Pandas how to calculate mean for second level multi index columns

Question

In Pandas, How to get mean for second level columns?

        bar                 baz                 foo          
        one       two       one       two       one       two

aaa       1         2         3         4         5         6
bbb      -1        -2        -3        -4        -5        -6

Expected:

             one       two

aaa            3         4
bbb           -3        -4

Answer 1

Use parameters level=1 and axis=1 for MultiIndex in columns:

df = df.groupby(level=1, axis=1).mean()