[R] nlme-problem

[Douglas Bates]

<nioksane at cc.jyu.fi> writes:

> I have a problem with fitting a nonlinear mixed-effects model to my data.
> I was able to fit quite simple exponential model but now I'm trying to fit
> the following Gompertz model:
> 
> > gomp<-nlme(paino~b0+b1*exp(-exp(-b2*(daydeg.scale-b3))), data=group1,
> fixed=b0+b1+b2+b3~1, start=c(b0=150, b1=3000, b2=0.6, b3=2500))
> 
> I'm getting the following error:
> 
> Error in chol((value + t(value))/2) : the leading minor of order 2 is not
> positive definite
> 
> Could you give me some ideas what might be wrong and how to deal with it?
> I know that the data and the syntax should be ok as I have tested it
> already with exponential model. The problem seems to be in the model-form.

Try using the optional arguments

 verbose = TRUE, control = list(msVerbose = TRUE)

in your call to nlme so that you can see how the parameter estimates
are being changed.

By the way, there is a self-starting model function SSgompertz that
you could try instead.  I would try to fit an nlsList model first with
that.