Skip to content

Record accuracy sweep results for fire, terrain_metrics, morphology#1401

Merged
brendancol merged 1 commit intomainfrom
accuracy-sweep-state-2026-04-30-batch2
May 1, 2026
Merged

Record accuracy sweep results for fire, terrain_metrics, morphology#1401
brendancol merged 1 commit intomainfrom
accuracy-sweep-state-2026-04-30-batch2

Conversation

@brendancol
Copy link
Copy Markdown
Contributor

Summary

State-file update for the 2026-04-30 accuracy sweep batch (three modules). Adds rows to .claude/sweep-accuracy-state.csv.

Module Severity Categories Notes
fire clean Per-pixel ops only; NaN guards (x != x); CUDA bounds strict <; division guarded; CPU/GPU/dask paths algorithmically identical.
terrain_metrics LOW 2;5 Inf input not rejected (propagates as Inf, consistent across backends); dask+cupy non-nan path double-pads (output still correct).
morphology HIGH 2;5 Two HIGH bugs fixed in #1397 / PR #1398 and #1399 / PR #1400.

Test plan

  • CI passes (state-file change only)

Three modules audited 2026-04-30.

- fire: clean (per-pixel ops only; NaN guards via x!=x; CUDA bounds
  use strict <; division guarded; CPU/GPU/dask paths line-for-line
  equivalent; no CRIT/HIGH/MEDIUM).

- terrain_metrics: LOW only (Inf input not rejected, propagates;
  dask+cupy non-nan boundary path double-pads with consistent output).

- morphology: HIGH x 2 fixed.
  - #1397 / PR #1398: morph_erode/dilate seeded centre cell into the
    running min/max even when kernel[centre]==0, contaminating the
    result with the excluded pixel (all four backends).
  - #1399 / PR #1400: _morph_chunk_numpy and _morph_chunk_cupy used
    result[hy:-hy, hx:-hx] which becomes the empty slice 0:-0 for
    1xN or Nx1 kernels, raising ValueError on the dask backends
    while numpy and cupy paths handled them.
@github-actions github-actions Bot added the performance PR touches performance-sensitive code label May 1, 2026
@brendancol brendancol merged commit c3190e3 into main May 1, 2026
11 checks passed
@brendancol brendancol deleted the accuracy-sweep-state-2026-04-30-batch2 branch May 4, 2026 13:04
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

performance PR touches performance-sensitive code

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant