🧪 This is a test repository for QuantEcon Actions development
Do not use for production purposes
Production repository: lecture-python-intro
An Undergraduate Lecture Series for the Foundations of Computational Economics
This repository is the experiment and integration sandbox for QuantEcon/actions, and its post-release canary — stage 2 of actions#100. It is a clone of lecture-python-intro wired to the composite actions at the floating @v0 tag, so that a real lecture-shaped build exercises the parts of the chain that the in-repo PR harness structurally cannot reach.
The question it answers is "does this work against a real-ish lecture repo?" — try a new action, a new version, a new workflow shape, and find out end to end. Its history is exactly that: PR-scoped build caching (#36), PDF and notebook builds in CI preview (#35), the @v0 migration (#26).
It cannot be, and that is deliberate rather than a gap to close here. A gate has to mean red stops the release, and three properties of this repo make red ambiguous — all three of which are the right properties for a sandbox:
- it runs
ghcr.io/quantecon/quantecon-build:latest, so the environment moves underneath it; - Dependabot is enabled and its PRs change build behaviour;
- it pins
@v0, so it only ever exercises a published release, never a candidate — and because a release moves that tag, consumers are on the new code at the same moment this repo is.
The first two are how upstream breakage gets discovered early, which is a sandbox's job and the opposite of a gate's. Release gating lives in a separate frozen fixture, test-actions-release — see actions#136 and actions#135.
So: experiment here; gate there. If you are about to add something to this repo to make a release safer, it probably belongs in test-actions-release instead.
It is the only place publish-gh-pages and preview-netlify are exercised at all. Neither is testable inside the actions repo: a composite action is all-or-nothing, so uses: ./publish-gh-pages dies at configure-pages before reaching any of its own logic, and testing the previews there would need live deploy credentials on a pull_request workflow.
The lecture set is deliberately small. lectures/_toc.yml carries five self-contained lectures rather than the full book, chosen mechanically: no build-time network access, no heavy or optional imports, no local {doc} cross-references, no citations. The other lectures stay on disk but are not built (only_build_toc_files: true).
That trim is the point, not a shortcut. Carrying the full book meant this repo went red in April 2026 because a cell in inequality.md raised — content drift, nothing to do with the actions — and it stayed red and unwatched for months. A canary whose failures are usually not about the thing it tests is a canary nobody reads. With the trimmed set, a red run means an action regression.
Because it pins @v0 rather than a fixed release, this repo sees each release as consumers see it — which also means it can only find a regression after the floating tag moves. Pre-release coverage lives in the actions repo's own PR harness.
Jupyter notebook versions of each lecture are available for download via the website.
To comment on the lectures please add to or open an issue in the issue tracker (see above). We welcome pull requests!
Please read the QuantEcon style guide first, so that you can match our style.