Rebuild it yourself

The book, its notebooks and the data are all in one GitHub repository, supertj/youtube-monetization-by-country. The book lives in the book folder and reads the CSV files from the rest of the repository by relative path, so nothing is downloaded while it builds.

Run the notebooks¶

You need Python 3.12. From a clone of the repository:

cd book
python3.12 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
pip install jupyterlab
jupyter lab

Open country-shares.ipynb or scoring-rule.ipynb and run all cells. Each one finishes in a few seconds. The shared chart colours are in plotstyle.py, next to the notebooks.

Build the book¶

This is a Jupyter Book 2 project, built on MyST. The same environment builds it. Jupyter Book 2 also needs Node.js 18 or newer; if node is not on your path, the jupyter book command offers to install a private copy.

jupyter book start --execute          # live preview on a local port
jupyter book build --html --execute   # static site in _build/html

--execute runs both notebooks during the build, so the figures and the numbers in the notebook chapters’ text come from the CSV files as they are at that moment. Change a CSV and rebuild, and the text follows.

Versions used for this edition¶

The same pins are in book/requirements.txt.

A longer analysis¶

The folder analysis/2026-10-membership-rule holds a standalone script that goes further than the last chapter of this book. Among other things it simulates how much spread between countries sampling noise alone would produce, and tests whether the share of unknowns differs between countries with and without channel memberships. Its README explains how to run it.