Kaustubh

Reminder: Make sure the masks functionality is being used correctly, if it's not, remove it within this PR: https://github.com/numba/numba/pull/8041#discussion_r920881476

I also wanted to add a similar support (of multiple user defined axes) for np.squeeze but I couldn't find the code where Numba has binded that function to it's respective...

Hi @stuartarchibald , So I tried using `@overload` for `np.squeeze` and I think I'd be a great idea to also add it as an bound method attribute to arrays too,...

> Can't overload_method be used instead? There are a few issues with `overload_method` too. For instance, We cannot have return types of different dimensions in a single overload method. Also...

We can go ahead for a review with the current functionality of this PR. (Implementing `np.squeeze` will probably require more discussion and planning than the scope of this particular PR.)

@bszollosinagy is right. This issue was initially discussed at the start of the `Generator` project in Numba, the older NumPy random API in Numba is not a exact replacement of...

> Is this enough for those interested to start work on implementing numpy.random.Generator more fully? @stuartarchibald, This particular implementation is definitely what's needed to get started on the newer `numpy.random`...

> In that if a RandomState object were to be created it then all these functions for creating distributions can be altered to take something Generator-like and wired up as...

Ah, I think Numba can't lower it directly yet. ```python import numba import numpy as np rng = np.random.default_rng(23) @numba.njit def sample(rng): return rng.normal() sample(rng) # 0.5532605888887387 ``` Also in...

I think the general progress is being tracked at: https://github.com/numba/numba/projects/19