A workaround for Version Control Jupyter Notebooks

Version control using git is quite simple for codes. However, since Jupyter notebook has a different format and often contains binary outputs, it is culumbersome to track those changes. Here is a workaround to save each notebook as script and version controllong those scripts instead. The idea is borrowed from stackoverflow.

Use Git (updated 2023)

If you use Jupyter Lab, there is a Git plugin that you can use to diff the Jupyter notebooks.

A workaround

• create a jupyter_notebook_config.py file. Alternatively, you can type jupyter --config-dir to see whether the file exists. If not, then create it by

jupyter notebook --generate-config

This will usually create jupyter_notebook_config.py under ~/.jupyter.

• add the following lines to the jupyter_notebook_config.py file

import os
from subprocess import check_call

def post_save(model, os_path, contents_manager):
    """post-save hook for converting notebooks to .py scripts"""
    if model['type'] != 'notebook':
        return # only do this for notebooks
    d, fname = os.path.split(os_path)
    check_call(['ipython', 'nbconvert', '--to', 'script', fname, '--output-dir', './src'], cwd=d)

c.FileContentsManager.post_save_hook = post_save

It will save a copy of the script into ./src directory everytime the notebook is saved.