# Standardize the data
data {
  for ( j in 1:n.x ) {
    x.mean[j]  <- mean(x[,j])
    x.sd[j] <-   sd(x[,j])
    for ( i in 1:n ) {
      zX[i,j] <- ( x[i,j] - x.mean[j] ) / x.sd[j]
    }
  }
}
# Specify the model for standardized data
model {
  for ( i in 1:n ) {
      y[i] ~ dcat( lambda.exp[1:q,i] )
    
    for ( j in 1:q ) {
      lambda.exp[j,i] <- exp( zbeta0[j] + sum( zbeta[j,1:n.x] * zX[i,1:n.x] ) )
    }
  }
  
  # Priors vague on standardized scale
  zbeta0[1] <- 0
  for ( k in 1:n.x ) { 
    zbeta[1,k] <- 0 
  }
  for ( l in 2:q ) {
    zbeta0[l] ~ dnorm(0 , 0.0001)
    for ( m in 1:n.x ) {
      zbeta[l,m] ~ dnorm(0 , 0.0001)
    }
  }
  
  # Transform to original scale
  for ( j in 1:q ) {
    beta[j,1:n.x] <- zbeta[j,1:n.x] / x.sd[1:n.x] 
    beta0[j] <- zbeta0[j] - sum( zbeta[j,1:n.x] * x.mean[1:n.x] / x.sd[1:n.x] )
  }
}