Package com.bayesserver.causal

Class BackdoorInference

java.lang.Object

com.bayesserver.causal.CausalInferenceBase

com.bayesserver.causal.BackdoorInference

All Implemented Interfaces:

Inference

public final class BackdoorInference
extends CausalInferenceBase

Estimates the causal effect, using the 'Backdoor Adjustment' formula to avoid confounding bias. For treatments (X) and outcomes (Y) this is denoted P(Y|Do(X)). Either the Total or Direct effect can be calculated. Multiple treatments, outcomes and unobserved nodes are supported as well as fixed evidence.

Constructor Summary

Constructors

Constructor	Description
BackdoorInference(Network network)	Initializes a new instance of the BackdoorInference class.

Method Summary

Modifier and Type	Method	Description
protected void	queryDistributionsInner(QueryOptions queryOptions, QueryOutput queryOutput)

Methods inherited from class com.bayesserver.causal.CausalInferenceBase

getBaseEvidence, getEvidence, getNetwork, getQueryDistributions, getQueryFunctions, getQueryLifecycle, query, setBaseEvidence, setEvidence, setNetwork, setQueryDistributions, setQueryFunctions, setQueryLifecycle

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

BackdoorInference

public BackdoorInference(Network network)

Initializes a new instance of the BackdoorInference class.

Parameters:

network - The Bayesian network.

Method Detail

queryDistributionsInner

protected void queryDistributionsInner(QueryOptions queryOptions,
                                       QueryOutput queryOutput)
                                throws InconsistentEvidenceException

Specified by:

queryDistributionsInner in class CausalInferenceBase

Throws:

InconsistentEvidenceException