TYP: DataFrame.__setitem__ with None
#1550
Conversation
cmp0xff
force-pushed
the
feature/setitem-none
branch
from
6f392cb to
746c760
Compare
| df.loc[:, ["x"]] = [[None], [pd.NA], [pd.NaT]] # type: ignore[assignment,index] | ||
| df.iloc[:, [0]] = [[None], [pd.NA], [pd.NaT]] # type: ignore[assignment,index] | ||
|
|
||
| df.loc[:, iter(["x"])] = [[None], [pd.NA], [pd.NaT]] # type: ignore[assignment,index] | ||
| df.iloc[:, iter([0])] = [[None], [pd.NA], [pd.NaT]] # type: ignore[assignment,index] |
There was a problem hiding this comment.
Not too sure why they don't work for mypy
There was a problem hiding this comment.
@Dr-Irv any insights on this one? I tried and got the same error.
There was a problem hiding this comment.
It's a mypy bug and a difference between the pyright setting of strictListInference, which we have turned on because we are using pyright in strict mode.
I created python/mypy#20420
Can you add a comment referring to the above issue for those tests?
| df[["col1"]] = [[None], [pd.NA]] # type: ignore[assignment,list-item] | ||
| df[iter(["col1"])] = [[None], [pd.NA]] # type: ignore[assignment] |
There was a problem hiding this comment.
Not too sure why they don't work for mypy
loicdiridollou
left a comment
loicdiridollou
left a comment
There was a problem hiding this comment.
Will approve preemptively, just one question for @Dr-Irv and then we should be good.
|
@loicdiridollou do you think the PR is good to merge? Also it would be good if we can make a new version after this PR, since we'll have all the regressions fixed, cc @Dr-Irv, thanks. |
3 participants
__setitem__should accept sequences containing None? #1538assert_type()to assert the type of any return value)AGENTS.md.