[R] lme problem

Douglas Bates bates at stat.wisc.edu
Thu Mar 25 20:45:39 CET 1999

>>>>> "John" == John Logsdon <j.logsdon at lancaster.ac.uk> writes:

  John> Hi I am trying to get to grips with lme and perhaps the
  John> euro-cent is not dropping.  I want to define a model with some
  John> terms random and some fixed so that:

  John> gd<-groupedData(y~x|c,df)

  John> defining by the group c in data frame df, whereupon

  John> mod<-lme(y~t+f,gd,~t)

  John> where I want the t as a random effect and the f as the fixed
  John> part.

  John> Without the f it works OK but when I add the f, I get an error
  John> message from model.frame saying there is an error in the
  John> lengths.  As far as I can see, all the lengths are the same
  John> (and I have checked it with length() of course).

  John> Is there an obvious reason for this?  I am probably doing
  John> something stupid but it escapes me at the moment.  I will try
  John> and do the same with MLn this evening ...

What you are doing looks reasonable.  Since the version of lme in the
R library is rather old it is possible that the fault lies in the
lme code rather than in your understanding of the lme code.

Jose' Pinheiro and I had a chance to talk with many of the R
developers at the Vienna conference and I think we understand much
better now how to modify the current lme code for use with R.  We hope
to bring the lme library in R up-to-date with the beta version of
nlme-3.0 for S-PLUS but we can't promise a release date on that yet.

Could you send us a transcript of your session including the output of
after you get the error condition?  That could help a lot tracking
down the problem.
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list