I have dataframe like that. I need to choose and count all distinct users, who have title "banner_click" and "order". So I don't understand how to do it in pandas, in SQL you do like UniqExactIf and go on, there I need to find all users, who had both these title.

     user         title
0  user_0  banner_click  # has both "banner_click" and "order"
1  user_0         order  #
2  user_1   banner_show
3  user_1         order
4  user_2         order  # also has both "banner_click" and "order"
5  user_2  banner_click  #

I have tried "in" but I guess it doesn't work correct

main = df.query("title in ('banner_click','order')").agg({'user':'nunique'})

Expected output: 2 (user_0 and user2 are a match)

Reproducible example:

df = pd.DataFrame({'user': ['user_0', 'user_0', 'user_1', 'user_1', 'user_2', 'user_2'],'title': ['banner_click', 'order', 'banner_show', 'order', 'order', 'banner_click']})

You can use isin() to retrieve the relevant rows and then use df.nunique().

df.loc[df['title'].isin(['banner_click', 'order'])]['user'].nunique()