GH1139 Series.rename inplace #1140
Conversation
|
@loicdiridollou let me know when you think you'll get to this. I'd like to do a new release soon, and want to know if I should wait for this change or not. |
|
I was planning to take care of it in the evening but if you need to make a
release feel free to go ahead !3;943 that.
…On Tue, Mar 4, 2025 at 2:58 PM Irv Lustig ***@***.***> wrote:
@loicdiridollou <https://github.com/loicdiridollou> let me know when you
think you'll get to this. I'd like to do a new release soon, and want to
know if I should wait for this change or not.
—
Reply to this email directly, view it on GitHub
<#1140 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AMQNSQCSSP6NGEOAKRHH2T32SYAXNAVCNFSM6AAAAABYFNRBHCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMOJYG43DKMZVGI>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
[image: Dr-Irv]*Dr-Irv* left a comment (pandas-dev/pandas-stubs#1140)
<#1140 (comment)>
@loicdiridollou <https://github.com/loicdiridollou> let me know when you
think you'll get to this. I'd like to do a new release soon, and want to
know if I should wait for this change or not.
—
Reply to this email directly, view it on GitHub
<#1140 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AMQNSQCSSP6NGEOAKRHH2T32SYAXNAVCNFSM6AAAAABYFNRBHCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMOJYG43DKMZVGI>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
loicdiridollou
force-pushed
the
gh1139_rename
branch
from
73a1f14 to
2a112e6
Compare
loicdiridollou
force-pushed
the
gh1139_rename
branch
from
2a112e6 to
470ccf4
Compare
tests/test_series.py
Outdated
| # TODO this should not raise an error, the lambda is matched with Hashable | ||
| # check( | ||
| # assert_type(pd.Series([1, 2, 3]).rename(lambda x: x**2, inplace=True), None), type(None) | ||
| # ) |
There was a problem hiding this comment.
I think the solution here is to not use Renamer, and have the first overload be for Callable[[Any], Label] with Literal[True] returning None, then have the second overload be for Hashable | None returning Self and the third overload just having Mapping[Any, Label] as the argument.
There was a problem hiding this comment.
If I do so mypy seems to match the return of the expression with Any instead of None, let me see how I can address
There was a problem hiding this comment.
I have pushed for you to take a look but not sure how to address, this is a mypy issue and not pyright, pyright is happy with the current setup and does not flag it.
There was a problem hiding this comment.
I got it to work. The issue is that Callable and Hashable are overlapping and mypy is having trouble with that. So instead of using Hashable, I used Scalar | tuple[Hashable, ...] which corresponds to the docs anyway.
So my review contains all the changes (as suggestions) that you should need to make.
pandas-stubs/core/series.pyi
Outdated
| @@ -1052,6 +1064,17 @@ class Series(IndexOpsMixin[S1], NDFrame): | |||
| errors: IgnoreRaise = ..., | |||
| ) -> None: ... | |||
| @overload | |||
| def rename( # type: ignore[overload-overlap] | |||
| self, | |||
| index: Hashable | None, | |||
There was a problem hiding this comment.
| index: Hashable | None, | |
| index: Scalar | tuple[Hashable,...] | None, |
assert_type()to assert the type of any return valueAdded some test framework migration as this was how I originally found the issue with
rename