[R-sig-ME] user defined covariance structure in lme4 /nlme (Tapan Mehta)

[Gregor Gorjanc]

Tapan Mehta <tapmehta at ...> writes:
...
>  I have tried using corSymm but it does not give me correct random
> effects estimates as compared to what I get through SAS. I am trying to
> estimate the genetic and error variance and based on a simulated data
> (through SAS) with genetic variance 7 and error variance 2 whereas the
> results I obtain from the code below gives 7.58 and 1.38. 

But this is not so bad! Try with several datasets and you will get the feeling
of the quality of estimates.

Gregor