Use the HierarchicalClusterAnalysis to identify homogeneous subgroups of cases in a population using the agglomerative hierarchical clustering method. The method starts out by treating every case as its own cluster. It then successively combines ("agglomerates") clusters that are closest to each other until the entire population is one cluster. The order in which clusters are combined specifies a hierarchy. The last remaining clusters determine membership of the homogeneous groups.

The DistanceMeasure determines how the distance between clusters is calculated. The default is to use the squared Euclidean distance. The DistanceMeasures class defines several common distance measures.

The LinkageMethod property determines how cluster distances are updated when two clusters are merged. The default is the centroid method. The Standardize property determines whether the variables are transformed to all have the same mean and standard deviation. The default is trueTruetruetrue (True in Visual Basic).

Once the clustering has been computed by calling the Compute(ParallelOptions) method, the GetClusterPartition(Int32) method can be used to partition the observations. This method returns a HierarchicalClusterCollection that provides detailed information about each of the clusters.

The results of a hierarchical cluster analysis are often presented in graphical form as a dendrogram, a tree-like structure that shows how clusters were combined. The DendrogramRoot property returns a DendrogramNode object that represents the root of the dendrogram. It provides all the information necessary to produce a dendrogram. See the DendrogramNode class for details.

Numerical Libraries