GEOPY-2781: Inversion stalls on tiling for large problems during redistribution of clusters

[domfournier]

GEOPY-2781 - Inversion stalls on tiling for large problems during redistribution of clusters

[Copilot]

Pull request overview

This PR updates the location-tiling behavior used to partition survey locations into tiles, apparently to remove the previous “even population” rebalancing step (which relied on linear_sum_assignment) and to adjust tests accordingly.

Changes:

• Simplifies tile_locations() to always use raw KMeans cluster labels (removing the redistribution/balancing step).
• Disables the tile-population balancing test by commenting it out and adding TODO notes.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
tests/locations_test.py	Comments out the population-balancing test for tile_locations() and adds TODO notes about a future scalable balancing approach.
simpeg_drivers/utils/nested.py	Removes the Hungarian-assignment-based rebalancing logic; tile_locations() now returns KMeans labels directly.

Comments suppressed due to low confidence (1)

simpeg_drivers/utils/nested.py:545

• When sorting is provided, grid_locs is permuted before fitting KMeans, but the returned tile indices are positions in that permuted array. Downstream slicing (e.g., create_survey() filters survey.ordering[:, 2] against the provided indices) expects receiver IDs in the original indexing used by ordering (often the geoh5/receiver index, not the permuted position). Please map the clustered indices back through sorting before returning (or avoid permuting grid_locs and instead pass weights/ordering differently), so tiles reference the same index space as survey.ordering.

    cluster_id = kmeans.labels_

    tiles = []
    for tid in set(cluster_id):
        tiles += [np.where(cluster_id == tid)[0]]

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 89.83%. Comparing base (b1ad7f4) to head (61854bb).
⚠️ Report is 7 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop     #365      +/-   ##
===========================================
- Coverage    89.85%   89.83%   -0.02%     
===========================================
  Files          125      125              
  Lines         6437     6425      -12     
  Branches       794      793       -1     
===========================================
- Hits          5784     5772      -12     
  Misses         449      449              
  Partials       204      204              

Files with missing lines	Coverage Δ
simpeg_drivers/utils/nested.py	95.16% <100.00%> (-0.27%)	⬇️

🚀 New features to boost your workflow:

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