barrier: Functions for defining the Barrier models

In inbo/INLA: Full Bayesian Analysis of Latent Gaussian Models using Integrated Nested Laplace Approximations

Description

Functions for defining Barrier models as an inla rgeneric model

Usage

 inla.barrier.pcmatern(mesh, barrier.triangles, prior.range,
                       prior.sigma, range.fraction=0.2)
 inla.barrier.polygon(mesh, barrier.triangles, Omega=NULL)
 inla.barrier.q(fem, ranges, sigma=1)
 inla.barrier.fem(mesh, barrier.triangles, Omega=NULL)
 

Arguments

mesh	The mesh to build the model on, from inla.mesh.2d
barrier.triangles	The numerical ids of the triangles that make up the barrier area
prior.range	2 parameters (range0,Prange) for the prior spatial range. If Prange is NA, then range0 is used as a fixed range value (TODO).
prior.sigma	2 parameters (sig0,Psig) for the prior marginal standard deviation sigma. If Psig is NA, then sig0 is used as a fixed sigma value (TODO).
range.fraction	The length of the spatial range inside the barrier area, as a fraction of the range parameter.
Omega	Advanced option for creating a set of permeable barriers (not documented)

Details

This model is described in the ArXiv preprint arXiv:1608.03787. For examples, see https://haakonbakka.bitbucket.io/btopic107.html.

Value

inla.barrier.pcmatern gives the (rgeneric) model object for fitting the model in INLA, inla.barrier.polygon gives the polygon around the barrier (mainly for plotting), inla.barrier.q is an internal method producing the Q matrix from a result of inla.barrier.fem, inla.barrier.fem is an internal method producing the Finite Element matrices.

Author(s)

Haakon Bakka [e-mail]

See Also

inla.spde2.pcmatern

inbo/INLA documentation built on Dec. 6, 2019, 9:51 a.m.