_output_552.png",
"application_vnd.jupyter.stderr_output_694.png",
"text_plain_output_355.png",
"application_vnd.jupyter.stderr_output_800.png",
"text_plain_output_15.png",
"application_vnd.jupyter.stderr_output_618.png",
"text_plain_output_133.png",
"application_vnd.jupyter.stderr_output_392.png",
"application_vnd.jupyter.stderr_output_806.png",
"application_vnd.jupyter.stderr_output_690.png",
"text_plain_output_771.png",
"text_plain_output_651.png",
"application_vnd.jupyter.stderr_output_666.png",
"application_vnd.jupyter.stderr_output_414.png",
"application_vnd.jupyter.stderr_output_436.png",
"application_vnd.jupyter.stderr_output_608.png",
"text_plain_output_437.png",
"application_vnd.jupyter.stderr_output_146.png",
"text_plain_output_699.png",
"text_plain_output_387.png",
"text_plain_output_555.png",
"application_vnd.jupyter.stderr_output_324.png",
"application_vnd.jupyter.stderr_output_528.png",
"application_vnd.jupyter.stderr_output_360.png",
"text_plain_output_759.png",
"application_vnd.jupyter.stderr_output_484.png",
"application_vnd.jupyter.stderr_output_674.png",
"text_plain_output_375.png",
"text_plain_output_831.png",
"text_plain_output_659.png",
"text_plain_output_515.png",
"text_plain_output_157.png",
"application_vnd.jupyter.stderr_output_190.png",
"application_vnd.jupyter.stderr_output_380.png",
"text_plain_output_773.png",
"application_vnd.jupyter.stderr_output_270.png",
"text_plain_output_317.png",
"text_plain_output_251.png",
"application_vnd.jupyter.stderr_output_344.png",
"application_vnd.jupyter.stderr_output_18.png",
"text_plain_output_423.png",
"application_vnd.jupyter.stderr_output_86.png",
"text_plain_output_9.png",
"application_vnd.jupyter.stderr_output_334.png",
"application_vnd.jupyter.stderr_output_526.png",
"text_plain_output_633.png",
"application_vnd.jupyter.stderr_output_754.png",
"application_vnd.jupyter.stderr_output_782.png",
"application_vnd.jupyter.stderr_output_38.png",
"application_vnd.jupyter.stderr_output_482.png",
"application_vnd.jupyter.stderr_output_568.png",
"text_plain_output_325.png",
"application_vnd.jupyter.stderr_output_240.png",
"text_plain_output_785.png",
"text_plain_output_203.png",
"text_plain_output_505.png",
"application_vnd.jupyter.stderr_output_272.png",
"application_vnd.jupyter.stderr_output_88.png",
"text_plain_output_603.png",
"text_plain_output_655.png",
"text_plain_output_119.png",
"text_plain_output_373.png",
"application_vnd.jupyter.stderr_output_148.png",
"text_plain_output_741.png",
"application_vnd.jupyter.stderr_output_520.png",
"text_plain_output_551.png",
"text_plain_output_583.png",
"application_vnd.jupyter.stderr_output_58.png",
"application_vnd.jupyter.stderr_output_638.png",
"application_vnd.jupyter.stderr_output_66.png",
"text_plain_output_131.png",
"text_plain_output_817.png",
"text_plain_output_343.png",
"application_vnd.jupyter.stderr_output_724.png",
"application_vnd.jupyter.stderr_output_718.png",
"text_plain_output_123.png",
"application_vnd.jupyter.stderr_output_68.png",
"text_plain_output_31.png",
"application_vnd.jupyter.stderr_output_106.png",
"text_plain_output_379.png",
"application_vnd.jupyter.stderr_output_224.png",
"text_plain_output_281.png",
"text_plain_output_639.png",
"application_vnd.jupyter.stderr_output_4.png",
"text_plain_output_557.png",
"application_vnd.jupyter.stderr_output_26.png",
"application_vnd.jupyter.stderr_output_178.png",
"text_plain_output_273.png",
"application_vnd.jupyter.stderr_output_322.png",
"application_vnd.jupyter.stderr_output_832.png",
"text_plain_output_263.png",
"text_plain_output_229.png",
"application_vnd.jupyter.stderr_output_384.png",
"text_plain_output_111.png",
"application_vnd.jupyter.stderr_output_406.png",
"application_vnd.jupyter.stderr_output_620.png",
"application_vnd.jupyter.stderr_output_238.png",
"application_vnd.jupyter.stderr_output_564.png",
"text_plain_output_753.png",
"text_plain_output_669.png",
"text_plain_output_461.png",
"application_vnd.jupyter.stderr_output_822.png",
"application_vnd.jupyter.stderr_output_650.png",
"application_vnd.jupyter.stderr_output_450.png",
"application_vnd.jupyter.stderr_output_524.png",
"text_plain_output_589.png",
"text_plain_output_101.png",
"application_vnd.jupyter.stderr_output_796.png",
"application_vnd.jupyter.stderr_output_490.png",
"text_plain_output_169.png",
"text_plain_output_531.png",
"text_plain_output_161.png",
"text_plain_output_489.png",
"application_vnd.jupyter.stderr_output_136.png",
"text_plain_output_305.png",
"text_plain_output_275.png",
"text_plain_output_779.png",
"application_vnd.jupyter.stderr_output_6.png",
"text_plain_output_795.png",
"text_plain_output_725.png",
"text_plain_output_301.png",
"application_vnd.jupyter.stderr_output_422.png",
"application_vnd.jupyter.stderr_output_162.png",
"application_vnd.jupyter.stderr_output_376.png",
"application_vnd.jupyter.stderr_output_676.png",
"application_vnd.jupyter.stderr_output_812.png",
"application_vnd.jupyter.stderr_output_794.png",
"application_vnd.jupyter.stderr_output_232.png",
"text_plain_output_691.png",
"application_vnd.jupyter.stderr_output_260.png",
"text_plain_output_467.png",
"text_plain_output_221.png",
"application_vnd.jupyter.stderr_output_576.png",
"application_vnd.jupyter.stderr_output_134.png",
"text_plain_output_155.png",
"application_vnd.jupyter.stderr_output_194.png",
"text_plain_output_65.png",
"text_plain_output_803.png",
"text_plain_output_419.png",
"application_vnd.jupyter.stderr_output_302.png",
"text_plain_output_215.png",
"application_vnd.jupyter.stderr_output_664.png",
"application_vnd.jupyter.stderr_output_546.png",
"application_vnd.jupyter.stderr_output_792.png",
"text_plain_output_189.png",
"text_plain_output_415.png",
"text_plain_output_637.png",
"application_vnd.jupyter.stderr_output_476.png",
"text_plain_output_13.png",
"application_vnd.jupyter.stderr_output_478.png",
"application_vnd.jupyter.stderr_output_656.png",
"text_plain_output_107.png",
"application_vnd.jupyter.stderr_output_336.png",
"application_vnd.jupyter.stderr_output_402.png",
"application_vnd.jupyter.stderr_output_542.png",
"application_vnd.jupyter.stderr_output_738.png",
"text_plain_output_567.png",
"application_vnd.jupyter.stderr_output_518.png",
"text_plain_output_695.png",
"application_vnd.jupyter.stderr_output_316.png",
"application_vnd.jupyter.stderr_output_468.png",
"application_vnd.jupyter.stderr_output_662.png",
"application_vnd.jupyter.stderr_output_750.png",
"text_plain_output_417.png",
"text_plain_output_707.png",
"text_plain_output_545.png",
"application_vnd.jupyter.stderr_output_714.png",
"text_plain_output_393.png",
"application_vnd.jupyter.stderr_output_762.png",
"application_vnd.jupyter.stderr_output_570.png",
"application_vnd.jupyter.stderr_output_404.png",
"application_vnd.jupyter.stderr_output_820.png",
"text_plain_output_243.png",
"application_vnd.jupyter.stderr_output_330.png",
"text_plain_output_611.png",
"application_vnd.jupyter.stderr_output_366.png",
"application_vnd.jupyter.stderr_output_278.png",
"text_plain_output_45.png",
"text_plain_output_599.png",
"application_vnd.jupyter.stderr_output_716.png",
"text_plain_output_665.png",
"application_vnd.jupyter.stderr_output_174.png",
"text_plain_output_257.png",
"text_plain_output_405.png",
"text_plain_output_353.png",
"application_vnd.jupyter.stderr_output_454.png",
"text_plain_output_277.png",
"text_plain_output_457.png",
"application_vnd.jupyter.stderr_output_510.png",
"text_plain_output_739.png",
"application_vnd.jupyter.stderr_output_12.png",
"text_plain_output_361.png",
"text_plain_output_171.png",
"text_plain_output_837.png",
"application_vnd.jupyter.stderr_output_720.png",
"application_vnd.jupyter.stderr_output_574.png",
"text_plain_output_561.png",
"text_plain_output_431.png",
"application_vnd.jupyter.stderr_output_644.png",
"application_vnd.jupyter.stderr_output_342.png",
"text_plain_output_159.png",
"text_plain_output_713.png",
"text_plain_output_29.png",
"text_plain_output_359.png",
"text_plain_output_529.png",
"text_plain_output_347.png",
"application_vnd.jupyter.stderr_output_82.png",
"text_plain_output_763.png",
"application_vnd.jupyter.stderr_output_288.png",
"text_plain_output_783.png",
"text_plain_output_129.png",
"application_vnd.jupyter.stderr_output_358.png",
"application_vnd.jupyter.stderr_output_398.png",
"application_vnd.jupyter.stderr_output_772.png",
"application_vnd.jupyter.stderr_output_388.png",
"text_plain_output_349.png",
"application_vnd.jupyter.stderr_output_332.png",
"application_vnd.jupyter.stderr_output_72.png",
"text_plain_output_483.png",
"text_plain_output_363.png",
"text_plain_output_289.png",
"application_vnd.jupyter.stderr_output_290.png",
"application_vnd.jupyter.stderr_output_586.png",
"text_plain_output_255.png",
"application_vnd.jupyter.stderr_output_8.png",
"text_plain_output_329.png",
"text_plain_output_49.png",
"application_vnd.jupyter.stderr_output_308.png",
"text_plain_output_791.png",
"text_plain_output_63.png",
"application_vnd.jupyter.stderr_output_394.png",
"application_vnd.jupyter.stderr_output_580.png",
"text_plain_output_27.png",
"text_plain_output_177.png",
"text_plain_output_607.png",
"application_vnd.jupyter.stderr_output_306.png",
"application_vnd.jupyter.stderr_output_604.png",
"application_vnd.jupyter.stderr_output_424.png",
"application_vnd.jupyter.stderr_output_534.png",
"text_plain_output_681.png",
"application_vnd.jupyter.stderr_output_834.png",
"text_plain_output_333.png",
"text_plain_output_581.png",
"application_vnd.jupyter.stderr_output_592.png",
"application_vnd.jupyter.stderr_output_824.png",
"application_vnd.jupyter.stderr_output_80.png",
"text_plain_output_269.png",
"application_vnd.jupyter.stderr_output_300.png",
"application_vnd.jupyter.stderr_output_818.png",
"text_plain_output_503.png",
"text_plain_output_735.png",
"text_plain_output_153.png",
"application_vnd.jupyter.stderr_output_784.png",
"application_vnd.jupyter.stderr_output_816.png",
"text_plain_output_57.png",
"application_vnd.jupyter.stderr_output_600.png",
"application_vnd.jupyter.stderr_output_728.png",
"text_plain_output_469.png",
"application_vnd.jupyter.stderr_output_10.png",
"application_vnd.jupyter.stderr_output_396.png",
"text_plain_output_357.png",
"text_plain_output_21.png",
"application_vnd.jupyter.stderr_output_464.png",
"application_vnd.jupyter.stderr_output_220.png",
"application_vnd.jupyter.stderr_output_810.png",
"text_plain_output_47.png",
"text_plain_output_623.png",
"application_vnd.jupyter.stderr_output_98.png",
"text_plain_output_121.png",
