PANDAS groupby 2 columns then count and mean-CodePudding

I have a data frame of users and each time they entered a website, it looks like this:

(if there are x row with same week and date it means the user entered the site x time that date).

ID	week	date
1	2	20/07/21
2	3	23/07/21
2	3	23/07/21
2	3	26/07/21
2	4	30/07/21
2	4	30/07/21
2	4	30/07/21
2	4	31/07/21

so far I've managed to do this:

ID	week	date	days number
1	2	20/07/21	1
2	3	23/07/21	2
2	3	26/07/21	1
2	4	30/07/21	3
2	4	31/07/21	1

using this code:

df.groupby(['ID','week','date']).agg({'date':['count']})

but I need to calculate the mean times each user used the site by week, so each user has a row for each week. Therefor the output I need looks like this:

ID	week	mean days number
1	2	1
2	3	1.5
2	4	2

Any ideas how to continue?

Thanks!!

CodePudding user response：

Use:

(df.groupby(['ID', 'week', 'date'], as_index=False)['date']
 .agg('count')
 .groupby(['ID', 'week'], as_index=False)
 .agg(**{'mean days number': ('date', 'mean')})
)

Output:

   ID  week  mean days number
0   1     2               1.0
1   2     3               1.5
2   2     4               2.0