datagen: datagen

In VBmix: Variational Bayesian Mixture Models

Description

generates data from a random multivariate Gaussian, and adds redundant dimensions by random linear combinations with noise.

Usage

datagen(dreal = 2, deff = 6, npts = 200, noise = 0.1, genmean = rep(0, dreal),
genspan = 6, iso = FALSE)

Arguments

dreal	dimensionality of the multivariate Gaussian.
deff	dimensionality of the returned sample.
npts	number of elements to be sampled.
noise	noise magnitude for the linear combination.
genmean	mean of the multivariate Gaussian.
genspan	maximal magnitude of the diagonal elements in the covariance matrix. Non-diagonal elements are sampled under constraints of positive-definiteness.
iso	sample from an isotropic multivariate Gaussian (i.e. diagonal covariance matrix).

Value

matrix of sampled row-elements.

Author(s)

Pierrick Bruneau

Examples

temp <- datagen()

VBmix documentation built on May 30, 2017, 2:34 a.m.