python groupby sum single colonnes

df.groupby(['A','C'], as_index=False)['B'].sum()
Panicky Pony