Bayesian Methods in Health Economics

178 Bayesian Methods in Health Economics

mean and standard deviation and then compute the precision (required by

JAGS/BUGS) as the reciprocal of the variance.

Finally, we use the function lognPar (cfr. §4.7) to compute the parameters

of the distribution of the length of time with influenza, which is modelled on

the log scale.

The model can now be run calling the function JAGS, which we do using

the following code.

library(R2jags)

dataJags <- list("S","H","r0","r1","n0","n1","x","m",

"mu.inf","tau.inf","mu.l","tau.l")

filein <- "EvSynth.txt"

params <- c("p0","p1","rho","l","c.inf","alpha","delta","gamma")

inits <- function(){

list(alpha=rnorm(S,0,1),delta=rnorm(S,0,1),mu.delta=rnorm(1),

sigma.delta=runif(1),gamma=rnorm(H,0,1),mu.gamma=rnorm(1),

sigma.gamma=runif ...