Spectral division methods for block generalized Schur decompositions (Q4813611)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 2091006
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Spectral division methods for block generalized Schur decompositions |
scientific article; zbMATH DE number 2091006 |
Statements
Spectral division methods for block generalized Schur decompositions (English)
0 references
13 August 2004
0 references
generalized eigenproblem
0 references
matrix sign and disc functions
0 references
spectral divide-and-conquer algorithms
0 references
comparison of methods
0 references
spectral division method
0 references
block generalized Schur decompositions
0 references
matrix pairs
0 references
subspace extraction technique
0 references
numerical experiments
0 references
0 references
0 references
0 references
0 references
0.7696801424026489
0 references
0.7611976265907288
0 references
0.7589651346206665
0 references
0.752325713634491
0 references
0.7511925101280212
0 references
A modified spectral division method for block generalized Schur decompositions of matrix pairs is provided. The new method bases on \textit{A. N. Malyshev}'s iteration [Linear Algebra Appl. 188/189, 489--520 (1993; Zbl 0782.65056)]. The given analysis exposes more algebraic structures of the successive matrix pairs in the spectral division iterations and reveals some potential computational difficulties in the subspace extraction stage of the existing algorithms. The authors present a subspace extraction technique that computes bases for both the left- and right-deflating subspace with a single Malyshev iteration. This reduces the arithmetic cost by nearly 50\%. The numerical experiments with three examples compare the algorithm of \textit{A. N. Malyshev} [loc. cit.], and of \textit{Z. Bai, J. Demmel} and \textit{M. Gu} [Num. Math., 76, No.3, 279--308 (1997; Zbl 0876.65021)] with the proposed algorithm. An internet address to download the matlab files is given.
0 references
