fit.lgcp: Function to fit a spatio-temporal log-Gaussian Cox process...

Description Usage Arguments Value

View source: R/fit_lgcp.r

Description

Function to fit a spatio-temporal log-Gaussian Cox process model with an intercept and covariates (optional)

Usage

1
2
3
4
5
6
fit.lgcp(mesh = NULL, mesh.pars = NULL, locs = NULL, temp = NULL,
  covariates = NULL, prior.rho = list(theta = list(prior = "pccor1", param =
  c(0, 0.9))), prior.range = c(400, 0.5), prior.sigma = c(1, 0.05),
  verbose = FALSE, control.inla = list(strategy = "gaussian", int.strategy =
  "eb"), control.fixed = list(prec.intercept = 0.001),
  return.attributes = FALSE, ns = NULL)

Arguments

mesh

a “mesh” object i.e. delauney triangulation of the domain, an object returned by make.mesh.

mesh.pars

a named vertor of mesh parameters, must contain cutoff length at which to cut off triangle edge lengths, min triangle edge length inside region, and max triangle edge length outside region.

locs

a matrix of observation locations, where each row corresponds to the observation.

temp

a numeric vector specifying a temporal index for each observation (starting at 1.....T).

covariates

a named data.frame of covariates

prior.rho

prior for the temporal correlation coefficient, by default a pcprior is used with param=c(0,0.9).

prior.range

pc prior for the range of the latent field (rnage0,Prange) (i.e., P(range < range0) = Prange NOTE should be changed to reflect range of the domain by default this is 400m

prior.sigma

pc prior for the sd of the latent field (sigma0,Psigma) by default c(1,0.05) i.e., prob sigma > 1 = 0.05

verbose

Logical if TRUE model fit is output to screen.

control.inla

a list to control model fitting (as per inla)

control.fixed

a list as per inla by default sets prior for precision intercept

response

a vector of response variable, each corresponds to the spatial locations in locs.

Value

A inla result object


cmjt/lgcpSPDE documentation built on July 25, 2019, 3:05 p.m.