Replace matlab_find with numpy functions in mapgen

[faridyagubbayli]

Greptile Summary

This PR replaces MATLAB-compatibility shim calls (matlab_find, ind2sub, sub2ind, matlab_assign) with native NumPy equivalents in make_line and the DIMENSION 1 block of make_bowl in mapgen.py. The index-arithmetic equivalence is correct: matlab_find(diff == min(diff))[0] → np.argmin(diff) preserves first-occurrence tie-breaking, and the 1-based ↔ 0-based offset adjustments are properly removed.

• In make_line, six identical index-selection sites are simplified: np.argmin(diff) replaces the matlab_find round-trip, and the downstream poss_x[index[0] - 1] accesses become plain poss_x[index].
• In make_bowl, the DIMENSION 1 forward/backward surface-extraction pipeline is rewritten using np.where + direct 3-D array assignment, correctly mapping the flipped backward index via (Nx - 1) - index_back_flipped. DIMENSION 2 and DIMENSION 3 are left unchanged, still relying on matlab_find.

Confidence Score: 4/5

The changed logic is a correct mechanical translation from MATLAB-style helpers to native NumPy — index equivalence holds in both make_line and the make_bowl DIMENSION 1 block. The only concern is that DIMENSION 2 and DIMENSION 3 of make_bowl were not migrated, leaving the function internally inconsistent.

The six make_line sites and the make_bowl DIMENSION 1 block all translate cleanly, and the 0-based/1-based offset adjustments are handled correctly throughout. The remaining make_bowl dimensions still use the old matlab_find path, which is not a correctness issue but does leave an incomplete job that will need a follow-up.

kwave/utils/mapgen.py — specifically the DIMENSION 2 (lines ~1745–1763) and DIMENSION 3 (lines ~1783–1801) blocks in make_bowl, which still use the old matlab_find pattern.

Important Files Changed

Filename	Overview
kwave/utils/mapgen.py	Replaces matlab_find-based index arithmetic with np.argmin in make_line and with np.where/direct array indexing in the DIMENSION 1 block of make_bowl; DIMENSION 2 and DIMENSION 3 in make_bowl still use the old matlab_find pattern

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A[pixel_map shape Nx x Ny x Nz] --> B[min along axis 0\nvalue_forw, index_forw]
    A --> C[flip along axis 0\nthen min along axis 0\nvalue_back, index_back]
    B --> D["np.where(value_forw < THRESHOLD)\ny_coords_forw, z_coords_forw"]
    C --> E["np.where(value_back < THRESHOLD)\ny_coords_back, z_coords_back"]
    D --> F["x_ind_forw = index_forw[y, z]"]
    E --> G["x_ind_back_flipped = index_back[y, z]\nx_ind_back = (Nx-1) - x_ind_back_flipped"]
    F --> H["bowl_sm[x_ind_forw, y_coords_forw, z_coords_forw] = 1"]
    G --> I["bowl_sm[x_ind_back, y_coords_back, z_coords_back] = 1"]

Loading

Comments Outside Diff (1)

1. kwave/utils/mapgen.py, line 1745-1763 (link)

   Incomplete migration — DIMENSION 2 still uses matlab_find

   The DIMENSION 1 block was fully migrated to native numpy operations, but DIMENSION 2 (lines 1745–1763) and DIMENSION 3 (lines 1783–1801) still use matlab_find, ind2sub, sub2ind, and matlab_assign. This leaves the refactoring half-done and makes the three dimension blocks inconsistent with each other.

_{Reviews (1): Last reviewed commit: "Merge branch 'master' into chore/do-not-..." | Re-trigger Greptile}

[codecov]

Codecov Report

❌ Patch coverage is 84.37500% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 75.04%. Comparing base (cf4ec8f) to head (40b2c4b).
⚠️ Report is 3 commits behind head on master.

Files with missing lines	Patch %	Lines
kwave/utils/mapgen.py	84.37%	3 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #609      +/-   ##
==========================================
- Coverage   75.06%   75.04%   -0.02%     
==========================================
  Files          57       57              
  Lines        8126     8128       +2     
  Branches     1582     1584       +2     
==========================================
  Hits         6100     6100              
  Misses       1405     1405              
- Partials      621      623       +2     

Flag	Coverage Δ
3.10	75.02% <84.37%> (-0.02%)	⬇️
3.11	75.02% <84.37%> (-0.02%)	⬇️
3.12	75.02% <84.37%> (-0.02%)	⬇️
3.13	75.02% <84.37%> (-0.02%)	⬇️
macos-latest	74.97% <84.37%> (-0.02%)	⬇️
ubuntu-latest	74.97% <84.37%> (-0.02%)	⬇️
windows-latest	74.91% <84.37%> (-0.02%)	⬇️

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

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

🚀 New features to boost your workflow:

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