Description
Describe the issue:
Passing fill_value=None results in the MaskedArray dtype changing to object. Is this behavior intended? I didn't see mention of this in any changelog.
>>> np.__version__
'1.25.2'
>>> np.ma.MaskedArray([1,1,1], [True, False, True], dtype='uint64').filled()
array([999999, 1, 999999], dtype=uint64)
>>> np.__version__
'2.1.0'
>>> np.ma.MaskedArray([1,1,1], [True, False, True], dtype='uint64').filled()
array([np.int64(999999), 1, np.int64(999999)], dtype=object)
Reproduce the code example:
import numpy as np
assert np.ma.MaskedArray([1,1,1], [True, False, True], dtype='uint64').filled().dtype == np.dtype("uint64")Error message:
No response
Python and NumPy Versions:
2.1.0
3.12.4 (main, Jun 6 2024, 18:26:44) [Clang 15.0.0 (clang-1500.3.9.4)]
Runtime Environment:
[{'numpy_version': '2.1.0',
'python': '3.12.4 (main, Jun 6 2024, 18:26:44) [Clang 15.0.0 '
'(clang-1500.3.9.4)]',
'uname': uname_result(system='Darwin', node='<redacted>', release='23.2.0', version='<redacted>', machine='arm64')},
{'simd_extensions': {'baseline': ['NEON', 'NEON_FP16', 'NEON_VFPV4', 'ASIMD'],
'found': ['ASIMDHP'],
'not_found': ['ASIMDFHM']}}]
Context for the issue:
No response