Dazbo's Customised Miniconda

Tagline

A Python3 Miniconda image with Conda dependencies and a minimal set of data science packages.

GitHub

https://github.com/derailed-dash/dazboconda

Overview

This is a modified Docker container image, based on continuumio/miniconda3. It's default action is to run Jupyter Lab.

• It pre-installs a minimal set of data science packages, including:
  • Matplotlib
  • Numpy
  • Pandas
  • Networkx
  • ImageIO
  • Pillow
  • Regex
  • Scipy
  • tqdm
• A docker-compose.yml for starting and serving your site through the container; it automatically sets up environment variables and volumes.

Usage Instructions

Running the Container

Run this from where we want to store our Jupyter Notebooks on the host.

Run the container:

# To launch JupyterLab
docker run -e "JUPYTER_TOKEN=docker" --rm -it -v "${PWD}:/opt/notebooks" -p 8888:8888 daz502/dazboconda:latest

# To run a shell
docker run -e "JUPYTER_TOKEN=docker" --rm -it -v "${PWD}:/opt/notebooks" -p 8888:8888 daz502/dazboconda:latest sh

Launching JupyterLab with Docker Compose

Simply launch the container as follows:

docker compose up

Rebuilding the Image

If the image is ever updated, we'll want to update the tag and push to docker.io. E.g.

# build
docker image build -t daz502/dazboconda:tagname .

# push
docker push daz502/dazboconda:tagname