Get the precision matrix of rSPDEobj1d objects

Function to get the precision matrix of a rSPDEobj1d object

Usage

# S3 method for class 'rSPDEobj1d'
precision(
  object,
  user_loc = NULL,
  user_nu = NULL,
  user_kappa = NULL,
  user_sigma = NULL,
  user_range = NULL,
  user_tau = NULL,
  user_m = NULL,
  ordering = c("field", "location"),
  ldl = FALSE,
  ...
)

Arguments

object

The covariance-based rational SPDE approximation, computed using matern.rational()

user_loc

If non-null, update the locations where to evaluate the model.

user_nu

If non-null, update the shape parameter of the covariance function.

user_kappa

If non-null, update the range parameter of the covariance function.

user_sigma

If non-null, update the standard deviation of the covariance function.

user_range

If non-null, update the range parameter of the covariance function.

user_tau

If non-null, update the parameter tau.

user_m

If non-null, update the order of the rational approximation, which needs to be a positive integer.

ordering

Return the matrices ordered by field or by location?

ldl

Directly build the LDL factorization of the precision matrix?

...

Currently not used.

Value

A list containing the precision matrix Q of the process and its derivatives if they exist, and a matrix A that extracts the elements corresponding to the process. If ldl=TRUE, the LDL factorization is returned instead of Q. If the locations are not ordered, the precision matrix is given for the ordered locations, but the A matrix returns to the original order.

See also

simulate.rSPDEobj1d(), matern.rational()

Examples

# Compute the covariance-based rational approximation of a
# Gaussian process with a Matern covariance function on R
sigma <- 1
nu <- 0.8
range <- 0.2

# create mass and stiffness matrices for a FEM discretization
x <- seq(from = 0, to = 1, length.out = 101)

op_cov <- matern.rational(
  loc = x, nu = nu,
  range = range, sigma = sigma, m = 2,
  parameterization = "matern"
)

# Get the precision matrix:
prec_matrix <- precision(op_cov)