PrepareINLAFuncContLikApprox: Prepare for running INLA function with continuous likelihood...

In PointProcess/SealPupProduction-JRSSC-code:

Description

Gathers and stacks the input to the INLA function

Usage

PrepareINLAFuncContLikApprox(
  mesh,
  intPoints,
  obsPoints,
  covGrid,
  spatial = TRUE,
  use.covariates = TRUE,
  covariate.fitting = "quadratic",
  additional.iid.term = FALSE,
  Matern.alpha = 2,
  covariates.type = "band1",
  INLA.theta.startval = NULL
)

Arguments

mesh	Mesh object, being the output from inla.mesh.2d
intPoints	Data frame with the integration point locations and their associated weights
obsPoints	Data frame with the point pattern locations (locations with at least one seal observed)
covGrid,	im object representing the covariate values on a dense grid where the counts live.
spatial	Logical indicating whether a spatial spde model should be used
use.covariates	Logical, indicating whether covariates should be used in the fitting (default is true)
covariate.fitting	String, indicating how to model covariates. "linear", quadratic (default) or "linAndLog", or FALSE for no covariates
additional.iid.term	Logical, indicating whether to include an additional iid (Gaussian) term in the latent field specification. FALSE is default
Matern.alpha	Numeric, corresponding to the alpha parameter in the Matern covariance function (2 is the default)
covariates.type	String equal to "band1" or "band2" indicating which of the bands from the satellite data should be used.
INLA.theta.startval	List containing the start values for the theta parameter in the INLA run. (NULL indicates that automatic start values should be used, and is the default)

Value

List with all variables necessary to run the INLA function, in addition to the spde object

PointProcess/SealPupProduction-JRSSC-code documentation built on Jan. 27, 2020, 10:06 p.m.