Repository navigation
Merge main's milestone 6 work into master - #67
Merged
Merged
Conversation
benchmark/scaling compares dense FPCA with dense FACE on grids up to 5000 points and PACE with sparse FACE on 2-10 points per curve, 20 cells x 200 replicates, run in resumable route/replicate chunks under GNU Parallel. The Discovery results fill the FPCA page's route note. No defaults change. Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Add long-grid scaling benchmark and route note (milestone 6, task 4)
An estimated σ² that truncates to zero made the conditional-expectation covariance singular, so sparse fits failed or returned extreme scores. Estimated σ² is now at least 0.025 times the maximum fitted pointwise signal variance; a supplied noise_variance is untouched. The raw estimate, floor and whether it bound are stored in settings and shown. Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Floor the estimated noise variance for PACE and FACE
master gained the audit follow-ups (#52-#65) and KernelSmoothing's new names while main gained the scaling benchmark (#53) and the estimated-σ² floor (#61). Conflicts keep both: the floor logic with master's max_k and basis_matrix names, and master's FACE score helpers. Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
mlakolar
changed the title
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## master #67 +/- ##
=======================================
Coverage 98.52% 98.53%
=======================================
Files 33 33
Lines 3793 3810 +17
=======================================
+ Hits 3737 3754 +17
Misses 56 56 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
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
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.
Requested by Mladen
Before: the repo had two diverged branches.
main(the default) held milestone 6's scaling benchmark (#53) and the estimated-σ² floor (#61).masterheld the audit follow-ups (#52–#65) and the move to KernelSmoothing's new names. Since KernelSmoothing removed its old names tonight,mainfails its own test suite.After:
mainhas both sides and passes again.masterwill then be moved to the same commit.How: a merge of
origin/masterintoorigin/main. There were conflicts insrc/face.jl,src/face_sparse.jlandsrc/pace.jl, all resolved to keep both sides: the floor logic with master'smax_k/basis_matrixnames, plus master's batched FACE score helpers. One test from master asserted that dense and FACE scores agree on noiseless data, which no longer holds now that the floor shrinks FACE scores. It now compares reconstructions, asmainalready did.The full suite passes locally on Julia 1.13 (12,326 tests). A clean Julia 1.10 run is in progress.
🤖 Generated with Claude Code