Package edu.cmu.tetrad.search.work_in_progress

Class Dci

java.lang.Object

edu.cmu.tetrad.search.work_in_progress.Dci

public class Dci
extends Object

Implements the DCI (Distributed Causal Inference) algorithm for learning causal structure over a set of variable from multiple datasets that each may only measure proper overlapping subsets of that sets, or datasets with some variables in common and others not. The algorithm currently takes as input a set of PAGs (presumably learned using a local learning algorithm such as FCI) and returns a complete set of PAGs over every variable included a dataset that are consistent with all the PAGs (same d-separations and d-connections)

Author:

Robert Tillman

Constructor Summary

Constructors

Constructor	Description
Dci(List<IndependenceTest> tests)
Dci(List<IndependenceTest> tests, ResolveSepsets.Method method)

Method Summary

Modifier and Type	Method	Description
int	getDepth()
long	getElapsedTime()
List<SepsetMapDci>	getSepset()	Gets the resulting sepsets
boolean	isSubtrek(List<Node> trek, List<Node> subtrek)	Determines whether one trek is a subtrek of another trek
List<Graph>	search()	Begins the DCI search procedure, described at each step
void	setDepth(int depth)

Methods inherited from class java.lang.Object

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

Constructor Details

Dci

public Dci(List<IndependenceTest> tests)

Dci

public Dci(List<IndependenceTest> tests,
 ResolveSepsets.Method method)

Method Details

getDepth

public int getDepth()

setDepth

public void setDepth(int depth)

getElapsedTime

public long getElapsedTime()

getSepset

public List<SepsetMapDci> getSepset()

Gets the resulting sepsets

search

public List<Graph> search()

Begins the DCI search procedure, described at each step

isSubtrek

public boolean isSubtrek(List<Node> trek,
 List<Node> subtrek)

Determines whether one trek is a subtrek of another trek