Skip to content

feat(data): the unit and the meaning of every parameter, in the metadata (T2 of #449) - #461

Merged
ocots merged 25 commits into
mainfrom
feat/parameter-info
Nov 20, 2026
Merged

ocots merged 25 commits into
mainfrom
feat/parameter-info

Conversation

@ocots

@ocots ocots commented Nov 19, 2026

Copy link
Copy Markdown
Member

Refs #449 (T2, option (a) chosen by the maintainer). Stacked on #460: its base is fix/t2-review; retarget it to main once #460 is merged. Applied conventions: ct-conventions v0.14.0, ct-agents v0.5.0.

One source of truth for the parameters. Before this PR, the values were written in the metadata and again in hand-written tables in 17 descriptions. The page of a problem already generated a ### Parameters section from the metadata, so 5 pages had two. Now the metadata hold the unit and the meaning of each parameter, and the page shows them.

Changes

  • :parameter_info, a required field of the metadata: one entry per parameter, in the order of :parameters.
    :parameter_info => (
        y_t0=(unit="m", meaning="initial altitude"),
        c0=(unit=nothing, meaning="zero-lift drag coefficient"),
        ...
    ),
    unit is nothing for a dimensionless parameter, or for one whose unit the sources do not give (the academic problems, jackson, robot, insurance…).
  • Checks (test_parameter_info): the keys are exactly the parameters, each entry is (unit, meaning) with a non-empty meaning.
  • The generated section (parameters_section in docs/problems.jl): Parameter | Default value | Unit | Meaning.
  • The descriptions lose their parameter sections. The outline test now refuses one. The one sentence of prose they held (balanced_field: values of a mid-size jet, from Dymos) moves to the introduction.
  • 358 parameters, one commit per problem. The meanings come from the hand-written tables, the descriptions and the models.

Found on the way (for T3)

Some parameters are not read by the models; their meaning says so:

  • brachistochrone: u_min, u_max (±π) are read, but the OptimalControl and JuMP models bound the control by ±1.57 written in the code (and tf by 0.1 and 20).
  • balanced_field: v_min, v_max are not read; α_min is read but unused (the models use 0 ≤ α, and v ≥ 1, tf ≥ 0.1 in the code).
  • space_shuttle: α_t0 is used neither by the models nor by the initial guess.

The formatter also wrapped the long :status lines of 4 metadata files.

Test plan

  • Locally (Julia 1.12): Pkg.test(; test_args=["core"]): 2812 passed, 1 broken (the existing @test_skip of test_discovery_cache.jl).
  • The generated section, rendered for rocket: the four columns, with – for no unit.
  • CI, and the documentation build.

🤖 Generated with Claude Code

ocots and others added 24 commits November 19, 2026 23:15
…problem page

A new metadata field `:parameter_info`, optional until every problem has
it; checked by test_parameter_info, shown by the generated Parameters
section.

Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…er sections

`:parameter_info` is required: every parameter has its unit and meaning
in the metadata, and the page shows them in the generated Parameters
section. The hand-written parameter sections of the descriptions are
removed (their values were a second copy, and 5 pages had two
"Parameters" sections); the outline test refuses one.

Refs #449

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@ocots ocots added the run ci github-runner Trigger the CI workflow on GitHub-hosted runners label Nov 19, 2026
…er-info

# Conflicts:
#	ext/MetaData/mountain_car.jl
Base automatically changed from fix/t2-review to main November 20, 2026 07:26
@ocots
ocots merged commit 7d0d6e2 into main Nov 20, 2026
11 of 12 checks passed
@ocots
ocots deleted the feat/parameter-info branch November 20, 2026 07:27
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

run ci github-runner Trigger the CI workflow on GitHub-hosted runners

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant