Distribution Theory for the Von Mises-Fisher Distribution and Its Application
The von Mises-Fisher distribution is the most important distribution in directional data analysis. We derive the sampling distributions of the sample resultant length, the sample mean direction and the component lengths. For the multi-sample case, the conditional distribution of the individual sample resultant lengths given the combined sample resultant length is derived. These results depend heavily on the corresponding distributions for the isotropic random walk on hypersphere. Using these results we investigate some optimum properties of various important tests. Most of these tests were formulated intuitively by Watson and Williams (1956). Mardia (1972) in his book concentrated on the optimum properties of the circular and spherical cases, and this paper extends and unifies some of the parametric work.
Key WordsDirectional data analysis von Mises-Fisher distribution multi-sample problems
Unable to display preview. Download preview PDF.
- Fisher, R. A. (1953). Proc. Roy. Soc. Lond. A 217, 295–305.Google Scholar
- Mardia, K. V. (1974a). Characterizations of directional distributions. Presented to the First International Conference on Characterizations. Calgary, Canada.Google Scholar
- Mardia, K. V. (1974b). Statistics of directional data. To be presented to the Roy. Statist. Soc.Google Scholar
- Mardia, K. V. and Zemroch, P. J. (1974a). Circular statistics. Algorithm to appear in J. Roy. Statist. Soc. Ser. C.Google Scholar
- Mardia, K. V. and Zemroch, P. J. (1974b). Spherical statistics. Algorithm to appear in J. Roy. Statist. Soc. Ser. C.Google Scholar
- Rao, J. S. (1969). Some contributions to the analysis of circular data. Ph.D. Thesis, Indian Statistical Institute, Calcutta.Google Scholar
- Stephens, M. A. (1962). The statistics of directions. Ph.D. Thesis, Toronto University, Canada.Google Scholar
- Watson, G. N. (1948). A Treatise on the Theory of Bessel Functions. 2nd. edn. Cambridge Univ. Press.Google Scholar