ars: Adaptive Rejection Sampler

In vdchoi/ars: Adaptive Rejection Sampler

Description

Sample from a log-concave density function，potentially unnormalized, using the Adaptive Rejection Sampling method described in Gilks et al (1992)

Usage

ars(g, a, b, N, n_per_step = 500)

Arguments

g	input density function
a	lower bound of function domain D (could be -Inf)
b	upper bound of function domain D (could be Inf)
N	number of observations required
n_per_step	Default = 500. Number of observations between each update of the hull function

Value

A one-dimensional vector of N independent observations from the the given input density

Examples

g <- function (x) {
return(dnorm(x))
}
ars_results <- ars(g, a = -Inf, b= Inf, N = 100000)

vdchoi/ars documentation built on Jan. 1, 2021, 12:36 p.m.