Distribution of the largest root of a matrix for Roy's test in multivariate analysis of variance

From MaRDI portal
(Redirected from Publication:900830)




Abstract: Let denote two independent real Gaussian mathsfpimesmathsfm and mathsfpimesmathsfn matrices with mathsfm,mathsfngeqmathsfp, each constituted by zero mean i.i.d. columns with common covariance. The Roy's largest root criterion, used in multivariate analysis of variance (MANOVA), is based on the statistic of the largest eigenvalue, Theta1, of , where and are independent central Wishart matrices. We derive a new expression and efficient recursive formulas for the exact distribution of Theta1. The expression can be easily calculated even for large parameters, eliminating the need of pre-calculated tables for the application of the Roy's test.









This page was built for publication: Distribution of the largest root of a matrix for Roy's test in multivariate analysis of variance

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q900830)