[R] Na action with Lowess smoothing

[ripley@stats.ox.ac.uk]

On Thu, 27 Mar 2003, kjetil brinchmann halvorsen wrote:

> On 27 Mar 2003 at 10:38, Wayne Jones wrote:
> 
> Check out the version og lowess in the gregmisc package, it has a
> formula interface and na.action argument.
> (Why cannot this be included in base R, to avoid having two othervise 
> identical functions?)

The one in R is a long-standing S-compatible one, and in S lowess was
superseded by loess over a decade ago.  So a better answer is `use loess
instead'.

> > I cant seem to find a way for the lowess smoothing function to handle "NA"
> > values. 

-- 
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