Python Data Science with Jupyter on Docker

Inspired by article “How to set up and run Python Data Science Development Environment with Jupyter on Docker” [Archived]

Data Scientists that write or use code must work with a variety of Data Science development tools, including the  Anaconda (or Miniconda) Python distribution platform, the Python programming language, the R programming language, Matplotlib library, Pandas Library, the Jupyter application, and many others.

Installing a version of Python on one’s computer can be challenging (see Understanding Python installation mess) and while it is possible to even install both Anaconda and Python separately, using a Docker method keeps things separate and even more modular that “environments.”

The article describes how to use pre-made  Docker images to run Jupyter Labs. the Jupyter Docker Stacks simplifies how to create a ready-to-use Jupyter application with Data Science libraries in a few commands. The article contains the commands to install and run the necessary software and contains an embedded ~9min video demonstration.

Table of Contents of article

• Introduction
• Introduction to Jupyter Docker Stacks
• Running the Jupyter Docker Scipy-Notebook Image
• How to use other Python Libraries
• What if I use Eikon Data API
• Demo prerequisite
• How to run the Examples
• Conclusion
• References
• GitHub

Here is a screen capture of the web browser that I started with command:

docker run -p 8888:8888 --name notebook -v "${PWD}":/home/jovyan/work -e JUPYTER_ENABLE_LAB=yes  -it jupyter/scipy-notebook

There are more options with the jupyter/datascience-notebook:

docker run -p 8888:8888 --name notebook -v "${PWD}":/home/jovyan/work -e JUPYTER_ENABLE_LAB=yes  -it jupyter/datascience-notebook

Quoting from the article:

Jupyter Docker Stacks provide various images for developers based on their requirements such as:

• jupyter/scipy-notebook: Jupyter Notebook/JupyterLab with conda/mamba , ipywidgets, and popular packages from the scientific Python ecosystem (Pandas, Matplotlib, Seaborn, Requests, etc.)
• jupyter/r-notebook: Jupyter Notebook/JupyterLab with R interpreter, IRKernel, and devtools.
• jupyter/datascience-notebook: Everything in jupyter/scipy-notebook and jupyter/r-notebook images with Julia support.
• jupyter/tensorflow-notebook: Everything in jupyter/scipy-notebook image with TensorFlow.

See more detail about all image types on Selecting a Docker Image page.

Short demonstration

The following short 8minute video shows how it works:

 

This other article has more details on the working of Docker:

How to Run Jupyter Notebook on Docker – [Archived]
No more Python env and package update

---

Top illustration background: Felipe on Pixabay