Package edu.cmu.tetrad.search.utils

Class ClusterSignificance

java.lang.Object
edu.cmu.tetrad.search.utils.ClusterSignificance
public class ClusterSignificance extends Object

Provides some methods to check significance of clusters for clustering algroithms. It is assumed that for each cluster there is a single latent that is a parent of all the nodes in the cluster. Significance of these models is returned.

Version:
$Id: $Id
Author:
josephramsey

  • Constructor Details

    • ClusterSignificance

      public ClusterSignificance(List<Node> variables, DataModel dataModel)
      Constructs a new cluster significance object.
      Parameters:
      variables - The variables in the model.
      dataModel - The data model.

  • Method Details

    • variablesForIndices

      public static List<Node> variablesForIndices(List<Integer> cluster, List<Node> variables)
      Converts a list of indices into a list of Nodes representing a cluster.
      Parameters:
      cluster - The indices of the variables.
      variables - The variables.
      Returns:
      The extracted node list.

    • variablesForIndices

      public static List<List<Node>> variablesForIndices(Set<List<Integer>> clusters, List<Node> _variables)
      Converts a list of indices into a list of Nodes representing a cluster.
      Parameters:
      clusters - The indices of the variables.
      _variables - The variables.
      Returns:
      The extracted node list.

    • printClusterPValues

      public void printClusterPValues(Set<List<Integer>> clusters)
      Prints the p-values for the given clusters.
      Parameters:
      clusters - The clusters.

    • setCheckType

      public void setCheckType(ClusterSignificance.CheckType checkType)
      Sets the type of check to perform.
      Parameters:
      checkType - The type of check.
      See Also:

    • significant

      public boolean significant(List<Integer> cluster, double alpha)
      Returns the p-value for the given cluster.
      Parameters:
      cluster - The cluster.
      alpha - The alpha level.
      Returns:
      The p-value.

    • getModelPValue

      public double getModelPValue(List<List<Integer>> clusters)
      Returns the p-value for the given cluster.
      Parameters:
      clusters - The clusters.
      Returns:
      The p-value.