User Guide

• 1. Constraint-based causal discovery
  • 1.1. Fundamental Assumptions of Constraint-Based Causal Discovery
  • 1.2. (Non-parametric) Markovian SCMs with Observational Data
  • 1.3. (Non-parametric) Semi-Markovian SCMs with Observational Data
  • 1.4. (Non-parametric) SCMs with Interventional Data
    • 1.4.1. dodiscover.constraint.PsiFCI
  • 1.5. Choosing the conditioning sets
    • 1.5.1. dodiscover.constraint.ConditioningSetSelection
  • 1.6. Hyperparameters and controlling overfitting
  • 1.7. Robust learning
• 2. Independence
  • 2.1. Conditional Mutual Information
    • 2.1.1. dodiscover.ci.CMITest
    • 2.1.2. dodiscover.ci.ClassifierCMITest
  • 2.2. Partial (Pearson) Correlation
    • 2.2.1. dodiscover.ci.FisherZCITest
  • 2.3. Discrete, Categorical and Binary Data
    • 2.3.1. dodiscover.ci.GSquareCITest
  • 2.4. Kernel-Approaches
    • 2.4.1. dodiscover.ci.KernelCITest
  • 2.5. Classifier-based Approaches
    • 2.5.1. dodiscover.ci.ClassifierCITest
• 3. Conditional Discrepancy
  • 3.1. Discrete, Categorical and Binary Data
  • 3.2. Kernel-Approaches
    • 3.2.1. dodiscover.cd.KernelCDTest
  • 3.3. Bregman-Divergences
    • 3.3.1. dodiscover.cd.BregmanCDTest