convenient known vs. observed inspector

[jGaboardi]

This PR:

• sees where buffer=clip limit gets us when calling voronoi_skeleton() within:
  • artifacts.multiple_remaining()
  • artifacts.one_remaining_c()
  • xref buffer=clip_limit for smoother angles? #254
• provides a convenient known vs. observed inspection notebook where we can easily
  • display the full PNG with diffs in red
  • see the known vs. observed discrepancies tabular data
  • explore the known vs. observed discrepancies
  • known vs. observed discrepancies are curated here in conftest.py if highest-level checks pass and only for dev.
  • attached here is some example data – Doula only

ci_artifacts.zip

---

Even if we need to remove the buffer=clip limit portion from this PR¹, the second bullet point is worth keeping for sure.

Footnotes

1. After reviewing the CI artifacts via the checker notebook it seems these buffer=clip limit changes lead to sharper angles, so we should probably leave as is. ↩

[jGaboardi]

I am going to break this PR into 2 parts:

1. Keep "try buffer=clip limit" part here (and probably close later)
2. Open a fresh PR for the "convenient known vs. observed inspector" portion, which is super handy

Decided to simply remove the buffer=clip limit stuff from the PR since it's a dead-end

[jGaboardi]

@martinfleis – I'd like to get this curated since it makes finding trouble makers much easier if high-level checks are passing (which they weren't even doing that with the pandas>=3 issues from yesterday)

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 98.8%. Comparing base (b7421ac) to head (78919c4).
⚠️ Report is 3 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##            main    #255     +/-   ##
=======================================
- Coverage   99.1%   98.8%   -0.2%     
=======================================
  Files          7       7             
  Lines       1274    1274             
=======================================
- Hits        1262    1259      -3     
- Misses        12      15      +3     

see 1 file with indirect coverage changes

🚀 New features to boost your workflow:

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

[jGaboardi]

@martinfleis Unless you have strong feelings against this I am going to proceed with merge. If you do have strong feelings against, I'd like to hear them and we can discuss (because this inspection method in this PR has been very helpful for debugging/comparing differences).