tree-praxis 0.0.14

Rashomon set tools via PRAXIS, including approximation, exact calculation, and variable importance utilities.

PRAXIS: Fast computation of near-optimal trees and the Rashomon set

PRAXIS is a high-performance C++/Python library for decision tree optimization. It includes (i) single decision tree optimization (both heuristic and optimal), (ii) Rashomon set computation (both approximate and optimal), and (iii) variable importance computation over all good models via the Rashomon Importance Distribution (both approximate and exact).

Installation

Building from Source

If you have access to C++ build tools, you can install directly from GitHub:

pip install "git+https://github.com/zakk-h/LicketyRESPLIT.git@praxis"

Prebuilt Binary Wheels

To make installation simpler, LicketyRESPLIT provides prebuilt binary wheels for all major platforms:

• Windows (64-bit)
• macOS (Intel and Apple Silicon)
• Linux (any distribution)

Supported Python versions: 3.9 – 3.13

Installation Steps

1. Go to the release page: https://github.com/zakk-h/LicketyRESPLIT/releases/tag/praxis-v0.0.8

2. Choose the wheel that matches:

   • Your Python version (cp39, cp310, cp311, cp312, cp313)
   • Your operating system

3. Install it with pip (replace the URL with your chosen wheel file from step 2):

macOS (Apple Silicon), Python 3.11:

pip install "https://github.com/zakk-h/LicketyRESPLIT/releases/download/praxis-v0.0.8/praxis-0.0.8-cp311-cp311-macosx_11_0_arm64.whl"

Example

See examples/example.ipynb for a complete walkthrough of using PRAXIS.