Kernel density estimation for (hyper-)spherical data using a von Mises-Fisher kernel
A von Mises-Fisher kernel is used for the non parametric density estimation.
vmf.kde(x, h, thumb = "none")
x |
A matrix with unit vectors, i.e. the data being expressed in Euclidean cordinates. |
h |
The bandwidth to be used. |
thumb |
If this is "none", the given bandwidth is used. If it is "rot" the rule of thumb suggested by Garcia-Portugues (2013) is used. |
A von Mises-Fisher kernel is used for the non parametric density estimation.
A list including:
h |
The bandwidth used. |
f |
A vector with the kernel density estimate calculated for each of the data points. |
Michail Tsagris
R implementation and documentation: Michail Tsagris mtsagris@uoc.gr and Giorgos Athineou <gioathineou@gmail.com>
Garcia Portugues, E. (2013). Exact risk improvement of bandwidth selectors for kernel density estimation with directional data. Electronic Journal of Statistics, 7, 1655-1685.
x <- rvmf(100, rnorm(5), 15) h <- vmfkde.tune(x)[1] f1 <- vmf.kde(x, h = h, thumb = "none") f2 <- vmf.kde(x, h = h, thumb = "rot") f1$h ; f2$h
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.