

   NNoorrmmaall kkeerrnneell ddeennssiittyy eessttiimmaattee

        nkden(data, bandwidth, n.points, grid)

   AArrgguummeennttss::

       data: A vector or matrix of oberservations. Rows are
             considerd to be independent random samples from a
             continous distribution.

   bandwidth: The bandwidht for the kernels estimates in the
             scale of standard deviation for the normal den-
             sity.

   n.points: Number of equally spaced points to evaluate a uni-
             variate density estimate.

       grid: A vector or matrix of values to evaluate the esti-
             mate.The defautl is to use the data.

   VVaalluuee::

        A list where x is the points used for evaluation, y the
        density estimates at these points and bandwidths and h
        the vector of bandwidths. If more than one bandwidth is
        given then the estimates are arranged as columns in the
        matrix y.

   SSeeee AAllssoo::

        nkreg, nkden.cv, ksmooth

   EExxaammpplleess::

        # univariate estimate with several bandwidths

        nkden( minitri$swim, c(2.0,4.0,6.0),n.points=150)-> look
        matplot( look$x, look$y, type="l") # plot all of them togther

        # a bivariate estimate

        nkden( minitri[,1:2], 2.0) -> look2
        # create gridded surface from values at data points.

        interp( look2$x[,1], look2$x[,2], look2$y) -> surface
        persp(surface, xlab="swim", ylab="bike")

