Package edu.cmu.tetrad.search.test

Class MsepTest

java.lang.Object

edu.cmu.tetrad.search.test.MsepTest

All Implemented Interfaces:

IndependenceTest

public class MsepTest
extends Object
implements IndependenceTest

Checks independence facts for variables associated with the nodes in a given graph by checking m-separation facts on the underlying nodes. We use the IndependenceTest interface here so that this m-separation test can be used in place of a statistical conditional independence test in algorithms to provide oracle information.

Version:

$Id: $Id

Author:

josephramsey

Constructor Summary

Constructors

Constructor	Description
MsepTest(IndependenceFacts facts)	Constructor.
MsepTest(IndependenceFacts facts, boolean keepLatents)	Constructor.
MsepTest(IndependenceFacts facts, List<Node> variables)	Constructor.
MsepTest(Graph graph)	Constructor.
MsepTest(Graph graph, boolean keepLatents)	Constructor.

Method Summary

Modifier and Type	Method	Description
IndependenceResult	checkIndependence(Node x, Node y, Set<Node> z)	Checks the independence between two nodes with respect to a set of conditioning nodes.
boolean	determines(List<Node> z, Node x1)	Determines if a node is m-separated from a set of conditioning nodes.
double	getAlpha()	Returns an alpha level, 0.5.
DataSet	getData()	Returns the data set used for the test.
Graph	getGraph()	Returns the underlying graph that is being used to calculate d-separation relationships.
Node	getVariable(String name)	Returns the Node object with the given name.
List<Node>	getVariables()	Return the list of TetradNodes over which this independence checker is capable of determinine independence relations-- that is, all the variables in the given graph or the given data set.
IndependenceTest	indTestSubset(List<Node> vars)	Conducts an independence test on a subset of variables.
boolean	isMSeparated(Node x, Node y, Set<Node> z)	Auxiliary method to calculate msep(x, y | z) directly from nodes instead of from variables.
boolean	isVerbose()	Returns True just in case verbose output should be printed.
void	setAlpha(double alpha)	Sets the alpha level for the independence test.
void	setPag(boolean pag)	True iff the graph is a PAG
void	setVerbose(boolean verbose)	Sets the verbosity level for the program.
String	toString()	Returns a string representation of this test.

Methods inherited from class java.lang.Object

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

Methods inherited from interface edu.cmu.tetrad.search.IndependenceTest

checkIndependence, determines, getCov, getDataSets, getSampleSize, getVariableNames

Constructor Details

MsepTest

public MsepTest(Graph graph)

Constructor.

Parameters:

graph - The graph for which m-separation facts should be checked. This may be a DAG, CPDAG, or PAG. In the latter case, m-separation results will be returned (same algorithm).

MsepTest

public MsepTest(IndependenceFacts facts,
 List<Node> variables)

Constructor.

Parameters:

facts - Independence facts to be used for direct calculations of m-separation.

variables - The variables for the facts, if different from those that independence facts would return.

See Also:

• IndependenceFacts

MsepTest

public MsepTest(IndependenceFacts facts)

Constructor.

Parameters:

facts - Independence facts to be used for direct calculations of m-separation.

See Also:

• IndependenceFacts

MsepTest

public MsepTest(Graph graph,
 boolean keepLatents)

Constructor.

Parameters:

graph - The graph for which m-separation facts should be checked. This may be a DAG, CPDAG, or PAG. In the latter case, m-separation results will be returned (same algorithm).

keepLatents - Whether latent in the graph should be used in conditional independence facts. If the graph is being marginalized, this should be false.

MsepTest

public MsepTest(IndependenceFacts facts,
 boolean keepLatents)

Constructor.

Parameters:

facts - Independence facts to be used for direct calculations of m-separation.

keepLatents - Whether latent in the graph should be used in conditional independence facts. If the graph is being marginalized, this should be false.

See Also:

• IndependenceFacts

Method Details

indTestSubset

public IndependenceTest indTestSubset(List<Node> vars)

Conducts an independence test on a subset of variables.

Specified by:

indTestSubset in interface IndependenceTest

Parameters:

vars - The sublist of variables to test independence on.

Returns:

This IndependenceTest object.

Throws:

IllegalArgumentException - If the subset is empty or contains variables that are not original variables.

checkIndependence

public IndependenceResult checkIndependence(Node x,
 Node y,
 Set<Node> z)

Checks the independence between two nodes with respect to a set of conditioning nodes.

Specified by:

checkIndependence in interface IndependenceTest

Parameters:

x - The first node to check independence for.

y - The second node to check independence for.

z - The set of conditioning nodes.

Returns:

The result of the independence test.

Throws:

NullPointerException - if the set of conditioning nodes is null or contains null elements.

IllegalArgumentException - if x or y is not an observed variable.

RuntimeException - if an undefined p-value is encountered during the test.

See Also:

• IndependenceResult

isMSeparated

public boolean isMSeparated(Node x,
 Node y,
 Set<Node> z)

Auxiliary method to calculate msep(x, y | z) directly from nodes instead of from variables.

Parameters:

x - a Node object

y - a Node object

z - a Set object

Returns:

True, if so.

getVariables

public List<Node> getVariables()

Return the list of TetradNodes over which this independence checker is capable of determinine independence relations-- that is, all the variables in the given graph or the given data set.

Specified by:

getVariables in interface IndependenceTest

Returns:

This list.

determines

public boolean determines(List<Node> z,
 Node x1)

Determines if a node is m-separated from a set of conditioning nodes.

Parameters:

z - The set of conditioning nodes.

x1 - The node to check independence for.

Returns:

True if the node is m-separated from the conditioning nodes, false otherwise.

Throws:

UnsupportedOperationException - if not implemented.

getAlpha

public double getAlpha()

Returns an alpha level, 0.5. This is an arbitrary number that will help decide whether a pseudo p-value returned by the test represents a dependence or an independence.

Specified by:

getAlpha in interface IndependenceTest

Returns:

0.5.

setAlpha

public void setAlpha(double alpha)

Sets the alpha level for the independence test.

Specified by:

setAlpha in interface IndependenceTest

Parameters:

alpha - The level of significance for the test.

getVariable

public Node getVariable(String name)

Returns the Node object with the given name.

Specified by:

getVariable in interface IndependenceTest

Parameters:

name - the name of the variable to retrieve

Returns:

the Node object with the given name if found, null otherwise

getGraph

public Graph getGraph()

Returns the underlying graph that is being used to calculate d-separation relationships.

Returns:

This graph.

toString

public String toString()

Returns a string representation of this test.

Specified by:

toString in interface IndependenceTest

Overrides:

toString in class Object

Returns:

"M-separation".

getData

public DataSet getData()

Returns the data set used for the test.

Specified by:

getData in interface IndependenceTest

Returns:

The data set used for the test.

See Also:

• DataSet
• ICovarianceMatrix
• DataModel

isVerbose

public boolean isVerbose()

Returns True just in case verbose output should be printed.

Specified by:

isVerbose in interface IndependenceTest

Returns:

True, if so.

setVerbose

public void setVerbose(boolean verbose)

Sets the verbosity level for the program.

Specified by:

setVerbose in interface IndependenceTest

Parameters:

verbose - True if verbose output should be printed, false otherwise.

setPag

public void setPag(boolean pag)

True iff the graph is a PAG

Parameters:

pag - True if the graph is a PAG, false if a DAG, MPDAG, or CPDAG.