Cache when rendering a notebook

[linogaliana]

I need to give access to jupyter notebook to beginners. I am writing that notebook using .qmd format and render it into .ipynb.

quarto render index.qmd --to ipynb

The notebook can be found here and uses some time consuming computations. I would like to skip them when re-rendering the document (to check, for instance, that some elements are okay) and caching seemed like a good approach.

I thus put the followiing elements in my header:

---
title: "Un Yuka avec Python"
execute:
  cache: true
format:
  html:
    code-fold: true
    code-summary: "Dérouler la solution 🟡"
---

But each time I re-render the document, everything starts from zero again. Am I using the option wrong ?

[cderv]

Does caching works when you rendering to html and not .ipynb ?

As caching with python is using jupyter system, and you are rendering to ipynb itself here, it is possible that there is no caching happening.
I am not clear about the code path for converting to ipynb - this a possible output format, but not sure how much of quarto feature are supported with this output format.

@jjallaire, do you have a more clear answer on this ? Thank you

[jjallaire]

The notebook cache is per-format (as the image formats and sizes can vary across pandoc formats) so you won't hit the cache on your first ipynb render after an HTML render (but you will hit it on subsequent ipynb renders)