"text_plain_output_25.png",
"text_plain_output_523.png",
"text_plain_output_401.png",
"text_plain_output_77.png",
"text_plain_output_421.png",
"application_vnd.jupyter.stderr_output_34.png",
"text_plain_output_535.png",
"text_plain_output_527.png",
"text_plain_output_183.png",
"application_vnd.jupyter.stderr_output_536.png",
"text_plain_output_149.png",
"text_plain_output_383.png",
"text_plain_output_207.png",
"application_vnd.jupyter.stderr_output_756.png",
"application_vnd.jupyter.stderr_output_444.png",
"application_vnd.jupyter.stderr_output_90.png",
"text_plain_output_391.png",
"application_vnd.jupyter.stderr_output_538.png",
"application_vnd.jupyter.stderr_output_352.png",
"text_plain_output_413.png",
"text_plain_output_709.png",
"application_vnd.jupyter.stderr_output_584.png",
"application_vnd.jupyter.stderr_output_144.png",
"application_vnd.jupyter.stderr_output_140.png",
"text_plain_output_663.png",
"text_plain_output_87.png",
"text_plain_output_3.png",
"text_plain_output_217.png",
"text_plain_output_657.png",
"text_plain_output_427.png",
"application_vnd.jupyter.stderr_output_214.png",
"application_vnd.jupyter.stderr_output_44.png",
"text_plain_output_141.png",
"application_vnd.jupyter.stderr_output_742.png",
"application_vnd.jupyter.stderr_output_590.png",
"text_plain_output_225.png",
"text_plain_output_701.png",
"text_plain_output_191.png",
"text_plain_output_609.png",
"application_vnd.jupyter.stderr_output_320.png",
"text_plain_output_737.png",
"application_vnd.jupyter.stderr_output_544.png",
"text_plain_output_821.png",
"text_plain_output_259.png",
"application_vnd.jupyter.stderr_output_440.png",
"text_plain_output_447.png",
"text_plain_output_801.png",
"application_vnd.jupyter.stderr_output_160.png",
"text_plain_output_283.png",
"text_plain_output_495.png",
"text_plain_output_247.png",
"application_vnd.jupyter.stderr_output_42.png",
"text_plain_output_835.png",
"text_plain_output_113.png",
"text_plain_output_371.png",
"application_vnd.jupyter.stderr_output_602.png",
"application_vnd.jupyter.stderr_output_298.png",
"application_vnd.jupyter.stderr_output_598.png",
"application_vnd.jupyter.stderr_output_192.png",
"application_vnd.jupyter.stderr_output_770.png",
"text_plain_output_827.png",
"text_plain_output_479.png",
"application_vnd.jupyter.stderr_output_678.png",
"application_vnd.jupyter.stderr_output_702.png",
"text_plain_output_81.png",
"text_plain_output_69.png",
"application_vnd.jupyter.stderr_output_670.png",
"application_vnd.jupyter.stderr_output_84.png",
"text_plain_output_667.png",
"application_vnd.jupyter.stderr_output_180.png",
"text_plain_output_175.png",
"text_plain_output_165.png",
"text_plain_output_767.png",
"text_plain_output_145.png",
"application_vnd.jupyter.stderr_output_230.png",
"text_plain_output_125.png",
"application_vnd.jupyter.stderr_output_428.png",
"application_vnd.jupyter.stderr_output_314.png",
"application_vnd.jupyter.stderr_output_120.png",
"text_plain_output_487.png",
"application_vnd.jupyter.stderr_output_778.png",
"application_vnd.jupyter.stderr_output_752.png",
"text_plain_output_595.png",
"text_plain_output_643.png",
"text_plain_output_575.png",
"application_vnd.jupyter.stderr_output_558.png",
"text_plain_output_197.png",
"application_vnd.jupyter.stderr_output_60.png",
"application_vnd.jupyter.stderr_output_648.png",
"application_vnd.jupyter.stderr_output_216.png",
"text_plain_output_777.png",
"text_plain_output_315.png",
"text_plain_output_429.png",
"application_vnd.jupyter.stderr_output_372.png",
"application_vnd.jupyter.stderr_output_202.png",
"text_plain_output_517.png",
"text_plain_output_823.png",
"text_plain_output_433.png",
"text_plain_output_7.png",
"application_vnd.jupyter.stderr_output_184.png",
"application_vnd.jupyter.stderr_output_594.png",
"text_plain_output_513.png",
"text_plain_output_745.png",
"application_vnd.jupyter.stderr_output_390.png",
"application_vnd.jupyter.stderr_output_596.png",
"text_plain_output_645.png",
"text_plain_output_411.png",
"text_plain_output_91.png",
"text_plain_output_787.png",
"application_vnd.jupyter.stderr_output_688.png",
"text_plain_output_245.png",
"application_vnd.jupyter.stderr_output_660.png",
"text_plain_output_497.png",
"text_plain_output_789.png",
"application_vnd.jupyter.stderr_output_514.png",
"application_vnd.jupyter.stderr_output_826.png",
"application_vnd.jupyter.stderr_output_30.png",
"text_plain_output_265.png",
"application_vnd.jupyter.stderr_output_416.png",
"application_vnd.jupyter.stderr_output_108.png",
"application_vnd.jupyter.stderr_output_62.png",
"text_plain_output_435.png",
"text_plain_output_689.png",
"application_vnd.jupyter.stderr_output_328.png",
"text_plain_output_59.png",
"text_plain_output_409.png",
"text_plain_output_103.png",
"text_plain_output_71.png",
"application_vnd.jupyter.stderr_output_470.png",
"text_plain_output_751.png",
"text_plain_output_539.png",
"application_vnd.jupyter.stderr_output_250.png",
"application_vnd.jupyter.stderr_output_686.png",
"text_plain_output_211.png",
"application_vnd.jupyter.stderr_output_242.png",
"application_vnd.jupyter.stderr_output_654.png",
"application_vnd.jupyter.stderr_output_294.png",
"text_plain_output_601.png",
"application_vnd.jupyter.stderr_output_588.png",
"text_plain_output_541.png",
"application_vnd.jupyter.stderr_output_612.png",
"application_vnd.jupyter.stderr_output_130.png",
"application_vnd.jupyter.stderr_output_28.png",
"application_vnd.jupyter.stderr_output_364.png",
"application_vnd.jupyter.stderr_output_448.png",
"application_vnd.jupyter.stderr_output_658.png",
"application_vnd.jupyter.stderr_output_680.png",
"text_plain_output_749.png",
"text_plain_output_653.png",
"text_plain_output_543.png",
"text_plain_output_781.png",
"text_plain_output_451.png",
"application_vnd.jupyter.stderr_output_256.png",
"text_plain_output_109.png",
"application_vnd.jupyter.stderr_output_46.png",
"application_vnd.jupyter.stderr_output_780.png",
"text_plain_output_459.png",
"text_plain_output_615.png",
"text_plain_output_41.png",
"application_vnd.jupyter.stderr_output_206.png",
"application_vnd.jupyter.stderr_output_456.png",
"text_plain_output_253.png",
"application_vnd.jupyter.stderr_output_234.png",
"application_vnd.jupyter.stderr_output_734.png",
"application_vnd.jupyter.stderr_output_312.png",
"text_plain_output_723.png",
"application_vnd.jupyter.stderr_output_682.png",
"application_vnd.jupyter.stderr_output_630.png",
"text_plain_output_291.png",
"application_vnd.jupyter.stderr_output_616.png",
"application_vnd.jupyter.stderr_output_606.png",
"application_vnd.jupyter.stderr_output_708.png",
"text_plain_output_241.png",
"text_plain_output_231.png",
"text_plain_output_533.png",
"text_plain_output_345.png",
"text_plain_output_649.png",
"application_vnd.jupyter.stderr_output_252.png",
"application_vnd.jupyter.stderr_output_64.png",
"application_vnd.jupyter.stderr_output_76.png",
"text_plain_output_209.png",
"text_plain_output_185.png",
"application_vnd.jupyter.stderr_output_262.png",
"text_plain_output_85.png",
"text_plain_output_765.png",
"text_plain_output_605.png",
"text_plain_output_549.png",
"text_plain_output_67.png",
"text_plain_output_797.png",
"text_plain_output_573.png",
"text_plain_output_297.png",
"text_plain_output_53.png",
"text_plain_output_313.png",
"application_vnd.jupyter.stderr_output_480.png",
"application_vnd.jupyter.stderr_output_798.png",
"application_vnd.jupyter.stderr_output_572.png",
"application_vnd.jupyter.stderr_output_386.png",
"application_vnd.jupyter.stderr_output_20.png",
"text_plain_output_635.png",
"text_plain_output_703.png",
"text_plain_output_711.png",
"text_plain_output_193.png",
"text_plain_output_441.png",
"text_plain_output_403.png",
"application_vnd.jupyter.stderr_output_338.png",
"application_vnd.jupyter.stderr_output_126.png",
"application_vnd.jupyter.stderr_output_560.png",
"text_plain_output_23.png",
"text_plain_output_761.png",
"application_vnd.jupyter.stderr_output_218.png",
"application_vnd.jupyter.stderr_output_446.png",
"application_vnd.jupyter.stderr_output_494.png",
"text_plain_output_173.png",
"application_vnd.jupyter.stderr_output_36.png",
"text_plain_output_683.png",
"application_vnd.jupyter.stderr_output_100.png",
"text_plain_output_235.png",
"application_vnd.jupyter.stderr_output_430.png",
"application_vnd.jupyter.stderr_output_266.png",
"text_plain_output_151.png",
"text_plain_output_89.png",
"application_vnd.jupyter.stderr_output_22.png",
"text_plain_output_299.png",
"text_plain_output_51.png",
"text_plain_output_677.png",
"application_vnd.jupyter.stderr_output_748.png",
"application_vnd.jupyter.stderr_output_166.png",
"text_plain_output_825.png",
"application_vnd.jupyter.stderr_output_508.png",
"text_plain_output_525.png",
"application_vnd.jupyter.stderr_output_318.png",
"text_plain_output_731.png",
"text_plain_output_705.png",
"application_vnd.jupyter.stderr_output_292.png",
"application_vnd.jupyter.stderr_output_726.png",
"text_plain_output_99.png",
"text_plain_output_381.png",
"text_plain_output_571.png",
"text_plain_output_163.png",
"text_plain_output_179.png",
"application_vnd.jupyter.stderr_output_744.png",
"text_plain_output_537.png",
"application_vnd.jupyter.stderr_output_408.png",
"application_vnd.jupyter.stderr_output_374.png",
"application_vnd.jupyter.stderr_output_500.png",
"text_plain_output_569.png",
"text_plain_output_239.png",
"application_vnd.jupyter.stderr_output_186.png",
"application_vnd.jupyter.stderr_output_168.png",
"text_plain_output_127.png",
"text_plain_output_559.png",
"text_plain_output_311.png",
"text_plain_output_719.png",
"text_plain_output_295.png",
"text_plain_output_279.png",
"text_plain_output_507.png",
"application_vnd.jupyter.stderr_output_56.png",
"application_vnd.jupyter.stderr_output_452.png",
"text_plain_output_509.png",
"application_vnd.jupyter.stderr_output_104.png",
"text_plain_output_337.png",
"application_vnd.jupyter.stderr_output_196.png",
"text_plain_output_499.png",
"application_vnd.jupyter.stderr_output_50.png",
"text_plain_output_807.png",
"text_plain_output_563.png",
"application_vnd.jupyter.stderr_output_736.png",
"application_vnd.jupyter.stderr_output_114.png",
"text_plain_output_97.png",
"text_plain_output_729.png",
"application_vnd.jupyter.stderr_output_828.png",
"application_vnd.jupyter.stderr_output_492.png",
"text_plain_output_717.png",
"text_plain_output_227.png",
"application_vnd.jupyter.stderr_output_226.png",
"text_plain_output_453.png",
"text_plain_output_1.png",
"text_plain_output_33.png",
"application_vnd.jupyter.stderr_output_128.png",
