From c57d6d72b3184268eb7d6a8643f887ac86da26f3 Mon Sep 17 00:00:00 2001 From: Matt McKay Date: Mon, 17 Aug 2026 11:13:44 +1000 Subject: [PATCH] =?UTF-8?q?Annotate=20the=20new=20bivariate=5Fdist=20yfina?= =?UTF-8?q?nce=20read=20=E2=80=94=20main=20is=20red=20without=20it?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit lecture-python-intro#837 added lectures/bivariate_dist.md, which downloads AMZN and COST monthly closes through yfinance in a solved exercise. The strict audit fails on any live-API read with no `api:` annotation — by design, so a new fetch in a lecture repo cannot appear here silently — so the audit-dashboard build on main has been failing since that lecture landed (run 31984234688, the sole warning being `missing_api_annotations: lecture-python-intro:bivariate_dist:yfinance`). Nothing to do with wave C1, which merged shortly before: the same run reported zero problems in every other bucket, and re-running the scan here with the annotation added exits 0 with all five buckets empty. Worth noting how the timing hid it. This repo's audit reads the CONSUMING repos, so a lecture-repo merge can turn main red here with no change to this repo at all, and nothing in lecture-python-intro reports it. The local run against a clone that predated the new lecture passed; it only failed once CI cloned intro fresh. `pedagogy: incidental` follows the sibling reads — the lecture points at observed_distributions for the same workflow, so the download is the illustration's input rather than its lesson. CATALOG.md is unaffected; api annotations do not feed the generated catalog. Co-Authored-By: Claude Opus 5 (1M context) --- scripts/audit_annotations.yml | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/scripts/audit_annotations.yml b/scripts/audit_annotations.yml index fe60754..232de0e 100644 --- a/scripts/audit_annotations.yml +++ b/scripts/audit_annotations.yml @@ -74,6 +74,13 @@ api: series: SI.POV.GINI, NY.GDP.PCAP.KD pedagogy: mixed note: demonstrates searching for the Gini series ID with wbgapi + lecture-python-intro:bivariate_dist:yfinance: + series: AMZN, COST + pedagogy: incidental + note: > + two downloads in a solved exercise, fitting a bivariate normal to monthly + returns; the lecture points at observed_distributions for the same + workflow, so the fetch is the illustration's input rather than its lesson lecture-python-intro:fitting_distributions:yfinance: series: AMZN pedagogy: incidental