py-bbn

A Python library for exact associational, interventional, and counterfactual reasoning using Bayesian Belief Networks (BBNs) and Structural Causal Models (SCMs).

Current release: 4.0.0.

The package includes exact marginal, joint, conditional, interventional, and counterfactual query APIs, array-backed and pandas-friendly result modes, and model serialization.

Installation

pip install pybbn

License

py-bbn is released under the Apache-2.0 license.

Contents

• 1. Quickstart
  • 1.1. Creating a model
  • 1.2. Associational query
  • 1.3. Interventional query
  • 1.4. Counterfactual query
  • 1.5. Graphical query
  • 1.6. Data sampling
  • 1.7. Serde
• 2. Huang Example
  • 2.1. Model
  • 2.2. Graphs
  • 2.3. Queries
  • 2.4. Save
• 3. BBN Generation
  • 3.1. Singly-connected BBN
  • 3.2. Multi-connected BBN
• 4. Benchmarks
  • 4.1. Methodology
  • 4.2. bnlearn Results
  • 4.3. pyAgrum Results
  • 4.4. Interpretation
• 5. Bibliography

API Documentation

• py-bbn
  • Common
  • Potential
  • Factory
  • Parameter
  • Reasoning
  • Associational Query
  • Interventional Query
  • Counterfactual Query
  • Graphical Query
  • Sampling
  • Generator
  • Serde
  • External

Indices and tables

• Index

• Module Index

• Search Page

Copyright

Software

Copyright |copyright_years| Rocket Vector

Art

Copyright 2020 Daytchia Vang

Citation

@misc{rocketvector_pybbn_2024,
title={PyBBN},
url={https://pybbn.rocketvector.io},
author={Vang, Jee},
year={2024},
month={Jan}}