Overview

This is a template for generating a reproducible Python environment for data science within a docker environment. The current base setup is like the following:

Package	Usage
docker	OS environment
poetry	Python package manager
jupyter notebook	Development environment

Steps for environment creation

0. Create environment files (e.g. copy examples)

cp env/.jupyter.env.example env/.jupyter.env

1. Build docker container

sudo docker-compose build .

2.1 Start python within docker container

sudo docker-compose run code python

2.2 Start jupyter notebook

sudo docker-compose up --build

Add a python package

1. Use poetry to add package and resolve dependencies

sudo docker-compose run code poetry add <package>

2. Commit changes to git

git add code/poetry.lock code/pyproject.toml