partial variance estimation in lme()

Matt Wand (mwand@hsph.harvard.edu)
Wed, 14 Jan 1998 13:36:03 -0500 (EST)


Dear S-PLUS'ers,

I would like to fit a linear mixed model
where some of the variance components are
estimated, while others are specified.

The argument "var.estimate" of lme()
suggests that you can either have all
of the variance components estimated,
or all of them fixed at their "start" values.

So my question is: is there a way around
this, i.e. have some variance components
estimated and others fixed at their "start" values?

Best regards,

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Matt Wand, Associate Professor, |
+ Department of Biostatistics, +
| Harvard School of Public Health, |
+ 677 Huntington Avenue, Boston, MA 02115, USA +
| phone: +1 (617) 432 1056 (work), +1 (617) 661 9583 (home) |
+ e-mail: mwand@hsph.harvard.edu, fax: +1 (617) 739 1781 +
| internet: http://www.biostat.harvard.edu/~mwand |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+