3

I am new to WinBUGS and OpenBUGS. I just fit a model to try an example. I was wondering whether I can get predictions generated by WinBUGS/OpenBUGS. If not, are there any convenient ways available (with the help of other applications, R, for example) to do this?

Thanks a lot.

Tomas
  • 5,735
  • 11
  • 52
  • 93
HelloL
  • 183
  • 1
  • 7
  • It is hard to know exactly what you are after. Could you post your model and state what you want to predict? WinBUGS and OpenBUGS are capable of giving distributions for missing data for example. – tristan Aug 19 '12 at 12:28
  • 1
    Also look here: http://stats.stackexchange.com/questions/29932/how-to-generate-predictions-with-rjags/29946#29946 – Glen Aug 20 '12 at 13:58

1 Answers1

1

Yes, you can. In Bayesian tools, it is very easy to get predictions. In your design matrix, just add new rows, with response variable set to NA. You can see concrete example here.

Community
  • 1
Tomas
  • 5,735
  • 11
  • 52
  • 93