BUG: first("2M") returning incorrect results #38446
Conversation
|
If we decide to backport, we could cherry pick this one on top of the other Commit. But I am not quite sure either if we should backport. |
|
we can just revert the original and then merge to master |
|
Oh right, that would work. Should I open a Revert PR? |
|
Ok, thx. Will adjust this one after the other is merged |
|
after the revert is merged, will reopen issue and change milestone. so can update the OP here to close issue again. |
…623_fix � Conflicts: � pandas/core/generic.py � pandas/tests/frame/methods/test_first_and_last.py
|
Done. This fixes both cases now. |
doc/source/whatsnew/v1.2.0.rst
Outdated
| @@ -607,6 +607,7 @@ Datetimelike | |||
| - Bug in :meth:`Series.isin` with ``datetime64[ns]`` dtype and :meth:`.DatetimeIndex.isin` failing to consider timezone-aware and timezone-naive datetimes as always different (:issue:`35728`) | |||
| - Bug in :meth:`Series.isin` with ``PeriodDtype`` dtype and :meth:`PeriodIndex.isin` failing to consider arguments with different ``PeriodDtype`` as always different (:issue:`37528`) | |||
| - Bug in :class:`Period` constructor now correctly handles nanoseconds in the ``value`` argument (:issue:`34621` and :issue:`17053`) | |||
| - Bug in :meth:`DataFrame.first` and :meth:`Series.first` returning two months for offset one month when first day is last calendar day (:issue:`29623`) | |||
There was a problem hiding this comment.
so we have three options
- include in 1.2
- change to 1.3
- hold-off merging for a week or so, and then include in 1.2.1
i'm ok with 2 or 3, not so keen on 1.
There was a problem hiding this comment.
3. hold-off merging for a week or so
actually, no harm in adding a 1.2.1 whatsnew now.
There was a problem hiding this comment.
I am fine with either one of them
doc/source/whatsnew/v1.2.0.rst
Outdated
| @@ -607,6 +607,7 @@ Datetimelike | |||
| - Bug in :meth:`Series.isin` with ``datetime64[ns]`` dtype and :meth:`.DatetimeIndex.isin` failing to consider timezone-aware and timezone-naive datetimes as always different (:issue:`35728`) | |||
| - Bug in :meth:`Series.isin` with ``PeriodDtype`` dtype and :meth:`PeriodIndex.isin` failing to consider arguments with different ``PeriodDtype`` as always different (:issue:`37528`) | |||
| - Bug in :class:`Period` constructor now correctly handles nanoseconds in the ``value`` argument (:issue:`34621` and :issue:`17053`) | |||
| - Bug in :meth:`DataFrame.first` and :meth:`Series.first` returning two months for offset one month when first day is last calendar day (:issue:`29623`) | |||
pandas/core/generic.py
Outdated
| if not isinstance(offset, Tick) and offset.is_on_offset(self.index[0]): | ||
| # GH#29623 if first value is end of period, remove offset with n = 1 | ||
| # before adding the real offset | ||
| end_date = end = self.index[0] - to_offset(offset.name) + offset |
There was a problem hiding this comment.
offset.base instead of to_offset(offset.name)?
|
thanks @phofl |
black pandasgit diff upstream/master -u -- "*.py" | flake8 --diffThis would fix #29623 for offsets with n > 1
cc @simonjayhawkins