Link Search Menu Expand Document
PyXAI
Papers Video GitHub In-the-Loop EXPEKCTATION Release Notes About

Modules

PyXAI exposes three top-level modules, each covering a distinct phase of the workflow:

Builder

The Builder module lets you construct tree-based models by hand using DecisionNode, LeafNode, DecisionTree, RandomForest, and BoostedTrees objects. This is useful for testing and verifying explanations on controlled examples without needing a dataset.

Learning

The Learning module wraps Scikit-learn, XGBoost, and LightGBM to handle the full experimental protocol: training, cross-validation, instance selection, preprocessing, and model import/export via ModelIO.

Explaining

The Explaining module provides methods for computing explanations on tree-based ML models. Its initialize function returns the appropriate Explainer subclass (ExplainerDT, ExplainerRF, ExplainerBT) depending on the model type.


Table of contents