PAWL (version 0.5)

createAdaptiveRandomWalkProposal: Adaptive Random Walk proposal distribution for MCMC algorithms

Description

Create the adaptive gaussian random walk proposal that is used as a default in adaptiveMH and pawl, whenever the target distribution is continuous.

Usage

createAdaptiveRandomWalkProposal(nchains, targetdimension, adaptiveproposal, adaptationrate, sigma_init)

Arguments

nchains

Object of class "numeric": it should be an integer representing the desired number of parallel chains.

targetdimension

Object of class "numeric": it should be an integer representing the dimension of the target distribution.

adaptiveproposal

Object of class "logical": specifies whether an adaptive proposal (Robbins-Monroe type of adaptation) should be used. Default is FALSE.

adaptationrate

Object of class "function": specifies the rate at which the adaptation of the proposal is performed. It should be a function defined on [0, + infty[ such that it is not integrable but its square is integrable, e.g. t -> 1/t for instance. The default is t -> t^-0.6.

sigma_init

Object of class "numeric": it should be a positive real number specifying the standard deviation of the proposal distribution at the first iteration. If the proposal is adaptive, it acts as a starting point for the adaptation. If it is not adaptive, then this value is used throughout all the iterations. Default is 1.

Value

The function returns an object of class proposal-class, to be used in calls to adaptiveMH and pawl.

See Also

proposal-class, adaptiveMH, pawl