@jku-icg/jupyterlab_nbprovenance

JupyterLab Notebook Provenance

An extension for JupyterLab (v2+) to track interaction provenance in Jupyter notebooks.
The provenance information is stored using the provenance tracking library trrack.

This extension listens to notebook change events from JupyterLab and stores these changes as a provenance graph with Trrack. The trrack-notebook-vis library takes the stored provenance information to visualize the provenacne graph in a sidepanel in JupyterLab. From the visualization, users can browse the notebook's history, filter it, and switch between states.

🚀 Try it out on Binder:

Install

Install JupyterLab if you haven't already.

To install the extension from the terminal, type:

jupyter labextension install @jku-icg/jupyterlab_nbprovenance

Binder

To use the extension in a JupyterLab instance from Binder, add a postBuild file with the above command to the root of the repository (see the demo branch for reference).

Development

1. Prepare Environment (optional)

   conda create --name notebook_provenance python=3.8.5
   conda activate notebook_provenance
   conda install jupyterlab

2. Clone this repo, cd into the folder

3. Then build the extension

   yarn install
   yarn build

4. Install the extension

   jupyter labextension install --minimize=False # install the current directory as an extension

   Disabling the minifier is optional, but minimizig the code is not necessary locally and it reduces the build time.

5. Start JuypterLab in watch mode

   jupyter lab --watch

6. Make code changes

7. Rebuild the extension with yarn build or watch for file changes and built automatically with yarn watch
   JupyterLab updates itself, due to the --watch parameter

8. Refresh the JupyterLab in your browser to load the updated files

Also see the JuypterLab Extension Developer Guide.

Publish

Make sure to npm login with the user icgbot first.

npm pack --dry-run  # verify files
npm publish

Attributions

This extension uses icons by fontawesome.com, available under CC BY 4.0.