Documentation & Publishing

jupyter-book

Turns a folder of data-science notebooks into an interactive online textbook, code and all.

Install it: pip install jupyter-book

What does it do?

Jupyter Book takes Jupyter notebooks and Markdown files, the mix of code, charts, and explanation that scientists and analysts already write, and compiles them into a polished, book-like website with chapters, citations, and a table of contents. Readers can click a button to actually run the code live in their browser instead of staring at a static screenshot of a graph. It’s built for what its creators call computational narratives, documents where the explanation and the working code are the same artifact. Underneath, it actually relies on Sphinx to build the pages.

See it in action

These are terminal commands that first create a new folder of starter book files and then compile the notebooks and text inside it into a finished, browsable online book.

jupyter-book create mybook/
jupyter-book build mybook/

Why would a non-developer care?

A lot of what you might picture as a boring PDF report from a data team could instead be a living document where tweaking a number updates a chart in real time. That distinction matters enormously in science and education, where re-running the calculation behind a claim is how you actually verify it.

Real-world examples

It’s widely used to publish open-access textbooks and scientific papers with reproducible results attached, across university data-science courses and research lab publications. Without something like it, a printed textbook’s code examples are dead text you retype by hand; with it, you click launch and the chapter runs in a cloud notebook.

Who uses it

Educators, scientists, and data teams publishing tutorials, textbooks, or research where the code itself is part of the story.

How it compares to alternatives

It sits alongside bookdown, the similar tool from the R programming community that directly inspired it, and competes loosely with plain Sphinx or GitBook for anyone who doesn’t need live code execution.

Fun fact

Jupyter Book was directly inspired by bookdown from the R community, and its team deliberately built it to bring the same reproducible-book workflow to Python and Jupyter users.

New to Python and want to actually try libraries like this yourself?

Find a beginner-friendly course

Related libraries