Docker packages an app with everything it needs into a container. You can run PostgreSQL, Jupyter or a colleague’s script without installing anything on your PC and without breaking your other software.

Image vs container

• Image — the recipe (read-only), e.g. postgres:11.
• Container — a running copy of an image. Stop it, delete it, start a fresh one in seconds.

First commands

docker run --rm hello-world                     # test the install
docker run -d --name db -e POSTGRES_PASSWORD=secret -p 5432:5432 postgres:11
docker ps                                       # what is running
docker logs db                                  # its output
docker stop db && docker rm db                  # clean up

You now have a real PostgreSQL server for practising SQL, removable in one line.

Jupyter notebooks in one command

docker run -p 8888:8888 -v "%cd%":/home/jovyan/work jupyter/scipy-notebook

-v shares your current folder with the container, so notebooks are saved on your PC.

Your own script: a Dockerfile

FROM python:3.7-slim
WORKDIR /app
COPY requirements.txt .
RUN pip install -r requirements.txt
COPY report.py .
CMD ["python", "report.py"]

docker build -t monthly-report .
docker run --rm monthly-report