metropolis_step: Metropolis Step

In lucas-castillo/SampleR: Tools To Compare Human Performance To Sampling Algorithms

Description

Helper function for MCMC-based samplers.Calculates a proposed move from the current position and decides whether the chain should move there based on a Metropolis-Hastings acceptance rule.

Usage

metropolis_step(current_x, pdf, proposal_f, beta = 1)

Arguments

current_x	vector. Current position of the chain
pdf	Probability density function of the target distribution
proposal_f	Function from which to draw proposals, given a mean
beta	Temperature of the chain. Acceptance of the proposed state becomes more likely as beta approaches 0.

Value

List with the position for the next step of the chain and a boolean representing whether the proposal has been accepted or not

Examples

# This function is used internally in MCMC based samplers

pdf_function <- function(x){mvtnorm::dmvnorm(x, mean = c(0,1), sigma = diag(2))}
proposal_f <- function(mu){mvtnorm::rmvnorm(1, mean = mu, sigma = diag(2))}
current_position = c(2,3)
metropolis_step(current_position,pdf_function,proposal_f)

lucas-castillo/SampleR documentation built on Jan. 1, 2021, 8:25 a.m.