Welcome to the documentation!

This library provides a comprehensive suite of tools for performing conditional independence testing. This documentation serves as a practical user guide and a research reference for the underlying statistical theory.

User Guides

• Installation
  • Quick Install
  • Development Setup (uv)
  • Optional LightGBM Extra
  • Optional Tigramite Extra
  • Local External Wrappers
  • Optional R-Based Setup
• Quickstart
  • Example: Running PC with a CITK Test
• How to Choose a Conditional Independence Test
  • Key Considerations
  • Summary Table
• Interpreting Test Results
  • The P-Value
  • Accessing the Results

Theoretical Foundations

• What is Conditional Independence?
  • Formal Definition
  • Intuitive Explanation
  • Why It Matters in Causal Discovery
• A Taxonomy of Conditional Independence Tests
  • 1. Correlation-Based Tests
  • 2. Contingency Table-Based Tests
  • 3. Regression-Based Tests
  • 4. Kernel-Based Tests
  • 5. Machine Learning-Based Tests

Conditional Independence Tests

• Conditional Independence Test Reference
  • Chi-Squared Test
  • G-Squared Test
  • Fisher’s Z Test
  • Spearman’s Rho Test
  • Kernel Conditional Independence (KCI) Test
  • Random Forest CI Test
  • Double Machine Learning (DML) CI Test
  • Conformalized Residual Independence Test (CRIT)
  • E-Value Double Machine Learning (EDML) CI Test

API Reference

• API Reference
  • Simple CI Tests
  • Regression-Based Tests
  • Nearest Neighbor Tests
  • Kernel Tests
  • Machine Learning-Based Tests
  • Adapter Tests

Development

• Contributing
• Changelog