de.lmu.ifi.dbs.elki.evaluation.histogram

Class ComputeOutlierHistogram

java.lang.Object

de.lmu.ifi.dbs.elki.evaluation.histogram.ComputeOutlierHistogram

All Implemented Interfaces:

Evaluator, ResultProcessor

public class ComputeOutlierHistogram
extends Object
implements Evaluator

Compute a Histogram to evaluate a ranking algorithm. The parameter -hist.positive specifies the class label of "positive" hits.

Author:

Lisa Reichert, Erich Schubert

Nested Class Summary

Nested Classes

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

Field Summary

Fields

Modifier and Type	Field and Description
private int	bins Number of bins
static OptionID	BINS_ID number of bins for the histogram Default value: EuclideanDistanceFunction Key: -comphist.bins
static OptionID	POSITIVE_CLASS_NAME_ID The object pattern to identify positive classes Key: -comphist.positive
private Pattern	positiveClassName Stores the "positive" class.
private ScalingFunction	scaling Scaling function to use
static OptionID	SCALING_ID Parameter to specify a scaling function to use.
private boolean	splitfreq Flag to make split frequencies
static OptionID	SPLITFREQ_ID Flag to count frequencies of outliers and non-outliers separately Key: -histogram.splitfreq

Constructor Summary

Constructors

Constructor and Description
ComputeOutlierHistogram(Pattern positive_class_name, int bins, ScalingFunction scaling, boolean splitfreq) Constructor.

Method Summary

Methods

Modifier and Type	Method and Description
HistogramResult<DoubleVector>	evaluateOutlierResult(Database database, OutlierResult or) Evaluate a single outlier result as histogram.
void	processNewResult(ResultHierarchy hier, Result newResult) Process a result.

Methods inherited from class java.lang.Object

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

Field Detail

POSITIVE_CLASS_NAME_ID

public static final OptionID POSITIVE_CLASS_NAME_ID

The object pattern to identify positive classes

Key: -comphist.positive

BINS_ID

public static final OptionID BINS_ID

number of bins for the histogram

Default value: EuclideanDistanceFunction

Key: -comphist.bins

SCALING_ID

public static final OptionID SCALING_ID

Parameter to specify a scaling function to use.

Key: -comphist.scaling

SPLITFREQ_ID

public static final OptionID SPLITFREQ_ID

Flag to count frequencies of outliers and non-outliers separately

Key: -histogram.splitfreq

positiveClassName

private Pattern positiveClassName

Stores the "positive" class.

bins

private int bins

Number of bins

scaling

private ScalingFunction scaling

Scaling function to use

splitfreq

private boolean splitfreq

Flag to make split frequencies

Constructor Detail

ComputeOutlierHistogram

public ComputeOutlierHistogram(Pattern positive_class_name,
                       int bins,
                       ScalingFunction scaling,
                       boolean splitfreq)

Constructor.

Parameters:

positive_class_name - Class name

bins - Bins

scaling - Scaling

splitfreq - Scale inlier and outlier frequencies independently

Method Detail

evaluateOutlierResult

public HistogramResult<DoubleVector> evaluateOutlierResult(Database database,
                                                  OutlierResult or)

Evaluate a single outlier result as histogram.

Parameters:

database - Database to process

or - Outlier result

Returns:

Result

processNewResult

public void processNewResult(ResultHierarchy hier,
                    Result newResult)

Description copied from interface: ResultProcessor

Process a result.

Specified by:

processNewResult in interface ResultProcessor

Parameters:

hier - The base of the result tree.

newResult - Newly added result subtree.