Eric Wieser

Actually, since the code to produce it is already there for you - can you replot that graph with weights of `[1, 2, 3, 4]` and maybe `[2, 1, 1,...

I don't agree with that - `linspace` on floats behaves as: ``` In [6]: np.linspace(0, np.nan, 5) Out[6]: array([nan, nan, nan, nan, nan]) ``` so passing in `NaT` should just...

It sounds like you have a bug to fix for when `stop` is nat then :)

Indeed I noted that this was a poorly performing solution when I wrote it in #10723 https://github.com/numpy/numpy/blob/72b98a0d42c9e52cd2650cf357bc0ccb20264b27/numpy/core/src/common/npy_longdouble.c#L114-L122 I figured it was best to have correct behavior first, and wait for...

@mhvk: This might be a good place to expose the new gufunc axis arguments

> Having a gufunc is not quite the same as having an arbitrary axis, It is now that @mhvk allowed gufuncs to specify axis arguments > One difficulty would be...

> I thought our policy was to not emit a warning where we can set a value instead. I don't remember seeing that conversation. If that is indeed our policy,...

Would be interesting to see how extensive the changes need to be to make CI pass, and whether this impacts the scipy or pandas test suites.

Looks like randomgen added more tests that rely on this behavior, should be straightforward to just remove them https://dev.azure.com/numpy/numpy/_build/results?buildId=6596&view=ms.vss-test-web.build-test-results-tab&runId=1076954&resultId=109094&paneView=debug

To me there's no question that if we do this, it will need to be after a deprecation period. I'm definitely in favor of tracking down bugs in downstream projects...