nextLOCsample: Generate a random sample of points

In qle: Simulation-Based Quasi-Likelihood Estimation

Description

Generate a random sample of points as a set of candidate points for evaluation

Usage

nextLOCsample(S, x, n, lb, ub, pmin = 0.05, invert = FALSE)

Arguments

S	variance matrix of sample points (usually chosen as the information matrix)
x	an approximate root as the mean value if the multivariate normal distribution
n	number of points to randomly sample
lb	vector of lower bounds of the (hyper)box
ub	vector of upper bounds of the (hyper)box
pmin	minimum required ratio of points falling inside the hyperbox of parameters
invert	optional, invert=FALSE (default) for no inversion of 'S'

Details

The function generates a random sample of points with mean and variance given by 'x' and 'S', respectively, according to a (truncated) multivariate normal distribution (using rejection sampling) to match the parameter space given by the lower and upper bound vectors.

Value

Matrix of sampled locations.

Author(s)

M. Baaske

Examples

 X <- nextLOCsample(matrix(c(1,0,0,1),nr=2), c(0,0), 10, c(-0.5,-0.5), c(0.5,0.5))
 

qle documentation built on May 2, 2019, 9:55 a.m.