[R] Re: initial values for non linear models? {was "Info_Request"}

[Martin Maechler]

>>>>> "millanti" == millanti  <millanti at libero.it> writes:

    millanti> Hi, I'm an R user and I have to answer a question
    millanti> about a problem I'm not able to solve by R code:
    millanti> the command nlm(....) to fit a non linear model
    millanti> requires an initial estimation of the model's
    millanti> unknown parameters.  But how can I have an initial
    millanti> estimation?

In general, this is a very hard problem, particularly if the
dimension of your unknown parameters is large. You don't say
anything about your model which would help to help you...

nls in package nls, {i.e. you type  library(nls)} has so called
self-starting models for a few classes of non-linear *regression* models.
I don't know if this helps, since nlm() is not only for regression models.

type   ?selfStart
and    apropos("^SS") to get a list of predefined selfstarting
models, and then, e.g.  ?SSlogis

Further you may need to learn more about non-linear regression in general.
E.g., 
    Bates, D.M. and Watts, D.G. (1988) Nonlinear Regression Analysis
         and Its Applications; Wiley.
and -- for selfstarting (and more general, see help(gls, package="nlme") )
models
    Jos\'e C. Pinheiro and Douglas M. Bates (2000)
    Mixed-Effects Models in S and S-Plus; Springer.

    millanti> Thanks a lot.

you're welcome.
Please, next time use R-help {if you've spent some time trying
to solve the problem yourself ..} instead,
[[also please use a more distinct "Subject:" line.]]

Martin Maechler <maechler at stat.math.ethz.ch>	http://stat.ethz.ch/~maechler/
Seminar fuer Statistik, ETH-Zentrum  LEO C16	Leonhardstr. 27
ETH (Federal Inst. Technology)	8092 Zurich	SWITZERLAND
phone: x-41-1-632-3408		fax: ...-1228			<><
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._