public class NeighborChainAgglomerativeClustering extends Object implements Clustering
SimilarityFunction.
This implementation is based on the following paper:
| Modifier and Type | Class and Description |
|---|---|
static class |
NeighborChainAgglomerativeClustering.ClusterLink
The similarity method used when comparing and merging compelte clusters.
|
static class |
NeighborChainAgglomerativeClustering.Link
A simple struct for storing links between two clusters.
|
| Constructor and Description |
|---|
NeighborChainAgglomerativeClustering()
Creates a
NeighborChainAgglomerativeClustering with no minimum
similarity threshold. |
NeighborChainAgglomerativeClustering(NeighborChainAgglomerativeClustering.ClusterLink method,
SimilarityFunction simFunc)
Creates a
NeighborChainAgglomerativeClustering with the specified
minimum similarity threshold. |
| Modifier and Type | Method and Description |
|---|---|
Assignments |
cluster(Matrix m,
int numClusters,
Properties props)
Returns the agglomerative clustering result using
NeighborChainAgglomerativeClustering.ClusterLink and
SimilarityFunction specified from the constructor. |
Assignments |
cluster(Matrix m,
Properties props)
Unsupported
|
static Assignments |
clusterAdjacencyMatrix(Matrix adj,
NeighborChainAgglomerativeClustering.ClusterLink method,
int numClusters)
Clusters the points represented as an adjacency matrix in
adj
using the supplied ClusterLink method into numCluster. |
static Assignments |
formAssignments(Collection<Set<Integer>> clusters,
int numPoints)
Returns the
Assignments corresponding to the set of clusters and
number of data points. |
public NeighborChainAgglomerativeClustering()
NeighborChainAgglomerativeClustering with no minimum
similarity threshold. When using this constructor, you must specifcy the
number of clusters or the results will be undefined.public NeighborChainAgglomerativeClustering(NeighborChainAgglomerativeClustering.ClusterLink method, SimilarityFunction simFunc)
NeighborChainAgglomerativeClustering with the specified
minimum similarity threshold. When clustering without specifying the
number of clusters, clustering stops when the most similar clusters have
a similarity below this threshold.public Assignments cluster(Matrix m, Properties props)
cluster in interface Clusteringm - the Matrix whose row data points are to be
clusteredprops - the properties to use for any parameters each clustering
algorithm may needAssignment instances that indicate zero or
more clusters to which each row belongs.UnsupportedOperationExceptionpublic Assignments cluster(Matrix m, int numClusters, Properties props)
NeighborChainAgglomerativeClustering.ClusterLink and
SimilarityFunction specified from the constructor. The SimilarityFunction will be used to build a symmetric adjacency matrix
and the NeighborChainAgglomerativeClustering.ClusterLink will determine how to udpate similarities
between newly nerged clusters.cluster in interface Clusteringm - the Matrix whose row data points are to be
clusterednumClusters - the number of clusters to generateprops - the properties to use for any parameters each clustering
algorithm may needAssignment instances that indicate zero or
more clusters to which each row belongs.public static Assignments clusterAdjacencyMatrix(Matrix adj, NeighborChainAgglomerativeClustering.ClusterLink method, int numClusters)
adj
using the supplied ClusterLink method into numCluster.adj - A symmetric Matrix recording similarities between
nodes in a graph.method - NeighborChainAgglomerativeClustering.ClusterLink method that determines how similarities
between clusters will be updated.numClusters - The desired number of clusters.Assignments from each data point to it's assigned
cluster.public static Assignments formAssignments(Collection<Set<Integer>> clusters, int numPoints)
Assignments corresponding to the set of clusters and
number of data points.Copyright © 2012. All Rights Reserved.