Don't recalcuate histogram if not necessary

[thorstenwagner]

Hi,

I'm making great use of this plugin, but if data gets huge (like 32 million datapoints) it becomes much slower.

For example, after the umap is loaded and I circle a cluster it takes like 17 seconds to highlight the clusters.

This is the first PR to improve this. With the proposed changes, the 2dhistogram is not recalculated if the data hasn't changed.

It already reduces the computational time from 17 seconds to like 12 seconds.

Seconds PR follows :-)

Best,
Thorsten

[codecov]

Codecov Report

Attention: 5 lines in your changes are missing coverage. Please review.

Comparison is base (2b47b7f) 76.52% compared to head (36f9a8c) 76.38%.

Files	Patch %	Lines
napari_clusters_plotter/_Qt_code.py	28.57%	5 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #282      +/-   ##
==========================================
- Coverage   76.52%   76.38%   -0.14%     
==========================================
  Files          16       16              
  Lines        1853     1859       +6     
==========================================
+ Hits         1418     1420       +2     
- Misses        435      439       +4     

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

[Cryaaa]

This looks good to me! Will
Merge once tests pass. By the way, we are restructuring the plotter code to tidy it up and improve it currently. If you have some suggestions for larger changes or more efficient approaches we should make please let us know and we can try to implement them in the next big release! Thanks for taking care of this improvement 😊

[thorstenwagner]

Fantastic! Thanks for reacting so quick!