[R-sig-ME] fixing variance components ratios

Juan Pedro Steibel steibelj at msu.edu
Mon Mar 9 14:04:57 CET 2009

Hello all,

I have a question on using lmer with know variance component ratios:

I want to fit a model of the type: y=X b+Z_1 u_1+...+Z_Q u_Q+e,
with u_q~N(0,V_q I), q=1...Q and e~N(0,V_e).
The values of these variance components are unknown, but we know the 
values of the variance ratios: R_q=V_q/V_e for q=1...Q.

Is there a way to specify this in lmer, such that lmer uses the variance 
ratios (R_q) and produces estimates of V_e, blue(b) and blup(u)?

I know that given those VC ratios the problem could be solved with GLS 
in a single iteration, but I was hoping to use the lmer computational 
machinery if this is feasible.

Thanks in advance!

Juan Pedro Steibel

Assistant Professor
Statistical Genetics and Genomics

Department of Animal Science & 
Department of Fisheries and Wildlife

Michigan State University
1205-I Anthony Hall
East Lansing, MI
48824 USA 

Phone: 1-517-353-5102
E-mail: steibelj at msu.edu

More information about the R-sig-mixed-models mailing list