The R Project SVN R

\name{princomp}
\alias{princomp}
\alias{princomp.formula}
\alias{princomp.default}
\alias{plot.princomp}
\alias{print.princomp}
\alias{predict.princomp}
\title{Principal Components Analysis}
\usage{
\method{princomp}{formula}(x, data = NULL, subset, na.action, ...)
\method{princomp}{default}(x, cor = FALSE, scores = TRUE, covmat = NULL,
         subset = rep(TRUE, nrow(as.matrix(x))), ...)
}
\arguments{
  \item{x}{a formula or matrix (or data frame) which provides the data for the
    principal components analysis.}
  \item{data}{an optional data frame containing the variables in the
    formula \code{x}. By default the variables are taken from
    \code{environment(x)}.}
  \item{subset}{an optional vector used to select rows (observations) of the
    data matrix \code{x}.}
  \item{na.action}{a function which indicates what should happen
    when the data contain \code{NA}s.  The default is set by
    the \code{na.action} setting of \code{\link{options}}, and is
    \code{\link{na.fail}} if that is unset. The ``factory-fresh''
    default is \code{\link{na.omit}}.}
  \item{cor}{a logical value indicating whether the calculation should
    use the correlation matrix or the covariance matrix.}
  \item{scores}{a logical value indicating whether the score on each
    principal component should be calculated.}
  \item{covmat}{a covariance matrix, or a covariance list as returned by
    \code{\link{cov.wt}}, \code{\link{cov.mve}} or \code{\link{cov.mcd}}.
    If supplied, this is used rather than the covariance matrix of
    \code{x}.}
  \item{\dots}{arguments passed to or from other methods. If \code{x} is
    a formula one might specify \code{cor} or \code{scores}.}
}
\description{
  \code{princomp} performs a principal components analysis on the given
  data matrix and returns the results as an object of class
  \code{princomp}.
}
\value{
  \code{princomp} returns a list with class \code{"princomp"}
  containing the following components:
  \item{sdev}{the standard deviations of the principal components.}
  \item{loadings}{the matrix of variable loadings (i.e., a matrix
    whose columns contain the eigenvectors).  This is of class
    \code{"loadings"}: see \code{\link{loadings}} for its \code{print}
    method.}
  \item{center}{the means that were subtracted.}
  \item{scale}{the scalings applied to each variable.}
  \item{n.obs}{the number of observations.}
  \item{scores}{if \code{scores = TRUE}, the scores of the supplied
    data on the principal components.}
  \item{call}{the matched call.}
  \item{na.action}{If relevant.}
}
\details{
  \code{princomp} is a generic function with \code{"formula"} and
  \code{"default"} methods.

  The calculation is done using \code{\link{eigen}} on the correlation or
  covariance matrix, as determined by \code{\link{cor}}.  This is done for
  compatibility with the S-PLUS result.  A preferred method of
  calculation is to use svd on \code{x}, as is done in \code{prcomp}.

  Note that the default calculation uses divisor \code{N} for the
  covariance matrix.

  The \code{\link{print}} method for the these objects prints the
  results in a nice format and the \code{\link{plot}} method produces
  a scree plot (\code{\link{screeplot}}).  There is also a
  \code{\link{biplot}} method.

  If \code{x} is a formula then the standard NA-handling is applied to
  the scores (if requested): see \code{\link{napredict}}.
}
\references{
  Mardia, K. V., J. T. Kent and J. M. Bibby (1979).
  \emph{Multivariate Analysis}, London: Academic Press.

  Venables, W. N. and B. D. Ripley (1997, 9).
  \emph{Modern Applied Statistics with S-PLUS}, Springer-Verlag.
}
\seealso{
  \code{\link{summary.princomp}}, \code{\link{screeplot}},
  \code{\link{biplot.princomp}},
  \code{\link{prcomp}}, \code{\link{cor}}, \code{\link{cov}},
  \code{\link{eigen}}.
}
\examples{
## The variances of the variables in the
## USArrests data vary by orders of magnitude, so scaling is appropriate
data(USArrests)
(pc.cr <- princomp(USArrests))  # inappropriate
princomp(USArrests, cor = TRUE) # =^= prcomp(USArrests, scale=TRUE)
## Similar, but different:
## The standard deviations differ by a factor of sqrt(49/50)

summary(pc.cr <- princomp(USArrests, cor = TRUE))
loadings(pc.cr)  ## note that blank entries are small but not zero
plot(pc.cr) # shows a screeplot.
biplot(pc.cr)

## Formula interface
princomp(~ ., data = USArrests, cor = TRUE)
# NA-handling
USArrests[1, 2] <- NA
pc.cr <- princomp(~ ., data = USArrests, na.action=na.exclude, cor = TRUE)
pc.cr$scores
}
\keyword{multivariate}