The following line (and similar other examples in the whole project where traceback is shown for test failures) needs change (x -> 'ACTUAL', y -> 'DESIRED')

Seems like numpy/testing/_private/utils.py", line 1004, in assert_array_almost_equal also requires the above change.

Searching for BLANKLINE in the entire project in VS Code, I found 7 examples which might require x -> ACTUAL and y -> DESIRED change. I haven't updated my master yet though.

I see them under, not sure if these should be updated

Seems like you have already made the replacement where the CI is showing mismatch error. It's happening only on macOS I think. I don't think there is anything else to be done.

In the following line. I think names aren't updated to ACUTAL and DESIRED. May be, that's why doc-tests are failing.

You can apply the following diff (use git apply file_containing_the_following.diff)

diff --git a/numpy/testing/_private/utils.py b/numpy/testing/_private/utils.py
index af9e31e77..162debfe5 100644
--- a/numpy/testing/_private/utils.py
+++ b/numpy/testing/_private/utils.py
@@ -741,7 +741,7 @@ def func_assert_same_pos(x, y, func=isnan, hasval='nan'):
             msg = build_err_msg([x, y],
                                 err_msg + '\nx and y %s location mismatch:'
                                 % (hasval), verbose=verbose, header=header,
-                                names=('x', 'y'), precision=precision)
+                                names=('ACTUAL', 'DESIRED'), precision=precision)
             raise AssertionError(msg)
         # If there is a scalar, then here we know the array has the same
         # flag as it everywhere, so we should return the scalar flag.
@@ -759,7 +759,7 @@ def func_assert_same_pos(x, y, func=isnan, hasval='nan'):
                                 err_msg
                                 + f'\n(shapes {x.shape}, {y.shape} mismatch)',
                                 verbose=verbose, header=header,
-                                names=('x', 'y'), precision=precision)
+                                names=('ACTUAL', 'DESIRED'), precision=precision)
             raise AssertionError(msg)
 
         flagged = bool_(False)

Other than that, I get the following error on your branch. Not sure if it's related. Just commenting anyways.

numpy.core.ndarray
------------------

File "build/testenv/lib/python3.9/site-packages/numpy/__init__.py", line ?, in ndarray.__class_getitem__
Failed example:
    np.ndarray[Any, np.dtype[Any]]
Expected:
    numpy.ndarray[typing.Any, numpy.dtype[Any]]
Got:
    numpy.ndarray[typing.Any, numpy.dtype[typing.Any]]


ERROR:  failed checking numpy.core

@czgdp1807 on which python version are you receiving this error?
This far I've been unable to reproduce this error on Python 3.9.1 and 3.9.7 with the latest upstream/main branch.

@czgdp1807 on which python version are you receiving this error?
This far I've been unable to reproduce this error on Python 3.9.1 and 3.9.7 with the latest upstream/main branch.

I get it with 3.9.6.

Oh, wait a minute, I see now what's going on here.
I'll open a PR in a bit.

In any case, don't the __class_getitem__ refguide failure get into the way of this PR.

@czgdp1807 @BvB93 are there any more changes required for this PR to be complete?

This looks good to me. @BvB93 @almostsquare Any thoughts on this?

Thanks for the changes, @sayantikabanik! Looks good!

LGTM.

Since this is user-facing, do we want to add a one-line release note about the changed behaviour?

Since this is user-facing, do we want to add a one-line release note about the changed behaviour?

I second that.

@sayantikabanik You can make the release note in doc/release/upcoming_changes. Look at the other notes to see how it is done.

This needs a clean up of git history. We will meet later today and help you clean up the stuff here.

This needs a clean up of git history. We will meet later today and help you clean up the stuff here.

Thanks :D

Could the CI be restarted from the backend? It's stuck I think.

It's Travis again. @rgommers It's get stuck very often. Can be ignored I think.

It looks like this was close and valuable for consistency.
@sayantikabanik would you like to merge main here? That should get CI going again.
I could also open a new PR to do this - preserving your commits - if you'd prefer. (Update: Done.)

@ngoldbaum looks like this can be closed now

Finished in #24931