Package edu.cmu.tetrad.search.work_in_progress

Class ISBicScore

java.lang.Object

edu.cmu.tetrad.search.work_in_progress.ISBicScore

All Implemented Interfaces:

ISScore

public class ISBicScore
extends Object
implements ISScore

A class representing the ISBicScore, which calculates BIC scores for a Bayesian network considering different structural changes and their impacts using an information-sharing mechanism.

Constructor Summary

Constructors

Constructor	Description
ISBicScore(DataSet dataSet, DataSet testCase)	Constructs an ISBicScore instance with the provided data sets.

Method Summary

Modifier and Type	Method	Description
boolean	determines(List<Node> z, Node y)	Determines if the given list of nodes can determine the specified node.
DataSet	getDataSet()	Retrieves the data set from the current context.
double	getKAddition()	Retrieves the current value of the k_addition parameter.
double	getKDeletion()	Retrieves the current value of the k_deletion parameter.
double	getKReorientation()	Retrieves the current value of the k_reorientation parameter.
int	getMaxDegree()	Retrieves the maximum degree value.
int[]	getParentValues(int nodeIndex, int rowIndex, int[] dims)	Calculates the parent values at given node and row indices based on the provided dimensions.
int[]	getParentValuesForCombination(int rowIndex, int[] dims)	Calculates the parent values for a given row index based on the provided dimensions.
double	getPenaltyDiscount()	Retrieves the penalty discount associated with this entity.
double	getSamplePrior()	Calculates and returns the sample prior value.
int	getSampleSize()	Retrieves the current sample size.
double	getStructurePrior()	Retrieves the prior probability of the given structure.
Node	getVariable(String targetName)	Retrieves a Node from the variables list that matches the specified target name.
List<Node>	getVariables()	Retrieves the list of variables associated with this ISBicScore instance.
boolean	isEffectEdge(double bump)	Must be called directly after the corresponding scoring call.
double	localScore(int node, int[] parents_is, int[] parents_pop, int[] children_pop)	Calculates the local score for a given node considering its parents.
double	localScore1(int node, int[] parents_is, int[] parents_pop, int[] children_pop)	Computes a local score based on the given node, its parents, and its children's populations.
double	localScoreDiff(int x, int y, int[] z, int[] z_pop, int[] child_pop)	Calculates the difference in local scores for a given variable when it is added to or removed from a set of parent variables.
void	setPenaltyDiscount(double penaltyDiscount)	Sets the penalty discount.
void	setSamplePrior(double samplePrior)	Sets the value of the sample prior.
void	setStructurePrior(double structurePrior)	Sets the structure prior value for the current configuration.
void	setVariables(List<Node> variables)	Sets the list of variables for the instance.

Methods inherited from class java.lang.Object

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

Constructor Details

ISBicScore

public ISBicScore(DataSet dataSet,
 DataSet testCase)

Constructs an ISBicScore instance with the provided data sets.

Parameters:

dataSet - The primary DataSet instance for training, must not be null.

testCase - The test DataSet instance used for validation or testing.

Throws:

NullPointerException - if dataSet is null.

IllegalArgumentException - if dataBox within dataSet is not an instance of VerticalIntDataBox.

Method Details

localScore

public double localScore(int node,
 int[] parents_is,
 int[] parents_pop,
 int[] children_pop)

Calculates the local score for a given node considering its parents.

Specified by:

localScore in interface ISScore

Parameters:

node - The index of the node to calculate the score for.

parents_is - Array of indices representing the parents of the node in the IS model.

parents_pop - Array of indices representing the parents of the node in the POP model.

children_pop - Array of indices representing the children of the node in the POP model.

Returns:

The local score for the given node.

localScoreDiff

public double localScoreDiff(int x,
 int y,
 int[] z,
 int[] z_pop,
 int[] child_pop)

Calculates the difference in local scores for a given variable when it is added to or removed from a set of parent variables.

Specified by:

localScoreDiff in interface ISScore

Parameters:

x - The variable to be added or removed.

y - The target variable whose score is being calculated.

z - The set of parent variables excluding the variable x.

z_pop - The set of population parent variables excluding the variable x.

child_pop - The set of population children variables.

Returns:

The difference in local scores after adding/removing the variable x.

getParentValues

public int[] getParentValues(int nodeIndex,
 int rowIndex,
 int[] dims)

Calculates the parent values at given node and row indices based on the provided dimensions.

Parameters:

nodeIndex - The index of the node for which parent values are being calculated.

rowIndex - The row index that needs to be converted to parent values.

dims - An array representing the dimensions of the parents.

