Disable resetting of y_min/y_max in layer artist

[astrofrog]

I need to figure out if this code is still needed, but for now it seems it definitely isn't for specviz, so this branch can be used with specviz to avoid rescaling all the time.

[codecov]

Codecov Report

Merging #247 (42c0f7f) into main (7e096c7) will decrease coverage by 0.05%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #247      +/-   ##
==========================================
- Coverage   88.63%   88.58%   -0.06%     
==========================================
  Files          82       82              
  Lines        4093     4136      +43     
==========================================
+ Hits         3628     3664      +36     
- Misses        465      472       +7     

Impacted Files	Coverage Δ
glue_jupyter/bqplot/profile/viewer.py	82.60% <ø> (-0.73%)	⬇️
glue_jupyter/bqplot/histogram/layer_artist.py	91.66% <100.00%> (ø)
glue_jupyter/bqplot/image/layer_artist.py	88.46% <100.00%> (ø)
glue_jupyter/bqplot/profile/layer_artist.py	92.30% <100.00%> (-0.90%)	⬇️
glue_jupyter/utils.py	62.38% <0.00%> (-3.04%)	⬇️
glue_jupyter/conftest.py	100.00% <0.00%> (ø)
glue_jupyter/common/slice_helpers.py	0.00% <0.00%> (ø)
glue_jupyter/tests/test_state_traitlets_helpers.py	100.00% <0.00%> (ø)
glue_jupyter/bqplot/scatter/layer_artist.py	99.42% <0.00%> (+<0.01%)	⬆️
glue_jupyter/tests/main_test.py	98.44% <0.00%> (+0.01%)	⬆️
... and 14 more

---

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 7e096c7...42c0f7f. Read the comment docs.

[pllim]

If it is somehow needed elsewhere, Jdaviz still need an option to turn it off.

[pllim]

I can confirm that this also fixes the issue downstream at spacetelescope/jdaviz#760 , in addition to spacetelescope/jdaviz#335

[pllim]

@camipacifici reported that this might have broken the "Home" button, as pressing Home now does not reset Y-axis zoom in Specviz, though it works fine for Imviz.

[astrofrog]

I've updated this to a point where I think I'd be happy to merge. I've also fixed the Home button.

At this point it would be helpful if someone could test this out with jdaviz/specviz though?

cc @camipacifici @pllim

[rosteen]

I just confirmed that selecting a subset in the profile viewer no longer resets the y-axis zoom, and that the home button now properly resets both axes again in the profile viewer. Looks good!

[pllim]

I'll have to double check that the markers no longer reset Imviz zoom/pan, unless someone gets to it first.

[javerbukh]

When using mosviz, selecting a new row does not change the Y-axis of the spectrum viewer at all. The scale of the first row persists for all the rest, even if the range is much different.

[pllim]

My comment at #247 (comment) is no longer true. This does not prevent pan/zoom reset in Imviz no more when markers are added. Not sure what changed.

But the Imviz issue is not important to fix right now.

[astrofrog]

This PR never had any impact on the image viewer so I wouldn't expect it to fix your issue @pllim - but if you open a separate issue I can look into it

[pllim]

This PR never had any impact on the image viewer

It did at one point... 😆

But don't worry about Imviz for now.

[pllim]

But since you asked, I did open an issue at spacetelescope/jdaviz#760

[pllim]

This actually breaks jdaviz... See spacetelescope/jdaviz#874 . Maybe we need this is a very localized section of Jdaviz only...

[astrofrog]

I am working on this in glue-viz/glue#2233