Testingequalityofcovariancematrixhaslongbeenaninterestingissueinstatisticsinference,Toovercomethesparsenessofdatapointsinhigh-dimensionalspaceanddealwiththegeneralcases,theauthorsuggestsseveralprojectionpursuittypestatistics.Someresultsonthelimitingdistidbutionsofthestatisticsareobtained.Somepropertiesofbootstrapapproximationareinvestigated.Furthermore,forcomputationalreasonsanapproximationforthestatisticsbasedonnumber-theoreticroethodisapplied.Severalsimulationexperimentsareperformed.
