java.lang.Object
org.palladiosimulator.monitorrepository.statisticalcharacterization.StatisticalCharacterizationAggregator
org.palladiosimulator.monitorrepository.statisticalcharacterization.GeometricMeanAggregator

public class GeometricMeanAggregator extends StatisticalCharacterizationAggregator
StatisticalCharacterizationAggregator corresponding to the GeometricMean model class which computes the geometric mean of a sequence of measurements (discrete case) or a (positive) function with respect to an interval (continuous case):
In the discrete case this is simply GM=(x1 * x2 * ... * xn) ^ 1/n with n being the number of measurements.
Note that it is not defined if any of the measurements is negative or exactly 0. In such a case this implementation returns 0.

In the continuous case, this aggregator approximates the integral GM=e^(∫log(f(x))dx ⁄ (b-a)) where a and b are upper and lower interval bound and the integration is carried out from a to b, and the (positive) function f is interpolated from a given (finite) set of samples (i.e., measurements).
See Also: