A crossproduct free JacobiDavidson type method for computing a partial generalized singular value decomposition (GSVD) of a large matrix pair
Abstract
A CrossProduct Free (CPF) JacobiDavidson (JD) type method is proposed to compute a partial generalized singular value decomposition (GSVD) of a large regular matrix pair $(A,B)$. It implicitly solves the mathematically equivalent generalized eigenvalue problem of $(A^TA,B^TB)$ but does not explicitly form the crossproduct matrices and thus avoids the possible accuracy loss of the computed generalized singular values and generalized singular vectors. The method is an innerouter iteration method, where the expansion of the right searching subspace forms the inner iterations that approximately solve the correction equations involved and the outer iterations extract approximate GSVD components with respect to the subspaces. Some convergence results are established for the inner and outer iterations, based on some of which practical stopping criteria are designed for the inner iterations. A thickrestart CPFJDGSVD algorithm with deflation is developed to compute several GSVD components. Numerical experiments illustrate the efficiency of the algorithm.
 Publication:

arXiv eprints
 Pub Date:
 April 2020
 DOI:
 10.48550/arXiv.2004.13975
 arXiv:
 arXiv:2004.13975
 Bibcode:
 2020arXiv200413975H
 Keywords:

 Mathematics  Numerical Analysis;
 65F15;
 15A18;
 15A12;
 65F10
 EPrint:
 31 pages, 2 figures