java - WEKA HierarchicalClusterer class always return 2 clusters -

here code:

import weka.clusterers.clusterevaluation; import weka.clusterers.hierarchicalclusterer; import weka.clusterers.em; import weka.core.converters.csvloader; import weka.core.converters.converterutils.datasource; import weka.core.neighboursearch.performancestats;  import java.io.file; import java.io.ioexception; import java.text.parseexception; import java.util.arraylist; import java.util.enumeration;  import weka.core.*;  public class wekasample1 {  public static void main(string[] args) {      instances data = null;     csvloader csvloader = new csvloader();     try {         csvloader.setsource(new file("d:\\weka\\numbers.csv"));          data = csvloader.getdataset();                 hierarchicalclusterer h = new hierarchicalclusterer();              distancefunction d = new distancefunction() {          @override         public void setoptions(string[] arg0) throws exception {          }          @override         public enumeration listoptions() {             return null;         }          @override         public string[] getoptions() {             return null;         }          @override         public void update(instance arg0) {          }          @override         public void setinvertselection(boolean arg0) {          }          @override         public void setinstances(instances arg0) {          }          @override         public void setattributeindices(string arg0) {          }          @override         public void postprocessdistances(double[] arg0) {          }          @override         public boolean getinvertselection() {             return false;         }          @override         public instances getinstances() {             return null;         }          @override         public string getattributeindices() {             return null;         }          @override         public double distance(instance arg0, instance arg1, double arg2,                 performancestats arg3) {             return 0;         }          @override         public double distance(instance arg0, instance arg1, double arg2) {             return 0;         }          @override         public double distance(instance arg0, instance arg1, performancestats arg2)                 throws exception {             return 0;         }          @override         public double distance(instance arg0, instance arg1) {              double s1 = arg0.value(0);             double s2 = arg1.value(0);              return double.positive_infinity;         }     };      h.setdistancefunction(d);     selectedtag s = new selectedtag(1, hierarchicalclusterer.tags_link_type);     h.setlinktype(s);      h.buildclusterer(data);   //      double[] arr; //      for(int i=0; i<data.size(); i++) { //           //          arr = h.distributionforinstance(data.get(i)); //          for(int j=0; j< arr.length; j++) //              system.out.print(arr[j]+","); //          system.out.println(); //           //      }          system.out.println(h.numberofclusters());     } catch (exception e) {         e.printstacktrace();     }  }  } 

now, output number of clusters generated 2 if modify distancefucntion method also. how know instance if of cluster? when uncomment code above written distribution instances, arrayoutofbound exception.

but in general, can explain how clustering done hierarchically weka here?

here data set, of length 10 , dimension 2:

1    1 2    2 3    3 4    4 5    5 6    6 7    7 8    8 9    9 10   10 

try real data set, not evenly spaced array of points.

because have same distance next! single link, should single cluster, maybe there rounding issues.

plus, distance function use 0/infinity, too!

try using weka ui first.