public class ExtendedClusterer extends Clusterer
BIC, GAUSS, MIN_DIST, SILHOUETTE
Constructor and Description |
---|
ExtendedClusterer(java.util.List<Clusterable> data) |
Modifier and Type | Method and Description |
---|---|
void |
clear(double scale,
double n_points)
This method removes points that are far from any cluster from the data.
|
java.util.Map<Clusterable,java.util.List<Clusterable>> |
cluster(java.util.List<Clusterable> centers,
int max,
int type,
java.lang.Number... parameters)
Overwrites Clusterer.cluster().
|
void |
contract(double scale,
double minimum_cluster_size)
This method performs a contraction of the data and clustering of the
contracted data.
|
java.util.Map<Clusterable,java.util.List<Clusterable>> |
run(java.util.List<Clusterable> centers,
int max_clusters,
int scale,
double minimum_cluster_size)
Convenience method to first clear isolated points, then contract data,
and finally cluster.
|
public ExtendedClusterer(java.util.List<Clusterable> data)
public java.util.Map<Clusterable,java.util.List<Clusterable>> run(java.util.List<Clusterable> centers, int max_clusters, int scale, double minimum_cluster_size)
max_clusters
- The maximum number of clusters.scale
- Maximum distance of neighbors.minimum_cluster_size
- Minimum number of neighbors. If this is < 1, it is
interpreted as a fraction of the number of data points.public void clear(double scale, double n_points)
scale
- Maximum distance of neighbors.n_points
- Minimum number of neighbors. If this is < 1, it is
interpreted as a fraction of the number of data points.public void contract(double scale, double minimum_cluster_size) throws NotImplementedException
scale
- Scale at which contraction seizes.minimum_cluster_size
- Minimum size of resultant clusters. If this
is < 1, it is interpreted as a fraction of
the number of data points.NotImplementedException
public java.util.Map<Clusterable,java.util.List<Clusterable>> cluster(java.util.List<Clusterable> centers, int max, int type, java.lang.Number... parameters) throws NotImplementedException
cluster
in class Clusterer
centers
- A list of initial centroids.max
- The maximum number of centroids at which the algorithm is stopped.type
- The method used to determine the optimal number of clusters.parameters
- Mode specific parameters.NotImplementedException