This repository demonstrates how to perform a full integration of a jupyter notebook in a web page.
The code is highlighted with pygments. Math equations are rendered with MathJax. The jupyter notebook is converted to HTML with nbconvert. The main style is from bootstrap. Custom CSS classes are provided in this repo for a consistent style between the included jupyter HTML file, and the mother HTML page.
This setup is fully explained in my article https://thedatafrog.com/jupyter-notebooks-web-pages.
The file templates/overfitting.html has been generated with
jupyter nbconvert --execute --template basic overfitting.ipynb
templates/main.html is the template web page in which we want to include
render.py is a simple custom python script that generates index.html from the template. This script is just for demonstration purposes. For real websites, I use django and its template engine.
the css directory contains custom CSS classes for the styling of the various elements in the web page:
- css/main.css : main styling
- css/notebook.css : styling of the jupyter notebook elements (e.g. input and output prompts, images, ...)
- css/pygments/notebook/colorful.css : highlighting style for code in the jupyter notebook
Just do
python render.py
And open the resulting index.html in your browser to see the results.
If you edit template.html do not forget to render again.
Here is what I get in my django website:
