Title: Projection-based uniformity tests for directional data
Authors: Eduardo Garcia-Portugues - Carlos III University of Madrid (Spain) [presenting]
Paula Navarro - Universidad de Cantabria (Spain)
Juan A Cuesta-Albertos - Universidad de Cantabria (Spain)
Abstract: Testing uniformity of a sample supported on the hypersphere is one of the first steps when analyzing multivariate data for which only the directions (and not the magnitudes) are of interest. A projection-based class of uniformity tests on the hypersphere is introduced. The new class allows for extensions of circular-only uniformity tests and introduces the first instance of an Anderson-Darling test in the context of directional data. Asymptotics and optimality of the new class of tests are discussed. A simulation study corroborates the theoretical findings. Finally, a real data example illustrates the usage of the new tests.