zuffy 0.5

Zuffy is a sklearn compatible open source python library that uses GP to induce Fuzzy Pattern Trees

Zuffy - Fuzzy Pattern Trees with Genetic Programming

An open source scikit-learn compatible library for introducing FPTs as an Explainability Tool

---

Zuffy is a python package for exploring Fuzzy Pattern Trees and it is compatible with scikit-learn.

It aims to provide a simple set of tools for the exploration of FPTs that are induced using genetic programming (GP) techniques. The GP functionality is provided by the gplearn library and the author is grateful to its creator, Trevor Stephens, for his work on this.

Refer to the documentation for further information.

Setup

Zuffy has been developed and tested with Python 3.11 and these library versions:

Library	Version
sklearn	1.5.2*
numpy	1.26.4
pandas	2.2.1
matplotlib	3.9.2
gplearn	0.4.2

Note that Scikit-learn version 1.6+ modified the API around its "tags" and, until the authors update all their estimators, Zuffy will not run with version 1.6+. The gplearn library, upon which Zuffy has a dependency, has not been updated since 2023.

To display the FPT you will need to install graphviz.

Zuffy Installation

Use pip:

pip install zuffy

or clone the repository:

git clone https://github.com/pxom/zuffy.git

and install the required dependencies:

pip install -r requirements.txt

Resources

• Documentation <https://zuffy.readthedocs.io/en/latest/?badge=latest>
• Source Code <https://github.com/zuffy-dev/zuffy/>
• Installation <https://github.com/zuffy-dev/zuffy#installation>

Examples

A collection of examples illustrating Zuffy features are available here.

How to cite Zuffy

Authors of scientific papers including results generated using Zuffy are asked to cite the following paper:

@article{ZUFFY_1, 
    author    = "PXOM",
    title     = { {Zuffy}: Induction of Fuzzy Pattern Trees through Genetic Programming },
    pages    = { 0--0 },
    volume    = { 1 },
    month     = { Jul },
    year      = { 2025 }
}