"application_vnd.jupyter.stderr_output_150.png",
"text_plain_output_631.png",
"text_plain_output_39.png",
"application_vnd.jupyter.stderr_output_556.png",
"text_plain_output_335.png",
"application_vnd.jupyter.stderr_output_142.png",
"application_vnd.jupyter.stderr_output_326.png",
"text_plain_output_233.png",
"text_plain_output_473.png",
"application_vnd.jupyter.stderr_output_304.png",
"application_vnd.jupyter.stderr_output_830.png",
"text_plain_output_385.png",
"text_plain_output_55.png",
"text_plain_output_293.png",
"text_plain_output_199.png",
"application_vnd.jupyter.stderr_output_530.png",
"text_plain_output_463.png",
"text_plain_output_319.png",
"text_plain_output_805.png",
"application_vnd.jupyter.stderr_output_138.png",
"text_plain_output_829.png",
"application_vnd.jupyter.stderr_output_412.png",
"application_vnd.jupyter.stderr_output_548.png",
"text_plain_output_93.png",
"application_vnd.jupyter.stderr_output_200.png",
"text_plain_output_19.png",
"text_plain_output_439.png",
"text_plain_output_341.png",
"application_vnd.jupyter.stderr_output_280.png",
"text_plain_output_105.png",
"text_plain_output_465.png",
"text_plain_output_491.png",
"text_plain_output_679.png",
"text_plain_output_641.png",
"text_plain_output_249.png",
"application_vnd.jupyter.stderr_output_122.png",
"application_vnd.jupyter.stderr_output_488.png",
"application_vnd.jupyter.stderr_output_624.png",
"application_vnd.jupyter.stderr_output_94.png",
"text_plain_output_619.png",
"application_vnd.jupyter.stderr_output_282.png",
"application_vnd.jupyter.stderr_output_730.png",
"text_plain_output_17.png",
"text_plain_output_323.png",
"application_vnd.jupyter.stderr_output_462.png",
"text_plain_output_755.png",
"application_vnd.jupyter.stderr_output_652.png",
"application_vnd.jupyter.stderr_output_182.png",
"application_vnd.jupyter.stderr_output_158.png",
"text_plain_output_597.png",
"application_vnd.jupyter.stderr_output_78.png",
"text_plain_output_11.png",
"application_vnd.jupyter.stderr_output_698.png",
"application_vnd.jupyter.stderr_output_370.png",
"text_plain_output_481.png",
"application_vnd.jupyter.stderr_output_276.png",
"application_vnd.jupyter.stderr_output_188.png",
"application_vnd.jupyter.stderr_output_696.png",
"application_vnd.jupyter.stderr_output_14.png",
"text_plain_output_267.png",
"application_vnd.jupyter.stderr_output_562.png",
"text_plain_output_553.png",
"text_plain_output_425.png",
"text_plain_output_591.png",
"text_plain_output_811.png",
"application_vnd.jupyter.stderr_output_790.png",
"application_vnd.jupyter.stderr_output_706.png",
"text_plain_output_625.png",
"application_vnd.jupyter.stderr_output_350.png",
"text_plain_output_577.png",
"application_vnd.jupyter.stderr_output_54.png",
"application_vnd.jupyter.stderr_output_118.png",
"application_vnd.jupyter.stderr_output_154.png",
"text_plain_output_727.png",
"application_vnd.jupyter.stderr_output_438.png",
"text_plain_output_747.png",
"application_vnd.jupyter.stderr_output_442.png",
"application_vnd.jupyter.stderr_output_198.png",
"text_plain_output_519.png",
"text_plain_output_733.png",
"text_plain_output_721.png",
"application_vnd.jupyter.stderr_output_712.png",
"text_plain_output_757.png",
"text_plain_output_303.png",
"application_vnd.jupyter.stderr_output_746.png",
"text_plain_output_621.png",
"text_plain_output_377.png",
"application_vnd.jupyter.stderr_output_460.png",
"text_plain_output_95.png",
"text_plain_output_339.png",
"application_vnd.jupyter.stderr_output_228.png",
"application_vnd.jupyter.stderr_output_614.png",
"application_vnd.jupyter.stderr_output_254.png",
"text_plain_output_547.png",
"text_plain_output_369.png",
"application_vnd.jupyter.stderr_output_582.png",
"application_vnd.jupyter.stderr_output_628.png",
"text_plain_output_587.png",
"application_vnd.jupyter.stderr_output_466.png",
"application_vnd.jupyter.stderr_output_340.png",
"text_plain_output_365.png",
"text_plain_output_815.png",
"application_vnd.jupyter.stderr_output_208.png",
"text_plain_output_61.png",
"text_plain_output_585.png",
"text_plain_output_83.png",
"application_vnd.jupyter.stderr_output_248.png",
"text_plain_output_647.png",
"application_vnd.jupyter.stderr_output_210.png",
"application_vnd.jupyter.stderr_output_92.png",
"application_vnd.jupyter.stderr_output_164.png",
"application_vnd.jupyter.stderr_output_102.png",
"text_plain_output_397.png",
"text_plain_output_661.png",
"text_plain_output_389.png",
"application_vnd.jupyter.stderr_output_788.png",
"application_vnd.jupyter.stderr_output_410.png",
"text_plain_output_351.png",
"application_vnd.jupyter.stderr_output_40.png",
"application_vnd.jupyter.stderr_output_532.png",
"application_vnd.jupyter.stderr_output_244.png",
"text_plain_output_135.png",
"text_plain_output_285.png",
"application_vnd.jupyter.stderr_output_264.png",
"application_vnd.jupyter.stderr_output_486.png",
"text_plain_output_675.png",
"application_vnd.jupyter.stderr_output_646.png",
"text_plain_output_493.png",
"application_vnd.jupyter.stderr_output_434.png"
] | # Install nb_black for autoformatting
!pip install nb_black --quiet | code |
121149832/cell_19 | [
"text_html_output_1.png"
] | from kaggle_secrets import UserSecretsClient
from logging import getLogger, INFO, FileHandler, Formatter, StreamHandler
from sklearn.metrics import accuracy_score
import json
import numpy as np
import os
import os
import pandas as pd
import random
import torch
import wandb
import warnings
import numpy as np
import pandas as pd
import math
import random
import time
from collections import OrderedDict
import tensorflow as tf
from tqdm import tqdm
import json
import os
import gc
from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split, GroupKFold, StratifiedGroupKFold
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Adam, SGD, AdamW
from torch.optim.optimizer import Optimizer
import torchvision.models as models
from torch.nn.parameter import Parameter
from torch.utils.data import DataLoader, Dataset
from torch.optim.lr_scheduler import CosineAnnealingWarmRestarts, CosineAnnealingLR, ReduceLROnPlateau
from torchinfo import summary
import onnx
import onnx_tf
from onnx_tf.backend import prepare
import warnings
warnings.filterwarnings('ignore')
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
VERSION = 12
def get_score(y_true, y_pred):
score = accuracy_score(y_true, y_pred)
return score
def seed_torch(seed=42):
random.seed(seed)
os.environ['PYTHONHASHSEED'] = str(seed)
np.random.seed(seed)
torch.manual_seed(seed)
torch.cuda.manual_seed(seed)
torch.backends.cudnn.deterministic = True
def load_relevant_data_subset_with_imputation(pq_path):
data_columns = ['x', 'y', 'z']
data = pd.read_parquet(pq_path, columns=data_columns)
data.replace(np.nan, 0, inplace=True)
n_frames = int(len(data) / CFG.rows_per_frame)
data = data.values.reshape(n_frames, CFG.rows_per_frame, len(data_columns))
return data.astype(np.float32)
def load_relevant_data_subset(pq_path):
data_columns = ['x', 'y']
data = pd.read_parquet(pq_path, columns=data_columns)
n_frames = int(len(data) / CFG.rows_per_frame)
data = data.values.reshape(n_frames, CFG.rows_per_frame, len(data_columns))
return data.astype(np.float32)
def read_dict(file_path):
path = os.path.expanduser(file_path)
with open(path, 'r') as f:
dic = json.load(f)
return dic
class CFG:
num_workers = 2
apex = False
scheduler = 'CosineAnnealingLR'
epochs = 500
print_freq = 200
cosanneal_params = {'T_max': 5, 'eta_min': 3 * 1e-05, 'last_epoch': -1}
reduce_params = {'mode': 'min', 'factor': 0.1, 'patience': 6, 'eps': 1e-06, 'verbose': True}
cosanneal_res_params = {'T_0': 3, 'eta_min': 1e-06, 'T_mult': 1, 'last_epoch': -1}
onecycle_params = {'pct_start': 0.1, 'div_factor': 10.0, 'max_lr': 0.001, 'steps_per_epoch': 3, 'epochs': 3}
momentum = 0.9
model_name = 'NN_ArcFace'
lr = 3 * 0.0001
weight_decay = 0.0001
gradient_accumulation_steps = 1
max_grad_norm = 1000
data_path = '../input/asl-signs/'
debug = False
arcface = True
use_aggregation_dataset = True
target_size = 250
rows_per_frame = 543
batch_size = 512
train = True
early_stop = True
target_col = 'label'
scale = 30.0
margin = 0.5
easy_margin = False
ls_eps = 0.0
fc_dim = 512
early_stopping_steps = 5
grad_cam = False
seed = 42
import os
OUTPUT_DIR = f'./{CFG.model_name}_version{VERSION}/'
if not os.path.exists(OUTPUT_DIR):
os.makedirs(OUTPUT_DIR)
def init_logger(log_file=OUTPUT_DIR + 'train.log'):
from logging import getLogger, INFO, FileHandler, Formatter, StreamHandler
logger = getLogger(__name__)
logger.setLevel(INFO)
handler1 = StreamHandler()
handler1.setFormatter(Formatter('%(message)s'))
handler2 = FileHandler(filename=log_file)
handler2.setFormatter(Formatter('%(message)s'))
logger.addHandler(handler1)
logger.addHandler(handler2)
return logger
LOGGER = init_logger()
train = pd.read_csv(f'{CFG.data_path}train.csv')
label_index = read_dict(f'{CFG.data_path}sign_to_prediction_index_map.json')
index_label = dict([(label_index[key], key) for key in label_index])
train['label'] = train['sign'].map(lambda sign: label_index[sign])
if CFG.debug:
CFG.epochs = 1
train = train.sample(n=4000, random_state=CFG.seed).reset_index(drop=True)
from kaggle_secrets import UserSecretsClient
user_secrets = UserSecretsClient()
wandb_api = user_secrets.get_secret('wandb_key')
import wandb
wandb.login(key=wandb_api)
def class2dict(f):
return dict(((name, getattr(f, name)) for name in dir(f) if not name.startswith('__')))
run = wandb.init(project='GISLR Competition', name=f'{CFG.model_name}_Version{VERSION}', config=class2dict(CFG), group=CFG.model_name, job_type='train') | code |
121149832/cell_3 | [
"text_html_output_4.png",
"text_html_output_2.png",
"text_html_output_5.png",
"text_html_output_1.png",
"application_vnd.jupyter.stderr_output_1.png",
"text_html_output_3.png"
] | !pip install onnx_tf
!pip install tflite-runtime
!pip install -q --upgrade wandb | code |
121149832/cell_17 | [
"text_plain_output_1.png"
] | from sklearn.metrics import accuracy_score
import json
import numpy as np
import os
import os
import pandas as pd
import random
import torch
import torch.nn as nn
import warnings
import numpy as np
import pandas as pd
import math
import random
import time
from collections import OrderedDict
import tensorflow as tf
from tqdm import tqdm
import json
import os
import gc
from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split, GroupKFold, StratifiedGroupKFold
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Adam, SGD, AdamW
from torch.optim.optimizer import Optimizer
import torchvision.models as models
from torch.nn.parameter import Parameter
