题解 | #牛客网用户练习的平均次日留存率#

# 类SQL方式
import pandas as pd

df = pd.read_csv('nowcoder.csv',sep=',')
df['date'] = pd.to_datetime(df['date']).dt.date
df=df.sort_values(['user_id','date'],ascending=True).reset_index(drop=True).drop_duplicates(['user_id','date'])

df['test'] = df.groupby('user_id')['date'].transform('shift')
df['diff'] = df['date'] - df['test']

total = df.user_id.count()
userid = df[df['diff'] == '1 days'].user_id.count()
print(round(userid/total,2))