Returns:

An array of integers where each value corresponds to a parent value at the given row index.

getParentValuesForCombination

public int[] getParentValuesForCombination(int rowIndex,
 int[] dims)

Calculates the parent values for a given row index based on the provided dimensions.

Parameters:

rowIndex - The row index that needs to be converted to parent values.

dims - An array representing the dimensions of the parents.

Returns:

An array of integers where each value corresponds to a parent value at the given row index.

getKAddition

public double getKAddition()

Retrieves the current value of the k_addition parameter.

Returns:

The value of the k_addition parameter.

getKDeletion

public double getKDeletion()

Retrieves the current value of the k_deletion parameter.

Returns:

The value of the k_deletion parameter.

getKReorientation

public double getKReorientation()

Retrieves the current value of the k_reorientation parameter.

Returns:

The value of the k_reorientation parameter.

getVariables

public List<Node> getVariables()

Retrieves the list of variables associated with this ISBicScore instance.

Specified by:

getVariables in interface ISScore

Returns:

a List of Node objects representing the variables.

setVariables

public void setVariables(List<Node> variables)

Sets the list of variables for the instance. All nodes in the new list must have the same names as the corresponding nodes in the existing list.

Parameters:

variables - the new list of variables to set

Throws:

IllegalArgumentException - if any variable in the new list has a different name from the corresponding variable in the existing list

getSampleSize

public int getSampleSize()

Retrieves the current sample size.

Specified by:

getSampleSize in interface ISScore

Returns:

the sample size as an integer.

isEffectEdge

public boolean isEffectEdge(double bump)

Must be called directly after the corresponding scoring call.

Specified by:

isEffectEdge in interface ISScore

Parameters:

bump - The threshold value to evaluate the significance of the effect on the edge.

Returns:

true if the edge has a significant effect; false otherwise.

getDataSet

public DataSet getDataSet()

Retrieves the data set from the current context.

Specified by:

getDataSet in interface ISScore

Returns:

DataSet object containing the data.

Throws:

UnsupportedOperationException - if the operation is not supported.

getStructurePrior

public double getStructurePrior()

Retrieves the prior probability of the given structure.

Specified by:

getStructurePrior in interface ISScore

Returns:

the prior probability of the structure as a double.

Throws:

UnsupportedOperationException - if the operation is not supported.

setStructurePrior

public void setStructurePrior(double structurePrior)

Sets the structure prior value for the current configuration.

Specified by:

setStructurePrior in interface ISScore

Parameters:

structurePrior - the value to be set as the structure prior.

getSamplePrior

public double getSamplePrior()

Calculates and returns the sample prior value.

Specified by:

getSamplePrior in interface ISScore

Returns:

The sample prior value as a double.

Throws:

UnsupportedOperationException - if the operation is not supported.

setSamplePrior

public void setSamplePrior(double samplePrior)

Sets the value of the sample prior.

Specified by:

setSamplePrior in interface ISScore

Parameters:

samplePrior - The value to be set as the sample prior.

Throws:

UnsupportedOperationException - if the operation is not supported.

getPenaltyDiscount

public double getPenaltyDiscount()

Retrieves the penalty discount associated with this entity.

Returns:

the penalty discount as a double

setPenaltyDiscount

public void setPenaltyDiscount(double penaltyDiscount)

Sets the penalty discount.

Parameters:

penaltyDiscount - The discount to be applied as a penalty, represented as a decimal.

getVariable

public Node getVariable(String targetName)

Retrieves a Node from the variables list that matches the specified target name.

Specified by:

getVariable in interface ISScore

Parameters:

targetName - the name of the variable to search for

Returns:

the Node that matches the target name, or null if no match is found

getMaxDegree

public int getMaxDegree()

Retrieves the maximum degree value.

Specified by:

getMaxDegree in interface ISScore

Returns:

the maximum degree, which is a constant value of 1000.

determines

public boolean determines(List<Node> z,
 Node y)

Determines if the given list of nodes can determine the specified node.

Specified by:

determines in interface ISScore

Parameters:

z - the list of nodes that are tested to determine the node y

y - the node whose determination is being tested

Returns:

true if the list of nodes z can determine the node y, otherwise false

localScore1

public double localScore1(int node,
 int[] parents_is,
 int[] parents_pop,
 int[] children_pop)

Computes a local score based on the given node, its parents, and its children's populations.

Specified by:

localScore1 in interface ISScore

Parameters:

node - an integer representing the current node for which the score is being computed

parents_is - an array of integers representing the indices of a node's parents

parents_pop - an array of integers representing the population of each parent

children_pop - an array of integers representing the population of each child

Returns:

a double representing the computed local score