adaptive.mcmc: Adaptive MCMC algorithm

In MJomaba/flu-evidence-synthesis: Flu Evidence Synthesis: Analysing Effectiveness of Vaccine Strategies

Adaptive MCMC algorithm

Description

MCMC which adapts its proposal distribution for faster convergence following: Sherlock, C., Fearnhead, P. and Roberts, G.O. The Random Walk Metrolopois: Linking Theory and Practice Through a Case Study. Statistical Science 25, no.2 (2010): 172-190.

Usage

adaptive.mcmc(lprior, llikelihood, nburn, initial, nbatch, blen = 1,
  outfun = NULL, acceptfun = NULL, verbose = FALSE, ...)

Arguments

lprior	A function returning the log prior probability of the parameters
llikelihood	A function returning the log likelihood of the parameters given the data
nburn	Number of iterations of burn in
initial	Vector with starting parameter values
nbatch	Number of batches to run (number of samples to return)
blen	Length of each batch
outfun	A function that is called for each batch. Can be useful to log certain values.
acceptfun	A function that is called whenever a sample is accepted.
verbose	Output debugging information
...	Extra parameters passed to the log likelihood function

Value

Returns a list with the accepted samples and the corresponding llikelihood values and the return of the optional outfun

See Also

adaptive.mcmc.cpp Used internally by this function.

MJomaba/flu-evidence-synthesis documentation built on April 26, 2022, 11:12 p.m.