topic Re: Obtaining minimum variance quadratic unbiased estimates as starting values for the covariance fa in SAS Programming

Obtaining minimum variance quadratic unbiased estimates as starting values for the covariance failed

superbug — Thu, 03 Jun 2021 18:29:21 GMT

I am using PROC GLIMMIX to do an analyses, My code is below

PROC GLIMMIX
DATA=mfirst METHOD=LAPLACE NOCLPRINT;
CLASS SCHOOLNUM;
MODEL PASS (EVENT="1")=peat timeaftergrad/CL DIST=BINARY LINK=LOGIT SOLUTION;
RANDOM INTERCEPT peat timeaftergrad/ SUBJECT=SCHOOLNUM S CL TYPE=VC;
/*parms (0.2) (1) / hold=2;*/
COVTEST /WALD;
ods output solutionR=mfirst_rest_mod2;
run;

I got the following error message

"Obtaining minimum variance quadratic unbiased estimates as starting values for the
covariance parameters failed"

Please help.

Thanks!

Re: Obtaining minimum variance quadratic unbiased estimates as starting values for the covariance fa

superbug — Thu, 03 Jun 2021 18:33:58 GMT

@Ksharp

You helped a similar problem posted in 2017.

By your suggestion, I added "parms (0.2) (1) / hold=2" as commented out in the code, but got the following error message

"ERROR: 3 PARMS must be given instead of 2."

Do you have any idea how to fix the error so that to make my code work?

Thank you!

Re: Obtaining minimum variance quadratic unbiased estimates as starting values for the covariance fa

Ksharp — Fri, 04 Jun 2021 11:55:56 GMT

Sorry ,I can't help you . Maybe @SteveDenham @lvm @StatDave could give you a hand .

Re: Obtaining minimum variance quadratic unbiased estimates as starting values for the covariance fa

lvm — Fri, 04 Jun 2021 13:04:38 GMT

This is because you have three variance components in your model. When you use the PARMS statement, you must give a starting value (or a held fixed value) for each variance (and/or covariance) component. Something like

parms (2) (2) (.1);

Note: in the old post, the last parameter (the second in that case) was being held fixed (not estimated). This is because you used "/ hold=2" option, where the "2" here refers to the second variance component. If you want to estimate now all three variances, don't put in that option.

With the error you got, you might have an overparameterized model. That is, possibly one of the variance components is actually 0 in your random-coefficients model.

Re: Obtaining minimum variance quadratic unbiased estimates as starting values for the covariance fa

superbug — Mon, 07 Jun 2021 15:48:15 GMT

@lvm

Thanks much for your reply!

Could you please explain the meaning of the numbers in the parenthesis?

parms (2) (2) (.1);

Re: Obtaining minimum variance quadratic unbiased estimates as starting values for the covariance fa

lvm — Mon, 07 Jun 2021 15:55:56 GMT

These are the initial estimates (guesses) of your variances, in the order in you random statements. You need to put something in, perhaps based on results from some other experiment. You can have sas try a range of guesses, and it will use the one that gives the minimum -2LL as the starting value for the optimization. Example
parms (.2 to 2.5 by .2) (1 to 5 by 1) (.5 to 5 by .5);
You would need ballpark estimates to start.

Re: Obtaining minimum variance quadratic unbiased estimates as starting values for the covariance fa

superbug — Mon, 07 Jun 2021 16:49:17 GMT

@lvm

Thanks much for your explanation!