from torch.utils.data import DataLoader, Dataset
from torch.optim.lr_scheduler import CosineAnnealingWarmRestarts, CosineAnnealingLR, ReduceLROnPlateau
from torchinfo import summary
import onnx
import onnx_tf
from onnx_tf.backend import prepare
import warnings
warnings.filterwarnings('ignore')
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
VERSION = 12
def get_score(y_true, y_pred):
score = accuracy_score(y_true, y_pred)
return score
def seed_torch(seed=42):
random.seed(seed)
os.environ['PYTHONHASHSEED'] = str(seed)
np.random.seed(seed)
torch.manual_seed(seed)
torch.cuda.manual_seed(seed)
torch.backends.cudnn.deterministic = True
def load_relevant_data_subset_with_imputation(pq_path):
data_columns = ['x', 'y', 'z']
data = pd.read_parquet(pq_path, columns=data_columns)
data.replace(np.nan, 0, inplace=True)
n_frames = int(len(data) / CFG.rows_per_frame)
data = data.values.reshape(n_frames, CFG.rows_per_frame, len(data_columns))
return data.astype(np.float32)
def load_relevant_data_subset(pq_path):
data_columns = ['x', 'y']
data = pd.read_parquet(pq_path, columns=data_columns)
n_frames = int(len(data) / CFG.rows_per_frame)
data = data.values.reshape(n_frames, CFG.rows_per_frame, len(data_columns))
return data.astype(np.float32)
def read_dict(file_path):
path = os.path.expanduser(file_path)
with open(path, 'r') as f:
dic = json.load(f)
return dic
lipsUpperOuter = [61, 185, 40, 39, 37, 0, 267, 269, 270, 409, 291]
lipsLowerOuter = [146, 91, 181, 84, 17, 314, 405, 321, 375, 291]
lipsUpperInner = [78, 191, 80, 81, 82, 13, 312, 311, 310, 415, 308]
lipsLowerInner = [78, 95, 88, 178, 87, 14, 317, 402, 318, 324, 308]
lips = lipsUpperOuter + lipsLowerOuter + lipsUpperInner + lipsLowerInner
class FeatureGen(nn.Module):
def __init__(self):
super(FeatureGen, self).__init__()
pass
def forward(self, x):
x = x[:, :, :2]
lips_x = x[:, lips, :].contiguous().view(-1, 43 * 2)
lefth_x = x[:, 468:489, :].contiguous().view(-1, 21 * 2)
pose_x = x[:, 489:522, :].contiguous().view(-1, 33 * 2)
righth_x = x[:, 522:, :].contiguous().view(-1, 21 * 2)
lefth_x = lefth_x[~torch.any(torch.isnan(lefth_x), dim=1), :]
righth_x = righth_x[~torch.any(torch.isnan(righth_x), dim=1), :]
x1m = torch.mean(lips_x, 0)
x2m = torch.mean(lefth_x, 0)
x3m = torch.mean(pose_x, 0)
x4m = torch.mean(righth_x, 0)
x1s = torch.std(lips_x, 0)
x2s = torch.std(lefth_x, 0)
x3s = torch.std(pose_x, 0)
x4s = torch.std(righth_x, 0)
xfeat = torch.cat([x1m, x2m, x3m, x4m, x1s, x2s, x3s, x4s], axis=0)
xfeat = torch.where(torch.isnan(xfeat), torch.tensor(0.0, dtype=torch.float32), xfeat)
return xfeat
feature_converter = FeatureGen()
X = np.load('/kaggle/input/isolated-sign-language-aggregation-preparation/feature_data.npy')
y = np.load('/kaggle/input/isolated-sign-language-aggregation-preparation/feature_labels.npy')
print(X.shape, y.shape) | code |
121149832/cell_14 | [
"text_plain_output_1.png"
] | from logging import getLogger, INFO, FileHandler, Formatter, StreamHandler
from sklearn.metrics import accuracy_score
import json
import numpy as np
import os
import os
import pandas as pd
import random
import torch
import warnings
import numpy as np
import pandas as pd
import math
import random
import time
from collections import OrderedDict
import tensorflow as tf
from tqdm import tqdm
import json
import os
import gc
from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split, GroupKFold, StratifiedGroupKFold
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.optim import Adam, SGD, AdamW
from torch.optim.optimizer import Optimizer
import torchvision.models as models
from torch.nn.parameter import Parameter
from torch.utils.data import DataLoader, Dataset
from torch.optim.lr_scheduler import CosineAnnealingWarmRestarts, CosineAnnealingLR, ReduceLROnPlateau
from torchinfo import summary
import onnx
import onnx_tf
from onnx_tf.backend import prepare
import warnings
warnings.filterwarnings('ignore')
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
VERSION = 12
def get_score(y_true, y_pred):
score = accuracy_score(y_true, y_pred)
return score
def seed_torch(seed=42):
random.seed(seed)
os.environ['PYTHONHASHSEED'] = str(seed)
np.random.seed(seed)
torch.manual_seed(seed)
torch.cuda.manual_seed(seed)
torch.backends.cudnn.deterministic = True
def load_relevant_data_subset_with_imputation(pq_path):
data_columns = ['x', 'y', 'z']
data = pd.read_parquet(pq_path, columns=data_columns)
data.replace(np.nan, 0, inplace=True)
n_frames = int(len(data) / CFG.rows_per_frame)
data = data.values.reshape(n_frames, CFG.rows_per_frame, len(data_columns))
return data.astype(np.float32)
def load_relevant_data_subset(pq_path):
data_columns = ['x', 'y']
data = pd.read_parquet(pq_path, columns=data_columns)
n_frames = int(len(data) / CFG.rows_per_frame)
data = data.values.reshape(n_frames, CFG.rows_per_frame, len(data_columns))
return data.astype(np.float32)
def read_dict(file_path):
path = os.path.expanduser(file_path)
with open(path, 'r') as f:
dic = json.load(f)
return dic
class CFG:
num_workers = 2
apex = False
scheduler = 'CosineAnnealingLR'
epochs = 500
print_freq = 200
cosanneal_params = {'T_max': 5, 'eta_min': 3 * 1e-05, 'last_epoch': -1}
reduce_params = {'mode': 'min', 'factor': 0.1, 'patience': 6, 'eps': 1e-06, 'verbose': True}
cosanneal_res_params = {'T_0': 3, 'eta_min': 1e-06, 'T_mult': 1, 'last_epoch': -1}
onecycle_params = {'pct_start': 0.1, 'div_factor': 10.0, 'max_lr': 0.001, 'steps_per_epoch': 3, 'epochs': 3}
momentum = 0.9
model_name = 'NN_ArcFace'
lr = 3 * 0.0001
weight_decay = 0.0001
gradient_accumulation_steps = 1
max_grad_norm = 1000
data_path = '../input/asl-signs/'
debug = False
arcface = True
use_aggregation_dataset = True
target_size = 250
rows_per_frame = 543
batch_size = 512
train = True
early_stop = True
target_col = 'label'
scale = 30.0
margin = 0.5
easy_margin = False
ls_eps = 0.0
fc_dim = 512
early_stopping_steps = 5
grad_cam = False
seed = 42
import os
OUTPUT_DIR = f'./{CFG.model_name}_version{VERSION}/'
if not os.path.exists(OUTPUT_DIR):
os.makedirs(OUTPUT_DIR)
def init_logger(log_file=OUTPUT_DIR + 'train.log'):
from logging import getLogger, INFO, FileHandler, Formatter, StreamHandler
logger = getLogger(__name__)
logger.setLevel(INFO)
handler1 = StreamHandler()
handler1.setFormatter(Formatter('%(message)s'))
handler2 = FileHandler(filename=log_file)
handler2.setFormatter(Formatter('%(message)s'))
logger.addHandler(handler1)
logger.addHandler(handler2)
return logger
LOGGER = init_logger()
train = pd.read_csv(f'{CFG.data_path}train.csv')
label_index = read_dict(f'{CFG.data_path}sign_to_prediction_index_map.json')
index_label = dict([(label_index[key], key) for key in label_index])
train['label'] = train['sign'].map(lambda sign: label_index[sign])
if CFG.debug:
CFG.epochs = 1
train = train.sample(n=4000, random_state=CFG.seed).reset_index(drop=True)
train.head() | code |
48166170/cell_13 | [
"text_html_output_1.png"
] | k_range = range(1, 21)
print('k range', k_range) | code |
48166170/cell_39 | [
"text_plain_output_1.png"
] | from sklearn import metrics
from sklearn.model_selection import GridSearchCV
from sklearn.neighbors import KNeighborsClassifier
val_results = []
train_results = []
k_range = range(1, 21)
for k in k_range:
clf_2 = KNeighborsClassifier(n_neighbors=k)
clf_2 = clf_2.fit(X_train, y_train)
pred_train = clf_2.predict(X_train)
train_score = metrics.accuracy_score(y_train, pred_train)
train_results.append(train_score)
pred_val = clf_2.predict(X_val)
val_score = metrics.accuracy_score(y_val, pred_val)
val_results.append(val_score)
clf_3 = KNeighborsClassifier()
param_grid = [{'weights': ['uniform'], 'n_neighbors': list(range(1, 21))}, {'weights': ['distance'], 'n_neighbors': list(range(1, 21))}]
gs = GridSearchCV(clf_3, param_grid, scoring='accuracy', cv=10)
gs = gs.fit(X_train, y_train)
clf_best = gs.best_estimator_
clf_best.fit(X_train, y_train)
y_pred = clf_best.predict(X_test)
print(metrics.accuracy_score(y_test, y_pred)) | code |
48166170/cell_26 | [
"image_output_1.png"
] | from sklearn import metrics
from sklearn.model_selection import cross_val_score, learning_curve, validation_curve
from sklearn.neighbors import KNeighborsClassifier
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
df2 = pd.read_csv('../input/diabetescsv/diabetes.csv')
X = df2.iloc[:, 0:8]
y = df2.iloc[:, 8]
val_results = []
train_results = []
k_range = range(1, 21)
for k in k_range:
clf_2 = KNeighborsClassifier(n_neighbors=k)
clf_2 = clf_2.fit(X_train, y_train)
pred_train = clf_2.predict(X_train)
train_score = metrics.accuracy_score(y_train, pred_train)
train_results.append(train_score)
pred_val = clf_2.predict(X_val)
val_score = metrics.accuracy_score(y_val, pred_val)
val_results.append(val_score)
clf_2 = KNeighborsClassifier()
def plot_validation_curve(clf, X, y, param_name, param_range):
train_scores, test_scores = validation_curve(clf, X, y, cv=10, scoring='accuracy', param_name=param_name, param_range=param_range, n_jobs=-1)
x_range = param_range
train_scores_mean = np.mean(train_scores, axis=1)
test_scores_mean = np.mean(test_scores, axis=1)
plot_validation_curve(clf_2, X_train, y_train, param_name='n_neighbors', param_range=range(1, 21)) | code |
48166170/cell_7 | [
"text_plain_output_1.png"
] | import pandas as pd
df2 = pd.read_csv('../input/diabetescsv/diabetes.csv')
X = df2.iloc[:, 0:8]
y = df2.iloc[:, 8]
print(X) | code |
48166170/cell_17 | [
"text_plain_output_1.png"
] | from sklearn import metrics
from sklearn.neighbors import KNeighborsClassifier
import matplotlib.pyplot as plt
val_results = []
train_results = []
k_range = range(1, 21)
for k in k_range:
clf_2 = KNeighborsClassifier(n_neighbors=k)
clf_2 = clf_2.fit(X_train, y_train)
pred_train = clf_2.predict(X_train)
train_score = metrics.accuracy_score(y_train, pred_train)
train_results.append(train_score)
pred_val = clf_2.predict(X_val)
val_score = metrics.accuracy_score(y_val, pred_val)
val_results.append(val_score)
plt.plot(k_range, val_results, 'b-', label='Val score')
plt.plot(k_range, train_results, 'r-', label='Train score')
plt.ylabel('Score')
plt.xlabel('Model complexity: k')
plt.legend()
plt.grid(True)
plt.show() | code |
48166170/cell_31 | [
"text_plain_output_1.png"
] | from sklearn.neighbors import KNeighborsClassifier
clf_3 = KNeighborsClassifier()
param_grid = [{'weights': ['uniform'], 'n_neighbors': list(range(1, 21))}, {'weights': ['distance'], 'n_neighbors': list(range(1, 21))}]
print(param_grid) | code |
48166170/cell_22 | [
"text_plain_output_1.png"
] | from sklearn import metrics
