[本週專題]WinBUGS: Undefined Real Result?

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

I want to analyze a multilevel multidimensional model in WinBUGS. the model is as below (N=2362 students responding to K=45 items of a test, students are nested within J=116 schools):

1. model{
   
2.         #responses
   
3. for(i in 1:N){
   
4.         for(j in 1:K){
   
5.         logit(p[i,j])<- a1[j]*t[i,1]+a2[j]*t[i,2]-b[j]
   
6.         y[i,j]~dbern(p[i,j]  }#for j
   
7. 
   
8.         t[i,1:2]~dmnorm(mu[sc,1:2],tau.p[1:2,1:2])
   
9.                 }#for i
   
10. 
    
11.         #school level
    
12.     for(j in 1:J){  
    
13.         mu[j,1:2]~dmnorm(m[j,1:2],tau.s[1:2,1:2])
    
14. 
    
15.                 }#for j of school
    
16.         #priors
    
17.         for(j in 1:J){
    
18.         m[j,1:2]~dmnorm(m0[1:2],cov[1:2,1:2])
    
19.                #m0=c(0,0) #cov=diag(2)
    
20.         }
    
21.         tau.p[1:2,1:2]~dwish(cov[1:2,1:2],2)
    
22.         tau.s[1:2,1:2]~dwish(cov[1:2,1:2],2)
    
23.         sigma.p[1:2,1:2]<-inverse(tau.p[,])
    
24.         sigma.s[1:2,1:2]<-inverse(tau.s[,])
    
25.         s2p<-sum(sigma.p[,])
    
26.         s2s<-sum(sigma.s[,])
    
27.         rho<-(s2s)/(s2s+s2p)
    
28.         a1[1]~dlnorm(0,4)
    
29.         a2[1]<-0
    
30.         b[1]~dnorm(0,1)
    
31.         for(s in 2:K) {
    
32.     a1~dlnorm(0,4)
    
33.     a2~dlnorm(0,4)
    
34.     b~dnorm(0,1)
    
35.                     }#for s of items   
    
36.         }#for model

复制代码

Also see next post

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：Undefined winbugs Defined DEFINE WINBUG schools nested within

I've set these functions as initial values:

1. ini<-function(){
   
2. list(tau.p=matrix(rgamma(4,100,100),2,2),
   
3. tau.s=matrix(rgamma(4,100,100),2,2),
   
4. t=rmvnorm(N,mean=c(0,0),sigma=diag(2)),
   
5. m=rmvnorm(J,mean=c(0,0),sigma=diag(2)),
   
6. mu=rmvnorm(J,mean=c(0,0),sigma=diag(2)),
   
7. a1=rlnorm(K,0, 0.4),
   
8. a2=c(NA,rlnorm(K-1,0, 0.4)),
   
9. b=rnorm(45,0,0.5))
   
10. }

复制代码

I use rube package in R to check and run my analysis and everything looks fine. When I run the model I receive "Trap 66 (postcondition violated)" or "undefined real result". I think the problem is from the initials but I have no idea how to solve it.

Any idea?