search for: corcompsym

Displaying 2 results from an estimated 2 matches for "corcompsym".

Did you mean: corcompsymm
2006 Jun 30
1
lme and SAS Proc mixed
...is using lme, but it is unclear from the documentation (and Pinheiro & Bates text) how the pdBlocked and compound symmetric structure can be combined. fit.lme<-lme(expression~End+logpgc,random=list(RefSeqID=pdBlocked(list (~1,~ProbeID-1),pdClass="pdSymm")),data=dataset,correlation=corCompSym m(form=~1|RefSeqID/ProbeID/ProbeNo)) The point estimates are essentially the same comparing R and SAS for the fixed effects, but the 95% confidence intervals are much shorter using lme(). In order to find the difference in the algorithms used by SAS and R I tried to extract the variance-covarianc...
2006 Jun 30
0
SAS Proc Mixed and lme
...is using lme, but it is unclear from the documentation (and Pinheiro & Bates text) how the pdBlocked and compound symmetric structure can be combined. fit.lme<-lme(expression~End+logpgc,random=list(RefSeqID=pdBlocked(list (~1,~ProbeID-1),pdClass="pdSymm")),data=dataset,correlation=corCompSym m(form=~1|RefSeqID/ProbeID/ProbeNo)) The point estimates are essentially the same comparing R and SAS for the fixed effects, but the 95% confidence intervals are much shorter using lme(). In order to find the difference in the algorithms used by SAS and R I tried to extract the variance-covarianc...