[R] start values for random effects in nlme

From: Lindsey Root <lroot_at_umiami.ir.miami.edu>
Date: Wed 02 Mar 2005 - 06:53:30 EST


I am attempting to run an exponential decay model in nlme with 3 fixed and random effects (an initial value, an asymptote, and a decay rate). I am having trouble getting convergence now that I have added the asymptote to the model, so I was attempting to specify start values for the random effects. I have 161 groups (SUBJECT) and have constructed a 161 x 3 matrix (E) for the random effects start values. However, I receive the following error message:

> fm3 <- nlme(model = RASCH16A ~ B + A*(exp(-K*DAYS)),

+ fixed = B + A + K ~ 1,
+ data = rasch,
+ random = B + A + K ~ 1 | SUBJECT, groups = ~SUBJECT, 
+ start = list(fixed = c(20, 68, 0.03), random= E))
Error in nlme.formula(model = RASCH16A ~ B + A * (exp(-K * DAYS)), fixed = B + :

         groups levels mismatch in random and starting values for random at level SUBJECT

Does anyone have any ideas as to what might be triggering this error message?

Thanks if you can help~
Lindsey Root



R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Received on Wed Mar 02 07:42:23 2005

This archive was generated by hypermail 2.1.8 : Fri 03 Mar 2006 - 03:30:39 EST