slmg/jupyter repository overview

⁠docker-jupyter

A minimal Jupyter Notebook image. Useful for quick testing.

This image ships Jupytext⁠ in addition to a bare Jupyter installation.

If you are after a more, Anaconda-like, full-featured installation, you may prefer to use one of the jupyter/docker-stacks⁠ image instead.

⁠Usage

Once you've spun Jupyter using one of the below methods, hit http://localhost:8888⁠ in your browser.

Tips:

• Use the argument --rm to automatically remove your container after it stops.
• If you want to expose you server to the network, so it can be accessed from another machine, use -p 8888:8888 instead of -p 127.0.0.1:8888:8888.
• If you want to save your notebooks after your container stops, use the argument -v <local path>:/home/moby/notebooks. The relevant directory will be mounted inside your container at /home/moby/notebooks, where Jupyter is configured to opperate.

Example using all the above

docker container run --rm -v $(pwd):/home/moby/notebooks -p 8888:8888 slmg/jupyter

⁠Detached mode

Keep Jupyter running after your shell exits.

docker container run -d --name jupyter -p 127.0.0.1:8888:8888 slmg/jupyter

You are going to need the auto-generated token to log-in. Fetch it from the logs as soon as the above command is run:

docker container logs --tail 1 jupyter

⁠Interactive mode

Customize your environment before starting the server. Ex install a Python lib, a notebook extension, modify Jupyter config, etc ...

docker container run -it --name jupyter -p 127.0.0.1:8888:8888 slmg/jupyter bash

Once inside the container, simply pip install what you want. Then run jupyter notebook to launch the server.

Note: it is also possible to install packages from Jupyter's web UI. Either run !pip install from an interactive cell, or spin up a terminal connected to your container from Jupyter's main page.