[R] absurd computiation times of lme

Christof Meigen christof at nicht-ich.de
Wed Oct 16 16:11:54 CEST 2002


Douglas Bates <bates at stat.wisc.edu> writes:
> You must
> ask yourself if you think that the ways in which these growth curves
> differ has that great a dimensionality.  In most cases I think it is a
> more effective modeling strategy to start with a few random effects
> and check residuals to see if the model needs to be made more complex
> instead of starting with an overly complex model.

I discussed that with Jim Ramsay, and he strongly discouraged me 
to "guess a reasonable, low dimensional basis", using either
known models for growth curves or something like PCA.

His argument was, that, in the first case, I put too much a priori
assumtions into the model, and in the other, that I use results
from the data to analyze the data, which would be some kind of
statistical "deadly sin".

He suggested, and that seemed plausible to me, to use indeed a
much too complex/flexible model like a quite high-dimensional
spline basis, and use lme to constrain that flexibility.

It seems to turn out that lme is not really made for this ...

> As Peter suggested, if you feel that lme is inadequate for your
> purposes we invite you to write better software.

...which does not mean I'd dare to say that I could write "better"
software, I had a look at the code when I was trying to write
a band-structured pdMat, and I'm really impressed ...

        Christof
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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