Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance. Project price only 1 $
You can buy this project and download/modify it how often you want.
///////////////////////////////////////////////////////////////////////////////
// For information as to what this class does, see the Javadoc, below. //
// Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, //
// 2007, 2008, 2009, 2010, 2014, 2015, 2022 by Peter Spirtes, Richard //
// Scheines, Joseph Ramsey, and Clark Glymour. //
// //
// This program is free software; you can redistribute it and/or modify //
// it under the terms of the GNU General Public License as published by //
// the Free Software Foundation; either version 2 of the License, or //
// (at your option) any later version. //
// //
// This program is distributed in the hope that it will be useful, //
// but WITHOUT ANY WARRANTY; without even the implied warranty of //
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the //
// GNU General Public License for more details. //
// //
// You should have received a copy of the GNU General Public License //
// along with this program; if not, write to the Free Software //
// Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA //
///////////////////////////////////////////////////////////////////////////////
package edu.cmu.tetrad.bayes;
import edu.cmu.tetrad.graph.Node;
import edu.cmu.tetrad.util.TetradSerializable;
import java.io.IOException;
import java.io.ObjectInputStream;
import java.util.*;
/**
* Calculates marginals of the form P(V=v') for an updated Bayes net for purposes of the CPT Invariant Updater.
*
* @author josephramsey
*/
public final class CptInvariantMarginalCalculator
implements TetradSerializable {
private static final long serialVersionUID = 23L;
/**
* @serial Cannot be null.
*/
private final BayesIm bayesIm;
/**
* @serial Cannot be null.
*/
private final Evidence evidence;
/**
* @serial Cannot be null.
*/
private final UpdatedBayesIm updatedBayesIm;
/**
* @serial Cannot be null.
*/
private double[][] storedMarginals;
//=============================CONSTRUCTORS==========================//
/**
* Constructs a new marginal calculator for the given updated Bayes IM. It is assumed that the first BayesIm
* encountered on calling the getParentIm() method recursively is the Bayes IM with respect to which conjunctions of
* the form P(V1=v1' and V2=v2' and ... and Vn=vn') should be calculated.
*/
public CptInvariantMarginalCalculator(BayesIm bayesIm, Evidence evidence) {
if (bayesIm == null) {
throw new NullPointerException();
}
if (evidence == null) {
throw new NullPointerException();
}
if (evidence.isIncompatibleWith(bayesIm)) {
throw new IllegalArgumentException("The variables for the given " +
"Bayes IM and evidence must be compatible.");
}
this.bayesIm = bayesIm;
this.evidence = evidence;
this.updatedBayesIm = new UpdatedBayesIm(bayesIm, evidence);
this.storedMarginals = initStoredMarginals();
}
/**
* Generates a simple exemplar of this class to test serialization.
*/
public static CptInvariantMarginalCalculator serializableInstance() {
MlBayesIm bayesIm = MlBayesIm.serializableInstance();
Evidence evidence = Evidence.tautology(bayesIm);
return new CptInvariantMarginalCalculator(bayesIm, evidence);
}
//=============================PUBLIC METHODS========================//
/**
* @return P(variable = category).
*/
public double getMarginal(int variable, int category) {
if (this.storedMarginals[variable][category] != -99.0) {
return this.storedMarginals[variable][category];
}
double marginal = 0.0;
boolean foundANumber = false;
for (int row = 0; row < this.bayesIm.getNumRows(variable); row++) {
double probability =
this.updatedBayesIm.getProbability(variable, row, category);
if (Double.isNaN(probability)) {
continue;
}
double probabilityOfRow = getProbabilityOfRow(variable, row);
if (Double.isNaN(probabilityOfRow)) {
continue;
}
marginal += probability * probabilityOfRow;
foundANumber = true;
}
if (!foundANumber) {
marginal = Double.NaN;
}
this.storedMarginals[variable][category] = marginal;
return marginal;
}
public UpdatedBayesIm getUpdatedBayesIm() {
return this.updatedBayesIm;
}
private double[][] initStoredMarginals() {
this.storedMarginals = new double[this.bayesIm.getNumNodes()][];
for (int i = 0; i < this.bayesIm.getNumNodes(); i++) {
this.storedMarginals[i] = new double[this.bayesIm.getNumColumns(i)];
Arrays.fill(this.storedMarginals[i], -99.0);
}
return this.storedMarginals;
}
private double getProbabilityOfRow(int variable, int row) {
int[] parents = this.bayesIm.getParents(variable);
int[] parentValues = this.bayesIm.getParentValues(variable, row);
double probabilityOfRow = 1.0;
for (int index = 0; index < parents.length; index++) {
if (noModifiedCpts(parents, index)) {
double marginal =
getMarginal(parents[index], parentValues[index]);
if (!Double.isNaN(marginal)) {
probabilityOfRow *= marginal;
}
} else {
Evidence evidence = new Evidence(this.evidence);
CptInvariantMarginalCalculator marginals =
new CptInvariantMarginalCalculator(this.bayesIm, evidence);
double marginal = marginals.getMarginal(parents[index],
parentValues[index]);
if (!Double.isNaN(marginal)) {
probabilityOfRow *= marginal;
}
}
}
return probabilityOfRow;
}
/**
* @return true iff conditioning on parents 0 through i - 1 would change any of the conditional probability tables
* relevant to calculating P(node i = value i).
*/
private boolean noModifiedCpts(int[] parents, int i) {
List target =
Collections.singletonList(this.bayesIm.getNode(parents[i]));
List conditioners = new LinkedList<>();
for (int j = 0; j < i; j++) {
conditioners.add(this.bayesIm.getNode(parents[j]));
}
List condAncestors = this.bayesIm.getDag().paths().getAncestors(conditioners);
List targetAncestor = this.bayesIm.getDag().paths().getAncestors(target);
Set intersection = new HashSet<>(condAncestors);
intersection.retainAll(targetAncestor);
return intersection.isEmpty();
}
/**
* Adds semantic checks to the default deserialization method. This method must have the standard signature for a
* readObject method, and the body of the method must begin with "s.defaultReadObject();". Other than that, any
* semantic checks can be specified and do not need to stay the same from version to version. A readObject method of
* this form may be added to any class, even if Tetrad sessions were previously saved out using a version of the
* class that didn't include it. (That's what the "s.defaultReadObject();" is for. See J. Bloch, Effective Java, for
* help.
*/
private void readObject(ObjectInputStream s)
throws IOException, ClassNotFoundException {
s.defaultReadObject();
if (this.bayesIm == null) {
throw new NullPointerException();
}
if (this.evidence == null) {
throw new NullPointerException();
}
if (this.storedMarginals == null) {
throw new NullPointerException();
}
if (this.updatedBayesIm == null) {
throw new NullPointerException();
}
}
}
