Package edu.cmu.tetrad.bayes

Class JunctionTreeAlgorithm

java.lang.Object

edu.cmu.tetrad.bayes.JunctionTreeAlgorithm

All Implemented Interfaces:

TetradSerializable, Serializable

public class JunctionTreeAlgorithm
extends Object
implements TetradSerializable

Junction Tree Algorithm.

This implementation follows the Weka's implementation.

Nov 8, 2019 2:22:34 PM

Author:

Kevin V. Bui (kvb2@pitt.edu)

See Also:

• MarginCalculator.java
• Serialized Form

Constructor Summary

Constructors

Constructor	Description
JunctionTreeAlgorithm(BayesIm bayesIm)
JunctionTreeAlgorithm(Graph graph, DataModel dataModel)

Method Summary

Modifier and Type	Method	Description
double	getConditionalProbabilities(int[] nodes, int[] values, int[] parents, int[] parentValues)	Get the joint probability of the nodes given their parents.
double[]	getConditionalProbabilities(int iNode, int[] parents, int[] parentValues)	Get the conditional probability of a node for all of its values.
double	getConditionalProbability(int iNode, int value, int[] parents, int[] parentValues)
double	getJointProbability(int[] nodes, int[] values)
double	getJointProbabilityAll(int[] nodeValues)	Get the joint probability of all nodes (variables).
double[]	getMarginalProbability(int iNode)
double	getMarginalProbability(int iNode, int value)
List<Node>	getNodes()
int	getNumberOfNodes()
void	setEvidence(int iNode, int value)
String	toString()

Methods inherited from class java.lang.Object

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

Constructor Details

JunctionTreeAlgorithm

public JunctionTreeAlgorithm(Graph graph,
 DataModel dataModel)

JunctionTreeAlgorithm

public JunctionTreeAlgorithm(BayesIm bayesIm)

Method Details

setEvidence

public void setEvidence(int iNode,
 int value)

getConditionalProbabilities

public double getConditionalProbabilities(int[] nodes,
 int[] values,
 int[] parents,
 int[] parentValues)

Get the joint probability of the nodes given their parents. Example: given x <-- z --> y, we can find P(x,y|z). Another example: given x <-- z --> y <-- w, we can find P(x,y|z,w)

getConditionalProbabilities

public double[] getConditionalProbabilities(int iNode,
 int[] parents,
 int[] parentValues)

Get the conditional probability of a node for all of its values.

getConditionalProbability

public double getConditionalProbability(int iNode,
 int value,
 int[] parents,
 int[] parentValues)

getJointProbabilityAll

public double getJointProbabilityAll(int[] nodeValues)

Get the joint probability of all nodes (variables). Given the nodes are X1, X2,...,Xn, then nodeValues[0] = value(X1), nodeValues[1] = value(X2),...,nodeValues[n-1] = value(Xn).

Parameters:

nodeValues - an array of values for each node

getJointProbability

public double getJointProbability(int[] nodes,
 int[] values)

getMarginalProbability

public double[] getMarginalProbability(int iNode)

getMarginalProbability

public double getMarginalProbability(int iNode,
 int value)

getNodes

public List<Node> getNodes()

getNumberOfNodes

public int getNumberOfNodes()

toString

public String toString()

Overrides:

toString in class Object