Repository navigation
feat(folds): validate P2Fold on Bermagui and document gauss4 - #349
Merged
Merged
Conversation
This was referenced Oct 20, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Validates the library P2Fold path on Bermagui S2 and S0 after the Phase 0 weight, NaN-row, CG convergence and update fixes, and the Phase 1-3 class, factory wiring and golden coverage.
Both decision gates pass using fewer P2 unknowns, fixed seeds, explicit buffer 0 and identical upstream DFI frames / fitted fold geometry:
The comparison also covers the requested full-resolution runs and both sampling rules. Unregularised centroid is unstable (about 60 degrees median). The largest gauss4 S0 run has 765,347 unknowns, 8.69 degrees median error, 16.1% above 30 degrees, 92.5 seconds solve time and 6.6 GiB peak RSS. The metric measures axial agreement with the fold event, not independent geological accuracy.
Adds reproducible comparison/summary scripts, a bundled synthetic example, usage docs, and the corrected handoff/work plans. Detailed generated outputs are retained locally. Corrects the mesh-size API/log wording:
nelementsis currently a grid-cell budget rather than a literal tet count. The report explains why the historical all-P2 upstream-chain numbers are not a direct before/after comparison.Validation: all ten candidate CG solves converged with finite inner probes; 172 interpolation tests passed (9 skipped, 2 expected failures); 9 gate/integration tests passed; the bundled example ran successfully. Sphinx HTML built with gallery execution disabled, with 286 warnings elsewhere in the documentation; the new page and example link were inspected.
Release follow-up: existing release-please PRs are scoped to the intended packages, but root dependency ranges currently exclude the pending loop-common/loop-interpolation 0.1.0 releases. Align requirements and lockfile before publishing the root release. Post-merge release checks and Phase 5 remain pending.
Validation summary and limitations