R help - Oct 2005 - problems with loess

Hi all,

I was unable to obtained a smoothed line using the loess function.
I used the following code reported in the examples of R documentation:

cars.lo <- loess(dist ~ speed, cars)

Then I tried to plot both the data and the smoothed line

plot(cars)
lines(cars.lo)

but what I obtained is simply a broken line joining all the data points.
I tried with different spans, but the results did not change.


Does anyone know what's going wrong? 


Thanks

Dimitri




platform i386-pc-mingw32
arch     i386           
os       mingw32        
system   i386, mingw32  
status                  
major    2              
minor    2.0    




-- 
Dimitri Giunchi
Dipartimento di Etologia Ecologia Evoluzione
Universit伱仩 di Pisa
Via A. Volta 6
56126 Pisa
ITALY
Tel: +39 050 2219036
Fax: +39 050 24653

On Fri, 7 Oct 2005, Dimitri Giunchi wrote:
> Hi all,
> I was unable to obtained a smoothed line using the loess function.I used
the following code reported in the examples of R documentation:
> cars.lo <- loess(dist ~ speed, cars)
> Then I tried to plot both the data and the smoothed line
> plot(cars)lines(cars.lo)
> but what I obtained is simply a broken line joining all the data points.I
tried with different spans, but the results did not change.
>
> Does anyone know what's going wrong?
Yes. lines(cars.lo) plots the original data stored in the loess fit.

You need to predict, or use scatter.smooth to do it for you.

-- 
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