from sklearn.neighbors import KNeighborsClassifier
val_results = []
train_results = []
k_range = range(1, 21)
for k in k_range:
clf_2 = KNeighborsClassifier(n_neighbors=k)
clf_2 = clf_2.fit(X_train, y_train)
pred_train = clf_2.predict(X_train)
train_score = metrics.accuracy_score(y_train, pred_train)
train_results.append(train_score)
pred_val = clf_2.predict(X_val)
val_score = metrics.accuracy_score(y_val, pred_val)
val_results.append(val_score)
clf_2 = KNeighborsClassifier()
print(clf_2) | code |
48166170/cell_37 | [
"image_output_1.png"
] | from sklearn.model_selection import GridSearchCV
from sklearn.neighbors import KNeighborsClassifier
clf_3 = KNeighborsClassifier()
param_grid = [{'weights': ['uniform'], 'n_neighbors': list(range(1, 21))}, {'weights': ['distance'], 'n_neighbors': list(range(1, 21))}]
gs = GridSearchCV(clf_3, param_grid, scoring='accuracy', cv=10)
gs = gs.fit(X_train, y_train)
clf_best = gs.best_estimator_
print('best model:', clf_best.get_params())
clf_best.fit(X_train, y_train) | code |
48166170/cell_5 | [
"text_plain_output_2.png",
"text_plain_output_1.png"
] | import pandas as pd
df2 = pd.read_csv('../input/diabetescsv/diabetes.csv')
df2.head(10) | code |
16163803/cell_13 | [
"image_output_1.png"
] | """his = model.fit_generator(train_gen,
epochs=10,
steps_per_epoch=len(X_train)/BATCH_SIZE,
validation_data=test_gen,
validation_steps=len(X_test)/BATCH_SIZE)""" | code |
16163803/cell_4 | [
"text_plain_output_1.png"
] | import cv2
import matplotlib.pyplot as plt
import os
import os
category = ['cat', 'dog']
EPOCHS = 50
IMGSIZE = 128
BATCH_SIZE = 32
STOPPING_PATIENCE = 15
VERBOSE = 1
MODEL_NAME = 'cnn_50epochs_imgsize128'
OPTIMIZER = 'adam'
TRAINING_DIR = '../input/dogs-vs-cats-redux-kernels-edition/train'
TEST_DIR = '../input/dogs-vs-cats-redux-kernels-edition/test'
for img in os.listdir(TRAINING_DIR)[7890:]:
img_path = os.path.join(TRAINING_DIR, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
plt.imshow(img_arr, cmap='gray')
plt.title(img.split('.')[0])
break | code |
16163803/cell_6 | [
"image_output_1.png"
] | import cv2
import matplotlib.pyplot as plt
import numpy as np # linear algebra
import os
import os
category = ['cat', 'dog']
EPOCHS = 50
IMGSIZE = 128
BATCH_SIZE = 32
STOPPING_PATIENCE = 15
VERBOSE = 1
MODEL_NAME = 'cnn_50epochs_imgsize128'
OPTIMIZER = 'adam'
TRAINING_DIR = '../input/dogs-vs-cats-redux-kernels-edition/train'
TEST_DIR = '../input/dogs-vs-cats-redux-kernels-edition/test'
for img in os.listdir(TRAINING_DIR)[7890:]:
img_path = os.path.join(TRAINING_DIR, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
break
def create_train_data(path):
X = []
y = []
for img in os.listdir(path):
if img == os.listdir(path)[7889]:
continue
img_path = os.path.join(path, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
img_arr = img_arr / 255.0
cat = np.where(img.split('.')[0] == 'dog', 1, 0)
X.append(img_arr)
y.append(cat)
X = np.array(X).reshape(-1, IMGSIZE, IMGSIZE, 1)
y = np.array(y)
return (X, y)
X, y = create_train_data(TRAINING_DIR)
print(f'features shape {X.shape}.\nlabel shape {y.shape}.') | code |
16163803/cell_2 | [
"text_plain_output_1.png"
] | import os
import os
print(os.listdir('../input')) | code |
16163803/cell_11 | [
"text_plain_output_1.png"
] | from keras.layers import Conv2D, MaxPooling2D, BatchNormalization, Dropout
from keras.layers import Dense, Flatten
from keras.models import Sequential
from keras.utils import to_categorical
import cv2
import matplotlib.pyplot as plt
import numpy as np # linear algebra
import os
import os
category = ['cat', 'dog']
EPOCHS = 50
IMGSIZE = 128
BATCH_SIZE = 32
STOPPING_PATIENCE = 15
VERBOSE = 1
MODEL_NAME = 'cnn_50epochs_imgsize128'
OPTIMIZER = 'adam'
TRAINING_DIR = '../input/dogs-vs-cats-redux-kernels-edition/train'
TEST_DIR = '../input/dogs-vs-cats-redux-kernels-edition/test'
for img in os.listdir(TRAINING_DIR)[7890:]:
img_path = os.path.join(TRAINING_DIR, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
break
def create_train_data(path):
X = []
y = []
for img in os.listdir(path):
if img == os.listdir(path)[7889]:
continue
img_path = os.path.join(path, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
img_arr = img_arr / 255.0
cat = np.where(img.split('.')[0] == 'dog', 1, 0)
X.append(img_arr)
y.append(cat)
X = np.array(X).reshape(-1, IMGSIZE, IMGSIZE, 1)
y = np.array(y)
return (X, y)
X, y = create_train_data(TRAINING_DIR)
y = to_categorical(y, 2)
model = Sequential()
model.add(Conv2D(32, (3, 3), activation='relu', input_shape=X.shape[1:]))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Conv2D(128, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Flatten())
model.add(Dense(256, activation='relu'))
model.add(Dropout(0.6))
model.add(Dense(2, activation='sigmoid'))
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history = model.fit(X, y, epochs=EPOCHS, batch_size=BATCH_SIZE, validation_split=1 / 3) | code |
16163803/cell_1 | [
"application_vnd.jupyter.stderr_output_1.png"
] | import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
import tensorflow as tf
from keras.preprocessing.image import ImageDataGenerator
from keras.utils import to_categorical
from keras.models import Sequential
from keras.layers import Conv2D, MaxPooling2D, BatchNormalization, Dropout
from keras.layers import Dense, Flatten
import cv2 | code |
16163803/cell_7 | [
"image_output_1.png"
] | from keras.utils import to_categorical
import cv2
import matplotlib.pyplot as plt
import numpy as np # linear algebra
import os
import os
category = ['cat', 'dog']
EPOCHS = 50
IMGSIZE = 128
BATCH_SIZE = 32
STOPPING_PATIENCE = 15
VERBOSE = 1
MODEL_NAME = 'cnn_50epochs_imgsize128'
OPTIMIZER = 'adam'
TRAINING_DIR = '../input/dogs-vs-cats-redux-kernels-edition/train'
TEST_DIR = '../input/dogs-vs-cats-redux-kernels-edition/test'
for img in os.listdir(TRAINING_DIR)[7890:]:
img_path = os.path.join(TRAINING_DIR, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
break
def create_train_data(path):
X = []
y = []
for img in os.listdir(path):
if img == os.listdir(path)[7889]:
continue
img_path = os.path.join(path, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
img_arr = img_arr / 255.0
cat = np.where(img.split('.')[0] == 'dog', 1, 0)
X.append(img_arr)
y.append(cat)
X = np.array(X).reshape(-1, IMGSIZE, IMGSIZE, 1)
y = np.array(y)
return (X, y)
X, y = create_train_data(TRAINING_DIR)
y = to_categorical(y, 2)
print(f'features shape {X.shape}.\nlabel shape {y.shape}.') | code |
16163803/cell_15 | [
"text_plain_output_1.png"
] | from keras.layers import Conv2D, MaxPooling2D, BatchNormalization, Dropout
from keras.layers import Dense, Flatten
from keras.models import Sequential
from keras.utils import to_categorical
import cv2
import matplotlib.pyplot as plt
import numpy as np # linear algebra
import os
import os
category = ['cat', 'dog']
EPOCHS = 50
IMGSIZE = 128
BATCH_SIZE = 32
STOPPING_PATIENCE = 15
VERBOSE = 1
MODEL_NAME = 'cnn_50epochs_imgsize128'
OPTIMIZER = 'adam'
TRAINING_DIR = '../input/dogs-vs-cats-redux-kernels-edition/train'
TEST_DIR = '../input/dogs-vs-cats-redux-kernels-edition/test'
for img in os.listdir(TRAINING_DIR)[7890:]:
img_path = os.path.join(TRAINING_DIR, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
break
def create_train_data(path):
X = []
y = []
for img in os.listdir(path):
if img == os.listdir(path)[7889]:
continue
img_path = os.path.join(path, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
img_arr = img_arr / 255.0
cat = np.where(img.split('.')[0] == 'dog', 1, 0)
X.append(img_arr)
y.append(cat)
X = np.array(X).reshape(-1, IMGSIZE, IMGSIZE, 1)
y = np.array(y)
return (X, y)
X, y = create_train_data(TRAINING_DIR)
y = to_categorical(y, 2)
model = Sequential()
model.add(Conv2D(32, (3, 3), activation='relu', input_shape=X.shape[1:]))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Conv2D(128, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Flatten())
model.add(Dense(256, activation='relu'))
model.add(Dropout(0.6))
model.add(Dense(2, activation='sigmoid'))
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history = model.fit(X, y, epochs=EPOCHS, batch_size=BATCH_SIZE, validation_split=1 / 3)
model.save_weights('CATSvsDOGS_model.h5')
model.save('CNN_CAT.model')
train_acc = model.evaluate(X_train, y_train, batch_size=32)
test_acc = (model.evaluate(X_test, y_test, batch_size=32),)
fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(18, 10))
ax1.plot(history.history['loss'], color='b', label='Training loss : {:0.4f}'.format(train_acc[0]))
ax1.plot(history.history['val_loss'], color='r', label='validation loss : {:0.4f}'.format(test_acc[0][0]))
ax1.set_xticks(np.arange(1, EPOCHS, 1))
ax1.set_yticks(np.arange(0, 1.0, 0.1))
ax1.legend()
ax2.plot(history.history['acc'], color='b', label='Training accuracy : {0:.4f}'.format(train_acc[1]))
ax2.plot(history.history['val_acc'], color='r', label='Validation accuracy : {0:.4f}'.format(test_acc[0][1]))
ax2.set_xticks(np.arange(1, EPOCHS, 1))
ax2.set_yticks(np.arange(0.4, 1.2, 0.1))
legend = plt.legend(loc='best', shadow=True)
plt.tight_layout()
plt.show() | code |
16163803/cell_16 | [
"text_plain_output_1.png"
] | from keras.layers import Conv2D, MaxPooling2D, BatchNormalization, Dropout
from keras.layers import Dense, Flatten
from keras.models import Sequential
from keras.utils import to_categorical
import cv2
import matplotlib.pyplot as plt
import numpy as np # linear algebra
import os
import os
category = ['cat', 'dog']
EPOCHS = 50
IMGSIZE = 128
BATCH_SIZE = 32
STOPPING_PATIENCE = 15
VERBOSE = 1
MODEL_NAME = 'cnn_50epochs_imgsize128'
OPTIMIZER = 'adam'
TRAINING_DIR = '../input/dogs-vs-cats-redux-kernels-edition/train'
TEST_DIR = '../input/dogs-vs-cats-redux-kernels-edition/test'
for img in os.listdir(TRAINING_DIR)[7890:]:
img_path = os.path.join(TRAINING_DIR, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
break
def create_train_data(path):
X = []
y = []
for img in os.listdir(path):
if img == os.listdir(path)[7889]:
continue
img_path = os.path.join(path, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
img_arr = img_arr / 255.0
cat = np.where(img.split('.')[0] == 'dog', 1, 0)
X.append(img_arr)
y.append(cat)
X = np.array(X).reshape(-1, IMGSIZE, IMGSIZE, 1)
y = np.array(y)
return (X, y)
X, y = create_train_data(TRAINING_DIR)
y = to_categorical(y, 2)
model = Sequential()
model.add(Conv2D(32, (3, 3), activation='relu', input_shape=X.shape[1:]))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Conv2D(128, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Flatten())
