【求助】WinBUGS不能设定初始值

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

问题是这样的，在data loaded和model compiled后，我人为设定了初始值list(pi=0,s1=0,s2=0,c1=0,c2=0,s3=0,c3=0)，然后摁了load units按钮，出现undefined real result窗口，还有一种情况是如果不把初始值全部设为0，会出现value of order of binomial x[8] must be greater than zero的语句，不知错在哪里，望各位达人指正。谢谢
model
{

x1.2<-pi*s1*s2*s3/(pi*s1*s2*s3+(1-pi)*(1-c1)*(1-c2)*(1-c3))

x[1]~dbin(u[1],x1.2)

x2.2<-pi*s1*s2*(1-s3)/(pi*s1*s2*(1-s3)+(1-pi)*(1-c1)*(1-c2)*c3)

x[2]~dbin(u[2],x2.2)

x3.2<-pi*s1*(1-s2)*(1-s3)/(pi*s1*(1-s2)+(1-pi)*(1-c1)*c2*c3)

x[3]~dbin(u[3],x3.2)

x4.2<-pi*s1*(1-s2)*s3/(pi*s1*(1-s2)*s3+(1-pi)*(1-c1)*c2*(1-c3))

x[4]~dbin(u[4],x4.2)

x5.2<-pi*(1-s1)*(1-s2)*(1-s3)/(pi*(1-s1)*(1-s2)*(1-s3)+(1-pi)*c1*c2*c3)

x[5]~dbin(u[5],x5.2)

x6.2<-pi*(1-s1)*(1-s2)*s3/(pi*(1-s1)*(1-s2)*s3+(1-pi)*c1*c2*(1-c3))

x[6]~dbin(u[6],x6.2)

x7.2<-pi*(1-s1)*s2*(1-s3)/(pi*(1-s1)*s2*(1-s3)+(1-pi)*c1*(1-c2)*c3)

x[7]~dbin(u[7],x7.2)

x8.2<-pi*(1-s1)*s2*s3/(pi*(1-s1)*s2*s3+(1-pi)*c1*(1-c2)*(1-c3))

x[8]~dbin(u[8],x8.2)

pi~dbeta(42.02,30.8)

s1~dbeta(29.67,6.29)

c1~dbeta(38.56,6.17)

s2~dbeta(25.77,5.17)

c2~dbeta(28.17,5.41)

s3~dbeta(23.06,6.95)

c3~dbeta(20.06,4.49)
}
list(u=c(110,16,30,19,58,1,21,4),x=c(110,15,22,15,54,0,15,2))
list(pi=0,s1=0,s2=0,c1=0,c2=0,s3=0,c3=0)

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：winbugs WINBUG BUGS Win bug 求助 winbugs

同楼主的问题，请求高人指点

Possible reasons include:
- initial values generated from a 'vague' prior distribution may be numerically extreme - specify
appropriate initial values;
- numerically impossible values such as log of a non-positive number - check, for example, that no
zero expectations have been given when Poisson modelling;
- numerical difficulties in sampling. Possible solutions include:
- better initial values;
- more informative priors - uniform priors might still be used but with their range restricted to
plausible values;
- better parameterisation to improve orthogonality;
- standardisation of covariates to have mean 0 and standard deviation 1.
- can happen if all initial values are equal.

使用手册第47页有介绍该问题

楼主问题解决了嘛？是怎么解决的呢