CAT_VS_DOG/src/split_data.py

import os
import argparse
import pandas as pd
from sklearn.model_selection import train_test_split
from get_data import read_params


def split_and_saved_data(config_path):
    config=read_params(config_path)
    test_data_path=config["split_data"]["test_path"]
    train_data_path=config["split_data"]["train_path"]
    raw_data_path=config["load_data"]["raw_dataset_csv"]
    split_ratio=config["split_data"]["test_size"]
    random_state=config["base"]["random_state"]
    df=pd.read_csv(raw_data_path,sep=",")
    train,test=train_test_split(df,test_size=split_ratio,random_state=random_state)
    train.to_csv(train_data_path,sep=',',index=False,encoding="utf-8")
    test.to_csv(test_data_path,sep=',',index=False,encoding="utf-8")





if __name__=="__main__":
    args=argparse.ArgumentParser()
    args.add_argument("--config",default="params.yaml")
    parsed_args=args.parse_args()
    split_and_saved_data(config_path=parsed_args.config)