de.lmu.ifi.dbs.elki.evaluation.clustering.internal

Class EvaluateSquaredErrors

java.lang.Object

de.lmu.ifi.dbs.elki.evaluation.clustering.internal.EvaluateSquaredErrors

All Implemented Interfaces:

Evaluator, ResultProcessor

public class EvaluateSquaredErrors
extends java.lang.Object
implements Evaluator

Evaluate a clustering by reporting the squared errors (SSE, SSQ), as used by k-means. This should be used with SquaredEuclideanDistanceFunction only (when used with other distances, it will manually square the values; but beware that the result is less meaningful with other distance functions).

For clusterings that provide a cluster prototype object (e.g. k-means), the prototype will be used. For other algorithms, the centroid will be recomputed.

TODO: support non-vector based clusterings, too, if the algorithm provided a prototype object (e.g. PAM).

TODO: when combined with k-means, detect if the distance functions agree (both should be using squared Euclidean), and reuse the SSQ values provided by k-means.

Since:

0.7.0

Author:

Erich Schubert

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	EvaluateSquaredErrors.Parameterizer Parameterization class.

Field Summary

Fields

Modifier and Type	Field and Description
private NumberVectorDistanceFunction<?>	distance Distance function to use.
private java.lang.String	key Key for logging statistics.
private static Logging	LOG Logger for debug output.
private NoiseHandling	noiseOption Handling of Noise clusters

Constructor Summary

Constructors

Constructor and Description
EvaluateSquaredErrors(NumberVectorDistanceFunction<?> distance, NoiseHandling noiseOption) Constructor.

Method Summary

Modifier and Type	Method and Description
double	evaluateClustering(Database db, Relation<? extends NumberVector> rel, Clustering<?> c) Evaluate a single clustering.
void	processNewResult(ResultHierarchy hier, Result result) Process a result.

Methods inherited from class java.lang.Object

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

Field Detail

LOG

private static final Logging LOG

Logger for debug output.

noiseOption

private NoiseHandling noiseOption

Handling of Noise clusters

distance

private NumberVectorDistanceFunction<?> distance

Distance function to use.

key

private java.lang.String key

Key for logging statistics.

Constructor Detail

EvaluateSquaredErrors

public EvaluateSquaredErrors(NumberVectorDistanceFunction<?> distance,
                             NoiseHandling noiseOption)

Constructor.

Parameters:

distance - Distance function to use.

noiseOption - Control noise handling.

Method Detail

evaluateClustering

public double evaluateClustering(Database db,
                                 Relation<? extends NumberVector> rel,
                                 Clustering<?> c)

Evaluate a single clustering.

Parameters:

db - Database

rel - Data relation

c - Clustering

Returns:

ssq

processNewResult

public void processNewResult(ResultHierarchy hier,
                             Result result)

Description copied from interface: ResultProcessor

Process a result.

Specified by:

processNewResult in interface ResultProcessor

Parameters:

hier - The base of the result tree.

result - Newly added result subtree.