Write your first package using literate programming
summary
Shows exactly what gets built and what tools actually wire it up: export notebook code to a package with Nbdev, build sdist and wheel with twine, push a release artifact to test Pipi (demoed upload but fork exercise hits a Windows rename script bug), run notebook tests via ipytest and nbdev CI, lint and types with nbqa running black, flake8 and mypy, render docs with quarto, and serve an interactive mini game via voila — the notebook→package→CI pipeline mostly works and where it fails the failure is specific.
AI-written summary · about
chapters (16)
- 0:00Literate programming intro
- 15:11Python modules and packages
- 32:00Publishing Python packages
- 32:41What is literate programming
- 39:18Jupyter and Quarto intro
- 42:07nbdev for package dev
- 1:08:13Semantic versioning best practices
- 1:13:49QA tools for Python notebooks
- 1:15:17CI/CD with GitHub Actions
- 1:26:27nbdev + CI/CD publishing
- 1:31:21Python packaging history
- 1:34:17Quarto documentation customization
- 1:36:49Dashboards with Voila and widgets
- 1:40:54ChatGPT in Jupyter
- 1:48:10Benefits of literate programming
- 1:50:18Q&A
related talks
17mLiterate Programming With Jupyter Notebooks and Quarto
30mCreate a Python Package From Zero to Hero
22mPublishing well formed Python packages
02h 36mPython Packaging Demystified
28mModern Python packaging without pain
11mHow To Create Your First PyPI Package
28mUnlock the Full Potential of Jupyter Notebooks
49mUntitled13.ipynb