Jupyter Notebooks

Jupyter Notebook lets you create and share computational documents. It offers a simple, streamlined, document-centric experience.

Deploying

You can deploy your notebooks from with the Jupyter Notebook environment - see the Connecting - Jupyter Notebook chapter to get started. Or, you can deploy from the command line with rsconnect-python - see the Connecting - Command Line with Python chapter.

Accessing the Notebook

Once your notebook has been deployed, you can view it as a document in Connect. If you selected with "Publish with Source Code" option in rsconnect-jupyter, you also have the option to re-run the notebook right inside Connect if you are the owner of the notebook or a collaborator. Look for the Refresh Report button in the toolbar:

You can also schedule a notebook to execute periodically. See the Scheduling chapter.

Limitations

• Interactive widgets in notebooks will not work. Notebooks are rendered with nbconvert, which runs a Jupyter kernel during rendering to execute your code. When visitors access the notebook on Connect, they see the rendered HTML version of the notebook but there is no active Jupyter kernel to interact with.
• Only Python 3 notebooks can be deployed. Notebooks using Python 2, R, or other Jupyter kernels cannot be deployed.