Add support for elliptical ROI

[javerbukh]

Pull Request Template

Description

Potentially fixes #31

I'm not sure if anything else needs to be done here to implement elliptical ROI in glue-astronomy, and by extension jdaviz. Testing on my machine does not look different but I'm not sure if further changes need to be made in glue-jupyter or jdaviz.

[codecov[bot]]

Codecov Report

Merging #32 (5376944) into master (930fd84) will decrease coverage by 0.86%. The diff coverage is 50.00%.

@@            Coverage Diff             @@
##           master      #32      +/-   ##
==========================================
- Coverage   97.27%   96.41%   -0.87%     
==========================================
  Files          13       15       +2     
  Lines         991     1003      +12     
==========================================
+ Hits          964      967       +3     
- Misses         27       36       +9     

Impacted Files	Coverage Δ
glue_astronomy/translators/regions.py	95.32% <50.00%> (-0.87%)	:arrow_down:
...onomy/io/spectral_cube/tests/test_spectral_cube.py	85.71% <0.00%> (-14.29%)	:arrow_down:
glue_astronomy/conftest.py	100.00% <0.00%> (ø)
glue_astronomy/__init__.py	71.42% <0.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 930fd84...5376944. Read the comment docs.

[pllim]

I also wonder how this will affect spacetelescope/jdaviz#697 🤔

[eteq]

(also that "build not found" error is weird, but will troubleshoot with @rosteen out-of-band) EDIT: seems to have been a one-off, but the real problem I'm trying to troubleshoot in #34

[pllim]

It's passing on windows though, so I'm pretty confident there's not a concern there.

See, Windows is useful sometimes. 😆