Question

I’m currently working on a dataframe that has a subscription date for every members. I would like to stats subscriptions per months but default behavior would counts each dates of every month separately.

I found a way of doing it modifying the date with slices and setting every dates day on 01 but i would rather use something that’s made by pandas.

Any suggestion on where i should head to ?

Answer 1

If your subscription date is a datetime.datetime instance, then you could use (untested) something like (where df is your DataFrame):

df.groupby(lambda L: (L.year, L.month))

You’ll need to adjust the groupby if the datetime isn’t your DataFrame’s index.