a numeric matrix of dimension <code>N</code> by <code>d</code> who's rows contain the points where the kernel will be evaluated.

a list with elements <code>alphas</code>, the set of projection directions, and <code>lambdas</code>, their associated bandwidth estimates.

bandwidths

a single numeric value containing a common scaling parameter.

a function for evaluating the univariate kernel.

kernel


Evaluate a <code>d</code> dimensional back projected kernel.


Nonparametric multivariate kernel density \
estimation using a back-projected kernel.