A convenient wrapper function (posterior) to sample the posterior using MCMC in rstan and returns a stanfit object for diagnostics. Use RCMstan to update the RCM and the enclosed operating model with MCMC samples..

posterior(x, ...)

# S4 method for RCModel
posterior(
  x,
  priors_only = FALSE,
  laplace = FALSE,
  chains = 2,
  iter = 2000,
  warmup = floor(iter/2),
  thin = 5,
  seed = 34,
  init = "last.par.best",
  cores = chains,
  ...
)

# S4 method for Assessment
posterior(x, priors_only = FALSE, ...)

RCMstan(RCModel, stanfit, sim, cores = 1)

Arguments

x

An object of class Assessment or RCModel.

...

Additional arguments to pass to rstan::sampling via tmbstan::tmbstan.

priors_only

Logical, whether to set the likelihood to zero and sample the priors only.

laplace

Logical, whether to do the Laplace approximation for random parameters.

chains

The numer of MCMC chains.

iter

The number of iterations for each chain, including warmup.

warmup

The number of burnin iterations

thin

The frequency at which iterations are kept (e.g., 5 saves every fifth iteration)

seed

Seed for random number generator during the MCMC.

init

The initial values of parameters for starting the MCMC chain. See tmbstan::tmbstan.

cores

The number of cores for running in parallel, e.g., one core per MCMC chain. Used in RCMstan for reconstructing the population.

RCModel

An object of class RCModel

stanfit

An object of class stanfit returned by posterior.

sim

A matrix of RCModel@OM@nsim rows and 2 columns that specifies the samples used to update the operating model. The first column specifies the chain and the second columns specifies the MCMC iteration.

Value

posterior returns an object of class stanfit. See class?stanfit.

RCMstan returns an updated RCModel.

Author

Q. Huynh