Package edu.cmu.tetrad.search.utils

Class EnsureMarkov

java.lang.Object

edu.cmu.tetrad.search.utils.EnsureMarkov

public class EnsureMarkov
extends Object

A helper class to encapsulate logic for ensuring a Markov property for subsequent testing after an initial local Markov graph has been found.

Version:

$Id: $Id

Author:

josephramsey

See Also:

• GraphUtils

Constructor Summary

Constructors

Constructor	Description
EnsureMarkov(Graph dag, IndependenceTest test)	Constructs an EnsureMarkov class for a given Markov dag.

Method Summary

Modifier and Type	Method	Description
static Map<org.apache.commons.lang3.tuple.Pair<Node,Node>,Set<Double>>	localMarkovAdjustPValues(Graph cpdag, boolean ensureMarkov, IndependenceTest test, Map<org.apache.commons.lang3.tuple.Pair<Node,Node>,Set<Double>> pValues, org.apache.commons.lang3.tuple.Pair<Node,Node> withoutPair)	Adjusts the p-values for a local Markov condition in a given constraint-based partially directed acyclic graph (CPDAG).
boolean	markovIndependence(Node x, Node y, Set<Node> z)	Checks the independence of two nodes given a set of conditioning nodes, and if Markov is to be preserved, checks to make sure the additional independence does not generate p-values that violate the Markov property.
void	setEnsureMarkov(boolean ensureMarkov)	Sets whether to ensure Markov property.

Methods inherited from class java.lang.Object

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

Constructor Details

EnsureMarkov

public EnsureMarkov(Graph dag,
 IndependenceTest test)

Constructs an EnsureMarkov class for a given Markov dag.

Parameters:

dag - The initial Markov dag. This dag should pass a local Markov check.

test - The independence test to use.

Method Details

localMarkovAdjustPValues

public static Map<org.apache.commons.lang3.tuple.Pair<Node,Node>,Set<Double>> localMarkovAdjustPValues(Graph cpdag,
 boolean ensureMarkov,
 IndependenceTest test,
 Map<org.apache.commons.lang3.tuple.Pair<Node,Node>,Set<Double>> pValues,
 org.apache.commons.lang3.tuple.Pair<Node,Node> withoutPair)
                                                                                                throws InterruptedException

Adjusts the p-values for a local Markov condition in a given constraint-based partially directed acyclic graph (CPDAG).

Parameters:

cpdag - the constraint-based partially directed acyclic graph (CPDAG) to adjust p-values for

ensureMarkov - a boolean flag indicating if the Markov condition should be ensured; should be true

test - the independence test to be used; must not be null and not an instance of MsepTest

pValues - a map of node pairs to sets of p-values used for adjustment

withoutPair - a pair of nodes for which adjustments are calculated without considering the edge between them

Returns:

a map of node pairs to sets of adjusted p-values

Throws:

IllegalArgumentException - if ensureMarkov is false or if the test is null or an instance of MsepTest

InterruptedException

setEnsureMarkov

public void setEnsureMarkov(boolean ensureMarkov)
                     throws InterruptedException

Sets whether to ensure Markov property. By default, false; this must be turned on.

Parameters:

ensureMarkov - True if the Markov property should be ensured.

Throws:

InterruptedException

markovIndependence

public boolean markovIndependence(Node x,
 Node y,
 Set<Node> z)
                           throws InterruptedException

Checks the independence of two nodes given a set of conditioning nodes, and if Markov is to be preserved, checks to make sure the additional independence does not generate p-values that violate the Markov property. Returns the independence-result, marked as 'true' if the nodes are independent or if the Markov property is preserved and the nodes are independent and the Markov property wouldn't be violated by the additional independence.

Parameters:

x - The first node.

y - The second node.

z - The set of conditioning nodes.

Returns:

True if the nodes are independent.

Throws:

InterruptedException