grouped = df.groupby('group') for group_name, group in grouped: print(' CREATE TABLE {}('.format(group_name)) print(group["column"])