de.jstacs.scoringFunctions.directedGraphicalModels.structureLearning.measures.btMeasures
Class BTMutualInformation

java.lang.Object
  extended by de.jstacs.scoringFunctions.directedGraphicalModels.structureLearning.measures.Measure
      extended by de.jstacs.scoringFunctions.directedGraphicalModels.structureLearning.measures.btMeasures.BTMutualInformation
All Implemented Interfaces:
Storable, Cloneable

public class BTMutualInformation
extends Measure

Structure learning Measure that computes a maximum spanning tree based on mutual information and uses the resulting tree structure as structure of a Bayesian tree (special case of a Bayesian network) in a BayesianNetworkScoringFunction.

Author:
Jan Grau

Field Summary
static int BG
          Compute mutual information only from background data
static int BOTH
          Use both data sets to compute the mutual information
static int FG
          Compute mutual information only from foreground data
 
Constructor Summary
BTMutualInformation(int clazz, double[] ess)
          Creates a new mutual information Bayesian tree Measure.
BTMutualInformation(StringBuffer buf)
          Re-creates a BTMutualInformation from is XML-representation as returned by toXML().
 
Method Summary
 BTMutualInformation clone()
           
 String getInstanceName()
          Returns the name of the Measure and possibly some additional information about the current instance.
 int[][] getParents(Sample fg, Sample bg, double[] weightsFg, double[] weightsBg, int length)
          Returns the optimal parents for the given data and weights.
 StringBuffer toXML()
          This method returns an XML-representation of an instance of the implementing class.
 
Methods inherited from class de.jstacs.scoringFunctions.directedGraphicalModels.structureLearning.measures.Measure
fillTensor, fillTensor, getCMI, getCMI, getEAR, getEAR, getMI, getMI, getStatistics, getStatisticsOrderTwo, isShiftable, sum, toParents, union
 
Methods inherited from class java.lang.Object
equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

FG

public static final int FG
Compute mutual information only from foreground data

See Also:
Constant Field Values

BG

public static final int BG
Compute mutual information only from background data

See Also:
Constant Field Values

BOTH

public static final int BOTH
Use both data sets to compute the mutual information

See Also:
Constant Field Values
Constructor Detail

BTMutualInformation

public BTMutualInformation(StringBuffer buf)
                    throws NonParsableException
Re-creates a BTMutualInformation from is XML-representation as returned by toXML().

Parameters:
buf - the XML-representation
Throws:
NonParsableException - is thrown if the XML-code could not be parsed

BTMutualInformation

public BTMutualInformation(int clazz,
                           double[] ess)
                    throws Exception
Creates a new mutual information Bayesian tree Measure.

Parameters:
clazz - the classes used for computation of mutual information, one of FG, BG, BOTH
ess - the equivalent sample sizes for both classes
Throws:
Exception - thrown if clazz is not one of the allowed values
Method Detail

clone

public BTMutualInformation clone()
                          throws CloneNotSupportedException
Overrides:
clone in class Measure
Throws:
CloneNotSupportedException

getInstanceName

public String getInstanceName()
Description copied from class: Measure
Returns the name of the Measure and possibly some additional information about the current instance.

Specified by:
getInstanceName in class Measure
Returns:
the name

getParents

public int[][] getParents(Sample fg,
                          Sample bg,
                          double[] weightsFg,
                          double[] weightsBg,
                          int length)
                   throws Exception
Description copied from class: Measure
Returns the optimal parents for the given data and weights. The array of parents p at each position i is build as follows:

Specified by:
getParents in class Measure
Parameters:
fg - the data of the current (foreground) class
bg - the data of the negative (background) class
weightsFg - the weights for the sequences of fg
weightsBg - the weights for the sequences of bg
length - the length of the model, must be equal to the length of the sequences
Returns:
the parents
Throws:
Exception - thrown if the lengths do not match or other problems concerning the data occur

toXML

public StringBuffer toXML()
Description copied from interface: Storable
This method returns an XML-representation of an instance of the implementing class.

Returns:
the XML-representation