The R Project SVN R

% file modreg/man/supsmu.Rd
% copyright (C) 1998 B. D. Ripley
%
\name{supsmu}
\alias{supsmu}
\title{Friedman's SuperSmoother}
\description{
  Smooth the (x, y) values by Friedman's ``super smoother''.
}
\usage{
supsmu(x, y, wt = rep(1, length(y)), span = "cv", periodic = FALSE,
       bass = 0)
}
\arguments{
  \item{x}{x values for smoothing}
  \item{y}{y values for smoothing}
  \item{wt}{case weights}
  \item{span}{the fraction of the observations in the span of the running
    lines smoother, or \code{"cv"} to choose this by leave-one-out
    cross-validation.}
  \item{periodic}{if \code{TRUE}, the x values are assumed to be in
    \code{[0, 1]} and of period 1.}
  \item{bass}{controls the smoothness of the fitted curve. Values of up
    to 10 indicate increasing smoothness.}
}
\details{
  \code{supsmu} is a running lines smoother which chooses between three
  spans for the lines. The running lines smoothers are symmetric, with
  \code{k/2} data points each side of the predicted point, and values of
  \code{k} as \code{0.5 * n}, \code{0.2 * n} and \code{0.05 * n}, where
  \code{n} is the number of data points.  If \code{span} is specified,
  a single smoother with span \code{span * n} is used.

  The best of the three smoothers is chosen by cross-validation for each
  prediction. The best spans are then smoothed by a running lines
  smoother and the final prediction chosen by linear interpolation. 
 
  The FORTRAN code says: ``For small samples (\code{n < 40}) or if there
  are substantial serial correlations between observations close in x -
  value, then a prespecified fixed span smoother (\code{span > 0})
  should be used.  Reasonable span values are 0.2 to 0.4.''
}
\value{
  A list with components
  \item{x}{the input values in increasing order with duplicates removed.}
  \item{y}{the corresponding y values on the fitted curve.}
}
\references{
  Friedman, J. H. (1984)
  SMART User's Guide.
  Laboratory for Computational Statistics, Stanford University Technical
  Report No. 1.

  Friedman, J. H. (1984)
  A variable span scatterplot smoother.
  Laboratory for Computational Statistics, Stanford University Technical
  Report No. 5.
}
\author{B. D. Ripley}
\seealso{\code{\link{ppr}}}

\examples{
data(cars)
attach(cars)
plot(speed, dist)
lines(supsmu(speed, dist))
lines(supsmu(speed, dist, bass=7), lty=2)
detach()
}
\keyword{smooth}