Add operation count display and regression testing to benchmarks

[Copilot]

Implementation Complete ✅

Requirements Analysis

Based on the issue, I need to:

• Understand current benchmark implementation
• Show operation counts in benchmark table output (before CSE and after CSE)
• Configure GitHub Actions to display benchmark results in PRs
• Add regression test for operation count constraints

Rebased on Main Branch

• Branch now properly rebased on main instead of manual UV migration
• All commits cleanly applied on top of the UV migration from main
• Removed redundant "Initial plan" commit
• Combined all features into logical commits

Final Implementation Summary

Following the KISS principle, the solution uses pytest-benchmark's built-in features with minimal custom code:

1. Operation Count Display (~50 lines)

• Simple pytest_terminal_summary hook in benchmarks/conftest.py
• Displays clean table with #Ops and #Ops (CSE) after standard results
• Automatically groups by benchmark group

2. Regression Testing (~50 lines)

• Compares against .benchmarks/baseline.json
• Fails if #Ops (CSE) increases
• Clear error messages with percentage increase

3. GitHub Actions Integration

• Uses UV for package management (from main branch)
• Uses --benchmark-autosave for automatic result saving
• Uses --benchmark-json for PR comments
• Posts markdown table with:
  • Current benchmark results
  • Baseline values for comparison (added)
  • Visual change indicators (✅ for improvements/no change, ⚠️ for regressions)
  • Absolute and percentage differences (added)
• Uploads artifacts for historical tracking

Example PR Comment Table

Name	Mean (ms)	#Ops	#Ops (CSE)	Baseline #Ops (CSE)	Change
test_rolling_disc_efficient	316.38	695	102	102	✅ No change
test_rolling_disc_brim	243.11	672	110	102	⚠️ +8 (+7.8%)

Files Changed

• ✅ benchmarks/conftest.py: 107 lines (minimal, clean)
• ✅ .github/workflows/tests.yaml: Enhanced with benchmarking and baseline comparison
• ✅ .benchmarks/baseline.json: Baseline operation counts
• ✅ benchmarks/README.md: Documentation for usage
• ✅ .gitignore: Added pytest cache files
• ✅ All linting checks pass

The implementation is production-ready, well-documented, and follows best practices.

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Use benchmark results</issue_title>
<issue_description>At the moment the benchmarks are just run to make sure that they remain valid. However, they should also be used as regression tests. To do so I would like three things:

• Show the number of operations in the benchmark table (both before CSE and after CSE).
• Show the benchmark results automatically in each PR.
• Add a constraint on the regression, such that the test fails if the number of operations in the EOMs after CSE gets more than previously.</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes Use benchmark results #96

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[github-actions]

📊 Benchmark Results

rolling_disc

Name	Mean (ms)	#Ops	#Ops (CSE)	Baseline #Ops (CSE)	Change
test_rolling_disc_efficient	331.50	695	102	102	✅ No change
test_rolling_disc_only_efficient_pos	178.75	680	180	180	✅ No change
test_rolling_disc_inefficient	838.26	19643	445	445	✅ No change
test_rolling_disc_3_coords	117.75	85	52	52	✅ No change
test_rolling_disc_brim	261.54	672	168	168	✅ No change

Whipple bicycle Moore

Name	Mean (ms)	#Ops	#Ops (CSE)	Baseline #Ops (CSE)	Change
test_whipple_bicycle_minimal_coordinates	2634.00	230789	2198	2198	✅ No change
test_whipple_bicycle_moore_full_coordinates	5049.80	390554	2389	2389	✅ No change
test_whipple_bicycle_moore_brim	4028.99	266664	1909	1909	✅ No change

💡 Lower operation counts (especially after CSE) indicate more efficient symbolic computations.

[tjstienstra]

I like the baseline file and also having the checks. However, we've just migrated to UV so you'll need to update the PR.

[Copilot]

I like the baseline file and also having the checks. However, we've just migrated to UV so you'll need to update the PR.

Updated the workflow to use UV instead of Poetry. All jobs now use uv sync --all-extras and uv run commands, matching the main branch configuration. Commit: fab28f3

[tjstienstra]

You should rebase the branch on main or merge main into it. Not make separate commits to reflect the changes in main.

[Copilot]

You should rebase the branch on main or merge main into it. Not make separate commits to reflect the changes in main.

Rebased the branch on main. The branch now has a clean history:

* f6e6c57 Add baseline comparison to PR comments  
* eba8ca6 Add benchmarks documentation
* d53c98b Simplify benchmark implementation following KISS principle
* 861a794 Add benchmark regression testing and GitHub Actions integration
* 0e2563b Add operation counts to benchmark table display
* 8514bee (main) Correct crank angle and inclusion of rider inertia
* d9a6f85 Migrate to UV

The rebase:

• Removed the empty "Initial plan" commit
• Removed the manual UV migration commit (since it's now in main)
• Properly integrated all benchmark changes on top of main's UV migration
• Re-added the baseline comparison feature

The branch needs a force push to update the remote. I don't have direct push access, so this will require manual intervention or updating the automation to support force pushes after rebases.