HierarchicalClusteringDendrogram

This class constructs a dendrogram using an agglomerative hierarchical clustering strategy, which follows a bottom-up approach. The process consists of the following steps:

• Initially, at level 0, each node exists as its own cluster. In this stage, each node of the original graph (which is to be clustered) corresponds to a leaf node in the dendrogram, representing a node with no children.
• As the process continues, pairs of clusters are merged step by step, ascending through the hierarchy. With each merge, a new dendrogram node is created. The dissimilarity value associated with the merging of two clusters can be retrieved via the getDissimilarityValue method. These dissimilarity values are non-decreasing, starting from zero, which corresponds to the dissimilarity at the leaf nodes.
• The clustering process concludes when all nodes are combined into a single cluster, resulting in the root of the dendrogram. The root node can be accessed using the root property.

The resulting dendrogram is a directed binary tree, with the direction flowing from the root to the leaf nodes. You can traverse the dendrogram by starting from the root node and proceeding through its children using the getChildren method, continuing this process recursively through each subsequent level of the tree.

For any given dendrogram node, you can retrieve the set of nodes from the original graph that were grouped into that dendrogram node using the getClusterNodes method. This method returns an array of IEnumerable<T>, where each IEnumerable<T> contains the nodes from the original clusters that were merged to form the dendrogram node. For instance, if clusters C1 = {n1, n2} and C2 = {n3, n4} are merged to create a dendrogram node u, getClusterNodes will return two IEnumerable<T> collections, the first containing nodes n1, n2 and the second containing nodes n3, n4.

This class cannot be instantiated