[R] fitting growth curves

Jose Pinheiro jcp at research.bell-labs.com
Thu Mar 29 16:50:59 CEST 2001


>> 
>> Dear R-list members,
>> 
>> Cynthia M. Jones wrote a paper (Fitting growth curves to retrospective
>> size-at-age data, Fisheries Research 46(2000):123-129; abstract at
>> http://www.elsevier.nl/gej-ng/10/19/44/70/24/37/abstract.html)where the
>> SAS procedure MIXED, Macro NLINMIX (Littell et. al., 1996)was used to
>> estimate the von Bertalanffy growth function parameters assuming that
>> data from the same fish are repeated-measures with autoregressive
>> correlation structure and fitting with a nonlinear repeated-measures
>> model.
>> 
>> Is there any similar procedure that I can use in R?
>
> elliptic in my growth library will do this even if the times are
> unequally spaced. (Unless it changed recently, nlme gives incorrect
> answers for the AR in this case.)

There is a continuous AR(1) class, corCAR1, in the nlme library which
should be used in the case when the times are unequally spaced and
measured on a continuous scale. This is equivalent to the correlation
structure implemented in elliptic. There are other correlation classes
in nlme that can be used with continuous, unequallly spaced times,
such as the spatial correlation structures. In the case of a
"classical" AR(1) model, because the autocorrelation parameter is
allowed to take negative values (unlike in the continuous AR(1) case,
where it is restricted to be non-negative), the times need to be
integer valued (as stated in the help page for corAR1, the correlation
class for AR(1) in nlme). Note that, even in this case, the times are
not required to be equally spaced.

Regards,
--Jose'

-----------------------------------------------------------------------------
 Jose' Pinheiro
 Bell Laboratories			jcp at research.bell-labs.com
 600 Mountain Avenue, Room 2C-258	office:	(908) 582-2390
 Murray Hill, NJ 07974	   		fax:	(908) 582-3340
-----------------------------------------------------------------------------

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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