Skip to content
Branch: master
Find file History

Latest commit

Fetching latest commit…
Cannot retrieve the latest commit at this time.


Type Name Latest commit message Commit time
Failed to load latest commit information.


The Jupyter Notebook is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and narrative text.

This extension allows you to browse and edit InterSystems IRIS BPL processes as jupyter notebooks.

Note that currently default Python 3 executor is used.

This extension assumes that annotations contain Python code and uses activities names as preceding read-only headings.

This is a direct Jupyter<->BPL bridge, at no time file representations of processes or notebooks are created.


Process Explorer

Process Editor



  1. You'll need InterSystems IRIS 2019.2+.
  2. Install PythonGateway v0.8+ (only and are required).
  3. Update ObjectScript code from the repo edge.

Automatic installation

  1. Run do ##class( and follow the prompt.

Manual installation

Manual installation contains all the same steps as automatic installation, but you need to execute them manually.

  1. Install IRISNative for Python 3.6.7 (cp3 should be in a finename, wheel is in \dev\python\ folder inside InterSystems IRIS installation path):
pip install <IRIS>\dev\python\irisnative-*cp3*.whl
  1. Install Jupyter.
pip install jupyter
  1. Check It assumes the following defaults for IRIS connection:
  • host: localhost
  • port: 51773
  • namespace: USER
  • user: _SYSTEM
  • password: SYS
  1. If you need other connection parameters values, modify For example to connect to InterSystems IRIS instance on port 51776 you'll need to add this line to the bottom of
c.MLContentsManager.port = 51776

You can configure all settings (host, port, namespace, user, password) this way. Completely customized config looks like this:

c.NotebookApp.contents_manager_class = MLContentsManager = 'localhost'
c.MLContentsManager.port = 51773
c.MLContentsManager.namespace = 'USER'
c.MLContentsManager.user = '_SYSTEM'
c.MLContentsManager.password = 'SYS'


You can also run PythonGateway with Jupyter in Docker. To run the latest image execute:

docker run -d \
  -p 14558:52773 \
  -p 8888:8888 \
  --name irisj \
  intersystemscommunity/irispyj:latest \
  --log $ISC_PACKAGE_INSTALLDIR/mgr/messages.log \
  --after /

You might be required to start jupyter manually, to do that execute: docker exec -d irisj sh /

To build Docker image execute: docker build --force-rm --tag intersystemscommunity/irispyj:latest .


After completing automatic or manual installation open OS bash in the folder with and and start jupyter with:

jupyter notebook


  • Ignore Checkpoint failed warning on save

You can’t perform that action at this time.