ResearchTrend.AI
  • Papers
  • Communities
  • Organizations
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 1401.3987
119
22
v1v2v3v4 (latest)

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

16 January 2014
M. Chiani
ArXiv (abs)PDFHTML
Abstract

Let X,Y{\bf X, Y} X,Y denote two independent real Gaussian p×m\mathsf{p} \times \mathsf{m}p×m and p×n\mathsf{p} \times \mathsf{n}p×n matrices with m,n≥p\mathsf{m}, \mathsf{n} \geq \mathsf{p}m,n≥p, 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, Θ1\Theta_1Θ1​, of (A+B)−1B{\bf{(A+B)}}^{-1} \bf{B}(A+B)−1B, where A=XXT{\bf A =X X}^TA=XXT and B=YYT{\bf B =Y Y}^TB=YYT are independent central Wishart matrices. We derive a new expression and efficient recursive formulas for the exact distribution of Θ1\Theta_1Θ1​. 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.

View on arXiv
Comments on this paper