Ensure Range streams and RangeToolLink respect subcoordinate axis range

[philippjfr]

This PR handles the previously poorly defined behavior of Range streams and RangeToolLink attached to a subcoordinate axis. Specifically both now pick up the range of the outer subcoordinate axis range rather than one of the inner subcoordinate ranges. The latter case is still something that will be needed in certain cases so that is still a problem to be solved.

Fixes #6010 and #6136

[philippjfr]

@droumis please test when you get a chance.

[codecov]

Codecov Report

Attention: Patch coverage is 89.36170% with 10 lines in your changes missing coverage. Please review.

Project coverage is 88.52%. Comparing base (9cc87fe) to head (683c604).
Report is 8 commits behind head on main.

Files	Patch %	Lines
holoviews/tests/ui/bokeh/test_callback.py	55.55%	8 Missing ⚠️
holoviews/plotting/bokeh/links.py	93.10%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6256      +/-   ##
==========================================
+ Coverage   88.47%   88.52%   +0.04%     
==========================================
  Files         323      323              
  Lines       67647    67894     +247     
==========================================
+ Hits        59853    60103     +250     
+ Misses       7794     7791       -3     

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

[ahuang11]

It seems to work. Awesome!

[ahuang11]

Actually the ranges look right, but when I zoom in and slice, I get a blank screen.

[droumis]

Looks good! Tested a few different versions, including multiple channel groups, with downsample1d, with wide df common index slice optimization, and with the large data pyramid approach.

Screen.Recording.2024-06-05.at.10.56.00.AM.mov

Screen.Recording.2024-06-05.at.11.01.52.AM.mov

[philippjfr]

@ahuang11 Remind me what the issue we encountered in the CZI call yesterday was?

[ahuang11]

After zooming in, then zooming out, the curves end up offset from the yticks.

[philippjfr]

Ah, yes, thanks!

[philippjfr]

@droumis @ahuang11 Should be fixed now.

[ahuang11]

Works for me!

[philippjfr]

@hoxbro This should be ready to review again.

[hoxbro]

I haven't tested the code. I hope either @droumis or @ahuang11 can do that.

Some small comments, mostly nit.

Also needs to update TestCurvePlot.test_batched_curve_subscribers_correctly_linked

[droumis]

This is amazing (combined with the hold render changes in #6265. Both for notebook and standalone.

video1254322944.mp4