JupyterLab Docker Container

This repository contains scrips for starting JupyterLab in Docker. The scripts are optimized for UBC DSCI and STAT working environments.

Startup

1. Run run.sh to start JupyterLab. Then use the printed link to access Jupyter from the browser
2. Put some files to the materials/ directory. They'll become available in Jupyter, so you can work on them there

Shutdown

1. Run uninstall.sh to stop the local JupyterLab

After you close Jupyter, the files you put in the materials/ directory are saved there.

Additional features

The JupyterLab image is based on the official UBC Docker images, but has some additions.

1. Jupyter spellcheck extension
2. The scripts work for any UBC DSCI and STAT courses that use Jupyter. Just change/uncomment the image variable in the .env file and start the container
3. Run update.sh to pull a newer UBC JupyterLab Docker image specified in the .env file. This image will be used for subsequent container startups. Only the latest uncommented image will be used

Troubleshooting

1. If you're working with private repositories, you can mount your ssh-key into the container's $HOME directory by adding this line to the docker-compose.yaml under volumes:

    - ~/.ssh/:/home/jovyan/.ssh/:ro

   Make sure that the $HOME directory of the container matches the one you specify in the docker-compose.yaml. In this case, it is /home/jovyan/, but it may change in newer versions of the JupyterLab image. To check the $HOME directory of the container, type

   # Get names of running containers
   docker container ls --format {{.Names}}
   
   # Get $HOME value of the container,
   # change <container-name> to your container
   docker container exec <container-name> bash -c 'echo $HOME'

2. If there are no files in JupyterLab after you add them to the materials/ directory, check the path to the container's $HOME directory in the docker-compose.yaml file under volumes using the steps above

Attributions

The JupyterLab images are based on UBC images from:

• ubcdsci
• ubcstat