Distribution Listener

The Distribution and the DistributionListener classes are designed to process series data considered as a set of random numbers. They maintain computations of basic and commonly used descriptive statistics. These statistics (quantitative characteristics of data distribution) are usually divided into two groups:

1. Average characteristics to measure the "central tendency" of the data set or the middle of the "expected" value of the data set.

2. Deviation characteristics to describe volatility over the data set and dispersion around the average of the distribution.

Average characteristics are presented by:

Distribution: ArithmeticMean, GeometricMean, HarmonicMean, ExpectedValue, QuadraticMean, Median, Midrange, FirstRawMoment, SecondRawMoment, ThirdRawMoment, FourthRawMoment.

DistributionListener: ArithmeticMean, GeometricMean, HarmonicMean, ExpectedValue, QuadraticMean, Median, Midrange, FirstRawMoment, SecondRawMoment, ThirdRawMoment, FourthRawMoment.

Deviation characteristics are:

Distribution: Min, Max, Range, Variance, StandardDeviation, Kurtosis, CoefficientOfVariation, Skewness, SecondCentralMoment, ThirdCentralMoment, FourthCentralMoment.

DistributionListener: Min, Max, Range, Variance, StandardDeviation, Kurtosis, CoefficientOfVariation, Skewness, SecondCentralMoment, ThirdCentralMoment, FourthCentralMoment.

Also, these classes provide some convenience calculations that may be useful in user-specific cases:

Distribution: Sum, SumOfAbsoluteValues, SumOfSquares, CumulativeDistribution, GetOrderStatistic, GetQuantile.

DistributionListener: Sum, SumOfAbsoluteValues, SumOfSquares, CumulativeDistribution, GetOrderStatistic, GetQuantile.

The classes differ only in their usage: 

• The Distribution is addressed mostly to process static or increasing data sets.

• The DistributionListener is intended for use in conjunction with containers implementing data queue model to process dynamic data sets corresponding to sliding time window.