ENH: tractometer-style volume compare

[AntoineTheb]

Quick description

Change the behavior of scil_volume_pairwise_comparison in the case of --single_compare.

If the option is not included, overreach is computed as the number/ratio of voxels in either volumes but not in both, a sort of logical xor. However, if comparing volumes against a reference one (ie using --single_compare), it would make more sense to compute the overreach as the number/ratio of voxels in the candidate volume that are not in the reference. This second behavior would be closer to a "tractometer-style" overreach.

This is a breaking change in the case of --single_compare.

Type of change

Check the relevant options.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

Provide data, screenshots, command line to test (if relevant)

TODO

Checklist

• My code follows the style guidelines of this project (run autopep8)
• I added relevant citations to scripts, modules and functions docstrings and descriptions
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I moved all functions from the script file (except the argparser and main) to scilpy modules
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

[codecov]

Codecov Report

❌ Patch coverage is 61.53846% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 72.44%. Comparing base (c70bb19) to head (844a30a).
⚠️ Report is 119 commits behind head on master.

❌ Your patch status has failed because the patch coverage (61.53%) is below the target coverage (90.00%). You can increase the patch coverage or adjust the target coverage.
❌ Your project status has failed because the head coverage (72.44%) is below the target coverage (75.00%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1242      +/-   ##
==========================================
- Coverage   72.45%   72.44%   -0.02%     
==========================================
  Files         293      293              
  Lines       25185    25194       +9     
  Branches     3527     3530       +3     
==========================================
+ Hits        18249    18253       +4     
- Misses       5445     5448       +3     
- Partials     1491     1493       +2     

Components	Coverage Δ
Scripts	75.29% <100.00%> (+<0.01%)	⬆️
Library	∅ <ø> (∅)

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[EmmaRenauld]

I think you should divide by the number of voxels in the reference volume, to follow explanations on tractometer.org:

Currently:
With my volume = 570 560 voxels and reference = 130 859

1. With --single_compare, no --ratio: 💯 This looks good. You indeed fixed the overreach raw ratio.

   • On master: 97 443 overlap, 506 533 overreach.
   • On your branch: 97 443 overlap, 473 117 overreach.
   • With scil_volume_math: 97 443 intersection, 473 117 difference (input - reference)

2. With --single_compare, with ratio: You seem to be dividing by the input volume's size.

   • Expecting overlap = 97,443/130,859 = 74.46% and overreach = 473 117/130 859 = 361.55%
   • On your branch, I get: 17% and 83%

[AntoineTheb]

@EmmaRenauld you're right, should be fixed now

[arnaudbore]

LGTM