public class ClusterLimitedIntRAC extends IntRAC
IntRAC but explicitly specify the limit the number of
clusters. Attempts to replace clusters which are "closer" with those that are
"further".
As with the IntRAC this is a combined clusterer, clusters and
assigner. See the IntRAC documentation for more details.
codebook, distances, nDims, threshold, totalSamplesCLUSTER_HEADER| Constructor and Description |
|---|
ClusterLimitedIntRAC()
Sets the expected number of clusters to 100 and radius to 128.
|
ClusterLimitedIntRAC(double radiusSquared)
Set the number of clusters to 100.
|
ClusterLimitedIntRAC(int[][] bKeys,
int subSamples,
int nClusters)
Attempt to learn the threshold and uses this as an expected number of
clusters.
|
| Modifier and Type | Method and Description |
|---|---|
ClusterLimitedIntRAC |
cluster(int[][] data)
Perform clustering on the given data.
|
asciiHeader, assign, assign, assignDistance, assignDistance, binaryHeader, calculateThreshold, cluster, defaultHardAssigner, getCentroids, numClusters, numDimensions, performClustering, readASCII, readBinary, size, writeASCII, writeBinarypublic ClusterLimitedIntRAC()
public ClusterLimitedIntRAC(double radiusSquared)
radiusSquared - public ClusterLimitedIntRAC(int[][] bKeys, int subSamples, int nClusters)
bKeys - subSamples - nClusters - public ClusterLimitedIntRAC cluster(int[][] data)
SpatialClusterercluster in interface SpatialClusterer<IntRAC,int[]>cluster in class IntRACdata - the data.