ium_458023/create-dataset.py

23 lines
751 B
Python
Raw Normal View History

2023-04-20 18:48:29 +02:00
import pandas
import os
from sklearn.model_selection import train_test_split
CUTOFF = int(os.environ['CUTOFF'])
salaries = pandas.read_csv('./ium_458023/ds_salaries.csv',engine='python',encoding='ISO-8859-1',sep=',')
salaries = salaries.dropna()
salaries = salaries.sample(100)
X,Y = salaries,salaries
# SPLIT BETWEEN DEV, TRAINS, AND TEST
salaries_train, salaries_temp, salaries_train, salaries_temp = train_test_split(X, Y, test_size=0.2, random_state=1)
salaries_dev, salaries_test, salaries_dev, salaries_test = train_test_split(salaries_temp, salaries_temp, test_size=0.2)
salaries_train.to_csv('salaries_train.csv', index=False)
salaries_dev.to_csv('salaries_dev.csv', index=False)
salaries_test.to_csv('salaries_test.csv', index=False)