I have 2000 strings in 670 000 dataset I would like to drop. They are not full duplicates (this is why I couldn\'t use df.drop_duplicates()), but they have the same id, this