strnorm: Simulate a Spatio-Temporal Random Variable

In michaeldumelle/DumelleEtAl2021STLMM: Fit Spatio-Temporal Linear Mixed Models

Description

Simulate a Spatio-Temporal Random Variable

Usage

strnorm(object, mu, size, condition = 1e-04, error, ...)

## S3 method for class 'matrix'
strnorm(object, mu, size, condition = 1e-04, error, ...)

## Default S3 method:
strnorm(
  object,
  mu,
  size,
  condition = 1e-04,
  error,
  xcoord,
  ycoord = NULL,
  tcoord,
  data,
  s_cor,
  t_cor,
  chol = FALSE,
  h_options = NULL,
  ...
)

Arguments

object	An covariance matrix or a "covparam" object.
mu	A mean vector with length equal to the number of rows in data or a scalar.
size	The number of independent simulations
condition	A small number added to the diagonals of matrices before inverting them to prevent ill-conditioning (defaults to 1e-4).
error	The random error type normal All random effects are Gaussian and mutually independent. component_squared Gaussian, mutually independent random effects are simulated, squared, and rescaled to match the sample variance on the Gaussian scale. sum_squared Gaussian, mutually independent random effects are simulated, summed, squared, and rescaled to match the sample variance on the Gaussian scale.
...	Additional arguments.
xcoord	A character vector specifying the column name of the x-coordinate variable in data.
ycoord	A character vector specifying the column name of the y-coordinate variable in data.
tcoord	A character vector specifying the column name of the t-coordinate (time) variable in data.
data	A data object containing all necessary variables.
s_cor	The spatial correlation exponential The exponential correlation (the default). spherical The spherical correlation. gaussian The Gaussian correlation.
t_cor	The temporal correlation exponential The exponential correlation (the default). spherical The spherical correlation. gaussian The Gaussian correlation. tent The tent (linear with sill) correlation.
chol	Should the Cholesky decomposition be used? If FALSE, efficient inversion algorithms are implemented. Defaults to FALSE.
h_options	A list containing options to compute distances if response, xcoord, ycoord, and tcoord are provided. Named arguments are h_t_distmetric The temporal distance matrix (defaults to "euclidean"). h_s_distmetric The spatial distance matrix (defaults to "euclidean").

Value

A vector of random variables (if size = 1) or a matrix of random variables (if size > 1) whose columns indicate seprate simulations. The row order corresponds to the rows of the covariance matrix (if object is a matrix) or the rows of data (if object is a covparam object.

michaeldumelle/DumelleEtAl2021STLMM documentation built on Dec. 21, 2021, 5:56 p.m.