Compute the M1 least-squares cross-validation score.
Usage
M1(grid, kern.fun, ...)
Arguments
grid
an object of class linbin.
kern.fun
the density function of the kernel. The first argument must be a matrix whose rows contain the points where the density function will be evaluated. See, for example, the dmvnorm function in the mvtnorm package.
...
additional arguments are passed to kern.fun.
Value
a single numeric value: the M1 least-squares cross-validation score.
Details
The computation is done using the Fourier transform of the kernel and the data as described in Silverman (1986).
References
Silverman, B. W. (1986) Density Estimation for Statistics and Data Analysis. London: Chapman and Hall.