dataframe的一些注意点筛选、去重合并

bank_user=t[['user_id']]#筛选数用户id这一列并进行去重处理：
bank_user=bank_user.drop_duplicates(subset='user_id',keep='first')
bank_user.info()

df1 = pd.merge(left = df_bill_train, right= df_loan_train, how = 'left',on = 'user_id')
t1 = df1[(df1['time_x']<=df1['time_y'])]
bill_user = t1[['user_id']]
bill_user = bill_user.drop_duplicates(subset = 'user_id',keep='first')
bill_user.info()
df2 = pd.merge(left = df_browse_train, right= df_loan_train, how = 'left',on = 'user_id')
t2= df2[(df2['time_x']<=df2['time_y'])]
browse_user = t2[['user_id']]
browse_user= browse_user.drop_duplicates(subset = 'user_id',keep='first')
browse_user.info()
user1 = pd.merge(left=bank_user,right=browse_user,how='inner',on = 'user_id')
user2 = pd.merge(left=user1,right=bill_user,how='inner',on = 'user_id')
user2.info()