https://communities.sas.com/t5/Statistical-Procedures/MCMC-iterations/m-p/248332#M13066 <P>Your convergence might be slow. Consider increasing your burn-in iterations and &nbsp;thining your chain by a factor of, let's say, 10 or 20.&nbsp;</P> <P>&nbsp;</P> <P><STRONG>PROC MCMC DATA = YourDATA &nbsp;nbi=50000 nmc=200000 thin =20 &nbsp;;</STRONG></P> <P>&nbsp;</P> <P>Your posterior estimates will be based on 200,000/20 = 10,000 observations.&nbsp;</P> Fri, 05 Feb 2016 16:52:01 GMT youtoub 2016-02-05T16:52:01Z https://communities.sas.com/t5/Statistical-Procedures/MCMC-iterations/m-p/248081#M13061 <P>How to decide number of Markov's Chain iterations while running proc MCMC for bayesian approach of multiple linear regression?</P><P>In my datasset it seems that I am unable to remove autocorrelation from the model. Please suggest.</P> Thu, 04 Feb 2016 20:20:58 GMT https://communities.sas.com/t5/Statistical-Procedures/MCMC-iterations/m-p/248081#M13061 piyushstats 2016-02-04T20:20:58Z https://communities.sas.com/t5/Statistical-Procedures/MCMC-iterations/m-p/248332#M13066 <P>Your convergence might be slow. Consider increasing your burn-in iterations and &nbsp;thining your chain by a factor of, let's say, 10 or 20.&nbsp;</P> <P>&nbsp;</P> <P><STRONG>PROC MCMC DATA = YourDATA &nbsp;nbi=50000 nmc=200000 thin =20 &nbsp;;</STRONG></P> <P>&nbsp;</P> <P>Your posterior estimates will be based on 200,000/20 = 10,000 observations.&nbsp;</P> Fri, 05 Feb 2016 16:52:01 GMT https://communities.sas.com/t5/Statistical-Procedures/MCMC-iterations/m-p/248332#M13066 youtoub 2016-02-05T16:52:01Z