remove use of deprecated np.int

[javierggt]

Description

This PR makes tiny changes to remove uses of a few deprecated dtypes (sot/skare3#753)

Testing

• Passes unit tests on MacOS, linux, Windows (at least one required)
• Functional testing

Fixes #

[taldcroft]

I thought the recommended change is np.int => int. From the release notes, "For a long time, np.int has been an alias of the builtin int".

[taldcroft]

And just to see for myself:

In [1]: np.int is int                                                           
Out[1]: True

In [2]: np.float is float                                                       
Out[2]: True

In [3]: isinstance(np.float64(10), float)                                       
Out[3]: True

[javierggt]

I would agree that perhaps np.int_ is not the right type. I changed it to int.

In other places (Annie, I believe) I wondered whether it should be np.int32 to be explicit. np.int32 is used in Annie in several places.

Also, the case is clear for bool, but if one uses int, the type might be platform dependent.

[taldcroft]

Should be bool.