org.metaqtl.algo
Class HClustAlgorithm

java.lang.Object
  extended by org.metaqtl.algo.HClustAlgorithm

public class HClustAlgorithm
extends java.lang.Object

This class implements stantard agglomerative hierarchical clustering methods.


Field Summary
static int CENTROID_METHOD
          Centroid method
static int METHOD
          A global variable to control the custering method.
static int WARD_METHOD
          Ward method
 
Constructor Summary
HClustAlgorithm()
           
 
Method Summary
static Tree CentroidMethod(double[] x, double[] sd)
          This method performs a hierarchical clustering of the given data points using a centroid method based on the mahalanobis distance.
static Tree run(double[] x, double[] sd)
          This method performs a agglomerative clustering of the given data points where x is the vector of observed values and sd the vector of their standard deviations.
static Tree WardMethod(double[] x, double[] sd)
          This method performs a hierarchical clustering of the given data points using the Ward's method.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

CENTROID_METHOD

public static final int CENTROID_METHOD
Centroid method

See Also:
Constant Field Values

WARD_METHOD

public static final int WARD_METHOD
Ward method

See Also:
Constant Field Values

METHOD

public static int METHOD
A global variable to control the custering method.

Constructor Detail

HClustAlgorithm

public HClustAlgorithm()
Method Detail

run

public static Tree run(double[] x,
                       double[] sd)
This method performs a agglomerative clustering of the given data points where x is the vector of observed values and sd the vector of their standard deviations. It returns an array of TreeNode which represents the hierarchical clustering.

Parameters:
x - the vector of data points.
sd - the vector of standard deviations.
Returns:
the tree which represents the hierarchical clustering.

CentroidMethod

public static Tree CentroidMethod(double[] x,
                                  double[] sd)
This method performs a hierarchical clustering of the given data points using a centroid method based on the mahalanobis distance.

Parameters:
x - the vector of data points.
sd - the vector of standard deviations.
Returns:
the tree which represents the hierarchical clustering.

WardMethod

public static Tree WardMethod(double[] x,
                              double[] sd)
This method performs a hierarchical clustering of the given data points using the Ward's method.

Parameters:
x - the vector of data points.
sd - the vector of standard deviations.
Returns:
the tree which represents the hierarchical clustering.