Accept the BMDB baseline after the SBML scope fixes - #1986
Merged
Conversation
Nightly run 32094317081 reported FAIL: 28 changed, the first run after #1983 (reaction-rate references) and #1985 (UNRESOLVED.initConc / UNRESOLVED.Size) landed. Every shard passed; the gate was refusing a baseline those fixes had made stale. Eighteen archives improved outright -- they now import AND simulate: 256 457 499 589 599 628 632 705 710 739 764 804 832 867 872 908 969 FAIL -> PASS 1065 SOLVER_FAILURE -> PASS Seven more get further than before, failing after import rather than during it: 534 535 536 537 547 1021 SBML_IMPORT_FAILURE -> SOLVER_FAILURE 613 SBML_IMPORT_FAILURE -> DIVIDE_BY_ZERO The remaining three look like a step backwards and are not. Each model has TWO unsupported features, and the reaction-rate check from #1983 now reports first: 63 UNSUPPORTED_NON_INT_STOCH -> SBML_IMPORT_FAILURE 245 UNSUPPORTED_NON_INT_STOCH -> SBML_IMPORT_FAILURE 627 UNSUPPORTED_NON_CONSTANT_COMPARTMENTS -> SBML_IMPORT_FAILURE Verified rather than assumed: 'Vhk' in 63 and 'r2' in 245 are both genuine <reaction> elements referenced from an assignmentRule, and neither is declared as a species, parameter or compartment, so these are not false positives. Both models also genuinely carry the non-integer stoichiometries the old entries named (0.5; 0.063, 0.127, ...). 627 is the same shape, on reaction 'vSTIM' alongside its non-constant compartment. Both old and new reasons are true; only which is reached first has changed, and the models fail either way. No archive went from PASS to FAIL. Taken verbatim from the run's test_cases.updated.ndjson artifact: 1182 rows before and after, 28 lines changed, no rows added, removed or reordered. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
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.
Accepts the BMDB baseline after the SBML scope fixes, so the nightly goes green again.
Nightly run 32094317081 reported
FAIL: 28 changed— the first run after #1983 (reaction-rate references) and #1985 (UNRESOLVED.initConc/UNRESOLVED.Size) landed. Every shard passed. The gate was refusing a baseline those fixes had made stale, which is what it is for.Eighteen archives improved outright
They now import and simulate:
Seven get further than before
Import now works; they fail later, at solve:
SBML_IMPORT_FAILURE→SOLVER_FAILURESBML_IMPORT_FAILURE→DIVIDE_BY_ZEROThree look like a step backwards and are not
UNSUPPORTED_NON_INT_STOCH→SBML_IMPORT_FAILUREUNSUPPORTED_NON_INT_STOCH→SBML_IMPORT_FAILUREUNSUPPORTED_NON_CONSTANT_COMPARTMENTS→SBML_IMPORT_FAILUREEach has two unsupported features, and the reaction-rate check from #1983 now reports first. Verified against the models rather than assumed:
Vhkin 63 andr2in 245 are both genuine<reaction>elements referenced from anassignmentRule, and neither is declared as a species, parameter or compartment — so these are not false positives.0.5;0.063,0.127, …).vSTIMalongside its non-constant compartment.Both the old and new reasons are true; only which one is reached first has changed. The models fail either way.
No archive went from PASS to FAIL.
The diff
Taken verbatim from the run's
test_cases.updated.ndjsonartifact, not hand-edited:No rows added, removed or reordered; only the 28
known_status/known_failure_typevalues changed.🤖 Generated with Claude Code