Allow mutations to be plotted per edge not per tree

[hyanwong]

Description

New parameter all_edge_mutations (default False, and only for Tree.draw_svg() not TreeSequence.draw_svg()) that plots "extra" mutations on a single tree. The default is to plot these in HotPink, and to extend the X axis (if show, which it isn't by default), to show the position of the extra mutations outside the tree interval:

Probably needs a few more tests, but opening this PR for feedback. I generalised the "background shading" function so that it could be used on a tree as well as a tree sequence.

Fixes #1253

PR Checklist:

• Tests that fully cover new/changed functionality.
• Documentation including tutorial content if appropriate.
• Changelogs, if there are API changes.

[codecov]

Codecov Report

Merging #1258 (5b384a5) into main (c84a89b) will increase coverage by 0.01%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #1258      +/-   ##
==========================================
+ Coverage   93.77%   93.78%   +0.01%     
==========================================
  Files          26       26              
  Lines       21939    21991      +52     
  Branches      971      987      +16     
==========================================
+ Hits        20573    20625      +52     
  Misses       1325     1325              
  Partials       41       41              

Flag	Coverage Δ
c-tests	92.49% <ø> (ø)
lwt-tests	92.97% <ø> (ø)
python-c-tests	95.03% <100.00%> (+0.02%)	⬆️
python-tests	98.67% <100.00%> (+0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
python/tskit/trees.py	97.44% <ø> (ø)
python/tskit/drawing.py	99.12% <100.00%> (+0.05%)	⬆️

---

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 c84a89b...5b384a5. Read the comment docs.

[benjeffery]

Will review later - just a quick comment to ask if we are at the point where a legend might be helpful?

[hyanwong]

Will review later - just a quick comment to ask if we are at the point where a legend might be helpful?

Thanks @benjeffery. No hurry on this.

Personally, I don't think we are quite at that point yet. The various fancy graphics I've added aren't always very suited to explaining in a legend (I'm thinking of background shading, and, to a lesser extent, ticks & sites on the X axis); also this is only relevant for a single tree, so the legend stuff in trees vs tree sequences could get messy. But other views welcome.

[grahamgower]

If you're open to some bike-shedding @hyanwong, I'd like to suggest that it would be better to have slightly more contrast between red/pink than HotPink provides.

[hyanwong]

After discussion with @jeromekelleher, I'm temporarily switching this to a draft PR again so that we can generalise the "edges above tree nodes" function in a separate PR first.

[hyanwong]

If you're open to some bike-shedding @hyanwong, I'd like to suggest that it would be better to have slightly more contrast between red/pink than HotPink provides.

How about PaleVioletRed @grahamgower?

I think LightPink may be too faded to be easily visible:

[grahamgower]

How about PaleVioletRed @grahamgower?

That's better, but the difference is still fairly small. I understand that red/green colourblindness is often driven by poor ability to distinguish reds from gray, so if you have any red-challenged friends, it might be worth asking them.

[awohns]

I just checked with two colour-blind folks, they can see the difference in the PaleVioletRed example clearly

[hyanwong]

I just checked with two colour-blind folks, they can see the difference in the PaleVioletRed example clearly

That's really helpful. Thanks @grahamgower @awohns

[hyanwong]

OK, this now uses the new tree._node_edges() method, so I'm re-open ing for review.

[jeromekelleher]

Looks good, minor comments.

[jeromekelleher]

Looks good, thanks.

[jeromekelleher]

Ready to merge I think - @benjeffery?

[benjeffery]

LGTM, one minor thing in the changelog.