CausalDiscoveryBase<T>

Base Classes · AiDotNet.CausalDiscovery

Abstract base class for causal discovery algorithms with shared statistical utilities.

Implements: ICausalDiscoveryAlgorithm<T>

Known derived types: BayesianCausalBase<T>, ConstraintBasedBase<T>, ContinuousOptimizationBase<T>, DeepCausalBase<T>, FunctionalBase<T>, GOBNILPAlgorithm<T>, HybridBase<T>, InfoTheoreticBase<T>, ScoreBasedBase<T>, TimeSeriesCausalBase<T>

For Beginners

This base class contains the “toolbox” of statistical tests and helper methods that all causal discovery algorithms share. Each specific algorithm (like NOTEARS or PC) extends this class and adds its own unique logic.

How It Works

Provides common functionality needed by most causal discovery algorithms: conditional independence testing, BIC scoring, partial correlation computation, covariance estimation, and OLS regression.

Properties

Property Summary
CausalDiscoveryCategory Category { get; }
bool SupportsLatentConfounders { get; }
bool SupportsMixedData { get; }
bool SupportsNonlinear { get; }
bool SupportsTimeSeries { get; }
string Name { get; }

Methods

Method Summary
CausalGraph<T> DiscoverStructure(Matrix<T> data, Vector<T> target, string[] featureNames)
CausalGraph<T> DiscoverStructure(Matrix<T> data, string[] featureNames)