题解 | #牛客网用户练习的平均次日留存率#

import pandas as pd
import numpy as np
df1 = pd.read_csv('nowcoder.csv',sep=',')
import datetime
sum_a = df1['user_id'].count()
df2 = pd.merge(df1,df1,on='user_id')
df2['date_x'] = pd.to_datetime(df2['date_x'],format='%Y-%m-%d').dt.date
df2['date_y'] = pd.to_datetime(df2['date_y'],format='%Y-%m-%d').dt.date
a = pd.to_timedelta(1,unit='d')
sum_b = df2.loc[df2['date_y']-df2['date_x']==a,'user_id'].count()
print(round(sum_b / sum_a,2))