neg2loglikDiagFactory {GeneralizedWendland}

Diagnostics for arbitrarily specified, likelihood-based Gaussian process models

Description

A helper function for rapidly exploring the parameter space around the maximum likelihood estimate

Usage

neg2loglikDiagFactory(y, X = data.frame(), distmat, covariance, ...)

Arguments

y

Dependent variable

X

Optional design matrix with covariates

distmat

Distance matrix. Can be provided either as a dense matrix or spam object.

covariance

Covariance function.

...

Other arguments to be passed on.

Details

theta_list

Named list of vectors with parameters to be passed to covariance.

cov.args_list (default = list())

Named list of vectors with arguments to be passed to covariance

chol.args_list (default = list())

Named list of vectors with arguments to be passed to choleskyFactory.

Value

Returns a function of the form function(theta_list, cov.args_list = list(), chol.args_list = list()) which returns a data.frame containing the neg2loglikelihood at all permutations of the provided arguments.

Note

The function manufactured by neg2loglikDiagFactory in principle also accepts covariance functions generated using covarianceFactory. However, the function is not yet compatible with the arguments fixed_range_value and fixed_nugget_value. For now, these should be left at default when using covarianceFactory.

Author(s)

Thomas Caspar Fischer

Examples

set.seed(63)
n <- 50
range <- 0.7
theta  <- c(range, 1, 1, 0, 0)

locs <- data.frame(x = runif(n), y = runif(n))
dmat  <- as.matrix(dist(locs))
Sigma <- cov.wendland(h = dmat, theta = theta)
y <- c(spam::rmvnorm(1, Sigma = Sigma))

neg2loglikIterator <- neg2loglikDiagFactory(y = y, distmat = dmat,
  covariance = cov.wendland)
theta_list <- list(range = 0.5, sill = 1, kappa = 0, mu = c(0, 0.25, 0.5),
  nugget = 0)
cov.args_list <- list(numint.abstol = c(1e-1, 1e-3, 1e-6), numint.reltol = c(1e-3))

results <- neg2loglikIterator(theta_list, cov.args_list = cov.args_list)