[R] bootstrapped CI for nonlinear models using nlsBoot from nlstools

[Bill Pikounis]

Hello Francisco:
I am not familiar with nlsBoot, but if things cannot be solved there,
another option is the example shown in Chapter 8, page 225-226 of the
Venables & Ripley MASS book (2002) which uses nls and the boot
recommended package from R. The MASS package has a scripts subfolder
for this book example application on the Stormer data set.

Kind Regards,
Bill


On Tue, Aug 14, 2012 at 2:58 PM, Francisco Mora Ardila
<fmora at oikos.unam.mx> wrote:
> Hi all
>
> I´m trying to get confidence intervals for parameters from nls modeling. I fitted a nls
> model to the following variables:
>
>> x
>  [1]   2   1   1   5   4   6  13  11  13 101 101 101
>> y
>  [1]  1.281055090  1.563609934  0.001570796  2.291579783  0.841891853
>  [6]  6.553951324 14.243274230 14.519899320 15.066473610 21.728809880
> [11] 18.553054450 23.722637370
>
> The model fitted was:
>
> model<-nls(y~SSgompertz(x,a,b,c))
>
> and it worked OK, with the following results:
>
> Formula: y ~ SSgompertz(x, a, b, c)
>
> Parameters:
>   Estimate Std. Error t value Pr(>|t|)
> a  21.2426     0.9689  21.924 4.03e-09 ***
> b   5.3330     1.4722   3.622  0.00555 **
> c   0.8045     0.0274  29.364 3.01e-10 ***
>
> Then, trying to get confidence intervals for the parameters using the nlsBoot function
> in the nlstools package I got this error:
>
>> modelboot<-nlsBoot(model)
> Error en data2[, var1] <- fitted1 + sample(scale(resid1, scale = FALSE),  :
>   objeto de tipo 'environment' no es subconjunto
>
> I´ve tried with another response variable and other self starting function (SSlogis) and
> got the same error.
>
> Any suggestions?
>
> Francisco
>
>
> ----------------------
> Francisco Mora Ardila
> Estudiante de Doctorado
> Centro de Investigaciones en Ecosistemas
> Universidad Nacional Autónoma de México
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.