model.add(Dense(256, activation='relu'))
model.add(Dropout(0.6))
model.add(Dense(2, activation='sigmoid'))
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history = model.fit(X, y, epochs=EPOCHS, batch_size=BATCH_SIZE, validation_split=1 / 3)
model.save_weights('CATSvsDOGS_model.h5')
model.save('CNN_CAT.model')
train_acc = model.evaluate(X_train, y_train, batch_size=32)
test_acc = (model.evaluate(X_test, y_test, batch_size=32),)
fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(18, 10))
ax1.plot(history.history['loss'], color='b', label="Training loss : {:0.4f}".format(train_acc[0]))
ax1.plot(history.history['val_loss'], color='r', label="validation loss : {:0.4f}".format(test_acc[0][0]))
ax1.set_xticks(np.arange(1, EPOCHS, 1))
ax1.set_yticks(np.arange(0, 1., 0.1))
ax1.legend()
ax2.plot(history.history['acc'], color='b', label="Training accuracy : {0:.4f}".format(train_acc[1]))
ax2.plot(history.history['val_acc'], color='r',label="Validation accuracy : {0:.4f}".format(test_acc[0][1]))
ax2.set_xticks(np.arange(1, EPOCHS, 1))
ax2.set_yticks(np.arange(0.4, 1.2, 0.1))
legend = plt.legend(loc='best', shadow=True)
plt.tight_layout()
plt.show()
for img in os.listdir(TEST_DIR)[800:]:
img_path = os.path.join(TEST_DIR, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
plt.imshow(img_arr, cmap='gray')
plt.title(img.split('.')[0])
break | code |
16163803/cell_14 | [
"text_plain_output_1.png"
] | from keras.layers import Conv2D, MaxPooling2D, BatchNormalization, Dropout
from keras.layers import Dense, Flatten
from keras.models import Sequential
from keras.utils import to_categorical
import cv2
import matplotlib.pyplot as plt
import numpy as np # linear algebra
import os
import os
category = ['cat', 'dog']
EPOCHS = 50
IMGSIZE = 128
BATCH_SIZE = 32
STOPPING_PATIENCE = 15
VERBOSE = 1
MODEL_NAME = 'cnn_50epochs_imgsize128'
OPTIMIZER = 'adam'
TRAINING_DIR = '../input/dogs-vs-cats-redux-kernels-edition/train'
TEST_DIR = '../input/dogs-vs-cats-redux-kernels-edition/test'
for img in os.listdir(TRAINING_DIR)[7890:]:
img_path = os.path.join(TRAINING_DIR, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
break
def create_train_data(path):
X = []
y = []
for img in os.listdir(path):
if img == os.listdir(path)[7889]:
continue
img_path = os.path.join(path, img)
img_arr = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)
img_arr = cv2.resize(img_arr, (IMGSIZE, IMGSIZE))
img_arr = img_arr / 255.0
cat = np.where(img.split('.')[0] == 'dog', 1, 0)
X.append(img_arr)
y.append(cat)
X = np.array(X).reshape(-1, IMGSIZE, IMGSIZE, 1)
y = np.array(y)
return (X, y)
X, y = create_train_data(TRAINING_DIR)
y = to_categorical(y, 2)
model = Sequential()
model.add(Conv2D(32, (3, 3), activation='relu', input_shape=X.shape[1:]))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Conv2D(128, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Flatten())
model.add(Dense(256, activation='relu'))
model.add(Dropout(0.6))
model.add(Dense(2, activation='sigmoid'))
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
history = model.fit(X, y, epochs=EPOCHS, batch_size=BATCH_SIZE, validation_split=1 / 3)
model.save_weights('CATSvsDOGS_model.h5')
model.save('CNN_CAT.model')
train_acc = model.evaluate(X_train, y_train, batch_size=32)
test_acc = (model.evaluate(X_test, y_test, batch_size=32),) | code |
105192890/cell_1 | [
"application_vnd.jupyter.stderr_output_1.png"
] | import os
import torch
from torch import nn
from torch.utils.data import TensorDataset, DataLoader
import pandas as pd
import numpy as np
import math
import matplotlib.pyplot as plt
from sklearn.metrics import r2_score
warnings.simplefilter('ignore')
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename)) | code |
121150886/cell_25 | [
"text_plain_output_1.png"
] | from torch.utils.data import TensorDataset, DataLoader, RandomSampler, SequentialSampler
from transformers import AutoTokenizer, AutoModel
import re
import torch
import torch
import unicodedata
if torch.cuda.is_available():
device = torch.device('cuda')
else:
device = torch.device('cpu')
tokenizer = AutoTokenizer.from_pretrained('asafaya/bert-mini-arabic')
def text_preprocessing(text):
text = unicodedata.normalize('NFC', text)
text = re.sub('(@.*?)[\\s]', ' ', text)
text = re.sub('&', '&', text)
text = re.sub('\\s+', ' ', text).strip()
text = re.sub('^https?:\\/\\/.*[\\r\\n]*', '<URL>', text)
return text
import emoji
import unicodedata
def preprocessing_for_bert(data, version='mini', text_preprocessing_fn=text_preprocessing):
input_ids = []
attention_masks = []
tokenizer = AutoTokenizer.from_pretrained('asafaya/bert-mini-arabic') if version == 'mini' else AutoTokenizer.from_pretrained('asafaya/bert-base-arabic')
for i, sent in enumerate(data):
encoded_sent = tokenizer.encode_plus(text=text_preprocessing_fn(sent), add_special_tokens=True, max_length=MAX_LEN, padding='max_length', return_attention_mask=True, truncation=True)
input_ids.append(encoded_sent.get('input_ids'))
attention_masks.append(encoded_sent.get('attention_mask'))
input_ids = torch.tensor(input_ids)
attention_masks = torch.tensor(attention_masks)
return (input_ids, attention_masks)
print('Tokenizing data...')
test_inputs, test_masks = preprocessing_for_bert(X_test)
test_dataset = TensorDataset(test_inputs, test_masks)
test_sampler = SequentialSampler(test_dataset)
test_dataloader = DataLoader(test_dataset, sampler=test_sampler, batch_size=32) | code |
121150886/cell_4 | [
"text_plain_output_2.png",
"application_vnd.jupyter.stderr_output_1.png"
] | from ydata_profiling import ProfileReport
import pandas as pd
df_reviews = pd.read_csv('/kaggle/input/arabic-100k-reviews/ar_reviews_100k.tsv', delimiter='\t')
profile = ProfileReport(df_reviews, title='Profiling Report')
profile.to_notebook_iframe() | code |
121150886/cell_6 | [
"text_plain_output_1.png"
] | import torch
if torch.cuda.is_available():
device = torch.device('cuda')
print(f'Using {torch.cuda.device_count()} GPU(s)!')
print(f'Device name: {torch.cuda.get_device_name(0)}')
else:
device = torch.device('cpu')
print('No GPU available.') | code |
121150886/cell_19 | [
"text_plain_output_1.png"
] | from torch.utils.data import TensorDataset, DataLoader, RandomSampler, SequentialSampler
from transformers import AdamW, get_linear_schedule_with_warmup
from transformers import AutoTokenizer, AutoModel
import numpy as np
import random
import re
import time
import torch
import torch
import torch.nn as nn
import unicodedata
if torch.cuda.is_available():
device = torch.device('cuda')
else:
device = torch.device('cpu')
tokenizer = AutoTokenizer.from_pretrained('asafaya/bert-mini-arabic')
def text_preprocessing(text):
text = unicodedata.normalize('NFC', text)
text = re.sub('(@.*?)[\\s]', ' ', text)
text = re.sub('&', '&', text)
text = re.sub('\\s+', ' ', text).strip()
text = re.sub('^https?:\\/\\/.*[\\r\\n]*', '<URL>', text)
return text
import emoji
import unicodedata
def preprocessing_for_bert(data, version='mini', text_preprocessing_fn=text_preprocessing):
input_ids = []
attention_masks = []
tokenizer = AutoTokenizer.from_pretrained('asafaya/bert-mini-arabic') if version == 'mini' else AutoTokenizer.from_pretrained('asafaya/bert-base-arabic')
for i, sent in enumerate(data):
encoded_sent = tokenizer.encode_plus(text=text_preprocessing_fn(sent), add_special_tokens=True, max_length=MAX_LEN, padding='max_length', return_attention_mask=True, truncation=True)
input_ids.append(encoded_sent.get('input_ids'))
attention_masks.append(encoded_sent.get('attention_mask'))
input_ids = torch.tensor(input_ids)
attention_masks = torch.tensor(attention_masks)
return (input_ids, attention_masks)
from torch.utils.data import TensorDataset, DataLoader, RandomSampler, SequentialSampler
train_labels = torch.tensor(y_train)
val_labels = torch.tensor(y_val)
batch_size = 16
train_data = TensorDataset(train_inputs, train_masks, train_labels)
train_sampler = RandomSampler(train_data)
train_dataloader = DataLoader(train_data, sampler=train_sampler, batch_size=batch_size)
val_data = TensorDataset(val_inputs, val_masks, val_labels)
val_sampler = RandomSampler(val_data)
val_dataloader = DataLoader(val_data, sampler=val_sampler, batch_size=batch_size)
from transformers import AdamW, get_linear_schedule_with_warmup
from torch.optim import SparseAdam, Adam
def initialize_model(epochs=4, version='mini'):
bert_classifier = BertClassifier(freeze_bert=False, version=version)
bert_classifier.to(device)
optimizer = AdamW(params=list(bert_classifier.parameters()), lr=5e-05, eps=1e-08)
total_steps = len(train_dataloader) * epochs
schedular = get_linear_schedule_with_warmup(optimizer, num_warmup_steps=0, num_training_steps=total_steps)
return (bert_classifier, optimizer, schedular)
import random
import time
loss_fn = nn.CrossEntropyLoss()
def set_seed(seed_value=42):
random.seed(seed_value)
np.random.seed(seed_value)
torch.manual_seed(seed_value)
torch.cuda.manual_seed_all(seed_value)
def train(model, train_dataloader, val_dataloader=None, epochs=4, evaluation=False):
for epoch_i in range(epochs):
t0_epoch, t0_batch = (time.time(), time.time())
total_loss, batch_loss, batch_counts = (0, 0, 0)
model.train()
for step, batch in enumerate(train_dataloader):
batch_counts += 1
b_input_ids, b_attn_mask, b_labels = tuple((t.to(device) for t in batch))
model.zero_grad()
logits = model(b_input_ids, b_attn_mask)
loss = loss_fn(logits, b_labels)
batch_loss += loss.item()
total_loss += loss.item()
loss.backward()
torch.nn.utils.clip_grad_norm_(model.parameters(), 1.0)
optimizer.step()
scheduler.step()
if step % 100 == 0 and step != 0 or step == len(train_dataloader) - 1:
time_elapsed = time.time() - t0_batch
batch_loss, batch_counts = (0, 0)
t0_batch = time.time()
avg_train_loss = total_loss / len(train_dataloader)
if evaluation == True:
val_loss, val_accuracy = evaluate(model, val_dataloader)
time_elapsed = time.time() - t0_epoch
set_seed(42)
bert_classifier, optimizer, scheduler = initialize_model(epochs=2)
train(bert_classifier, train_dataloader, val_dataloader, epochs=2, evaluation=True) | code |
121150886/cell_1 | [
"text_plain_output_1.png"
] | !pip install Arabic-Stopwords
!pip install emoji
# !pip install Tashaphyne
# !pip install qalsadi
# !pip install np_utils
!pip install ydata-profiling | code |
121150886/cell_8 | [
"text_plain_output_1.png"
] | from transformers import AutoTokenizer, AutoModel
tokenizer = AutoTokenizer.from_pretrained('asafaya/bert-mini-arabic') | code |
121150886/cell_15 | [
"application_vnd.jupyter.stderr_output_2.png",
"text_plain_output_3.png",
"text_plain_output_1.png"
] | import torch
import torch.nn as nn
from transformers import BertModel
class BertClassifier(nn.Module):
def __init__(self, freeze_bert=False, version='mini'):
super(BertClassifier, self).__init__()
D_in = 256 if version == 'mini' else 758
H = 50
D_out = 2
self.bert = AutoModel.from_pretrained('asafaya/bert-mini-arabic') if version == 'mini' else AutoModel.from_pretrained('asafaya/bert-base-arabic')
self.classifier = nn.Sequential(nn.Linear(D_in, H), nn.ReLU(), nn.Dropout(0.5), nn.Linear(H, D_out))
if freeze_bert:
for param in self.bert.parameters():
param.requires_grad = False
def forward(self, input_ids, attention_mask):
outputs = self.bert(input_ids=input_ids, attention_mask=attention_mask)
last_hidden_state_cls = outputs[0][:, 0, :]
logits = self.classifier(last_hidden_state_cls)
return logits | code |
121150886/cell_12 | [
"text_html_output_1.png"
] | from sklearn.model_selection import train_test_split
from transformers import AutoTokenizer, AutoModel
import pandas as pd
import re
import torch
import unicodedata
df_reviews = pd.read_csv('/kaggle/input/arabic-100k-reviews/ar_reviews_100k.tsv', delimiter='\t')
label_mapping = {'Positive': 1, 'Negative': 0}
df_reviews = df_reviews[df_reviews.label != 'Mixed']
df_reviews.label = df_reviews.label.map(label_mapping)
df_reviews.label.value_counts()
if torch.cuda.is_available():
device = torch.device('cuda')
else:
device = torch.device('cpu')
tokenizer = AutoTokenizer.from_pretrained('asafaya/bert-mini-arabic')
def text_preprocessing(text):
text = unicodedata.normalize('NFC', text)
text = re.sub('(@.*?)[\\s]', ' ', text)
text = re.sub('&', '&', text)
text = re.sub('\\s+', ' ', text).strip()
text = re.sub('^https?:\\/\\/.*[\\r\\n]*', '<URL>', text)
return text
import emoji
import unicodedata
def preprocessing_for_bert(data, version='mini', text_preprocessing_fn=text_preprocessing):
input_ids = []
attention_masks = []
tokenizer = AutoTokenizer.from_pretrained('asafaya/bert-mini-arabic') if version == 'mini' else AutoTokenizer.from_pretrained('asafaya/bert-base-arabic')
for i, sent in enumerate(data):
encoded_sent = tokenizer.encode_plus(text=text_preprocessing_fn(sent), add_special_tokens=True, max_length=MAX_LEN, padding='max_length', return_attention_mask=True, truncation=True)
input_ids.append(encoded_sent.get('input_ids'))
attention_masks.append(encoded_sent.get('attention_mask'))
input_ids = torch.tensor(input_ids)
attention_masks = torch.tensor(attention_masks)
return (input_ids, attention_masks)
from sklearn.model_selection import train_test_split
X = df_reviews.text.values
y = df_reviews.label.values
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
X_test, X_val, y_test, y_val = train_test_split(X_test, y_test, test_size=0.5, random_state=42)
MAX_LEN = 280
token_ids = list(preprocessing_for_bert([X[0]])[0].squeeze().numpy())
print(f'Original: {X[0]}')
print(f'Token IDs: {token_ids}')
print('Tokenizing data...')
train_inputs, train_masks = preprocessing_for_bert(X_train)
val_inputs, val_masks = preprocessing_for_bert(X_val) | code |
121150886/cell_5 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import pandas as pd
df_reviews = pd.read_csv('/kaggle/input/arabic-100k-reviews/ar_reviews_100k.tsv', delimiter='\t')
label_mapping = {'Positive': 1, 'Negative': 0}
df_reviews = df_reviews[df_reviews.label != 'Mixed']
print(df_reviews.shape)
df_reviews.label = df_reviews.label.map(label_mapping)
df_reviews.label.value_counts() | code |
128034284/cell_4 | [
"application_vnd.jupyter.stderr_output_1.png"
] | import pandas as pd
def get_nan_dummy(series):
"""Given a Series containing NaN and several classes return a dummy Series
indicating 0 for NaN and 1 for non-NaN data
Parameters
----------
- series : pd.Series, input series or col to dummify
Return
------
- s : pd.Series the dummy Series
"""
s = series.notna().astype(int).astype('category')
s.name = f'{series.name}_abs_pres'
return s
data = pd.read_csv('./train.csv')
data = data.drop('Id', axis=1)
new_cols = data.columns.to_list()
new_cols.reverse()
data = data[new_cols]
all_nan = data.isnull().sum().sort_values(ascending=False)
all_nan = all_nan[all_nan > 0]
cols = ['PoolQC', 'Alley', 'Fence', 'FireplaceQu', 'LotFrontage', 'MasVnrArea', 'GarageQual', 'BsmtQual']
abs_pres_series = []
for col in cols:
abs_pres_series.append(get_nan_dummy(data[col]))
abs_pres_series = pd.concat(abs_pres_series, axis=1)
data = pd.concat([data, abs_pres_series], axis=1)
a = data['BsmtExposure'].isna()[data['BsmtExposure'].isna() == True].index
b = data['BsmtCond'].isna()[data['BsmtCond'].isna() == True].index
mystery_row = [x for x in set(a) - set(b)]
mystery_row = mystery_row[0]
data.loc[mystery_row, ['BsmtExposure', 'BsmtFinType2', 'BsmtCond', 'BsmtQual']]
data.drop(mystery_row, axis=0, inplace=True) | code |
128034284/cell_2 | [
"application_vnd.jupyter.stderr_output_1.png"
] | import seaborn as sns
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from scipy.stats import norm, zscore
from sklearn.compose import make_column_transformer
from sklearn.compose import make_column_selector
from sklearn.ensemble import GradientBoostingRegressor, HistGradientBoostingRegressor, HistGradientBoostingClassifier
from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split, cross_val_score, RandomizedSearchCV
from sklearn.model_selection import cross_validate
from sklearn.pipeline import make_pipeline
from sklearn.preprocessing import StandardScaler, OrdinalEncoder
from sklearn.utils import shuffle
from sklearn.experimental import enable_halving_search_cv
from sklearn.model_selection import HalvingGridSearchCV, HalvingRandomSearchCV
sns.set_theme() | code |
2019512/cell_9 | [
"image_output_5.png",
"image_output_7.png",
"image_output_4.png",
"image_output_8.png",
"image_output_6.png",
"image_output_3.png",
"image_output_2.png",
"image_output_1.png",
"image_output_10.png",
"image_output_9.png"
] | from sklearn.cluster import KMeans, SpectralClustering, AgglomerativeClustering
from sklearn.feature_selection import VarianceThreshold
from sklearn.manifold import MDS
from sklearn.preprocessing import StandardScaler
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 numpy as np
import pandas as pd
from subprocess import check_output
frame = pd.read_csv('../input/HR_comma_sep.csv').rename(columns={'sales': 'position'})
pd.DataFrame({'dtypes': frame.dtypes, 'isnull': pd.isnull(frame).any(), 'count distinct': [len(frame[name].unique()) for name in frame.columns], 'max': frame.select_dtypes(exclude=['object']).max(), 'min': frame.select_dtypes(exclude=['object']).min(), 'std': frame.select_dtypes(exclude=['object']).std()}, index=frame.columns).sort_values('count distinct', ascending=False)
unique_cat = set(frame['position'].unique()) | {'unknown'}
frame['position'] = pd.Categorical(frame['position'], categories=unique_cat)
frame['position_'] = frame['position'].cat.codes
frame['salary'] = pd.Categorical(frame['salary'], categories=['low', 'medium', 'unknown', 'high'], ordered=True)
frame['salary_'] = frame['salary'].cat.codes
from sklearn.manifold import MDS
from sklearn.preprocessing import StandardScaler
from sklearn.feature_selection import VarianceThreshold
frame_ = pd.get_dummies(frame.drop(['salary_', 'position_'], axis=1)).sample(3000)
sampled_index = frame_.index
selector = VarianceThreshold(threshold=0.95 * (1 - 0.95))
frame_ = selector.fit_transform(frame_)
mds = MDS(n_components=2)
scaler = StandardScaler(with_mean=False)
frame_ = scaler.fit_transform(frame_)
frame_ = mds.fit_transform(frame_)
pc_1 = [frame_[i][0] for i in range(len(frame_))]
pc_2 = [frame_[i][1] for i in range(len(frame_))]
### Clustering with kmeans ###
from sklearn.cluster import KMeans, SpectralClustering, AgglomerativeClustering
def plot_pc (label, frame):
pc_1 = frame[frame['label']==label].loc[:,'pc1']
pc_2 = frame[frame['label']==label].loc[:,'pc2']
plt.scatter(pc_1, pc_2, label=label)
plt.legend()
def plot_clusters(function, n_clusters, function_kwargs=None):
if function_kwargs==None:
function_kwargs = dict()
model = function(n_clusters=n_clusters, **function_kwargs)
labels = model.fit_predict(frame_)
results = pd.DataFrame({'label':labels, 'pc1':pc_1, 'pc2':pc_2})
for i in range(n_clusters):
plot_pc(i, results)
plt.title('{} ({} clusters)'.format(function.__name__, n_clusters))
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2,2, figsize=(15,13))
for i, ax in zip(range(2,10,2), fig.axes):
plt.subplot(ax)
plot_clusters(KMeans, n_clusters=i, function_kwargs={'init': 'random'});
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2,2, figsize=(15,13))
for i, ax in zip(range(2,10,2), fig.axes):
plt.subplot(ax)
plot_clusters(AgglomerativeClustering, n_clusters=i, function_kwargs={'linkage': 'ward',
'affinity':'euclidean'})
n_clusters = list(range(1, 20))
score = []
for n in n_clusters:
model = KMeans(n_clusters=n).fit(frame_)
score.append(model.inertia_)
plt.figure(figsize=(10, 8))
plt.plot(n_clusters, score)
plt.xlabel('number of clusters')
plt.ylabel('score')
plt.title('Elbow Method')
sns.despine()
ax = plt.gca()
ax.xaxis.get_ticklabels() | code |
2019512/cell_6 | [
"image_output_1.png"
] | from sklearn.feature_selection import VarianceThreshold
from sklearn.manifold import MDS
from sklearn.preprocessing import StandardScaler
import matplotlib.pyplot as plt
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
from subprocess import check_output
frame = pd.read_csv('../input/HR_comma_sep.csv').rename(columns={'sales': 'position'})
pd.DataFrame({'dtypes': frame.dtypes, 'isnull': pd.isnull(frame).any(), 'count distinct': [len(frame[name].unique()) for name in frame.columns], 'max': frame.select_dtypes(exclude=['object']).max(), 'min': frame.select_dtypes(exclude=['object']).min(), 'std': frame.select_dtypes(exclude=['object']).std()}, index=frame.columns).sort_values('count distinct', ascending=False)
unique_cat = set(frame['position'].unique()) | {'unknown'}
frame['position'] = pd.Categorical(frame['position'], categories=unique_cat)
frame['position_'] = frame['position'].cat.codes
frame['salary'] = pd.Categorical(frame['salary'], categories=['low', 'medium', 'unknown', 'high'], ordered=True)
frame['salary_'] = frame['salary'].cat.codes
from sklearn.manifold import MDS
from sklearn.preprocessing import StandardScaler
from sklearn.feature_selection import VarianceThreshold
frame_ = pd.get_dummies(frame.drop(['salary_', 'position_'], axis=1)).sample(3000)
sampled_index = frame_.index
selector = VarianceThreshold(threshold=0.95 * (1 - 0.95))
