bym: Fit a Besag-York-Mollie Model using INLA

In diseasemappingDevel: Functions to compliment diseasemapping

Arguments

formula	the formula used to specify the model, default is "observed ~ 1", intercept only model
family	Distribution of responses, default is "poisson"
spPolygon	name of spatial polygon, with data attribute, containing observed and expected case per region
offset	offest variable name in the dataset, default is "logExpected"
ID	Identification column of the regions, which follows GMRF, default is "ID"
prior.Spatial	A string specify the prior distribution for spatail random effect, default is "c(1,0.01)"
prior.iid	A string specify the prior distribution for iid effect, default is "c(1,0.01)"
exceed	A number bigger than 1 for exceedance probability, default is 1.5, compute 50 \itemdebugPrint Debug information, default is FALSE \itemverbosePrint program excution process, default is TRUE \itemkeepKeep the data and result files, default is TRUE

It calls INLA packages, obtain the marginal distribution of the spatial random effect and the iid+spatial random effects A list with a spatial polygon object and an INLA object Patrick Brown data(popdata) data(casedata) model = getRates(casedata, popdata, ~age*sex, breaks=seq(0, 90, by=10) ) ontario = getSMR(popdata,model, casedata,regionCode="CSDUID", regionCodeCases="CSD2006") result =bym(formula=observed ~ 1,ID="CSDUID", family="poisson", spPolygon=ontario@data,offset = "logExpected") )

###Run BYM on grid library(spdep) nb8<-cell2nb(nrow=50, ncol=50, type="queen", torus=T) nb2inla("germany.graph",nb8) #Write graph file, then run bym function

INLA bym

diseasemappingDevel documentation built on May 2, 2019, 5:53 p.m.