Class hypervolume

Description

A class used to store a stochastic description of a hypervolume.

Objects from the Class

Objects can be created by calls of the form new("Hypervolume", ...).

Slots

Name:

Object of class "character". A string naming the hypervolume, used in plotting.

Data:

Object of class "matrix". If available, the raw data used to construct the hypervolume. Defaults to a one-row NaN vector for hypervolumes returned by set operations.

Dimensionality:

Object of class "numeric". The dimensionality of the hypervolume.

Volume:

Object of class "numeric". The volume of the hypervolume, in units of the product of all dimensions.

PointDensity:

Object of class "numeric". The number density of the uniformly sampled random points characterizing the hypervolume.

Bandwidth:

Object of class "numeric". If available, the bandwidth vector used to construct the hypervolume. Defaults to a one-row NaN vector for hypervolumes returned by set operations.

DisjunctFactor:

Object of class "numeric". The ratio of the inferred volume to the volume of a hypervolume constructed from the same data with disjunct data points (i.e. no kernels overlap). Varies from zero to one. High values suggest that bandwidth should be increased.

RepsPerPoint:

Object of class "numeric". If available, the number of random points used per observation to construct the hypervolume. Defaults to NaN for hypervolumes returned by set operations.

QuantileThresholdDesired:

Object of class "numeric". If available, the quantile requested by the user and used to construct the hypervolume. Defaults to NaN for hypervolumes returned by set operations.

QuantileThresholdObtained:

Object of class "numeric". If available, the quantile obtained by the hypervolume algorithm. Defaults to NaN for hypervolumes returned by set operations.

RandomUniformPointsThresholded:

Object of class "matrix" A set of uniformly random points guaranteed to be in the hypervolume.

ProbabilityDensityAtRandomUniformPoints:

Object of class "numeric" A vector of integers proportional to the probability density at each uniformly random point in the hypervolume. Defaults to a 1-valued vector for hypervolumes returned by set operations because set operations are well defined for volumes and not for probability density functions.

Methods

Summary and plot methods are available.