frame_ = selector.fit_transform(frame_)
print('{} features used'.format(frame_.shape[1]))
mds = MDS(n_components=2)
scaler = StandardScaler(with_mean=False)
frame_ = scaler.fit_transform(frame_)
frame_ = mds.fit_transform(frame_)
pc_1 = [frame_[i][0] for i in range(len(frame_))]
pc_2 = [frame_[i][1] for i in range(len(frame_))]
plt.figure(figsize=(10, 8))
plt.scatter(pc_1, pc_2, color='green')
plt.xlabel('pc 1')
plt.ylabel('pc 2') | code |
2019512/cell_2 | [
"image_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
from subprocess import check_output
frame = pd.read_csv('../input/HR_comma_sep.csv').rename(columns={'sales': 'position'})
frame.iloc[:, 5:].head() | code |
2019512/cell_1 | [
"text_html_output_1.png"
] | from subprocess import check_output
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
from subprocess import check_output
frame = pd.read_csv('../input/HR_comma_sep.csv').rename(columns={'sales': 'position'})
frame.iloc[:, :5].head() | code |
2019512/cell_7 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from sklearn.cluster import KMeans, SpectralClustering, AgglomerativeClustering
from sklearn.feature_selection import VarianceThreshold
from sklearn.manifold import MDS
from sklearn.preprocessing import StandardScaler
import matplotlib.pyplot as plt
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
from subprocess import check_output
frame = pd.read_csv('../input/HR_comma_sep.csv').rename(columns={'sales': 'position'})
pd.DataFrame({'dtypes': frame.dtypes, 'isnull': pd.isnull(frame).any(), 'count distinct': [len(frame[name].unique()) for name in frame.columns], 'max': frame.select_dtypes(exclude=['object']).max(), 'min': frame.select_dtypes(exclude=['object']).min(), 'std': frame.select_dtypes(exclude=['object']).std()}, index=frame.columns).sort_values('count distinct', ascending=False)
unique_cat = set(frame['position'].unique()) | {'unknown'}
frame['position'] = pd.Categorical(frame['position'], categories=unique_cat)
frame['position_'] = frame['position'].cat.codes
frame['salary'] = pd.Categorical(frame['salary'], categories=['low', 'medium', 'unknown', 'high'], ordered=True)
frame['salary_'] = frame['salary'].cat.codes
from sklearn.manifold import MDS
from sklearn.preprocessing import StandardScaler
from sklearn.feature_selection import VarianceThreshold
frame_ = pd.get_dummies(frame.drop(['salary_', 'position_'], axis=1)).sample(3000)
sampled_index = frame_.index
selector = VarianceThreshold(threshold=0.95 * (1 - 0.95))
frame_ = selector.fit_transform(frame_)
mds = MDS(n_components=2)
scaler = StandardScaler(with_mean=False)
frame_ = scaler.fit_transform(frame_)
frame_ = mds.fit_transform(frame_)
pc_1 = [frame_[i][0] for i in range(len(frame_))]
pc_2 = [frame_[i][1] for i in range(len(frame_))]
from sklearn.cluster import KMeans, SpectralClustering, AgglomerativeClustering
def plot_pc(label, frame):
pc_1 = frame[frame['label'] == label].loc[:, 'pc1']
pc_2 = frame[frame['label'] == label].loc[:, 'pc2']
plt.scatter(pc_1, pc_2, label=label)
plt.legend()
def plot_clusters(function, n_clusters, function_kwargs=None):
if function_kwargs == None:
function_kwargs = dict()
model = function(n_clusters=n_clusters, **function_kwargs)
labels = model.fit_predict(frame_)
results = pd.DataFrame({'label': labels, 'pc1': pc_1, 'pc2': pc_2})
for i in range(n_clusters):
plot_pc(i, results)
plt.title('{} ({} clusters)'.format(function.__name__, n_clusters))
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, figsize=(15, 13))
for i, ax in zip(range(2, 10, 2), fig.axes):
plt.subplot(ax)
plot_clusters(KMeans, n_clusters=i, function_kwargs={'init': 'random'}) | code |
2019512/cell_8 | [
"image_output_1.png"
] | from sklearn.cluster import KMeans, SpectralClustering, AgglomerativeClustering
from sklearn.feature_selection import VarianceThreshold
from sklearn.manifold import MDS
from sklearn.preprocessing import StandardScaler
import matplotlib.pyplot as plt
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
from subprocess import check_output
frame = pd.read_csv('../input/HR_comma_sep.csv').rename(columns={'sales': 'position'})
pd.DataFrame({'dtypes': frame.dtypes, 'isnull': pd.isnull(frame).any(), 'count distinct': [len(frame[name].unique()) for name in frame.columns], 'max': frame.select_dtypes(exclude=['object']).max(), 'min': frame.select_dtypes(exclude=['object']).min(), 'std': frame.select_dtypes(exclude=['object']).std()}, index=frame.columns).sort_values('count distinct', ascending=False)
unique_cat = set(frame['position'].unique()) | {'unknown'}
frame['position'] = pd.Categorical(frame['position'], categories=unique_cat)
frame['position_'] = frame['position'].cat.codes
frame['salary'] = pd.Categorical(frame['salary'], categories=['low', 'medium', 'unknown', 'high'], ordered=True)
frame['salary_'] = frame['salary'].cat.codes
from sklearn.manifold import MDS
from sklearn.preprocessing import StandardScaler
from sklearn.feature_selection import VarianceThreshold
frame_ = pd.get_dummies(frame.drop(['salary_', 'position_'], axis=1)).sample(3000)
sampled_index = frame_.index
selector = VarianceThreshold(threshold=0.95 * (1 - 0.95))
frame_ = selector.fit_transform(frame_)
mds = MDS(n_components=2)
scaler = StandardScaler(with_mean=False)
frame_ = scaler.fit_transform(frame_)
frame_ = mds.fit_transform(frame_)
pc_1 = [frame_[i][0] for i in range(len(frame_))]
pc_2 = [frame_[i][1] for i in range(len(frame_))]
### Clustering with kmeans ###
from sklearn.cluster import KMeans, SpectralClustering, AgglomerativeClustering
def plot_pc (label, frame):
pc_1 = frame[frame['label']==label].loc[:,'pc1']
pc_2 = frame[frame['label']==label].loc[:,'pc2']
plt.scatter(pc_1, pc_2, label=label)
plt.legend()
def plot_clusters(function, n_clusters, function_kwargs=None):
if function_kwargs==None:
function_kwargs = dict()
model = function(n_clusters=n_clusters, **function_kwargs)
labels = model.fit_predict(frame_)
results = pd.DataFrame({'label':labels, 'pc1':pc_1, 'pc2':pc_2})
for i in range(n_clusters):
plot_pc(i, results)
plt.title('{} ({} clusters)'.format(function.__name__, n_clusters))
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2,2, figsize=(15,13))
for i, ax in zip(range(2,10,2), fig.axes):
plt.subplot(ax)
plot_clusters(KMeans, n_clusters=i, function_kwargs={'init': 'random'});
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, figsize=(15, 13))
for i, ax in zip(range(2, 10, 2), fig.axes):
plt.subplot(ax)
plot_clusters(AgglomerativeClustering, n_clusters=i, function_kwargs={'linkage': 'ward', 'affinity': 'euclidean'}) | code |
2019512/cell_3 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
from subprocess import check_output
frame = pd.read_csv('../input/HR_comma_sep.csv').rename(columns={'sales': 'position'})
pd.DataFrame({'dtypes': frame.dtypes, 'isnull': pd.isnull(frame).any(), 'count distinct': [len(frame[name].unique()) for name in frame.columns], 'max': frame.select_dtypes(exclude=['object']).max(), 'min': frame.select_dtypes(exclude=['object']).min(), 'std': frame.select_dtypes(exclude=['object']).std()}, index=frame.columns).sort_values('count distinct', ascending=False) | code |
2019512/cell_10 | [
"text_html_output_1.png"
] | from sklearn.cluster import KMeans, SpectralClustering, AgglomerativeClustering
from sklearn.feature_selection import VarianceThreshold
from sklearn.manifold import MDS
from sklearn.preprocessing import StandardScaler
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 numpy as np
import pandas as pd
from subprocess import check_output
frame = pd.read_csv('../input/HR_comma_sep.csv').rename(columns={'sales': 'position'})
pd.DataFrame({'dtypes': frame.dtypes, 'isnull': pd.isnull(frame).any(), 'count distinct': [len(frame[name].unique()) for name in frame.columns], 'max': frame.select_dtypes(exclude=['object']).max(), 'min': frame.select_dtypes(exclude=['object']).min(), 'std': frame.select_dtypes(exclude=['object']).std()}, index=frame.columns).sort_values('count distinct', ascending=False)
unique_cat = set(frame['position'].unique()) | {'unknown'}
frame['position'] = pd.Categorical(frame['position'], categories=unique_cat)
frame['position_'] = frame['position'].cat.codes
frame['salary'] = pd.Categorical(frame['salary'], categories=['low', 'medium', 'unknown', 'high'], ordered=True)
frame['salary_'] = frame['salary'].cat.codes
from sklearn.manifold import MDS
from sklearn.preprocessing import StandardScaler
from sklearn.feature_selection import VarianceThreshold
frame_ = pd.get_dummies(frame.drop(['salary_', 'position_'], axis=1)).sample(3000)
sampled_index = frame_.index
selector = VarianceThreshold(threshold=0.95 * (1 - 0.95))
frame_ = selector.fit_transform(frame_)
mds = MDS(n_components=2)
scaler = StandardScaler(with_mean=False)
frame_ = scaler.fit_transform(frame_)
frame_ = mds.fit_transform(frame_)
pc_1 = [frame_[i][0] for i in range(len(frame_))]
pc_2 = [frame_[i][1] for i in range(len(frame_))]
### Clustering with kmeans ###
from sklearn.cluster import KMeans, SpectralClustering, AgglomerativeClustering
def plot_pc (label, frame):
pc_1 = frame[frame['label']==label].loc[:,'pc1']
pc_2 = frame[frame['label']==label].loc[:,'pc2']
plt.scatter(pc_1, pc_2, label=label)
plt.legend()
def plot_clusters(function, n_clusters, function_kwargs=None):
if function_kwargs==None:
function_kwargs = dict()
model = function(n_clusters=n_clusters, **function_kwargs)
labels = model.fit_predict(frame_)
results = pd.DataFrame({'label':labels, 'pc1':pc_1, 'pc2':pc_2})
for i in range(n_clusters):
plot_pc(i, results)
plt.title('{} ({} clusters)'.format(function.__name__, n_clusters))
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2,2, figsize=(15,13))
for i, ax in zip(range(2,10,2), fig.axes):
plt.subplot(ax)
plot_clusters(KMeans, n_clusters=i, function_kwargs={'init': 'random'});
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2,2, figsize=(15,13))
for i, ax in zip(range(2,10,2), fig.axes):
plt.subplot(ax)
plot_clusters(AgglomerativeClustering, n_clusters=i, function_kwargs={'linkage': 'ward',
'affinity':'euclidean'})
n_clusters = list(range(1, 20))
score = []
for n in n_clusters:
model = KMeans(n_clusters=n).fit(frame_)
score.append(model.inertia_)
sns.despine()
ax = plt.gca()
ax.xaxis.get_ticklabels()
n_clusters = 5
model = AgglomerativeClustering(n_clusters=n_clusters, linkage='ward', affinity='euclidean')
model.fit(frame_)
labels = model.labels_
results = pd.DataFrame({'label': labels, 'pc1': pc_1, 'pc2': pc_2})
plt.figure(figsize=(10, 8))
for i in range(n_clusters):
plot_pc(i, results) | code |