Auxiliar function to obtain domain-based initial values for log-likelihood optimization in rSPDE models with a latent stationary Gaussian Matern model
get.initial.values.rSPDE(
mesh = NULL,
mesh.range = NULL,
graph.obj = NULL,
n.spde = 1,
dim = NULL,
B.tau = NULL,
B.kappa = NULL,
B.sigma = NULL,
B.range = NULL,
nu = NULL,
parameterization = c("matern", "spde"),
include.nu = TRUE,
log.scale = TRUE,
nu.upper.bound = NULL
)
A vector of the form (theta_1,theta_2,theta_3) or where theta_1 is the initial guess for tau, theta_2 is the initial guess for kappa and theta_3 is the initial guess for nu.
An in INLA mesh
The range of the mesh.
A metric_graph
object. To be used in case both mesh
and mesh.range
are NULL
.
The number of basis functions in the mesh model.
The dimension of the domain.
Matrix with specification of log-linear model for \(\tau\). Will be used if parameterization = 'spde'
.
Matrix with specification of log-linear model for \(\kappa\). Will be used if parameterization = 'spde'
.
Matrix with specification of log-linear model for \(\sigma\). Will be used if parameterization = 'matern'
.
Matrix with specification of log-linear model for \(\rho\), which is a range-like parameter (it is exactly the range parameter in the stationary case). Will be used if parameterization = 'matern'
.
The smoothness parameter.
Which parameterization to use? matern
uses range, std. deviation and nu (smoothness). spde
uses kappa, tau and nu (smoothness). The default is matern
.
Should we also provide an initial guess for nu?
Should the results be provided in log scale?
Should an upper bound for nu be considered?