markusand / jupyterlab-data-science

Mount a Docker container with essential python packages ready for data science and data visualization

Data Science Jupyter Lab

Mount a Docker container with essential python packages ready for data science and data visualization.

• All packages in @jupyter/scipy-notebook: numpy, pandas, scipy, matplotlib, sqlalchemy, ...
• Geospatial packages: geopandas, shapely, geopy, geoplot
• Data viz: ipyleaflet, plotly

Install

git clone https://github.com/snowplanner/jupyterlab-data-science.git

In case you need another package that is not included in the container, just add it to the requirements.txt file and launch the following command on terminal to rebuild the image.

docker compose build

Usage

Create a .env file, even if empty.

After launching Docker, run the following command in terminal from the project directory root.

# Use -d flag for detached mode
docker compose up -d

# Access inside shell with
docker compose exec jupyter-lab sh

# Shutdown the container
docker compose down

Jupyter Lab is accessible on your browser at http://localhost:8888.

Alternatively, you can connect to the Docker container with VS Code. Despite this method adds some performance overhead, you'll be able to use the toolbox of your choice (linters, formatters, tests, etc.).

Contribute

⚠️ KEEP THIS REPO TIDY.
Use folders and subfolders to structure your notebooks and modules. Use appropriate semantic names following python naming conventions.

There is some basic sample code in the folder examples to help you get started.

Use conventional commits whenever possible; that means, always.

Always use a feature branch to develop your code and avoid future merge conflicts. If you'd like to integrate your code to the basecode, issue a new pull request to the main branch.

A code review process is a must to ensure standard compliant code.

🚫 Never ever EVER commit sensitive data or secrets.
Use environment variables to store tokens, passwords and other information that must be kept safe. The Docker container automatically embeds all environment variables from a .env file (not trackable with git) in the root of the repo. Access variables as follows:

import os

# Considering .env file to contain the API_KEY variable
# API_KEY=pk.XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
api_key = os.environ['API_KEY']