从多索引 df 获取特定部分，而不会丢失 0 级索引

Question

Imagine you have the following multiindexed dataframe

pd.DataFrame({('0-Engajados', 'affiliate'): 119,
     ('0-Engajados', 'attendance bot'): 7,
     ('1-Onboarding + Retorno', 'affiliate'): 118,
     ('1-Onboarding + Retorno', 'attendance bot'): 7})

And you wanted to grab everyone who has a level 0 index equal '0-Engajados' but without losing, the index part. It is worh noting that the second index values vary, so iloc not quite an option.

Wanted end result

pd.DataFrame({('0-Engajados', 'affiliate'): 119,
     ('0-Engajados', 'attendance bot'): 7})

I tried df.loc['0-Engajados']

but that loses me the first index is there a way to grab it without losing it?

Answer 1

You can wrap your selection in a list to preserve its level:

import pandas as pd

df = pd.DataFrame({('0-Engajados', 'affiliate'): [119],
     ('0-Engajados', 'attendance bot'): [7],
     ('1-Onboarding + Retorno', 'affiliate'): [118],
     ('1-Onboarding + Retorno', 'attendance bot'): [7]})

print(df)
  0-Engajados                1-Onboarding + Retorno               
    affiliate attendance bot              affiliate attendance bot
0         119              7                    118              7


print(df.loc[:, ['0-Engajados']])
  0-Engajados               
    affiliate attendance bot
0         119              7