#Data Preprocessing Code.py

#Data Preprocessing Code = 
import numpy as np
import dask.dataframe as dd
import pandas as pd

#This part of the code loads datasets
df_full = pd.read_csv("Trips_Full Data.csv")
df = pd.read_csv("Trips_By_Distance.csv")

#This part of the code is responsible for data cleaning 
df["Population staying at home"] = df["Population staying at home"].fillna(0)

#This Show types of columns and count non-null values
df.dtypes
df.notnull().sum().compute()

df.describe().compute()
df['Population staying at home'].describe().compute()