実対称三重対角固有値問題に対する多分割の分割統治法の改良(理論,行列・固有値問題の解法とその応用,<特集>平成18年研究部会連合発表会)
An Improvement of Multiple Division Divide-and-Conquer for Real Symmetric Tridiagonal Eigenproblem(Theory,Algorithms for Matrix/Eigenvalue Problems and their Application, "Joint Symposium of JSIAM Activity Groups 2006")

Yutaka Kuwajima, Takaomi Shigehara
2006 Nihon Oyo Suri Gakkai ronbunshi  
Abstr α ct . We improve divide − and − conquer with multiple divisions for real symmetric tridiagonal eigenproblem prQposed ill [ 13 } . The main improvements are the following two . The first is that we succeed in developing an algorithm for keeping the orthogonality among eigenvectors in double − precision floating − point number processing without a substantial increase of numerical cost , As a result , all the calculations are done without quadruple − precision 且oating − point number
more » ... point number processing , which is required in [ 13 ] 。 The secQnd is that we implement a deflation effect which substantially decreases numerical cost . As a result of these two improvements , we succeed in developing a program for real symmetric tridiagonal eigenproblem which is even faster than a LAPACK routine DSTEVD , while keeping a numerical accuracy comparable to DSTEVD .
doi:10.11540/jsiamt.16.4_453 fatcat:7lpycnsq7bd7djl3x22wrki5l4