de.lmu.ifi.dbs.elki.database
Class LabelsFromClustering
java.lang.Object
de.lmu.ifi.dbs.elki.database.LabelsFromClustering
public class LabelsFromClustering
- extends Object
This class generates a labeled database based on an existing clustering result.
- Author:
- Erich Schubert
Method Summary |
|
makeDatabaseFromClustering(Database<O> olddb,
R clustering,
Class<L> classLabel)
Retrieve a cloned database that
- does not contain noise points
- has labels assigned based on the given clustering
Useful for e.g. training a classifier based on a clustering. |
Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
label_prefix
private String label_prefix
LabelsFromClustering
public LabelsFromClustering()
makeDatabaseFromClustering
public <O extends DatabaseObject,R extends Clustering<M>,M extends Model,L extends ClassLabel> Database<O> makeDatabaseFromClustering(Database<O> olddb,
R clustering,
Class<L> classLabel)
throws UnableToComplyException
- Retrieve a cloned database that
- does not contain noise points
- has labels assigned based on the given clustering
Useful for e.g. training a classifier based on a clustering.
- Type Parameters:
O
- Database object typeR
- Clustering ResultM
- Cluster Model classL
- Label class- Parameters:
olddb
- database the original objects come fromclustering
- clustering to work onclassLabel
- label class to use.
- Returns:
- new database
- Throws:
UnableToComplyException