| Tuning of the bandwidth parameter in the von Mises-Fisher kernel {Directional} | R Documentation |
Tuning of the bandwidth parameter in the von Mises-Fisher kernel for (hyper-)spherical data
Description
Tuning of the bandwidth parameter in the von Mises-Fisher kernel for (hyper-)spherical data whit cross validation.
Usage
vmfkde.tune(x, low = 0.1, up = 1)
Arguments
x |
A matrix with the data in Euclidean cordinates, i.e. unit vectors. |
low |
The lower value of the bandwdith to search. |
up |
The upper value of the bandwdith to search. |
Details
Fast tuning of the bandwidth parameter in the von Mises-Fisher kernel for (hyper-)spherical data via cross validation.
Value
A vector including two elements:
Optimal h |
The best H found. |
cv |
The value of the maximised pseudo-likelihood. |
Author(s)
Michail Tsagris.
R implementation and documentation: Michail Tsagris mtsagris@uoc.gr and Giorgos Athineou <gioathineou@gmail.com>.
References
Garcia P.E. (2013). Exact risk improvement of bandwidth selectors for kernel density estimation with directional data. Electronic Journal of Statistics, 7, 1655–1685.
Wand M.P. and Jones M.C. (1994). Kernel smoothing. Crc Press.
See Also
vmf.kde,vmf.kerncontour, vm.kde, vmkde.tune
Examples
x <- rvmf(100, rnorm(3), 15)
vmfkde.tune(x)