Ilhan Polat

Our assumption is that `np` is a well-established shortcut for numpy and evident that it is required. But it is an assumption. This allows us skip a few hundreds of...

If I put my PM hat on and assess the priorities, this is probably at the bottom of our priority list. Even a full PEP8 compliance cleanup is above this...

@tylerjereddy @rgommers @ev-br Could you enable the discussions tab? Look for the "Discussions" checkbox in the repository settings page. I think we can use that instead of the ML which...

At least would be better than ML imho. We needed to have some issues converted to discussions anyways so if not for this then for other things we'll need it...

> I'm not sure about abs(). It depends on where you're coming from mathematically. If you're looking at the RMS as a [generalized mean](https://en.wikipedia.org/wiki/Generalized_mean) with p=2, then abs is wrong....

> I don't understand this. It's exactly what's in the Wiki page for p=2, isn't it? It is missing the brackets around the elements (note the abs in `√(Σ |x|²/n)`),...

Caching and downloading parts can be done without the dependencies. Requests is quite a hefty dependence for a light weight lib in my opinion. We can simply check a file...

> urllib may get deprecated and itself recommends requests That's really good news for me. Miraculously I missed the whole PSF takeover. Makes many things so much simpler.

We really shouldn't do sparse array operations with float16 it makes things just complicated and adds lots of _if float16 then raise notimplementederror_ guards for no good reason.

Scaling up an approximation uniformly, is not possible regardless of segment number but of course it gets better. It is also known that the approximation error is worst at the...