Dockerized 🐳 implementation of a TabPy server.

TabPy is a Python server that lets you execute Python code on the fly and display results in Tableau visualizations. For more information, visit the official documentation or the official GitHub repository.

We use poetry to manage the Python dependencies in conjunction with docker-compose to manage the Docker services.

Execute the following command to build and run the Docker services:

docker compose down && docker compose up -d --build

• docker compose down stops and removes the Docker services.
• docker compose up -d --build builds and runs the Docker services in the background, ensuring it runs on system startup.

The TabPy container is exposed on port 8001 at http://localhost:8001/. Proxying to a web server is recommended.

The TabPy password management system is enabled by default (this is necessary for Tableau Online and Tableau Server).

Manage TabPy user accounts using the tabpy-user command-line utility.

Configure TabPy to use the passwords file for authentication. Create a configuration file (tabpy_auth.conf):

[TabPy]
TABPY_PWD_FILE = <path_to_password_file>

tabpy-user add -u <username> -f <path_to_password_file> -p <password>

tabpy-user update -u <username> -f <path_to_password_file> -p <new_password>

Manually delete the user's line in the passwords file.

1. Go to Help > Settings and Performance > Manage External Service Connection.
2. Set connection type to TabPy/External API.
3. Enter Server (host) and Port.
4. Check Sign with username and password and enter credentials.
5. Click Test Connection.

TODO

TabPy is not supported in Tableau Public.

TODO

TODO

Several scripts are available to test out the TabPy server:

Implementation of a live, and fairly performant, graph visualization using the networkx library. The script generates the x, y, and weight attributes for each node and edge in the graph.

This script, meant for Tableau Prep, takes the Week # column and explodes it into a range of weeks. For example, one row with 1-3 will be exploded into three rows: 1, 2, and 3 & c.

Poor-mans implementation of a radar chart using edge bundling. The script generates the x, y, and weight attributes for each node and edge in the graph.