[R] about MCMC pack again...

[Doran, Harold]

Let's maybe back up a bit on this. You said you are interested in
learning about the application of the Gibbs sampler for IRT models. I
don't think opening the C++ code would be the best approach for this.

Let me recommend the following article

    Patz, R. J., and Junker, B. W.  (1999). A straightforward approach 
    to Markov chain Monte Carlo for item response models.  Journal of 
    Educational and Behavioral Statistics, 24, 146-178.

This will give you what you need to know. Richard Patz also developed a
program written in S that follows the models presented in the article.
You can find this somewhere on the statlib cmu website. Also, I don't
know how mcmcirt works under the hood exactly, but Gibbs sampler is a
special case of the MH algorithm when the acceptance rate is 1.

Harold


> -----Original Message-----
> From: r-help-bounces at stat.math.ethz.ch 
> [mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of 
> Mariagiulia Matteucci
> Sent: Friday, August 11, 2006 5:55 AM
> To: Barry Rowlingson
> Cc: r-help at stat.math.ethz.ch
> Subject: Re: [R] about MCMC pack again...
> 
> Hello, I am using Windows, I tried to use th File Search and 
> also the Windows Grep but I cannot find any file! In the list 
> you showed me there are some useful , I really don't know how 
> can I find them! I tried in the R folder, src folder, MCMC 
> pack folder and I dowloaded the .tar file about MCMC pack 
> where there are the codes, I really don't know what to do!
> 
> Mariagiulia
> 
> 
> On Aug 11, 2006 10:51 AM, Barry Rowlingson 
> <B.Rowlingson at lancaster.ac.uk> wrote:
> 
> > Mariagiulia Matteucci wrote:
> > > Hello, thank you very much for your previous answers 
> about the C++ 
> > > code.
> > > I am interested in the application of the Gibbs Sampler 
> in the IRT 
> > > models, so in the function MCMCirt1d and MCMCirtkd. I've found the
> > > C++
> > > source codes, as you suggested, but I cannot find 
> anything about the 
> > > Gibbs Sampler. All the files are for the Metropolis algorithm.
> > 
> >   $ cd MCMCpack/
> >   $ grep -ir gibbs .
> > 
> > produces loads of output, including:
> > 
> > ./src/MCMCfactanal.cc:    } // end Gibbs loop
> > ./src/MCMChierEI.cc:// and slice sampling and Gibbs 
> sampling to sample 
> > from the posterior
> > ./src/MCMCirt1d.cc:    } // end Gibbs loop
> > ./src/MCMCmixfactanal.cc:  // Gibbs Sampler //
> > ./src/MCMCmixfactanal.cc:  } // end Gibbs loop
> > ./src/MCMCoprobit.cc:    // Gibbs loop
> > ./src/MCMCordfactanal.cc:  // Gibbs Sampler // 
> ./src/MCMCpanel.cc:// 
> > simulate from posterior density and return a Gibbs by parameters 
> > matrix
> > ./src/MCMCpanel.cc: const int* burnin, const int* gibbs, const
> > int* thin,
> > ./src/MCMCpanel.cc:   int Mgibbs = gibbs[0];
> > ./src/MCMCpanel.cc:   int Mtotiter = Mburnin + Mgibbs;
> > ./src/MCMCpanel.cc:   Matrix<double> beta_holder(Mgibbs/Mthin,Mp);
> > ./src/MCMCpanel.cc:   Matrix<double> D_holder(Mgibbs/Mthin,Mq*Mq);
> > ./src/MCMCpanel.cc:   Matrix<double> sigma2_holder(Mgibbs/Mthin, 1);
> > ./src/MCMCpanel.cc:   // gibbs loop
> > ./src/MCMCregress.cc:     // Gibbs sampler
> > ./src/MCMCregress.cc:       // second set of Gibbs scans
> > ./src/MCMCSVDreg.cc:    /////////////////// Gibbs sampler 
> > ///////////////////
> > 
> >   Perhaps some of these are useful?
> > 
> > For your info, I know nothing about MCMCpack, I just know 
> how to use 
> > grep to search for things. If you are on Windows, you can 
> probably use 
> > the Windows File Explorer Search option to look for it. But give me 
> > grep anyday...
> > 
> > Barry
> >
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>