Hierarchical-Clustering

A java implementation of hierarchical clustering. No external dependencies needed, generic implementation.

Supports different Linkage approaches:

• Average Linkage
• Single Linkage
• Your Linkage function (if you extend the Linkage class)

Example usage with single Linkage

List<Double> vals= Arrays.asList(1d,2d,10d); //your data
Linkage la = new SingleLinkage<Double>((o, o2) -> Math.abs(o-o2));  //lamda distance function
ClusterAlgorithm<Double> h = new ClusterAlgorithm<Double>(vals,la); 
h.cluster(); //run 
System.out.println(h.getFirstCluster());
System.out.println("Top Cluster distance: "+h.getFirstCluster().distance);

Sample output for example above

((2.0 , 1.0) , 10.0)
Top Cluster distance: 8.0

Implementation Details:

• Clusters are modeled similar to a binary tree (a cluster contains right,left child)
• Linkage values between all Pairs are cached