Repository navigation
OMBackend 0.11.3: variable structure systems on the arrays-kept route; CI depot cache - #108
Merged
Merged
Conversation
- scalarizeModel: every partition and each structural mode scalarized by the backend - IMTKGen.BUILT_AT_SIMULATE: structural models build at simulate, translate takes them - Elements.structural: loop ranges over structural parameters (1:N) folded - recompilation: solver kwargs to init only, not reinit! - test/vssScalarizationTests.jl: two modes with an array, an array grown by recompilation Assisted-by: Claude Opus 5.5
ODEProblem(sys, u0Dict, tspan, p) took sys as the model function under MTK 11 and threw: every transition fell back to the vector u0 (vssTransferDAEInit). Now ODEProblem(sys, op, tspan), op = carried values + the mode's parameters. Assisted-by: Claude Opus 5.5
… report PR caches (0.8-2.6 GB per OS and run) took the repository over 10 GB and evicted master's. The report step prints which of our packages the restored depot has precompiled and why a cache is rejected (JULIA_DEBUG=loading). Assisted-by: Claude Opus 5.5
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## master #108 +/- ##
===========================================
+ Coverage 24.78% 83.14% +58.36%
===========================================
Files 28 114 +86
Lines 3430 34233 +30803
===========================================
+ Hits 850 28464 +27614
- Misses 2580 5769 +3189 ☔ 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Since 0.11.0 OM's translate takes the arrays-kept route only (no second flatten). VSS models with arrays failed there
(OM.jl vssTests + backendExtensions: 18 errors).
scalarizeModel: the backend scalarizes every partition and each structural mode with its own variables;scalarizeArraysrefused models with modes.generateIMTKCodeskips the build at translate): recorded inIMTKGen.BUILT_AT_SIMULATE, so_translateKeptno longer declines them for a missingBUILTentry.for i in 1:N, N a recompilation target):Elements.structuralholds the structural Integer parameters' values (
structuralValues); loop ranges fold them (iterationRange).reltol,abstol) went toreinit!too, which takes none: a MethodError forany recompiled model simulated with solver options.
ODEProblem(sys, u0Dict, tspan, p)was the pre-MTK-10 call; MTK 11 tooksysas themodel function and threw, so every such transition fell back to the vector u0 (
vssTransferDAEInit). NowODEProblem(sys, op, tspan)with the carried values and the mode's parameter values (_parameterValues).test/vssScalarizationTests.jl(two modes with an array; an array grown by recompilation), analytic values;no
vssTransferDAEInitfallback.CI:
actions/cache/restore, same paths and key prefix asjulia-actions/cache). PR caches (0.8-2.6 GB per OS and run) took the repository to 12.5 GB of 10 GB, and GitHubevicted master's.
rejected (
JULIA_DEBUG=loading). In OMBackend 0.11.2: suite time — no deprecated calls, one build per comparison, platform group on mac/Windows #107's run mac and Windows precompiled ImmutableList to OMFrontend again (mac438 s, Windows 245 s) at unchanged commits; Ubuntu took them from the cache. The cause is not known yet.
OM.jl vssTests + backendExtensions: 54 passed, 0 errors, 2 broken (
@test_brokensince 11-09).OMBackend suite (fresh process, 1 thread,
--depwarn=yes, macOS arm64): 2658/2658 (15 new), 836 s, peak RSS 11.39 GiB.