MCMC_fit_single: Bottom level MCMC wrapper function

In jameshay218/mcmcJH: Generic random walk MCMC algorithm for parameter inference

Description

Calls the run_metropolis_MCMC function to carry out the MCMC algorithm. This function handles parallelisation of chains and some basic formatting.

Usage

MCMC_fit_single(data, param_table, iterations = 1000, opt_freq = 50,
  thin = 50, burnin = 100, adaptive_period = 1, nchain = 1,
  popt = 0.44, tmp_filename, LIKELIHOOD_FUNCTION, MODEL_FUNCTION,
  VERBOSE = FALSE, PARALLEL = FALSE)

Arguments

data	the data over which to calculate likelihoods
param_table	table of parameter data as specified in load_param_table
iterations	number of iterations for the MCMC algorithm
opt_freq	how frequently the acceptance rate is adapted. Defaults to 50
thin	thinning value for the MCMC chain. Default is 50
burnin	the length of the burn in period. Defaults to 100
adaptive_period	length of the adaptive period. Defaults to 1
nchain	number of chains to run
popt	the desired acceptance rate. Defaults to 0.44
tmp_filename	the generic file name/location at which to save the MCMC chains. Note that the file extension will be appended to this string
LIKELIHOOD_FUNCTION	a valid pointer to an R function which returns a single, log likelihood of the data given the current parameters
MODEL_FUNCTION	a valid pointer to an R function which is used to evaluate the model for the current set of parameters
VERBOSE	boolean flag for additional output. Defaults to FALSE
PARALLEL	OPTIONAL boolean flag indicating if each MCMC chain should be run in parallel using doPar. Defaults to FALSE

Value

returns a list containing the whole MCMC chains, MCMC summary objects (from coda), the file locations of the MCMC chains, and the maximum likelihood parameters

See Also

run_metropolis_MCMC

jameshay218/mcmcJH documentation built on May 18, 2019, 11:20 a.m.