evmSim function

MCMC simulation around an evmOpt fit

evmSim(
  o,
  priorParameters,
  prop.dist,
  jump.const,
  jump.cov,
  iter,
  start,
  thin,
  burn,
  chains,
  export = NULL,
  verbose,
  trace,
  theCall,
  ...
)

Arguments

• o: a fit evmOpt object

• priorParameters: A list with two components. The first should be a vector of means, the second should be a covariance matrix if the penalty/prior is "gaussian" or "quadratic" and a diagonal precision matrix if the penalty/prior is "lasso", "L1" or "Laplace". If method = "simulate" then these represent the parameters in the Gaussian prior distribution. If method = 'optimize' then these represent the parameters in the penalty function. If not supplied: all default prior means are zero; all default prior variances are $10^4$; all covariances are zero.

• prop.dist: The proposal distribution to use, either multivariate gaussian or a multivariate Cauchy.

• jump.const: Control parameter for the Metropolis algorithm.

• jump.cov: Covariance matrix for proposal distribution of Metropolis algorithm. This is scaled by jump.const.

• iter: Number of simulations to generate

• start: Starting values for the chain; if missing, defaults to the MAP/ML estimates in o.

• thin: The degree of thinning of the resulting Markov chains.

• burn: The number of initial steps to be discarded.

• chains: The number of Markov chains to run. Defaults to 1. If you run more, the function will try to figure out how to do it in parallel using the same number of cores as chains.

• export: Character vector of names of variables to export. See the help file for parallel::export. Defaults to export = NULL

  and most users will never need to use it. Only matters on Windows.

• verbose: Whether or not to print progress to screen. Defaults to verbose=TRUE.

• trace: How frequently to talk to the user

• theCall: (internal use only)

• ...: ignored

Returns

an object of class evmSim:

• call: The call to evmSim that produced the object.

• threshold: The threshold above which the model was fit.

• map: The point estimates found by maximum penalized likelihood and which were used as the starting point for the Markov chain. This is of class evmOpt and methods for this class (such as resid and plot) may be useful.

• burn: The number of steps of the Markov chain that are to be treated as the burn-in and not used in inferences.

• thin: The degree of thinning used.

• chains: The entire Markov chain generated by the Metropolis algorithm.

• y: The response data above the threshold for fitting.

• seed: The seed used by the random number generator.

• param: The remainder of the chain after deleting the burn-in and applying any thinning.

Note

it is not expected that the user should call this directly