Conversation
POST /api/optimization was a stub raising NotImplementedError since the legacy Linkage bridge went away with pylinkage.joints. Mechanism has carried the optimizer contract (get/set_constraints, get/set_coords, step, get_rotation_period) since pylinkage 0.8.0, so the service now hands the Mechanism straight to pylinkage's PSO, differential evolution, Nelder-Mead and grid search, with the request's objective as eval function and bounds from generate_bounds(bounds_factor). Each result is rebuilt from the request dict with the optimized constraints and solved once with dt=0, so its mechanism_dict (geometry is joint positions only) previews and loads in the Design tab. Unassemblable results keep their score and get a warning instead. ValueError (bad joint_index, nothing to optimize) answers 400. Requires pylinkage >= 1.2.2: before it Mechanism.set_constraints did not change the solved geometry (pylinkage#58), so every optimizer returned the input mechanism. The smoke test asserts the optimized preview has a wider stroke than the input for all four algorithms; the wasm check does the same in Pyodide, and accepts PYLINKAGE_WHEEL= to test an unreleased pylinkage.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Makes the Optimize tab work, natively and on the Pages demo. Depends on pylinkage 1.2.2 (HugoFara/pylinkage#58); CI stays red until it is on PyPI, then
server/uv.lockgets bumped here.What the stub was about
POST /api/optimizationraisedNotImplementedErrorsince the server's legacy joint-basedLinkagebridge went away withpylinkage.joints.Mechanismhas had the optimizer contract since pylinkage 0.8.0, so the service now passes theMechanismstraight toparticle_swarm_optimization/differential_evolution_optimization/minimize_linkage(Nelder-Mead) /trials_and_errors_optimization, with the request's objective as eval function andgenerate_bounds(center, bounds_factor, bounds_factor).Doing so surfaced that
Mechanism.set_constraintsnever changed the solved geometry (link lengths are cached at build time; the nextstep()snapped everything back) — every optimizer on aMechanismwas a no-op. Fixed in pylinkage#58.Results
Each result is rebuilt from the request dict with the optimized constraints and solved once with
dt=0(drivers keep their angle), so itsmechanism_dict— geometry is joint positions only — previews on the Optimize canvas and loads with "Send to Design Tab". Results that cannot be assembled keep their score and produce a warning.constraint_namesare"<link> (length)"/"<driver> (radius)"inget_constraints()order.ValueErrors (badjoint_index, nothing to optimize) answer 400.Tests
test_optimization_changes_the_mechanism: all four algorithms return a preview whose simulated x-extent is wider than the input's and matches the reported score.check-wasm-backend.mjs: same assertion under Pyodide;PYLINKAGE_WHEEL=…installs a local pylinkage wheel instead of the PyPI pin, which is how this was verified before the release.npm run dev+npm run server: load four-bar → Optimize → Run (PSO 30×100, 1.2 s) → 1 result, score 136.46 vs 35 baseline → Send to Design → 62-frame simulation.