[R] confidence interval

[Prof Brian Ripley]

On Fri, 22 Oct 2004, lu kan wrote:

> Is there the similar thing for non-linear 'nls'? 

Do read the documentation for predict.nls, which also has an argument
"interval", which is what you asked.

And this time, read even more carefully.  Then look in the recent
archives, as the posting guide asks, for this question was discussed
recently.

> Uwe Ligges <ligges at statistik.uni-dortmund.de> wrote:
> lu kan wrote:
> 
> > I used a dataset to fit a linear model. For each new sample, a precition can be obtained using the fitted model. Now I wonder is there any way to calculate the confidence interval of the precition for new samples? Any suggestion or pointing to a reference would be very appreciated.
> 
> You already found predict(), so read ?predict.lm carefully and find that 
> there is a argument "interval".

> PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html

-- 
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 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595