s-news
[Top] [All Lists]

Loess standard error curves

To: <s-news@lists.biostat.wustl.edu>
Subject: Loess standard error curves
From: Eduardo Nogueira <nogueira@hc.unicamp.br>
Date: Fri, 23 Jan 2004 13:22:37 -0200
Organization: Unicamp
Reply-to: nogueira@hc.unicamp.br
User-agent: KMail/1.4.3
Dear Splus users

On a trellis graphic, I plotted   loess smooth curves of blood
pressure along time conditioned by the presence of age(4 levels)  and
hypertension (2 levels): 


xyplot(pressure~time|hypertension,data=data1,layout=c(4,2,1),
       panel=function(x,y){
         panel.xyplot(x,y )
         panel.loess(x,y,span=0.25) })

Is it possible to  add (superpose)  the standard error curves of the loess fit 
to this graphic?


Thanks

Eduardo
----------------------------------------------------------
Eduardo A Nogueira, MD, PhD
Cardiac Catheterization Laboratory
University of Campinas Medical School

<Prev in Thread] Current Thread [Next in Thread>
  • Loess standard error curves, Eduardo Nogueira <=