[R] Weighted Savitzky-Golay?

Liaw, Andy andy_liaw at merck.com
Mon Oct 4 18:48:33 CEST 2004


> From: Matt Hart 
> 
> Thanks. Yes, I think this is right. As I understand it S-G 
> does a local 
> unweighted polynomial regression. I think other smoothers 
> already weigh 
> the data by the distance from the estimation point. I'm not sure what 
> happens when you mix the two...

Makes perfect sense to statisticians.  The two sets of weights are for
different reasons.  The distance weights (kernels) are for smoothness and
bias/variance properties of the estimates.

Andy
 
> m.
> 
> Liaw, Andy wrote:
> 
> >If I'm not mistaken, S-G is essentially a local 
> (even-degree) polynomial
> >smoother with constant bandwidth.  You can use a constant 
> bandwidth local
> >polynomial smoother that allows weights; e.g., in the locfit package.
> >
> >HTH,
> >Andy
> >
> >  
> >
> >>From: Matt Hart
> >>
> >>Hi,
> >>
> >>Does anyone know how to use weights and generate error bounds for 
> >>Savitzky-Golay? I have a (smallish) set of points y equally 
> >>spaced each 
> >>with a known error and would like to smooth them using S-G 
> >>but so as to 
> >>take into account the error already have and construct new 
> >>error bounds 
> >>around them that take into account the errors they had at the 
> >>beginning 
> >>and the erros they get as a result of the smoothing.
> >>
> >>thanks, m.
> >>
> >>______________________________________________
> >>R-help at stat.math.ethz.ch mailing list
> >>https://stat.ethz.ch/mailman/listinfo/r-help
> >>PLEASE do read the posting guide! 
> >>http://www.R-project.org/posting-guide.html
> >>
> >>
> >>    
> >>
> >
> >______________________________________________
> >R-help at stat.math.ethz.ch mailing list
> >https://stat.ethz.ch/mailman/listinfo/r-help
> >PLEASE do read the posting guide! 
> http://www.R-project.org/posting-guide.html
> >
> >  
> >
> 
> 
>




More information about the R-help mailing list