03-08-2016 12:56 PM
Can anyone tell me what the following error means?
WARNING: The R matrix depends on observation order within subjects. Omitting observations from
the analysis because of missing values can affect this matrix. Consider using a
classification effect in the RANDOM _RESIDUAL_ statement to determine ordering in the R
I am attempting to run repeated measures on some data. The design was 20 animals blocked into light and heavy bodyweights, then assigned to 1 of 2 treatments withing each block. The animals had biopsies at 5 time points and gene expression was measured on 8 genes at each point. Attached is the SAS input as well as the output. I am not sure my denominator DF are correct and this may be associated with the error? Any help would be greatly appreciated.
03-29-2016 02:24 PM
When you have missing observations on the RANDOM _RESIDUAL_ statement in GLIMMIX then the procedure generates the warning message. Considering change the RANDOM _RESIDUAL_ to -
Proc glimmix data = JessicaPCR;
Class ID TRT BLOCK DAY;
Model AMPK = TRT|DAY/ddfm = KR;
random intercept / subject=BLOCK;
random intercept / subject=ID*TRT;
random day /subject = ID*TRT type=arh(1) residual;
LSMEANS TRT|DAY/ pdiff lines;
The above RANDOM statement should make note go away.