Python 3.8.7 (tags/v3.8.7:6503f05, Dec 21 2020, 17:59:51) [MSC v.1928 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy as np
>>> x = np.bool_(1)
>>> x
True
>>> type(x)
<class 'numpy.bool_'>
>>> isinstance(x, bool)
False
>>>

Could be related to the fact that np.bool_ doesn't seem to be a subtype of bool.

Quoting PEP: https://peps.python.org/pep-0636/#matching-specific-values

Literal values are compared with the == operator except for the constants True, False and None which are compared with the is operator.

So, it works as it should :)

Thanks for the reference, I couldn't find it!
Do you have any info on why these constant should behave differently? It seems a bit strange without furhter insight.

🙏

Because they are three special singletons :)

For example, 1 == 1.0 == True
Probably, matching True case with 1 value is not something users want.

Ok thanks for your answer, I guess it makes sense.

I was wondering then what would the most natural way of matching a numpy boolean array be. I think the answer is to view the array with dtype=object, since

>>> np.True_.astype(object) is True
True

Cheers!