pydata-london-2024-lightning-talk

Lightning talk @ PyData London 2024 - Version Control + Notebooks (The tale of a group project)

This repo contains the slides and some Jupytext examples.

• Setup
• Example 1 - Jupyter notebook -> Python file
• Example 2 - Python file -> Jupyter notebook

Setup

• install the requirements (jupyterlab, seaborn, jupytext, pre-commit)

  pip install -r requirements.txt

• install the pre-commit hooks

  pre-commit install

• start Jupyter

  jupyter-lab

Example 1 - Jupyter notebook -> Python file

In this example the notebook example-1.ipynb was created in Jupyter. The Python file (example-1.py) was not created manually, but instead was created by Jupytext the first time the notebook was saved.

In Jupyter open the notebook example-1, make changes to the notebook and see the same changes reflected into the .py file as soon as the notebook is saved.

Example 1 also shows how the pre-commit hooks will be applied to the .py file.

Example 2 - Python file -> Jupyter notebook

In this example the notebook was not created in Jupyter. The example shows how it is possible to re-create a notebook (locally) from a Jupytext .py files (for example, if working on the same notebook and only the Python file is commited to the git repo).

In Jupyter right-click on the example-2.py file, select Open With and then Notebook. The .ipynb file will be created, after which it will be possible to close the .py file and start working directly on the notebook.

Example 2 also shows how changes to the notebook file wil be displayed as diff when you look at the commit history for example-2.py. See for example the diff between 44333b27 and 3edca83c here.