LOS_Preprocessing.py

import pandas as pd
import numpy as np

## Complete Preprocessing Steps
def preprocess_data(df):
    # Columns to drop
    cols_to_drop = ['HOSPITAL_CODE','PATIENTID','ADMISSION_DATE','DISCHARGE_DATE']
    df = df.drop(cols_to_drop,axis=1)
    
    # Setting the index
    df.set_index('CASE_ID',inplace=True)

    # Fixing the data types issue
    num_columns = ['AVAILABLE_EXTRA_ROOMS_IN_HOSPITAL','VISITORS_WITH_PATIENT','ADMISSION_DEPOSIT','LOS']
    cat_columns = [col for col in df.columns.tolist() if col not in num_columns]
    
    for col in cat_columns:
        df[col] = df[col].astype(object)    
    
    for col in num_columns:
        df[col] = df[col].astype(int)
    
    # One-Hot Encoding
    df_final = pd.get_dummies(df)    
    
    return df_final