zuffy 0.2

Zuffy is a sklearn compatible open source python library for explainable machine learning through Fuzzy Pattern Trees

Zuffy - Fuzzy Pattern Trees with Genetic Programming

A Scikit-learn compatible Open Source library for introducing FPTs as an Explainability Tool

---

(NOTE THAT THIS PROJECT IS UNDER DEVELOPMENT AND LIKELY TO CHANGE SIGNIFICANTLY UNTIL THE FIRST RELEASE. USE AT YOUR OWN RISK.)

Zuffy is an open source python library for explainable machine learning models. It is compatible with scikit-learn.

It aims to provide a simple set of tools for the exploration of FPTs that are inferred using genetic programming techniques.

Refer to the documentation for further information.

Setup

It may work with other versions but Zuffy has been tested with Python 3.11.9 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+.

To display the FPT you will need to install graphviz:

Unix

sudo apt install graphviz

$ sudo apt install graphviz

Windows

???

Installation

Clone the repository:

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

To see more elaborate examples, look here.

import pandas as pd
from sklearn.datasets import load_iris
from zuffy import ZuffyClassifier, functions, visuals
from zuffy.wrapper import ZuffyFitIterator

iris = load_iris()
dataset = pd.DataFrame(data=iris.data, columns=iris.feature_names)
dataset['target'] = iris.target
targetNames = iris.target_names
X = dataset.iloc[:,0:-1]
y = dataset.iloc[:,-1]

fuzzy_X, fuzzy_features_names = functions.fuzzify_data(X)

zuffy = ZuffyClassifier(generations=15, verbose=1)
res = ZuffyFitIterator(zuffy, fuzzy_X, y, n_iter=3, split_at=0.25)

visuals.plot_evolution(
    res.getBestEstimator(),
    targetNames,
    res.getPerformance(),
    outputFilename='sample1_analysis')

visuals.graphviz_tree(
    res.getBestEstimator(),
    targetNames,
    featureNames=fuzzy_features_names,
    treeName=f"Iris Dataset (best accuracy: {res.getBestScore():.3f})",
    outputFilename='sample1_fpt')

* TBD *

In an sklearn Pipeline <https://scikit-learn.org/stable/modules/generated/sklearn.pipeline.Pipeline.html>_:

    from sklearn.pipeline import Pipeline
    from sklearn.preprocessing import StandardScaler

    pipe = Pipeline([
        ('scale', StandardScaler()),
        ('net', net),
    ])

    pipe.fit(X, y)
    y_proba = pipe.predict_proba(X)

How to cite Zuffy

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

@article{ZUFFY_1, 
    author    = "POM",
    title     = { {Zuffy}: Open Source inference of FPT using GP },
    pages    = { 0--0 },
    volume    = { 1 },
    month     = { Apr },
    year      = { 2025 },
    journal   = { Journal of Unknown }
}