[R] nls question

Prof Brian Ripley ripley at stats.ox.ac.uk
Mon Oct 9 17:34:02 CEST 2000

> From: "Mark M. Span" <span at psy.uva.nl>
> Date: Mon, 9 Oct 2000 16:27:35 +0200
> Just to add a note to the question asked by Bill Simpson..
> I am confronted with a similar problem, but the model I try to fit on my
> data is nonlinear.
> I have a four-parameter, biexponential model describing the response times
> of a group of subjects performing a set of task. What I want to evaluate is
> whether the parameter values differ between (groups of) tasks.
> myfunc    <-formula(x ~ a*exp(-b*age) + (c*exp(d*age)) )
> both x (representing response time) and age are variables in the current
> scope. Part of the data frame is presented below.  At the moment I fit the
> model for each task seperately. But now I cannot test the parameters for
> equality. A solution like in Bills case would be great.

Use a model parametrizing a,b,c by groups and one without, and use
anova to compare the models.  A worked example is in V&R3, page 249.
I'd try your example, but as you only have two age points (if age is
the final column), it is non-identifiable (four parameters for two
predictions only) so I assume you have much more data.

Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

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