Jon Carifio

After looking into this a bit, it looks to be a Python 3.10 issue. In particular, the culprit is [this change](https://bugs.python.org/issue37999), which breaks the implicit `float` -> `int` conversion that...

@astrofrog I've rebased and added an entry in the changelog

@astrofrog I've just rebased

Just rebasing to resolve a conflict

This PR definitely fell off of my radar. @astrofrog I had used the `get_cdelt()` method since the [astropy docs](https://docs.astropy.org/en/stable/api/astropy.wcs.Wcsprm.html?highlight=get_cdelt#astropy.wcs.Wcsprm.get_cdelt) say that it should work for CD or CROTA as well....

@dhomeier Thanks for the review! My main motivation for ensuring exact identity was making sure that unit conversions would work correctly. I wasn't really sure what sorts of unit classes...

@dhomeier I think the sanity check is a good idea. I'm trying to imagine what it would look like for a plugin/config to overwrite this kind of behavior in a...

@dhomeier I've updated this to make the changes that you suggested.

I would guess that changing the offending line to `polar = getattr(layer._viewer_state, 'using_polar', False)` should be enough to fix this, but I admittedly haven't tried it.

Ah yeah, looks like there are similar calls for other scatter viewer state properties (full sphere, degrees)