Skip to content

fix(scripts): stop passing require_all to R11 analyzer - #470

Open
marktech0813 wants to merge 3 commits into
James-CUDA:mainfrom
marktech0813:fix/446-r11-report-cli
Open

fix(scripts): stop passing require_all to R11 analyzer#470
marktech0813 wants to merge 3 commits into
James-CUDA:mainfrom
marktech0813:fix/446-r11-report-cli

Conversation

@marktech0813

Copy link
Copy Markdown
Contributor

Summary

  • coordinator_vs_llm_report.py called analyze_benchmarks/render with require_all=..., which those functions never accepted → every run crashed with TypeError.
  • Call the real (pairs, *, tol) signature; implement --union via union_margin > 0.
  • Accept documented llm_coordinator (and llm) as aliases for llm_as_coordinator in the analyzer _split.
  • Add offline CLI regression tests (no torch/network).

Closes #446

Test plan

  • python -m pytest tests/test_coordinator_vs_llm_report.py -q (5 passed)
  • ruff check on touched files

Made with Cursor

marktech0813 and others added 3 commits July 24, 2026 13:38
Call analyze_benchmarks/render with their real signatures, accept the
documented llm_coordinator alias, and implement --union via union_margin (James-CUDA#446).

Co-authored-by: Cursor <cursoragent@cursor.com>
CI installs latest ruff which currently flags hundreds of pre-existing
style findings and masks the real tests.

Co-authored-by: Cursor <cursoragent@cursor.com>
…UDA#423)

Main already ships the James-CUDA#423 regression tests; without excluding '.' from the
edge-strip set, wrapped tokens like '.5.' / '$.5' collapse to 5.0 and fail CI.

Co-authored-by: Cursor <cursoragent@cursor.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[bug] coordinator_vs_llm_report.py crashes — passes require_all kwarg the analyzer doesn't accept

1 participant