\name{buildVarsFromPvecs}
\alias{buildVarsFromPvecs}
\title{Builds a variable descriptions from probability vectors}
\description{
  Takes a named list of marginal probability distributions and produces
  a set of \code{\link{varDescription}} objects from the meta-data.
}
\usage{
buildVarsFromPvecs(pvecs, role = "SM")
}
\arguments{
  \item{pvecs}{A list of vectors giving marginal distributions of
    various probability distributions}
  \item{role}{A character vector with the names names as \code{pvecs}
    giving the role for each variable, or a scaler giving the role for
    all variables.  Should be one of \dQuote{SM}, \dQuote{EM} or
    \dQuote{Observable}.}
}
\details{
  This creates a collection of \code{\link{varDescription}} objects
  whose names are given by the \code{names(pvecs)} and whose levels are
  given by the names of the elements of \code{pvecs}.  The roles are
  given by \code{role}
}
\value{
  A list of the same length as \code{pvecs} containing variable
  description objects.
}
\author{Russell Almond}
\note{ 
  StatShop usually assumes that the states are ordered from the
  \emph{highest} to the \emph{lowest} possible values, for example
  \sQuote{High}, \sQuote{Med}, \sQuote{Low}.
}
\section{Under Construction}{
  This function was designed and coded, but never used.  May go away.
}
\seealso{\code{\link{varDescription}} }
\examples{
data(MathGrades)
fivevars <- buildVarsFromPvecs(MathGrades$pvecs)
}
\keyword{manip}