Package edu.cmu.tetrad.search

Class PcMb

java.lang.Object

edu.cmu.tetrad.search.PcMb

All Implemented Interfaces:

IGraphSearch, IMbSearch

public final class PcMb
extends Object
implements IMbSearch, IGraphSearch

Searches for a CPDAG representing all the Markov blankets for a given target T consistent with the given independence information. This CPDAG may be used to generate the actual list of DAG's that might be Markov blankets. Note that this code has been converted to be consistent with the CPC algorithm. The reference is here:

Bai, X., Padman, R., Ramsey, J., & Spirtes, P. (2008). Tabu search-enhanced graphical models for classification in high dimensions. INFORMS Journal on Computing, 20(3), 423-437.

This class is configured to respect knowledge of forbidden and required edges, including knowledge of temporal tiers.

Version:

$Id: $Id

Author:

josephramsey

See Also:

• FgesMb
• Knowledge

Constructor Summary

Constructors

Constructor	Description
PcMb(IndependenceTest test, int depth)	Constructs a new search.

Method Summary

Modifier and Type	Method	Description
Set<Node>	findMb(Node target)	Given the target, this returns all the nodes in the Markov Blanket.
String	getAlgorithmName()	Returns "PC-MB."
Set<Triple>	getAmbiguousTriples()	Returns the set of triples identified as ambiguous by the CPC algorithm during the most recent search.
int	getDepth()	Returns the depth of the search--that is, the maximum number of variables conditioned on in any conditional independence test.
long	getElapsedTime()	Returns the elapsed time of the most recent search.
Knowledge	getKnowledge()	Returns The knowledge used in search.
int	getNumIndependenceTests()	Returns the number of independence tests performed during the most recent search.
List<Node>	getTargets()	Return the targets of the most recent search.
IndependenceTest	getTest()	Returns the test used in search.
Graph	resultGraph()	Returns the result graph.
Graph	search()	Searches for the Markov blanket CPDAG for the given targets.
Graph	search(List<Node> targets)	Searches for the MB CPDAG for the given targets.
void	setDepth(int depth)	Sets the maximum number of conditioning variables for any conditional independence test.
void	setFindMb(boolean findMb)	Setter for the field findMb.
void	setKnowledge(Knowledge knowledge)	Sets knowledge, to which the algorithm is in fact sensitive.
void	setMeekPreventCycles(boolean meekPreventCycles)	Sets whether cycles should be prevented, using a cycle checker.
void	setVariables(List<Node> variables)	Setter for the field variables.
void	setVerbose(boolean verbose)	Sets the verbosity level of the search.

Methods inherited from class java.lang.Object

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

Constructor Details

PcMb

public PcMb(IndependenceTest test,
 int depth)

Constructs a new search.

Parameters:

test - The source of conditional independence information for the search.

depth - The maximum number of variables conditioned on for any

Method Details

setMeekPreventCycles

public void setMeekPreventCycles(boolean meekPreventCycles)

Sets whether cycles should be prevented, using a cycle checker.

Parameters:

meekPreventCycles - True, if so.

search

public Graph search(List<Node> targets)
             throws InterruptedException

Searches for the MB CPDAG for the given targets.

Parameters:

targets - The targets variable.

Returns:

a Graph object

Throws:

InterruptedException

search

public Graph search()
             throws InterruptedException

Searches for the Markov blanket CPDAG for the given targets.

Specified by:

search in interface IGraphSearch

Returns:

The Markov blanket CPDAG as a Graph object.

Throws:

InterruptedException - if any.

getAmbiguousTriples

public Set<Triple> getAmbiguousTriples()

Returns the set of triples identified as ambiguous by the CPC algorithm during the most recent search.

Returns:

This set.

getNumIndependenceTests

public int getNumIndependenceTests()

Returns the number of independence tests performed during the most recent search.

Specified by:

getNumIndependenceTests in interface IMbSearch

Returns:

This number.

getTargets

public List<Node> getTargets()

Return the targets of the most recent search.

Returns:

This list.

getElapsedTime

public long getElapsedTime()

Returns the elapsed time of the most recent search.

Returns:

This time.

getAlgorithmName

public String getAlgorithmName()

Returns "PC-MB."

Specified by:

getAlgorithmName in interface IMbSearch

Returns:

This string.

getDepth

public int getDepth()

Returns the depth of the search--that is, the maximum number of variables conditioned on in any conditional independence test.

Returns:

This depth.

setDepth

public void setDepth(int depth)

Sets the maximum number of conditioning variables for any conditional independence test.

Parameters:

depth - This depth.

resultGraph

public Graph resultGraph()

Returns the result graph.

Returns:

This graph.

findMb

public Set<Node> findMb(Node target)
                 throws InterruptedException

Given the target, this returns all the nodes in the Markov Blanket.

Returns the Markov blanket variables (not the Markov blanket DAG).

Specified by:

findMb in interface IMbSearch

Parameters:

target - a Node object

Returns:

a Set object

Throws:

InterruptedException

getTest

public IndependenceTest getTest()

Returns the test used in search.

Returns:

This test.

getKnowledge

public Knowledge getKnowledge()

Returns The knowledge used in search.

Returns:

This knowledge.

setKnowledge

public void setKnowledge(Knowledge knowledge)

Sets knowledge, to which the algorithm is in fact sensitive.

Parameters:

knowledge - This knowledge.

See Also:

• Knowledge

setVariables

public void setVariables(List<Node> variables)

Setter for the field variables.

Parameters:

variables - a List object

setFindMb

public void setFindMb(boolean findMb)

Setter for the field findMb.

Parameters:

findMb - a boolean

setVerbose

public void setVerbose(boolean verbose)

Sets the verbosity level of the search.

Parameters:

verbose - true if verbose output should be enabled, false otherwise