Package edu.cmu.tetrad.search.utils

Class PcCommon

java.lang.Object

edu.cmu.tetrad.search.utils.PcCommon

All Implemented Interfaces:

IGraphSearch

public final class PcCommon
extends Object
implements IGraphSearch

Prodies some common implementation pieces of variaous PC-like algorithms, with options for collider discovery type, FAS type, and conflict rule.

Author:

josephramsey

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static enum	PcCommon.ColliderDiscovery	Give the options for the collider discovery algroithm to use--FAS with sepsets reasoning, FAS with conservative reasoning, or FAS with Max P reasoning.
static enum	PcCommon.ConflictRule	Gives the type of conflict to be used, priority (when there is a conflict, keep the orientation that has already been made), bidirected (when there is a conflict, orient a bidirected edge), or overwrite (when there is a conflict, use the new orientation).
static enum	PcCommon.FasType	Gives the type of FAS used, regular or stable.
static enum	PcCommon.PcHeuristicType	NONE = no heuristic, PC-1 = sort nodes alphabetically; PC-1 = sort edges by p-value; PC-3 = additionally sort edges in reverse order using p-values of associated independence facts.

Constructor Summary

Constructors

Constructor	Description
PcCommon(IndependenceTest independenceTest)	Constructs a CPC algorithm that uses the given independence test as oracle.

Method Summary

Modifier and Type	Method	Description
Set<Edge>	getAdjacencies()	Returns The edges in the search graph.
Set<Triple>	getAmbiguousTriples()
Set<Triple>	getColliderTriples()
int	getDepth()
long	getElapsedTime()
IndependenceTest	getIndependenceTest()
Knowledge	getKnowledge()
Set<Triple>	getNoncolliderTriples()
boolean	isMeekPreventCycles()
static void	orientCollider(Node x, Node y, Node z, PcCommon.ConflictRule conflictRule, Graph graph)	Orient a single unshielded triple, x*-*y*-*z, in a graph.
Graph	search()	Runs the search and returns the search graph.
Graph	search(List<Node> nodes)	Runs the search over the given list of nodes only, returning the serach graph.
void	setColliderDiscovery(PcCommon.ColliderDiscovery colliderDiscovery)	Sets the type of collider discovery to do.
void	setConflictRule(PcCommon.ConflictRule conflictRule)	Sets the conflict rule to use.
void	setDepth(int depth)	Sets the maximum number of variables conditioned on in any conditional independence test.
void	setFasType(PcCommon.FasType fasType)
void	setKnowledge(Knowledge knowledge)	Sets the knowledge specification used in the search.
void	setMaxPathLength(int maxPathLength)
void	setMeekPreventCycles(boolean meekPreventCycles)	Sets to true just in case edges will not be added if they would create cycles.
void	setPcHeuristicType(PcCommon.PcHeuristicType pcHeuristic)
void	setVerbose(boolean verbose)	Sets whether verbose output should be printed.

Methods inherited from class java.lang.Object

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

Constructor Details

PcCommon

public PcCommon(IndependenceTest independenceTest)

Constructs a CPC algorithm that uses the given independence test as oracle. This does not make a copy of the independence test, for fear of duplicating the data set!

Parameters:

independenceTest - The independence test to use.

Method Details

orientCollider

public static void orientCollider(Node x,
 Node y,
 Node z,
 PcCommon.ConflictRule conflictRule,
 Graph graph)

Orient a single unshielded triple, x*-*y*-*z, in a graph.

Parameters:

conflictRule - The conflict rule to use.

graph - The graph to orient.

See Also:

• PcCommon.ConflictRule

setMaxPathLength

public void setMaxPathLength(int maxPathLength)

Parameters:

maxPathLength - The max path length for the max p collider orientation heuristic.

setFasType

public void setFasType(PcCommon.FasType fasType)

Parameters:

fasType - The type of FAS to be used.

setPcHeuristicType

public void setPcHeuristicType(PcCommon.PcHeuristicType pcHeuristic)

Parameters:

pcHeuristic - Which PC heuristic to use (see Causation, Prediction and Search). Default is PcHeuristicType.NONE.

See Also:

• PcCommon.PcHeuristicType

isMeekPreventCycles

public boolean isMeekPreventCycles()

Returns:

true just in case edges will not be added if they would create cycles.

setMeekPreventCycles

public void setMeekPreventCycles(boolean meekPreventCycles)

Sets to true just in case edges will not be added if they would create cycles.

Parameters:

meekPreventCycles - True just in case edges will not be added if they would create cycles.

search

public Graph search()

Runs the search and returns the search graph.

Specified by:

search in interface IGraphSearch

Returns:

This result graph.

search

public Graph search(List<Node> nodes)

Runs the search over the given list of nodes only, returning the serach graph.

Parameters:

nodes - The nodes to search over.

Returns:

The result graph.

setColliderDiscovery

public void setColliderDiscovery(PcCommon.ColliderDiscovery colliderDiscovery)

Sets the type of collider discovery to do.

Parameters:

colliderDiscovery - This type.

setConflictRule

public void setConflictRule(PcCommon.ConflictRule conflictRule)

Sets the conflict rule to use.

Parameters:

conflictRule - This rule.

See Also:

• PcCommon.ConflictRule

getElapsedTime

public long getElapsedTime()

Returns:

The elapsed time of search in milliseconds, after search() has been run.

getKnowledge

public Knowledge getKnowledge()

Returns:

The knowledge specification used in the search. Non-null.

setKnowledge

public void setKnowledge(Knowledge knowledge)

Sets the knowledge specification used in the search. Non-null.

getIndependenceTest

public IndependenceTest getIndependenceTest()

Returns:

the independence test used in the search, set in the constructor. This is not returning a copy, for fear of duplicating the data set!

getDepth

public int getDepth()

Returns:

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

setDepth

public void setDepth(int depth)

Sets the maximum number of variables conditioned on in any conditional independence test. If set to -1, the value of 1000 will be used. May not be set to Integer.MAX_VALUE, due to a Java bug on multi-core systems.

Parameters:

depth - The depth.

setVerbose

public void setVerbose(boolean verbose)

Sets whether verbose output should be printed.

Parameters:

verbose - True iff the case.

getAmbiguousTriples

public Set<Triple> getAmbiguousTriples()

Returns:

The set of ambiguous triples found during the most recent run of the algorithm. Non-null after a call to search().

getColliderTriples

public Set<Triple> getColliderTriples()

Returns:

The set of collider triples found during the most recent run of the algorithm. Non-null after a call to search().

getNoncolliderTriples

public Set<Triple> getNoncolliderTriples()

Returns:

The set of noncollider triples found during the most recent run of the algorithm. Non-null after a call to search().

getAdjacencies

public Set<Edge> getAdjacencies()

Returns The edges in the search graph.

Returns:

These edges.