R help - Jun 2012 - nls: how do you know if the model is significant?

Hi all,
I'm struggling with nls. How do you know if your model is significant? For a
lm, you get a p-value, but you don't get it for a nls. Is there a way to
calculate it?

For a lm I use this:
  a<-summary(lm(model ~obs))
  f.stat<-a$fstatistic
 
p.value<-1-pf(f.stat["value"],f.stat["numdf"],f.stat["dendf"])

Is there something similar for a nls?


The kind of output that I get is:


Formula: y ~ exp.f(x, a, b)

Parameters:
   Estimate Std. Error t value Pr(>|t|)    
a 1.381e+02  1.192e+01  11.583 3.19e-08 ***
b 1.790e-02  2.459e-03   7.279 6.19e-06 ***
---
Signif. codes:  0 ?***? 0.001 ?**? 0.01 ?*? 0.05 ?.? 0.1 ? ? 1 

Residual standard error: 13.21 on 13 degrees of freedom

Number of iterations to convergence: 6 
Achieved convergence tolerance: 9.123e-06 


I know that my parameters are significant but I need to say something about
the whole model.

Many thanks,
Nerak




--
View this message in context:
http://r.789695.n4.nabble.com/nls-how-do-you-know-if-the-model-is-significant-tp4632401.html
Sent from the R help mailing list archive at Nabble.com.

I suggest you consult a local statistician. You could also post to a
statistical help list like stats.stackexchange.com. Your query has
nothing to do with R, but is rather about the meaningfulness (or lack
thereof) of statistical significance in nonlinear modeling.

-- Bert

On Tue, Jun 5, 2012 at 6:49 AM, Nerak <nerak.t at hotmail.com>
wrote:
> Hi all,
> I'm struggling with nls. How do you know if your model is significant?
For a
> lm, you get a p-value, but you don't get it for a nls. Is there a way
to
> calculate it?
>
> For a lm I use this:
> ?a<-summary(lm(model ~obs))
> ?f.stat<-a$fstatistic
>
?p.value<-1-pf(f.stat["value"],f.stat["numdf"],f.stat["dendf"])
>
> Is there something similar for a nls?
>
>
> The kind of output that I get is:
>
>
> Formula: y ~ exp.f(x, a, b)
>
> Parameters:
> ? Estimate Std. Error t value Pr(>|t|)
> a 1.381e+02 ?1.192e+01 ?11.583 3.19e-08 ***
> b 1.790e-02 ?2.459e-03 ? 7.279 6.19e-06 ***
> ---
> Signif. codes: ?0 ?***? 0.001 ?**? 0.01 ?*? 0.05 ?.? 0.1 ? ? 1
>
> Residual standard error: 13.21 on 13 degrees of freedom
>
> Number of iterations to convergence: 6
> Achieved convergence tolerance: 9.123e-06
>
>
> I know that my parameters are significant but I need to say something about
> the whole model.
>
> Many thanks,
> Nerak
>
>
>
>
> --
> View this message in context:
http://r.789695.n4.nabble.com/nls-how-do-you-know-if-the-model-is-significant-tp4632401.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> 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.


-- 

Bert Gunter
Genentech Nonclinical Biostatistics

Internal Contact Info:
Phone: 467-7374
Website:
http://pharmadevelopment.roche.com/index/pdb/pdb-functional-groups/pdb-biostatistics/pdb-ncb-home.htm