Quantity 2.0

[nstarman]

PDF link: https://github.com/nstarman/astropy-APEs/blob/units-quantity-2.0/APE25/report.pdf

[pllim]

I totally missed this one, so I am confused why there is a APE 23 and APE 25 here. And why report is in PDF instead of Markdown.

Is this related to the following?

• units and constants and NEP 42 astropy#13460

[nstarman]

so I am confused why there is a APE 23 and APE 25 here

I'll clean that up.

The report is in PDF instead of Markdown.

The report was for the Moore grant, not the APE. I'm including the PDF to eventually turn into an APE, written in Markdown.

[byrdie]

@nstarman, I think this looks really good so far, I'm quite excited to work with other arrays inside Quantity!

I was wondering if you have given any thought to supporting arrays with labeled axes (e.g. xarray) during this effort? I think supporting xarray (or the other way around, xarray supporting Quantity) is really important going forward, and I think it's worth considering adding first-class support this feature for Quantity 2.0.

[MridulS]

xarray already has support for units via https://github.com/xarray-contrib/pint-xarray, so making astropy.units compatible with xarray would solve this right?

[byrdie]

I guess it depends on how we want to interact with the rest of the astropy ecosystem, for example: how can we make astropy.coordinates work with and propagate labeled axes? Is it easier to convert the underlying data buffers of SkyCoord etc. to instances of xarray, or make Quantity wrap xarray so only have to make minimal changes to astropy.coordinates?

[nstarman]

Support for x-array has to use their framework, which is how pint does it.
There already are unofficial libraries, like https://github.com/KaoruNishikawa/xarray-quantity to do this. With Quantity 2.0 we should make a coordinated package official version.