Python Jupyter devcontainer

Provides a devcontainer running Python 3.9 and Juptyerlab. Additional Python packages are managed using Poetry.

Requirements

The following software packages are required:

VS Code (see https://code.visualstudio.com/download)
Docker Desktop (see https://docs.docker.com/get-docker)

Within VS Code you will also need to install the following extensions locally:

Remote-Development
Docker

Setup

Start Docker Desktop (if not already running)

Clone the repository to your local machine

cd ~/repositories
git clone git@github.com:johnbullnz/python-jupyter-devcontainer.git

Open the folder in VS Code
Reopen the folder in a container. VS Code will automatically recognise that the folder contains a devcontainer configuration and will prompt you to reopen the folder in a container. Alternatively, open the Command Palette (Ctrl+Shift+P, F1) and select "Remote-Containers: Reopen in Container".

It will take a several minutes to create the devcontainer the first time you select "Reopen in Container".

Usage

Open an integrated terminal window. If the .venv virtual environment is not enabled, restart the terminal window to enable it.

Jupyterlab

Start Jupyterlab by running the following command in a terminal window:

./launch.sh

The URL of the Jupyterlab session will be printed on the screen. Click on the link to open Jupyterlab in a web browser.

Note that VS Code might map the devcontainer port to a different port on your local machine so copying the link may not work.

Working directory

Notebooks and local data files should be kept in the ./working directory.

Helper modules

Helper functions and class can be added to the helpers package (the ./helpers directory) and accessed within a notebook using import helpers.

Installing additional Python packages

The following Python packages are installed by default:

Pandas
Numpy
Matplotlib

Additional Python packages can be installed using Poetry in a VS Code integrated terminal window. For example:

poetry add scipy

Poetry will modify the pyproject.toml and poetry.lock files meaning that any changes will be applied should the devcontainer be rebuilt.

Data persistance

Only files stored in /workspaces/python-jupyter-devcontainer will persist on your local hard drive. Any other changes made to the devcontainer filesystem will be lost when the container is rebuilt.