Fix unit handling in continuum and moment calculations

[rosteen]

Fixes calculating moment 0 when display unit is in Hz, and more generally uses display unit more consistently in all continuum calculations.

[kecnry]

I feel like @cshanahan1 had just removed this. Was there a reason for that or did it just not seem to affect that use-case (see https://github.com/spacetelescope/jdaviz/pull/3211/files#r1787615549)

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 88.56%. Comparing base (99baa91) to head (7911b26).
Report is 2 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #3216   +/-   ##
=======================================
  Coverage   88.56%   88.56%           
=======================================
  Files         125      125           
  Lines       18751    18755    +4     
=======================================
+ Hits        16606    16610    +4     
  Misses       2145     2145           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[gibsongreen]

In testing I am still seeing the traceback occur for moment zero (and it is occurring if you chose any spectral unit we allow).

[rosteen]

In testing I am still seeing the traceback occur for moment zero (and it is occurring if you chose any spectral unit we allow).

Can you give some more detail about your workflow? I see everything working fine on my end:

Screen.Recording.2024-10-15.at.9.53.42.AM.mov

[gibsongreen]

I went through the workflow you showed in the video, and in the video below, and I double checked in a second conda environment and in both I was seeing the traceback:

Screen.Recording.2024-10-15.at.10.09.17.AM.mov

[rosteen]

Oh, that's not the same workflow as I used, you don't have a spectral region or continuum selected. I can't reproduce the error you're seeing though even without a region or continuum, although I do see an IndexError if I select a spectral region but have None for continuum in Hz. So I'll at least start looking into that.

[rosteen]

Ok, I just tried in a fresh environment and still only saw the error with the subset + no continuum combo, so I'll work on fixing that for now.

[rosteen]

Turns out the error I was seeing is due to a bug in specutils, see astropy/specutils#1187. We'll need to update the specutils pin after I release 1.18.0, I can do that here before we merge this.

[gibsongreen]

The tracebacks I was seeing came from my end, all resolved and working now! With the workflow you provided I tried many combinations with the example notebook and with a flux cube. The main notes I had I think are all outside of scope and are being addressed in follow ups and deal with the spectral density or pix2 handling.

[cshanahan1]

I tested this with a cube in Jy / pix2 as well as Jy / sr and both work, including the continuum preview. My only take it or leave it suggestion would be to include changing spectral units and slice value in tests to make sure that those changes are reflected in the calculation.

[rosteen]

Thanks for the reviews, I just bumped the specutils pin to the new release. I'll merge once tests pass.