Question

I am new to WinBUGS and I could not get the code below to work. The model is syntactically correct (it is a hierarchical logit model with random effects), but when I click load data the error message that appears is expected key word structure. What does that mean? Can anyone please help me work the code below please? My data set is bigger, but to simplify the question I am only working here with n=2 (number of groups) and k=5 (number of subjects in each group).

model{
for(i in 1:n){
  for(j in 1:k){
    yij[i,j] ~ dbern(p[i,j]) 
    logit(p[i,j]) <- alpha + beta*xij[i,j] + ui[i]
  }
  ui[i] ~ dnorm(0,tau)
}
alpha ~ dnorm(0,0.001)
beta ~ dnorm(0,0.001)
tau ~ dunif(0,100) 
}

with data:

list(n=2, k=5, yij=structure(.Data=c(1, 1, 1, 1, 1, 0, 0, 0, 0, 0), .Dim=c(2,5)),
xij = c(0.0494063719, -0.078101264, 0.2748560749, 0.1245743393, -2.531242809, .6849338859, 0.5302062384, 0.7828560148, -0.012341452, 0.5128471157), 
ui = c(0.5031197054, 0.5031197054, 0.5031197054, 0.5031197054, 0.5031197054, -2.13785637, -2.13785637, 2.13785637, -2.13785637, -2.13785637))
Était-ce utile?

La solution

xij looks to be matrix in the BUGS model, but you have at as a vector in the data. This should work:

list(n=2, k=5, yij=structure(.Data=c(1, 1, 1, 1, 1, 0, 0, 0, 0, 0), .Dim=c(2,5)),
xij = structure(.Data=c(0.0494063719, -0.078101264, 0.2748560749, 0.1245743393, -2.531242809, .6849338859, 0.5302062384, 0.7828560148, -0.012341452, 0.5128471157), .Dim=c(2,5)),
ui = c(0.5031197054, 0.5031197054, 0.5031197054, 0.5031197054, 0.5031197054, -2.13785637, -2.13785637, 2.13785637, -2.13785637, -2.13785637))
Licencié sous: CC-BY-SA avec attribution
Non affilié à StackOverflow
scroll top