R help - Nov 2011 - Please Help

HiI want to construct a logliikelood function in RHere is the situationy=number
of particles emitted in 1 hr period~pois(30)p=probability of detection of
radiation particlesx=number of particles detected by a radiation
detector~pois(30p)where p~beta(a,1)I have to calculate the loglikehood for a for
the range a(2,50)I wish to simulate 100 random samples for each aHere is my
code:-m=481n=100x = c(15, 36, 29, 28, 37, 32, 25, 27, 31, 21, 25, 27, 28, 31,
28, 20, 34, 25, 20, 34, 15,21, 28, 24, 31, 19, 34, 29, 18, 25, 16, 19, 44, 26,
34, 31, 21, 28, 11, 31, 21, 34, 25, 25,30, 23, 21, 35, 36, 21, 27, 29, 30, 22,
25, 30, 24, 27, 28, 22, 36, 29, 33, 35, 30, 32, 27,26, 25, 27, 23, 21, 39, 33,
24, 21, 19, 34, 32, 28, 27, 28, 23, 20, 24, 29, 21, 22, 31, 28,27, 28, 29, 21,
30, 28, 31, 22, 29,
18)a=c(1:m)*0.1+1.9p<-rbeta(n,a,1)loglik<-numeric(m)for(i in
1:n){p[i]<-rbeta(n,a[i],1)x[i]<-rpois(n,p[i]*30)loglik<-function(x,p)logLik(a=a(i))<-n*log(a)+(log(30)*sum(x))-sum(log(factorial(x)))-30*(sum(p[i]))+sum(x[i]*log(p[i]))+(a[j]-1)*sum(log(p[i]))}But
somehow I am not getting the answer please HELPP.S. the data for x is a random
sample of number of particles detected in 1 hr (given in the question)Really
appreciate the helpThanks


 		 	   		  
	[[alternative HTML version deleted]]

Two problems with your post:

a) you used a format that is not accepted by the mail server and  
resulted in all of your linefeeds being removed, so it is now  
basically unreadable. As the Posting Guide says, please use plain  
text. (And please read the Posting Guide for yourself ... and do it  
now before reading further.) All the free mailing services provide  
mechanisms for constructing plain text emails.

b) it appears to be homework, and homework submission are discouraged.  
You can attach a plea describing why you should be exempted from this  
rule, but it might require identifying your self more completely and  
explaining why you are not using the services of the educational  
institution you are studying under.


On Nov 12, 2011, at 8:09 AM, avish kamdar wrote:
>
> HiI want to construct a logliikelood function in RHere is the  
> situationy=number of particles emitted in 1 hr  
> period~pois(30)p=probability of detection of radiation  
> particlesx=number of particles detected by a radiation  
> detector~pois(30p)where p~beta(a,1)I have to calculate the  
> loglikehood for a for the range a(2,50)I wish to simulate 100 random  
> samples for each aHere is my code:-m=481n=100x = c(15, 36, 29, 28,  
> 37, 32, 25, 27, 31, 21, 25, 27, 28, 31, 28, 20, 34, 25, 20, 34,  
> 15,21, 28, 24, 31, 19, 34, 29, 18, 25, 16, 19, 44, 26, 34, 31, 21,  
> 28, 11, 31, 21, 34, 25, 25,30, 23, 21, 35, 36, 21, 27, 29, 30, 22,  
> 25, 30, 24, 27, 28, 22, 36, 29, 33, 35, 30, 32, 27,26, 25, 27, 23,  
> 21, 39, 33, 24, 21, 19, 34, 32, 28, 27, 28, 23, 20, 24, 29, 21, 22,  
> 31, 28,27, 28, 29, 21, 30, 28, 31, 22, 29, 18)a=c(1:m)*0.1+1.9p<- 
> rbeta(n,a,1)loglik<-numeric(m)for(i in 1:n){p[i]<-rbeta(n,a[i], 
>
1)x[i]<-rpois(n,p[i]*30)loglik<-function(x,p)logLik(a=a(i))<-n*log(a)
> +(log(30)*sum(x))-sum(log(factorial(x)))-30*(!
> sum(p[i]))+sum(x[i]*log(p[i]))+(a[j]-1)*sum(log(p[i]))}But somehow I  
> am not getting the answer please HELPP.S. the data for x is a random  
> sample of number of particles detected in 1 hr (given in the  
> question)Really appreciate the helpThanks
>
>
> 		 	   		
> 	[[alternative HTML version deleted]]
>
David Winsemius, MD
West Hartford, CT