May I please ask how would I address the problem below? I'm able to
get the VarianceA only if I exclude certain fixed effects or
alternatively comment out the 1st random statement. But I'd like to
understand what this notes actually mean. By the way, I'm trying to do
a 3-level multilevel modeling with var1-var5 binary 0/1 coded to see
by adjusting to this 5 variables, how they affect log_exp.
Your advice is greatly appreciated. Many thanks and have a lovely
163 proc mixed data=out covtest method=reml cl;
164 where type=0;
165 class A B;
166 model log_exp = var1 var2 var3 var4 var5
167 / DDFM=SATTERTH solution;
168 random int / subject=A;
169 random int / subject=A*B;
NOTE: Convergence criteria met.
NOTE: Estimated G matrix is not positive definite.
NOTE: Asymptotic variance matrix of covariance parameter estimates has
been found to be
singular and a generalized inverse was used. Covariance
parameters with zero variance do
not contribute to degrees of freedom computed by DDFM=SATTERTH.
NOTE: PROCEDURE MIXED used (Total process time):
real time 0.04 seconds
cpu time 0.04 seconds
Covariance Parameter Estimates
Cov Parm Subject Estimate Error
Intercept A 0 .
Intercept A*B 0.2249 0.1548
Residual 0.9380 0.1722