Package xal.tools.statistics

Class RunningWeightedStatistics

java.lang.Object

xal.tools.statistics.UnivariateStatistics

xal.tools.statistics.RunningWeightedStatistics

public class RunningWeightedStatistics
extends UnivariateStatistics

Calculate running statistics using weighted averaging

Field Summary

Fields

Modifier and Type	Field	Description
protected final int	asymptoticPopulation	population beyond which we should use the asymptotic weight
protected final double	asymptoticWeight	weight for new samples

Fields inherited from class xal.tools.statistics.UnivariateStatistics

mean, meanSquare, population

Constructor Summary

Constructors

Constructor	Description
RunningWeightedStatistics(double weight)	Constructor

Method Summary

Modifier and Type	Method	Description
void	addSample(double value)	Add a new sample measurement.
void	clear()	Clear the samples
protected final double	getWeight()	Get the current weight.
protected final double	getWeight(double population)	Get the weight to apply to the next new sample given the specified population.
double	sampleVarianceOfMean()	Get the variance of the mean from the actual value assuming the supporting data is a random subset of all the data.
double	varianceOfMean()	Get the variance of the mean from the actual value.

Methods inherited from class xal.tools.statistics.UnivariateStatistics

mean, meanSquare, population, sampleStandardDeviation, sampleStandardDeviationOfMean, sampleVariance, standardDeviation, standardDeviationOfMean, variance

Methods inherited from class java.lang.Object

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

Field Details

asymptoticWeight

protected final double asymptoticWeight

weight for new samples

asymptoticPopulation

protected final int asymptoticPopulation

population beyond which we should use the asymptotic weight

Constructor Details

RunningWeightedStatistics

public RunningWeightedStatistics(double weight)

Constructor

Parameters:

weight - asymptotic weight to apply to new samples in large populations

Method Details

clear

public void clear()

Clear the samples

getWeight

protected final double getWeight()

Get the current weight.

Returns:

the current weight

getWeight

protected final double getWeight(double population)

Get the weight to apply to the next new sample given the specified population. When we first start adding samples, we must trust the newest samples most and gradually shift trust to the older samples. Ultimately we will trust new samples at least at the asymptotic weight.

addSample

public void addSample(double value)

Add a new sample measurement.

Parameters:

value - The new sample measurement

varianceOfMean

public double varianceOfMean()

Get the variance of the mean from the actual value.

Overrides:

varianceOfMean in class UnivariateStatistics

Returns:

the variance of the mean from the actual value

sampleVarianceOfMean

public double sampleVarianceOfMean()

Get the variance of the mean from the actual value assuming the supporting data is a random subset of all the data.

Overrides:

sampleVarianceOfMean in class UnivariateStatistics

Returns:

the sample variance of the mean