Use this template
This repository is a template, not a library: it is not published to PyPI. A simulation code forks it once and keeps it as the home of its profiling results.
1. Fork it
Section titled “1. Fork it”Create a repository from this template (or fork it). In the new repository:
- In
docs/astro.config.mjs, setrepoto the repository’s name,ownerif it is notmax-models, and thetitleanddescription. The base path of the site follows fromrepo. - In
catalogue.yaml, setbaseto/plus the same name. The generated pages need the base path for their links, because the site’s link check rejects relative ones. - In
pyproject.toml, rename the project. - Enable GitHub Pages with the source GitHub Actions (Settings, Pages).
2. Describe your runs
Section titled “2. Describe your runs”Write catalogue.yaml: the manifest lists the cases and the profiling files of
each. Check it with
uv run scope-profiler site catalogue.yaml --check3. Get the files
Section titled “3. Get the files”scripts/fetch_data.sh puts the files into data/ before the site is built. In the template it
writes synthetic samples with scripts/make_sample_data.py; replace it with the way your files
reach the repository, and delete the sample script. See data and the workflow.
4. Build it
Section titled “4. Build it”make install # uv sync with the docs extramake docs-dev # fetches the data, writes the pages, serves http://localhost:4321/scope-profiler-catalog/Push to devel and the docs workflow builds the site and deploys it.
What stays the same
Section titled “What stays the same”The pages in docs/src/content/docs/catalogue/ and the chart data in docs/public/catalogue/
are generated and ignored by git; everything else is yours to change, including the
report components in docs/src/components/report/, which draw the pages from their data.