test: add failing repro for skyline solver reporting non-convergence as success#871
Closed
ivan-aksamentov wants to merge 1 commit into
Closed
test: add failing repro for skyline solver reporting non-convergence as success#871ivan-aksamentov wants to merge 1 commit into
ivan-aksamentov wants to merge 1 commit into
Conversation
…convergence as success Counterexample for #869: solve_log_tc returns Ok(z) unconditionally after its Newton loop, so max_iter = 0 returns the un-optimized decoupled warm start, which for the default positive stiffness is not the regularized optimum. optimize_skyline still yields a successful SkylineResult, so the pipeline's fail-loud contract cannot fire. The test asserts max_iter = 0 is an error and fails on the current head. Red test for author consideration; not intended to merge as-is.
ivan-aksamentov
deleted the
repro/skyline-nonconvergence-reported-as-success
branch
July 24, 2026 15:25
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Red counterexample for #869, base is the PR head branch so CI shows it failing against the code under review. Not intended to merge as-is: the author decides the fix, then adapts or removes the test.
solve_log_tcreturnsOk(z)unconditionally after its Newton loop. Withmax_iter = 0it takes no step and returns the decoupled warm start, which for the default positive stiffness is not the regularized optimum.optimize_skylinestill yields a successfulSkylineResult, so a caller cannot distinguish an optimized skyline from an un-optimized one, and the pipeline's fail-loud contract cannot fire because no error is raised. The same silent path covers an exhaustedmax_iterand an Armijo stall atalpha < 1e-12.The test runs a three-segment skyline with
max_iter = 0and asserts an error. On the current head it returnsOkwith the un-optimized warm start.Work items
optimize_skylinewithmax_iter = 0on a multi-segment regularized solve returns an errorPossible improvements
convergedflag, when the final gradient exceeds tolerance), extend to an exhausted-iteration and a stalled-line-search case