PROC PLM curiosity

[SteveDenham]

This new procedure in v9.22 uses stored information for scoring and post hoc comparisons, which would be useful in a lot of situations. However, I came across a quirk (reported to support) that others might want to check out.

With the code:
proc plm source=dat.storedstuff ddfm=kr seed=1 ;
show fit;
run;

where dat.storedstuff comes from PROC GLIMMIX, I thought I would get information criteria. Instead, I got:

Fit Statistics

Error df 155

I was wondering if anyone else has run into this, and if so, what I am missing.

TIA.

SteveDenham