Package edu.cmu.tetrad.search

Class Ion

java.lang.Object

edu.cmu.tetrad.search.Ion

public class Ion
extends Object

Implements the ION (Integration of Overlapping Networks) algorithm for distributed causal inference. The algorithm takes as input a set of PAGs (presumably learned using a local learning algorithm) over variable sets that may have some variables in common and others not in common. The algorithm returns a complete set of PAGs over every variable form an input PAG_of_the_true_DAG that are consistent (same d-separations and d-connections) with every input PAG_of_the_true_DAG.

Author:

Robert Tillman

Constructor Summary

Constructors

Constructor	Description
Ion(List<Graph> pags)	Constructs a new instance of the ION search from the input PAGs

Method Summary

Modifier and Type	Method	Description
List<Integer>	getIterations()
double	getMaxMemUsage()
List<String>	getRuntime()
String	getStats()
List<Graph>	search()	Begins the ION search procedure, described at each step
void	setAdjacencySearch(boolean b)	Sets adjacency search on or off
void	setPathLengthSearch(boolean b)	Sets path length search on or off
static List<List<Node>>	treks(Graph graph, Node node1, Node node2)

Methods inherited from class java.lang.Object

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

Constructor Details

Ion

public Ion(List<Graph> pags)

Constructs a new instance of the ION search from the input PAGs

Parameters:

pags - The PAGs to be integrated

Method Details

setPathLengthSearch

public void setPathLengthSearch(boolean b)

Sets path length search on or off

setAdjacencySearch

public void setAdjacencySearch(boolean b)

Sets adjacency search on or off

search

public List<Graph> search()

Begins the ION search procedure, described at each step

getRuntime

public List<String> getRuntime()

getMaxMemUsage

public double getMaxMemUsage()

getIterations

public List<Integer> getIterations()

getStats

public String getStats()

treks

public static List<List<Node>> treks(Graph graph,
 Node node1,
 Node node2)