Appendix G — Data Sources, Licences and Credits

Keywords

Quarto, reproducible publishing, coding agents, Markdown, Python, EPUB, data analysis, k-means clustering

Every number in this book is computed, when the book is built, from the data files listed here, or carries its own citation in the text. Each data row in the London files records its source, address and retrieval date; the water data’s full provenance, including every table’s licence and cut-off date, is in examples/water/data/README.md in the workshop repository.

London (Part II)

  • Visits. Association of Leading Visitor Attractions (ALVA), Visits made in 2025 to visitor attractions in membership with ALVA, https://www.alva.org.uk/details.cfm?p=423, and the same tables for 2015 to 2024 (details.cfm?p= 606, 607, 608, 609, 610, 615, 616, 617, 618 and 619), retrieved 2026-10-30. ALVA counts its members only, so the book writes “ALVA member attractions”.
  • Facts about the sights (data/sights.csv): each sight’s official visitor page, and for the National Gallery’s nearest station, Transport for London’s StopPoint service, checked 2026-11-01.
  • Churches. Visit London, Cathedrals and churches in London, retrieved 2026-10-31.
  • Places. Coordinates from Wikidata, CC0, retrieved 2026-11-01.
  • Borough boundaries. Office for National Statistics, Local Authority Districts (December 2024) boundaries, Open Government Licence v3.0; contains OS data © Crown copyright and database right 2024. Retrieved 2026-11-01.

Photographs

All via Wikimedia Commons, each checked on its file page on 2026-10-30 for its licence and the absence of people:

Table G.1: The London photographs.
Photo Author Licence
Big Ben in the sun John Dillenbeck CC0 1.0
British Museum Mike Peel CC BY-SA 4.0
Natural History Museum Julian Herzog CC BY 4.0
Tate Modern Diego Delso CC BY-SA 4.0
National Gallery Diego Delso CC BY-SA 4.0
Tower of London Dietmar Rabich CC BY-SA 4.0

Colombia (the water analysis)

The data were retrieved on 2026-10-30. Nothing in the water analysis is synthetic except the toy points that illustrate k-means in Section 20.1, which are labelled as such.

The evaluation (the fast path and Part III)

  • Bank Marketing, UCI Machine Learning Repository, dataset 222, licensed under Creative Commons Attribution 4.0 (CC BY 4.0): S. Moro, P. Rita and P. Cortez (2014), Bank Marketing [Dataset], https://doi.org/10.24432/C5K306, retrieved on 2026-11-18 and kept unchanged as data/bank-full.csv. The study behind it: S. Moro, P. Cortez and P. Rita, “A data-driven approach to predict the success of bank telemarketing”, Decision Support Systems 62 (2014), 22–31, https://doi.org/10.1016/j.dss.2014.03.001.
  • Acme Signals is fictional, and so are its releases and their months. Its monthly business figures, data/synthetic_business.csv, are synthetic, generated with a fixed seed by scripts/make_synthetic_business.py and labelled so wherever they are used. Every model metric in the book is computed from the real data.

Other material

  • Pelican drawings (Section 2.2): Simon Willison’s pelican riding a bicycle series, one run per model and effort, from his published transcripts for Haiku 5.5, Sonnet 5.5, Opus 5.5 and Fable 5.1. No licence is stated on them; they are reproduced with credit and links.
  • Software. Quarto (MIT License), Pandoc, Typst, Python, pandas, matplotlib, scikit-learn and three.js (MIT License), each under its own open-source licence.
  • Animations and diagrams were made for this book or for the workshop’s slides. The chat, agent and Windows animations are illustrations, as their captions say; the other terminal animations, the preview and the 3-D scenes are real recordings, made with a demo account where a terminal is shown.