Package elki.math.statistics.distribution.estimator

Class LogNormalLogMADEstimator

java.lang.Object

elki.math.statistics.distribution.estimator.LogNormalLogMADEstimator

All Implemented Interfaces:

DistributionEstimator<LogNormalDistribution>, LogMADDistributionEstimator<LogNormalDistribution>

@Reference(authors="F. R. Hampel",
           title="The Influence Curve and Its Role in Robust Estimation",
           booktitle="Journal of the American Statistical Association, June 1974, Vol. 69, No. 346",
           url="https://doi.org/10.2307/2285666",
           bibkey="doi:10.2307/2285666")
public class LogNormalLogMADEstimator
extends java.lang.Object
implements LogMADDistributionEstimator<LogNormalDistribution>

Estimator using Medians. More robust to outliers, and just slightly more expensive (needs to copy the data for partial sorting to find the median).

References:

F. R. Hampel
The Influence Curve and Its Role in Robust Estimation
Journal of the American Statistical Association, June 1974, Vol. 69, No. 346

P. J. Rousseeuw, C. Croux
Alternatives to the Median Absolute Deviation
Journal of the American Statistical Association, December 1993, Vol. 88, No. 424, Theory and Methods

Since:

0.6.0

Author:

Erich Schubert

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	LogNormalLogMADEstimator.Par	Parameterization class.

Field Summary

Fields

Modifier and Type	Field	Description
static LogNormalLogMADEstimator	STATIC	Static estimator, more robust to outliers by using the median.

Constructor Summary

Constructors

Modifier	Constructor	Description
private	LogNormalLogMADEstimator()	Constructor.

Method Summary

Modifier and Type	Method	Description
LogNormalDistribution	estimateFromLogMedianMAD(double median, double mad, double shift)	General form of the parameter estimation
java.lang.Class<? super LogNormalDistribution>	getDistributionClass()	Get the class that is produced by the estimator.
java.lang.String	toString()

Methods inherited from class java.lang.Object

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

Methods inherited from interface elki.math.statistics.distribution.estimator.DistributionEstimator

estimate

Methods inherited from interface elki.math.statistics.distribution.estimator.LogMADDistributionEstimator

estimate

Field Detail

STATIC

public static final LogNormalLogMADEstimator STATIC

Static estimator, more robust to outliers by using the median.

Constructor Detail

LogNormalLogMADEstimator

private LogNormalLogMADEstimator()

Constructor. Private: use static instance!

Method Detail

estimateFromLogMedianMAD

public LogNormalDistribution estimateFromLogMedianMAD(double median,
                                                      double mad,
                                                      double shift)

Description copied from interface: LogMADDistributionEstimator

General form of the parameter estimation

Specified by:

estimateFromLogMedianMAD in interface LogMADDistributionEstimator<LogNormalDistribution>

Parameters:

median - Median lof log values.

mad - Median absolute deviation from median (in logspace).

shift - Shift offset that was used to avoid negative values.

Returns:

Estimated distribution

getDistributionClass

public java.lang.Class<? super LogNormalDistribution> getDistributionClass()

Description copied from interface: DistributionEstimator

Get the class that is produced by the estimator.

Specified by:

getDistributionClass in interface DistributionEstimator<LogNormalDistribution>

Returns:

Distribution class

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object