python dataframe average by group by column name code example

df.groupby(['A', 'B']).mean()