Hypothesis test for von Mises-Fisher distribution over Kent distribution
The null hypothesis is whether a von Mises-Fisher distribution fits the data well, where the altenrative is that Kent distribution is more suitable.
fishkent(x, B = 999)
x |
A numeric matrix containing the data as unit vectors in Euclidean coordinates. |
B |
The number of bootstrap re-samples. By default is set to 999. If it is equal to 1, no bootstrap is performed and the p-value is obtained throught the asymptotic distribution. |
Essentially it is a test of rotational symmetry, whether Kent's ovalness parameter (beta) is equal to zero. This works for spherical data only.
A vector including:
test |
The value of the test statistic |
p-value or Bootstrap p-value |
The p-value of the test. |
Michail Tsagris.
R implementation and documentation: Michail Tsagris mtsagris@uoc.gr and Giorgos Athineou <gioathineou@gmail.com>.
Rivest, L. P. (1986). Modified Kent's statistics for testing goodness of fit for the Fisher distribution in small concentrated samples. Statistics & probability letters, 4(1): 1-4.
x <- rvmf(100, rnorm(3), 15) fishkent(x) fishkent(x, B = 1) iagesag(x)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.