Hi Everyone,
proc mixed returned the statement "Covariance Parameter Values At Last Iteration" which is an indication that I don't have convergence solution even though some covariance parameter estimates are displayed. Unfortunately, It could not give me "Fit statistics" from where I would extract "-2 Res Log Likelihood" to carry out likelihood ratio test. Please how do I improve efficient of my proc mixed procedure to have convergence so as NOT to display "Covariance Parameter Values At Last Iteration"
Meanwhile the result from other traits seem to be good.
See my script below
ods output Fitstatistics = &dsn._dm_fitstat(where=(descr="-2 Res Log Likelihood") rename=(value=full_loglik));
ods output Dimensions = &dsn._dm_dim(rename=(Descr=Descr1 value=full_dim)where=(Descr1="Covariance Parameters"));
proc mixed data = &dsn noitprint noclprint update;
by trait;
class trial rep gen;
model y = trial gen gen*trial/ddfm = satterth;
random rep(trial);
repeated /group = trial;
where trait in("dm");
run;
quit;
One route might be adjusting the default tolerance up so that you get convergence (convf, convg or convh options)?
-unison
Build your skills. Make connections. Enjoy creative freedom. Maybe change the world. Registration is now open through August 30th. Visit the SAS Hackathon homepage.
Register today!Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.