iprior (version 0.6.5)

fbmOptim: Find the Hurst coefficient of a FBM I-prior model

Description

From an ipriorKernel object, a golden section search of (0,1) is performed using optimize().

Usage

fbmOptim(object, method = c("ipriorOptim", "iprior"), maxit = 3,
  interval = c(0, 1), silent = FALSE)

Arguments

object

An object of class ipriorKernel.

method

One of c("ipriorOptim", "iprior") for model fitting of the final I-prior model.

maxit

If method is ipriorOptim, then this controls the number of initial EM steps before running a direction optimisation in optim. Defaults to 3 steps.

interval

It is possible to search within a specific range of values other than c(0, 1).

silent

(logical) Run the optimisation silently or not.

Value

An ipriorMod object.

Details

Currently, fbmOptim() will only search for a single optimum Hurst coefficient. If the model specified has multiple Hurst coefficients, this will be overridden with the single optimal Hurst value.

Examples

mod <- kernL(y ~ ., datfbm, model = list(kernel = "FBM"))
(mod.iprior <- fbmOptim(mod))