Django queryset group by and count distincts

Question

In my django application, I have a set of donation, which I want to make statistics on. For each donation, a person is linked with his email address.

I want to compute for each month, the number of donors, that is the number of unique email addess.

So far I have:

# filter all donations by month
truncate_date = connection.ops.date_trunc_sql('month', 'date')
qs = Donation.objects.extra({'month': truncate_date})

# group by months and annotate each with the count of emails
donors = qs.values('month').annotate(count=Count('email')).order_by('month')

This works well, I get a monthly report of all email donors, but as I got duplicate in email, it does not filter email by unique values.

The current database used is Postgresql.

How this can be done?

Answer 1

Count takes a distinct argument, so you can do:

donors = qs.values('month').annotate(count=Count('email', distinct=True)).order_by('month')