Fast calculation of the spherical and hyperspherical median
It calculates, very fast, the (hyper-)spherical median of a sample.
mediandir(x) mediandir_2(x)
x |
The data, a numeric matrix with unit vectors. |
The "mediandir" employes a fixed poit iterative algorithm stemming from the first derivative (Cabrera and Watson, 1990) to find the median direction as described by Fisher (1985) and Fisher, Lewis and Embleton (1987). In the big samples this is much much faster than "mediandir_2", since the search is based on iterations.
The median direction.
Michail Tsagris
R implementation and documentation: Michail Tsagris mtsagris@uoc.gr and Giorgos Athineou <gioathineou@gmail.com>
Fisher N. I. (1985). Spherical medians. Journal of the Royal Statistical Society. Series B, 47(2): 342-348.
Fisher N. I., Lewis T. and Embleton B. J. (1987). Statistical analysis of spherical data. Cambridge university press.
Cabrera J. and Watson G. S. (1990). On a spherical median related distribution. Communications in Statistics-Theory and Methods, 19(6): 1973-1986.
m <- rnorm(3) m <- m / sqrt( sum(m^2) ) x <- rvmf(100, m, 10) mediandir(x) mediandir_2(x)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.