v0.5.1 — gapminder data

Ships the gapminder data in the Python workshop container. Closes nesi/intro-python#5.

Content

Pins nesi/intro-python v0.5.0, which adds gapminder in both shapes:

• data/gapminder.csv — tidy long, 1704 rows (country, year, pop, continent, lifeExp, gdpPercap), for plotnine and groupby work
• data/gapminder_all.csv + data/gapminder_gdp_{africa,americas,asia,europe,oceania}.csv — wide, matching Software Carpentry python-novice-gapminder so its episodes teach unchanged
• data/README.md — columns, row counts, provenance

Learners find these in ~/intro-python/data/ alongside the existing Portal surveys.csv and species.csv.

Container

• Content repo now pinned by release tag (ARG PYTHON_VERSION="v0.5.0" with git clone --depth 1 --branch) rather than a commit SHA, matching the other training-environment app repos.
• Added a no-clobber wget fallback for the seven gapminder CSVs, pinned to the upstream Carpentries lesson commits. It is a no-op whenever the content pin already carries them; it exists so a rolled-back pin still yields a usable image.
• submit.yml.erb image bumped to v0.5.1. It had been left on v0.4.0, two releases behind this repo's own tags.

Supersedes v0.5.0

The v0.5.0 tag was pushed before the submit.yml.erb bump landed on main. At that tag the app still requests image v0.4.0, so a deploy pinned to v0.5.0 would launch the old container with no gapminder data. Use v0.5.1; do not deploy v0.5.0.

Verification

Built from the tag and inspected: /opt/intro-python/data carries all seven gapminder CSVs plus surveys/species, tidy (1704, 6), all (142, 38), surveys (35549, 9), .git stripped, files mode 644, pandas/plotnine/scikit-learn/scipy/numpy/matplotlib import, jupyter lab 4.6.3. The fallback block was tested both ways — downloads all seven into an empty dir, no-ops when files exist.

Deploy

nesi/training-environment PR #22 sets intro_python k8s_container and version to v0.5.1. Note that app is still enabled: false in that file.