R2OpenBUGS——矩阵、列表和向量的几个问题

Question

我是 R2OpenBUGS 的新手，非常神秘的错误非常令人沮丧。

我尝试运行一个非常简单的模型。我以前曾成功运行ning 类似的模型。我的问题是因为我有一个二维数组（矩阵）吗？

我尝试简化模型但没有成功。

错误如下：

model is syntactically correct
expected the collection operator c error pos 11
model compiled
expected a number or an NA error pos 1449
initial values generated, model initialized
model is updating
200 updates took 0 s
tau.0 is not a variable in the model
tau.1 is not a variable in the model
model is updating
****** Sorry something went wrong in procedure StdMonitor.Update in module DeviancePlugin ******

这是我使用的代码

rm(list=ls(all=TRUE))
cat("4") 
library(R2OpenBUGS)
rat.dat<- read.table("BigRatDat.txt",header=FALSE);

dose = data.matrix(rat.dat[1])
weight = data.matrix(rat.dat[3:13])
N<- length(dose);

cat("
model{
for(i in 1:50){
  for(j in 1:11){
        weight[i,j]~dnorm(mu[i,j],tau[i])
    mu[i,j]<-b.0[i]+b.1[i]*j
    }
    b.0[i]~dnorm(mu.0[i],tau.0)
    b.1[i]~dnorm(mu.1[i],tau.1)
  mu.0[i] <-b.00+b.01*dose[i]
  mu.1[i] <-b.00+b.01*dose[i]
    tau[i]~dgamma(0.01,0.01)
  dose[i]~dnorm(0,1)
}
b.00~dnorm(0,0.001)
b.01~dnorm(0,0.001)
b.10~dnorm(0,0.001)
b.11~dnorm(0,0.001)
tau.0~dgamma(0.01,0.01)
tau.1~dgamma(0.01,0.01)
}
",file="Rats2OpenBugs.txt")


data <- list("dose","weight")

inits <- function(){
  b.0<-rnorm(n=N,0);
  b.1<-rnorm(n=N,0);
  b.00<-rnorm(1,0);
  b.01<-rnorm(1,0);
  b.10<-rnorm(1,0);
  b.11<-rnorm(1,0);
  tau = rep(1,N);
  tau.0 = 1;
  tau.1 = 1;
  list(b.0=b.0,b.1=b.1,b.00=b.00,b.01=b.01,b.10=b.10,b.11=b.11,tau=tau,tau.0=1,tau.1=1)
}

params <- c("b.0","b.1","b.00","b.01","b.10","b.11","tau","tau.0","tau.1");

output.sim <- bugs(data,inits,params,model.file="Rats2OpenBugs.txt",
                   n.chains=1, n.iter=5000, n.burnin=200, n.thin=1
                   ,debug=TRUE)

一个数据文件：

0    1   54  60  63  74   77   89   93 100  108  114 124
0    2   69  75  81  90   97  120  114 119  126  138 143
0    3   77  81  87  94  101  110  117 124  134  141 151
0    4   64  69  77  83   88   96  104 109  120  123 131
0    5   51  58  62  71   74   81   88  93   99  103 113
0    6   64  71  77  89   90  100  106 114  122  134 139
0    7   80  91  97  101 111  119  129 131  137  147 154
0    8   79  85  89  99  104  105  116 121  132  139 147
0    9   77  82  88  92  101  109  119 127  135  144 158
0    10  79  84  91  98  107  114  119 131  137  146 155
.5   1   62  71  75  79   87   91  100 105  111  121 124
.5   2   68  73  81  89   94  101  110 114  123  132 139
.5   3   94 102 109 110  128  133  147 151  153  171 184
.5   4   81  90  95 102  109  120  128 137  141  154 160
.5   5   64  69  72  76   84   89   97 103  108  114 124
.5   6   67  74  81  81   84   95  100 109  119  128 130
.5   7   73  80  86  89   97  101  110 116  117  135 141
.5   8   71  74  82  84   93   97  102 113  119  124 131
.5   9   69  74  79  89   94  100  107 113  124  134 139
.5   10  60  62  67  74   78   85   92 103  112  121 130
 1   1   59  63  66  75   80   87   99 104  110  115 124
 1   2   56  66  70  81   77   88   96 100  113  120 130
 1   3   71  77  84  80   97  106  111 109  128  133 140
 1   4   59  64  69  76   85   88   96 104  110  119 126
 1   5   65  70  73  77   85   92   96 101  111  118 121
 1   6   61  69  77  81   89   92  107 111  118  127 132
 1   7   80  86  95  99  106  113  127 131  142  150 160
 1   8   74  80  84  90   99  101  108 117  126  133 140
 1   9   71  79  88  90   98  102  116 121  127  139 142
 1   10  69  75  80  86   96   97  104 113  122  129 138

Answer 1

问题是我试图使用只有一列的矩阵作为向量。 R 对此没有问题，但在将数据导出到 OpenBUGS 时它不起作用。该程序期望对矩阵的引用具有 2 个索引（用于行和列）。

我只需要更换：

dose = data.matrix(rat.dat[1])

与：

dose = unlist(as.vector(rat.dat[1]))

R2OpenBUGS——矩阵、列表和向量的几个问题

R2OpenBUGS - several problems with matrices, list and vectors

r

openbugs