Describe the issue:

np.argmin preserves the type of out (if passed), but it could be more precise than that and be narrowed down to integer or boolean arrays

Same for np.ma.argmin. Opening this as a follow-up to #28638 (comment)

Reproduce the code example:

from typing import Any
import numpy as np
import numpy.typing as npt

arr = np.array([1,2,3])
out = np.empty(())
reveal_type(np.argmin(arr, out=out))

Error message:

/home/marcogorelli/numpy-dev/t.py:7:13 - information: Type of "np.argmin(arr, out=out)" is "ndarray[tuple[()], dtype[float64]]"

Python and NumPy Versions:

2.3.0.dev0+git20250403.ec1cfa6
3.12.9 | packaged by conda-forge | (main, Mar 4 2025, 22:48:41) [GCC 13.3.0]

Type-checker version and settings:

pyright 1.1.398
spin run pyright t.py

Additional typing packages.

No response