\name{get.h}
\alias{get.h}
%- Also NEED an '\alias' for EACH other topic documented here.
\title{Estimate the critical bandwidth for specific number of modes }
\description{
  get.h finds the critical bandwidth for specific number of modes.  That is, it finds the smallest bandwidth for which "m" modes are present for a kernel density estimator. 
}
\usage{
get.h(x, m = 1, prec = 0.001, hmin = 0, hmax = 1)
}
%- maybe also 'usage' for other objects documented here.
\arguments{
  \item{x}{ the data vector in which to find the critical bandwidth }
  \item{m}{ the number of modes for the critical bandwidth }
  \item{prec}{ the precision for the resulting bandwidth }
  \item{hmin}{ the minimum value to start searching for the critical bandwidth, h }
  \item{hmax}{ the maximum value to start searching for the critical bandwidth, h }
}
\details{
  get.h uses the Gaussian kernel to estimate the density of a data vector given by x.  The bandwidth determines the spread of each data point.  Thus a larger bandwidth leads to a smoother density estaimate.  get.h finds the smallest bandwidth in which "m" modes are still present.
}
\value{

  \item{h}{ the critical bandwidth, rescaled for the standardized x-values for direct comparison }

}
\references{B.W. Silverman (1981),Using Kernel Density Estimates to
  Investigate Multimodatlity. J.R. Statist. Soc. B,43,1,97-99.}
\author{ Kevin Rader }


\seealso{ \code{\link{get.p}}, \code{\link{emp.f}}, \code{\link{get.num.modes}}}
\examples{
 set.seed(12345)
 x <- c(rnorm(20,0),rnorm(20,3))
 get.h(x)
}
\keyword{distribution}