Repository navigation
TYP: False positives in ndarray.__setitem__ with object_ dtype in NumPy 2.2 #27964
Description
Describe the issue:
Under NumPy 2.2, assignment of a slice to an object array results in a type check failure
Reproduce the code example:
import numpy as np
x = np.zeros(1, dtype=object)
x[0] = slice(None)Error message:
tmp.py:3: error: No overload variant of "__setitem__" of "ndarray" matches argument types "int", "slice" [call-overload]
tmp.py:3: note: Possible overload variants:
tmp.py:3: note: def __setitem__(self, str | list[str], Buffer | _SupportsArray[dtype[Any]] | _NestedSequence[_SupportsArray[dtype[Any]]] | bool | int | float | complex | str | bytes | _NestedSequence[bool | int | float | complex | str | bytes], /) -> None
tmp.py:3: note: def __setitem__(self, SupportsIndex | slice | EllipsisType | _SupportsArray[dtype[numpy.bool[builtins.bool]] | dtype[integer[Any]]] | _NestedSequence[_SupportsArray[dtype[numpy.bool[builtins.bool]] | dtype[integer[Any]]]] | builtins.bool | int | _NestedSequence[builtins.bool | int] | tuple[SupportsIndex | slice | EllipsisType | _SupportsArray[dtype[numpy.bool[builtins.bool]] | dtype[integer[Any]]] | _NestedSequence[_SupportsArray[dtype[numpy.bool[builtins.bool]] | dtype[integer[Any]]]] | builtins.bool | int | _NestedSequence[builtins.bool | int] | None, ...] | None, Buffer | _SupportsArray[dtype[Any]] | _NestedSequence[_SupportsArray[dtype[Any]]] | bool | int | float | complex | str | bytes | _NestedSequence[bool | int | float | complex | str | bytes], /) -> None
Found 1 error in 1 file (checked 1 source file)Python and NumPy Versions:
>>> import sys, numpy; print(numpy.__version__); print(sys.version)
2.2.0
3.12.3 (v3.12.3:f6650f9ad7, Apr 9 2024, 08:18:47) [Clang 13.0.0 (clang-1300.0.29.30)]
Runtime Environment:
>>> import numpy; numpy.show_runtime()
[{'numpy_version': '2.2.0',
'python': '3.12.3 (v3.12.3:f6650f9ad7, Apr 9 2024, 08:18:47) [Clang 13.0.0 '
'(clang-1300.0.29.30)]',
'uname': uname_result(system='Darwin', node='jmdg-macbookpro.roam.internal', release='23.6.0', version='Darwin Kernel Version 23.6.0: Thu Sep 12 23:35:29 PDT 2024; root:xnu-10063.141.1.701.1~1/RELEASE_ARM64_T6000', machine='arm64')},
{'simd_extensions': {'baseline': ['NEON', 'NEON_FP16', 'NEON_VFPV4', 'ASIMD'],
'found': ['ASIMDHP'],
'not_found': ['ASIMDFHM']}}]Context for the issue:
Related to #27957; this is a reduction of an error we're seeing in the JAX CI under NumPy 2.2
Reacted by Joren Hammudoglu
Activity
[-]BUG: type checking error in NumPy 2.2[/-][+]TYP: False positives in ``ndarray.__setitem__`` with ``object_`` dtype in NumPy 2.2[/+]jorenham commentedon Dec 10, 2024
caused by #27755
jorenham commentedon Dec 10, 2024
The problem is that
ndarray.__setitem__currently only acceptsnumpy.typing.ArrayLikevalues, causing it to reject valid values likeNonefornp.inexactdtypes, andobject()innp.object_dtypes.