2023-04-20 20:38:39 +02:00
|
|
|
import pandas
|
|
|
|
|
|
|
|
|
|
X_train = pandas.read_csv('X_train.csv',
|
|
|
|
|
engine='python',
|
|
|
|
|
encoding='ISO-8859-1',
|
|
|
|
|
sep=',')
|
|
|
|
|
X_dev = pandas.read_csv('X_dev.csv',
|
|
|
|
|
engine='python',
|
|
|
|
|
encoding='ISO-8859-1',
|
|
|
|
|
sep=',')
|
|
|
|
|
X_test = pandas.read_csv('X_test.csv',
|
|
|
|
|
engine='python',
|
|
|
|
|
encoding='ISO-8859-1',
|
|
|
|
|
sep=',')
|
|
|
|
|
|
|
|
|
|
X_train.describe(include='all').to_csv('X_train_stats.csv', index=True)
|
|
|
|
|
X_dev.describe(include='all').to_csv('X_dev_stats.csv', index=True)
|
2023-04-20 21:08:39 +02:00
|
|
|
X_test.describe(include='all').to_csv('X_test_stats.csv', index=True)
|
|
|
|
|
print(X_train.describe(include='all'))
|
|
|
|
|
print(X_dev.describe(include='all'))
|
|
|
|
|
print(X_test.describe(include='all'))
|
