Compute a binned approximation of the data on a regularly spaced grid using the multivariate linear binning rule described in Wand (1994).
Usage
mvlinbin(X, r = 7, padding)
Arguments
X
a numeric matrix.
r
a positive integer value. The number of grid points $M$ in each dimension is given by $M = 2^r$.
padding
a numeric vector of positive values with length equal to the number of columns of X specifying the amount of zero-padding added to each coordinate direction. No padding is added when this argument is missing.
Value
a list with class mvlinbin containing the following elements.
axes
a numeric matrix whose columns contain the grid points used along each axis to bin the data.
xi
a numeric array containing the binned approximation of the data.
X
a numeric matrix containing the input data.
deltas
a numeric vector containing the grid spacing.
M
an integer value giving the number of grid points used in each coordinate direction.
n
an integer value containing the number of data points binned.
d
an integer value giving the dimensionality of the data.
References
Wand, M. P. (1994). Fast Computation of Multivariate Kernel Estimators. Journal of Computational and Graphical Statistics, 3, 433-445.