This project is read-only.

poster

Topics: IRIS in General
Oct 2, 2014 at 8:56 AM
Edited Oct 2, 2014 at 9:34 AM
Dear all,

I am new to IRIS and have been able to estimate a structural model using full Bayesian procedures.
In the examples I use
s=stats(pos,theta,logpost)
to obtain basic statistics from the simulated chains.
By default it produces
s =
chain: [1x1 struct]
 mean: [1x1 struct]
  std: [1x1 struct]
 hist: [1x1 struct]
  mdd: -35162.5132733484
and I would like to obtain the highest probability density estimates.

How do I set the poster object to export to s the hpd's?

I appreciate very much your help.

Best regards,

Juan Manuel
Oct 2, 2014 at 3:18 PM
Hi Juan

Use the option 'hpdi=' N in the function stats( ), where N is the requested percent coverage of the HPDIs. For instance
s = stats(po,theta,logpost,'hpdi=',90)
will return a 90% HPDI for each parameter. Note that you need to enter 90 for 90%, not 0.90.

Best,
Jaromir
Marked as answer by jaromirbenes on 10/2/2014 at 7:18 AM
Oct 2, 2014 at 7:02 PM
Thanks a lot Jaromir.

It worked perfectly fine.

Best Regards,

Juan Manuel