Gibbs sampling
in [SAS]
|
From: Makkena on 26 Oct 2006 14:59 Dear Sas-lers, It's been a long time I visited the group. I need your help again. Does anyone have any code which would implement Gibbs sampling routine in SAS? I would really appreciate if anybody could share. TIA, Ramesh Makkena
From: toby989 on 26 Oct 2006 16:57 Here is some R code that I wrote yesterday. I think its not too difficult to translate that into SAS IML and to extend/modify for your needs. x = c() y = c() y[1] = 0.5 x[1] = 0 for (i in 2:100000) { x[i] = rbinom(1,10,y[i-1]) y[i] = rbeta(1,x[i]+1,10-x[i]+2) } xa = x[seq(1000, 100000, 100)] ya = y[seq(1000, 100000, 100)] u = xa*ya mean(u) > mean(u) [1] 1.618170 Makkena wrote: > Dear Sas-lers, > > It's been a long time I visited the group. I need your help again. Does > anyone have any code which would implement Gibbs sampling routine in > SAS? I would really appreciate if anybody could share. > > TIA, > Ramesh Makkena >
From: Lex Jansen on 26 Oct 2006 16:29 Hi Ramesh: Go to www.lexjansen.com/sugi and search for: Gibbs sampling This will give you a few proceedings: http://www2.sas.com/proceedings/sugi24/Stats/p266-24.pdf http://www2.sas.com/proceedings/sugi25/25/st/25p259.pdf http://www2.sas.com/proceedings/sugi22/STATS/PAPER264.PDF http://www2.sas.com/proceedings/sugi31/185-31.pdf HTH, Lex Jansen Please reply to the newsgroup. In order to fight spam and virusses replies to my email address listed in this post will be deleted without reading. If you want to reply to me personally, please replace "spam" with "admin" in the e-mail address. An alternative means of contact is the form at: http://www.lexjansen.com/feedback On Thu, 26 Oct 2006 11:59:26 -0700, Makkena <ramesh.makkena(a)NOVARTIS.COM> wrote: >Dear Sas-lers, > >It's been a long time I visited the group. I need your help again. Does >anyone have any code which would implement Gibbs sampling routine in >SAS? I would really appreciate if anybody could share. > >TIA, >Ramesh Makkena
From: Makkena on 28 Oct 2006 13:28 Thanks for my friend Lex Jansen and Toby for your respones. Ramesh Lex Jansen wrote: > Hi Ramesh: > > Go to www.lexjansen.com/sugi and search for: Gibbs sampling > > This will give you a few proceedings: > > http://www2.sas.com/proceedings/sugi24/Stats/p266-24.pdf > http://www2.sas.com/proceedings/sugi25/25/st/25p259.pdf > http://www2.sas.com/proceedings/sugi22/STATS/PAPER264.PDF > http://www2.sas.com/proceedings/sugi31/185-31.pdf > > HTH, > > Lex Jansen > > Please reply to the newsgroup. In order to fight spam and virusses replies > to my email address listed in this post will be deleted without reading. > If you want to reply to me personally, please replace "spam" with "admin" > in the e-mail address. > An alternative means of contact is the form at: > http://www.lexjansen.com/feedback > > > On Thu, 26 Oct 2006 11:59:26 -0700, Makkena <ramesh.makkena(a)NOVARTIS.COM> > wrote: > > >Dear Sas-lers, > > > >It's been a long time I visited the group. I need your help again. Does > >anyone have any code which would implement Gibbs sampling routine in > >SAS? I would really appreciate if anybody could share. > > > >TIA, > >Ramesh Makkena
|
Pages: 1 Prev: SPEC to test for heteroscedasticity Next: SPEC to test for heteroscedasticity |