colomb8

> While this is a bug, I would say that the second test actually gives the correct results (because you're adding one month at a time, when you add 1...

> hi I would add that the dataframe gives 2021-06-30 if it has 1 row and 2021-06-28 in case of 2 identical rows... there are several bugs here, however imho...

> Other similar instances of this operation in other related classes to `DateOffset` use the method where they calculate the combined offset of all `n` multiplications and move the date...