Diff from mypy_primer, showing the effect of this PR on type check results on a corpus of open source code:

pandas (https://github.com/pandas-dev/pandas)
+ pandas/core/reshape/concat.py:936: error: Generator has incompatible item type "ndarray[tuple[int, ...], dtype[Any]]"; expected "ndarray[tuple[int], dtype[Any]]"  [misc]
+ pandas/core/reshape/concat.py:940: error: Argument 1 to "append" of "list" has incompatible type "ndarray[tuple[int, ...], dtype[Any]]"; expected "ndarray[tuple[int], dtype[Any]]"  [arg-type]

Diff from mypy_primer, showing the effect of this PR on type check results on a corpus of open source code:

pandas (https://github.com/pandas-dev/pandas)
+ pandas/core/reshape/concat.py:936: error: Generator has incompatible item type "ndarray[tuple[int, ...], dtype[Any]]"; expected "ndarray[tuple[int], dtype[Any]]"  [misc]
+ pandas/core/reshape/concat.py:940: error: Argument 1 to "append" of "list" has incompatible type "ndarray[tuple[int, ...], dtype[Any]]"; expected "ndarray[tuple[int], dtype[Any]]"  [arg-type]

relevant pandas code:
https://github.com/pandas-dev/pandas/blob/cf6de581e1c27f06107b5be9e7f2923d7bbc98a6/pandas/core/reshape/concat.py#L918-L940

The change to np.repeat will apparently cause new typing errors in pandas.

@mroeschke since you worked on pandas-dev/pandas#61265, perhaps you are willing to share your thoughts on this change?

I'm fine to adapt to these typing changes in pandas (mostly by adding ignores or simple annotations until someone from the pandas community would like to improve the annotations internally)

I'm fine to adapt to these typing changes in pandas (mostly by adding ignores or simple annotations until someone from the pandas community would like to improve the annotations internally)

Good to hear, thanks!

Thanks Marco