df.drop_duplicate()