\name{stam.fit}
\alias{stam.fit}
\title{Fit StAM Model to Training Data}
\description{
Using the whole expression data provided fit one StAM model
according to the chosen shrinkage level.
}
\usage{
stam.fit(cv, expression.matrix, collapse.scnodes = FALSE,
alpha = 0.5, delta = NULL, max.nodes = 100)
}
\arguments{
\item{cv}{result stam.cv on the same data, must be of class \code{stamCV}}
\item{expression.matrix}{holds the expression levels. It may be of
class \code{exprSet} or \code{ExpressionSet}, or a
plain numeric matrix. In the first case \code{exprs} is used to extract the
expression levels. The matrix is expected to hold one column per sample and one row
per probeset.}
\item{collapse.scnodes}{if set to TRUE replace single children nodes after shrinkage}
\item{alpha}{root performance vs. mean redundancy weight. If set to NULL the root error
rate is used exclusively to determine the best shrinkage level. If a numeric vector
is provided, all alternatives are computed and the user is given an interactive
choice. Values between 0 and 1 are valid, 0 meaning exclusive weight on redundancy
and 1 putting exclusive weight on performance.}
\item{delta}{overrule alpha and set shrinkage level explicitely.}
\item{max.nodes}{choose default shrinkage level such that no more than this number of
nodes remain after shrinkage. }
}
\details{
In a first step \code{stam.fit} must choose a shrinkage level. In order to do so it
uses results stored in the cv. If the user provides a shrinkage level explicitly this
\code{delta} is used. If he specifies a single weighting factor \code{alpha} the
corresponding weighted score is used to determine the best shrinkage level. If
\code{alpha} is set to a vector of values, the corresponding scores are computed and a
default \code{delta} is suggested using the median value of the \code{alphas}. If this
shrinkage level leads to more than \code{max.nodes} nodes remaining the shrinkage level
is increase until no more than \code{max.nodes} remain after shrinkage.
Using the thus determined shrinkage level a weighting of nodes is computed using the
leaf node results from \code{cv}. Thereby, the whole dataset supplied is used.
}
\value{
An object of class stamFit is returned. You may use the print and plot methods
to further investigate the returned value.
}
\author{Claudio Lottaz}
\seealso{\code{\link{stam.cv}}, \code{\link{stamFit-class}}, \code{\link{plot.stamFit}},
\code{\link{stam.graph.plot}}, \code{\link{stam.writeHTML}}}
\examples{
\dontrun{
# prepare data
library(golubEsets)
data(Golub_Merge)
# load and prepare some data
golubTrain <- Golub_Merge[,1:38]
data(golubTrain.cv)
# compute fit
golubTrain.fit <- stam.fit(golubTrain.cv, golubTrain, alpha=seq(0, 1, 0.1))
# investigate
print(golubTrain.fit)
plot(golubTrain.fit)
}
# show clickable web-page
\dontrun{
map <- stam.graph.plot(golubTrain.fit, outfile="golubTrain")
cat("",
"
\n",
file="graph_plot.html")
browseURL(paste("file://", getwd(), "/graph_plot.html", sep=""))
}}
\keyword{classif}