Linear binning for multivariate data

Linear binning for 1- to 4-dimensional data.

Usage

binning(x, H, h, bgridsize, xmin, xmax, supp=3.7, w, gridtype="linear")

Arguments

x: matrix of data values
H,h: bandwidth matrix, scalar bandwidth
xmin,xmax: vector of minimum/maximum values for grid
supp: effective support for standard normal is [-supp,supp]
bgridsize: vector of binning grid sizes
w: vector of weights. Default is a vector of all ones.
gridtype: not yet implemented

Value

Returns a list with 2 fields

counts: linear binning counts
eval.points: vector (d=1) or list (d>=2) of grid points in each dimension

Details

For ks \(\geq\) 1.10.0, binning is available for unconstrained (non-diagonal) bandwidth matrices. Code is used courtesy of A. & J. Gramacki, and M.P. Wand. Default bgridsize are d=1: 401; d=2: rep(151, 2); d=3: rep(51, 3); d=4: rep(21, 4).

References

Gramacki, A. & Gramacki, J. (2016) FFT-based fast computation of multivariate kernel estimators with unconstrained bandwidth matrices. Journal of Computational & Graphical Statistics, 26, 459–462.

Wand, M.P. & Jones, M.C. (1995) Kernel Smoothing. Chapman & Hall. London.

Examples

data(unicef)
ubinned <- binning(x=unicef)