inla.spde.make.index: SPDE model index vector generation

In andrewzm/INLA: Functions which allow to perform full Bayesian analysis of latent Gaussian models using Integrated Nested Laplace Approximaxion

Description

Generates a list of named index vectors for an SPDE model.

Usage

inla.spde.make.index(name,
                     n.spde,
                     n.group = 1,
                     n.repl = 1,
                     ...)

Arguments

name	A character string with the base name of the effect.
n.spde	The size of the model, typically from spde$n.spde.
n.group	The size of the group model.
n.repl	The number of model replicates.
...	Additional parameters. Currently unused.

Value

A list of named index vectors.

name	Indices into the vector of latent variables
name.group	'group' indices
name.repl	Indices for replicates

Author(s)

Finn Lindgren [e-mail]

See Also

inla.spde.make.A, inla.spde2.result

Examples

loc = matrix(runif(100*2),100,2)
mesh = inla.mesh.create.helper(points.domain=loc, max.edge=c(0.1,0.5))
spde = inla.spde2.matern(mesh)
index = inla.spde.make.index("spatial", spde$n.spde, n.repl=2)
spatial.A = inla.spde.make.A(mesh, loc,
                             index=rep(1:nrow(loc), 2),
                             repl=rep(1:2, each=nrow(loc)))
y = 10+rnorm(100*2)
stack = inla.stack(data=list(y=y),
                   A=list(spatial.A),
                   effects=list(c(index, list(intercept=1))),
                   tag="tag")
data = inla.stack.data(stack, spde=spde)
formula = y ~ -1 + intercept + f(spatial, model=spde,
                                 replicate=spatial.repl)
result = inla(formula, family="gaussian", data=data, 
              control.predictor=list(A=inla.stack.A(stack)))
spde.result = inla.spde2.result(result, "spatial", spde)

andrewzm/INLA documentation built on May 10, 2019, 11:12 a.m.