Skip to content

Rewrite dated intro commentary; fill matplotlib and seaborn coverage - #30

Merged
tdhopper merged 3 commits into
masterfrom
content-refresh
Aug 5, 2026
Merged

Rewrite dated intro commentary; fill matplotlib and seaborn coverage#30
tdhopper merged 3 commits into
masterfrom
content-refresh

Conversation

@tdhopper

@tdhopper tdhopper commented Aug 5, 2026

Copy link
Copy Markdown
Owner

TL;DR

The intro's library commentary was written around 2017 — plotnine "a young product," plotly described via its retired online-publishing model, an apology for having "little experience" with interactive libraries that now have full columns. Rewritten for 2026 with every tagline re-verified against current docs. Coverage gaps also filled: matplotlib goes from 2 to 12 of 13 plots, classic seaborn from 5 to 13 of 13, bringing the site to 119 example cells.

Files to review (4):

File Why
INTRO.md (start here) Rewritten library commentary; legacy opener/closing sections tightened (two long-standing subject-verb errors fixed). Voice and section headers preserved.
Examples.ipynb 18 new cells: 10 matplotlib, 8 seaborn.
tests/test_plots.py Regenerated from notebook tags.

Reviewer notes

  • matplotlib skips one plot: scatter-with-regression would be ~12 lines of statsmodels and 2 of matplotlib, demonstrating the wrong library. The loess example stays because sm.nonparametric.lowess is two lines.
  • Six new cells carry site-visible notes where the library lacks a primitive (matplotlib faceting/stacking/dodging/KDE/loess; seaborn's missing stacked-bar and CI-band-around-lowess).
  • Checkable claims in the intro are true as of this branch: 18-line matplotlib faceted scatter vs 4-line ggplot2; five plots with no pandas column; two plots absent from seaborn.objects and hvplot.
  • Prose passed a PROSE.md audit; remaining flags were deliberate keeps (the site's original quirky section headers).

Verified: notebook executes clean, 119/119 cells emit PNGs, render + pytest pass.

Matplotlib covered 2 of the 13 plots and classic seaborn 5. Add cells so
matplotlib covers 12 and seaborn all 13, matching each reference plot's
titles and labels. Where matplotlib has no primitive (faceting, stacking,
dodging, KDE, loess) the cell does the work by hand and says so in a note.

Matplotlib skips scatter-with-regression: an OLS fit with a 95% confidence
band is a dozen lines of statsmodels and two of matplotlib, so it would
demonstrate statsmodels rather than matplotlib.
The library notes still described a 2017 site: plotnine as a young product,
plotly as an online publishing service, and an apology for having no
interactive examples that have since been added. Update the quotes against
each project's current docs (matplotlib, seaborn, plotnine, Vega-Altair,
plotly, Lets-Plot, Bokeh), point plotnine at plotnine.org and plotly at
plotly.com, drop the bqplot mention since the site has no bqplot column,
and say what each library actually covers in the grid below.
Fix two subject-verb errors, drop puffery and needless words in the
opener and closing sections. Library commentary untouched.
@tdhopper
tdhopper marked this pull request as ready for review August 5, 2026 11:48
@tdhopper
tdhopper merged commit 53eabd3 into master Aug 5, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant