jagam_cartae.txt

model {
  eta <- X %*% b ## linear predictor
  for (i in 1:n) { mu[i] <-  exp(eta[i]) } ## expected response
  for (i in 1:n) { y[i] ~ dpois(mu[i]) } ## response 
  ## Parametric effect priors CHECK tau=1/29^2 is appropriate!
  for (i in 1:2) { b[i] ~ dnorm(0,0.0012) }
  ## prior for s(DOY)... 
  K1 <- S1[1:3,1:3] * lambda[1] 
  b[3:5] ~ dmnorm(zero[3:5],K1) 
  ## prior for te(DOY,precip_2weeks)... 
  K2 <- S2[1:22,1:22] * lambda[2]  + S2[1:22,23:44] * lambda[3] + S2[1:22,45:66] * lambda[4]
  b[6:27] ~ dmnorm(zero[6:27],K2) 
  ## prior for te(plot17aspect,plot17slope)... 
  K3 <- S3[1:24,1:24] * lambda[5]  + S3[1:24,25:48] * lambda[6] + S3[1:24,49:72] * lambda[7]
  b[28:51] ~ dmnorm(zero[28:51],K3) 
  ## prior for te(LAI_1718avg,trap_CHM)... 
  K4 <- S4[1:24,1:24] * lambda[8]  + S4[1:24,25:48] * lambda[9] + S4[1:24,49:72] * lambda[10]
  b[52:75] ~ dmnorm(zero[52:75],K4) 
  ## prior for s(collectDate)... 
  for (i in 76:99) { b[i] ~ dnorm(0, lambda[11]) }
  ## prior for s(col_year_fac)... 
  for (i in 100:103) { b[i] ~ dnorm(0, lambda[12]) }
  ## prior for s(plot_trap)... 
  for (i in 104:143) { b[i] ~ dnorm(0, lambda[13]) }
  ## prior for s(plotID)... 
  for (i in 144:153) { b[i] ~ dnorm(0, lambda[14]) }
  ## smoothing parameter priors CHECK...
  for (i in 1:14) {
    lambda[i] ~ dgamma(.05,.005)
    rho[i] <- log(lambda[i])
  }
}