1

I am trying to understand how the dpois command in WinBUGS differs from the dpois command in R. I am attempting to predict estimates for counts within different areas. The code works in WinBUGS and results in a count estimation, however for some of my distributions the code is running much too long. I can export the eta information in the code below to R to generate predictions, but I don't think dpois in R does the same thing as it does in WinBUGS.

In R, ?dpois provides help. Is there a WinBUGS equivalent that would provide background about what the command is doing?

What does dpois in WinBUGS provide? If I attempt to use a similar command in R, what x value needs to be provided to dpois(x, lambda) to generate the equivalent result?

Thank you.

for(i in 1:16){ELU.abundance[i]~dpois(eta[i])}

park.total<-sum(ELU.abundance[1:16])

user72603
  • 11
  • 2
  • 1
    http://journal.r-project.org/archive/2013-1/lebauer-dietze-bolker.pdf (although now I notice that we skipped the Poisson! Do you have any reason to believe that `dpois` *is* different? (If `eta[i]` is your linear predictor, you may need to exponentiate it ...) – Ben Bolker Apr 02 '15 at 22:52
  • I don't believe Poisson would be different than any of the given examples. My code runs for a very long time when I attempt to predict in WinBUGS, and I'm trying to moving the prediction portion to R. Consequently, I want to understand exactly how the dpois command works so I can be sure it operates the same way in R when predicting. Thanks for the heads up on eta - it has previously been exponentiated. – user72603 Apr 03 '15 at 12:46

0 Answers0