Skip to main content

ipypy notebook format

Project description


Store Jupyter Notebooks in a more repo and coding friendly way.

Have you ever noticed that a Jupuyter Notebook nowadays is a json file? Containing together the metadata of the notebook itself, the list of cells, and for each cell, not only the source (what was written on it), but also the outputs, and metadata for each cell.

This project attemps to propose an alternative for that.


You can pick between a SplitCodeManager, which stores each notebook in:

  • my_file.ipynb (the usual notebook file, but with source information extracted)
  • my_file.ipypy (a pure code file that stores only the actual source code)

Or, a SplitOutputManager, which stores each notebook in:

  • my_file.ipynb (the usual notebook file, but without the cells output)
  • my_file.nbout (a json file that stores only the outputs of each cell)


  • You can now import your notebook from another file
  • You can now use standard coding tools and practices for manipulating Notebooks code:
    • testing
    • import code defined in there
    • editable naturally by any editor
    • refactoring
    • tracking changes
    • ...
  • The source code of your notebook, can be versioned in a repository, where diffs, and history is readable. Now makes sense. It's code.
  • (Work In Progress) You can choose to simply ignore the metadata files (.ipynb) in the repository, or keep them versioned. It should be up to you.


We are in beta. Once you open a notebook with this extension enabled, and later save it, your notebook will be saved in a format a bit incompatible.


$ pip install ipypy

You will also need to configure your jupyter so it uses ipypy, by editing your jupyter config file, or from command line

$ jupyter lab --NotebookApp.contents_manager_class="ipypy.SplitCodeManager"

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ipypy-0.1.2.tar.gz (7.2 kB view hashes)

Uploaded source

Built Distribution

ipypy-0.1.2-py3-none-any.whl (9.7 kB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page