Package org.tribuo.clustering.hdbscan

Class HdbscanOptions

java.lang.Object

org.tribuo.clustering.hdbscan.HdbscanOptions

All Implemented Interfaces:

com.oracle.labs.mlrg.olcut.config.Options

public final class HdbscanOptions
extends Object
implements com.oracle.labs.mlrg.olcut.config.Options

OLCUT Options for the HDBSCAN* implementation.

Field Summary

Fields

Modifier and Type	Field	Description
DistanceType	distType	Distance function in HDBSCAN*.
int	k	The number of nearest-neighbors to use in the initial density approximation.
int	minClusterSize	The minimum number of points required to form a cluster.
NeighboursQueryFactoryType	nqFactoryType	The nearest neighbour implementation factory to use.
int	numThreads	Number of threads to use for training the hdbscan model.

Fields inherited from interface com.oracle.labs.mlrg.olcut.config.Options

header

Constructor Summary

Constructors

Constructor	Description
HdbscanOptions()

Method Summary

Modifier and Type	Method	Description
String	getOptionsDescription()
HdbscanTrainer	getTrainer()	Gets the configured HdbscanTrainer using the options in this object.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

minClusterSize

@Option(longName="hdbscan-minimum-cluster-size",
        usage="The minimum number of points required to form a cluster.")
public int minClusterSize

The minimum number of points required to form a cluster. Defaults to 5.

distType

@Option(longName="hdbscan-distance-type",
        usage="The type of distance function to use for various distance calculations.")
public DistanceType distType

Distance function in HDBSCAN*. Defaults to L2 (EUCLIDEAN).

k

@Option(longName="hdbscan-k-nearest-neighbors",
        usage="The number of nearest-neighbors to use in the initial density approximation. The value includes the point itself.")
public int k

The number of nearest-neighbors to use in the initial density approximation. Defaults to 5.

numThreads

@Option(longName="hdbscan-num-threads",
        usage="Number of threads to use for training the hdbscan model.")
public int numThreads

Number of threads to use for training the hdbscan model. Defaults to 2.

nqFactoryType

@Option(longName="hdbscan-neighbour-query-factory-type",
        usage="The nearest neighbour implementation factory to use.")
public NeighboursQueryFactoryType nqFactoryType

The nearest neighbour implementation factory to use. Defaults to NeighboursQueryFactoryType.KD_TREE.

Constructor Details

HdbscanOptions

public HdbscanOptions()

Method Details

getOptionsDescription

public String getOptionsDescription()

Specified by:

getOptionsDescription in interface com.oracle.labs.mlrg.olcut.config.Options

getTrainer

public HdbscanTrainer getTrainer()

Gets the configured HdbscanTrainer using the options in this object.

Returns:

A HdbscanTrainer.