Jupyter kernel for DeltaQ

A customized Jupyter iHaskell kernel for use with the deltaq tool.

Running a Jupyter server using Nix

Launch Jupyter with the iHaskell kernel including the deltaq and probability-polynomial packages:

nix run

In fact, this Jupyter kernel can be launced without directly cloning this repository:

nix run github:functionally/deltaq-jupyter

Docker

Building a docker image using Nix

First build the image.

$ nix build -o deltaq-jupyter-docker.tar.gz ".#docker"

$ docker load --quiet < deltaq-jupyter-docker.tar.gz
Loaded image: localhost/jupyter-deltaq:2mf9zcagms0isg5n94xsml34kj4bn42k

Push to docker registry

You can push the image to a registry.

docker push localhost/jupyter-deltaq:2mf9zcagms0isg5n94xsml34kj4bn42k \
            docker.io/bwbush/jupyter-deltaq:2mf9zcagms0isg5n94xsml34kj4bn42k

Run the server in docker.

You can run the local image, exposing the service on port 9999 for example. The container should have at least 4 GB of memory and two CPUs, but larger computations require more memory. If you deploy this on kubernetes, you can use the HTTP path /api as the health check.

Locally,

docker run --publish 9999:8888 localhost/jupyter-deltaq:2mf9zcagms0isg5n94xsml34kj4bn42k

or from DockerHub,

docker run --publish 9999:8888 ghcr.io/functionally/deltaq-jupyter:latest

The default password is deltaq: you can change this by visiting http://localhost:9999/lab in a web browser. Alternatively, you can use with the default password by visiting http://localhost:9999/lab?token=deltaq.

Use the "upload" and "download" features to move notebooks in and out of the container.

Updates

In order to use the latest version of the deltaq packages, update the nix flake.

nix flake lock --update-input deltaq