I have read the mixed-effects models in S and S-Plus book (nlme::gls
code is based directly on this) and this problem is specified on page
204 eq (5.5). I have also calculated sigma based on (5.7) -after the
transformation documented (5.2) -and i do not get the same value as
either the package or my implementation.
Any advice would be most welcomed. Is there a bug in the estimation of
sigma in this package?
Ecosystem Dynamics & Assessment Branch
Northeast Fisheries Science Center
NOAA Fisheries Service
166 Water Street
Woods Hole, MA 02543