A Parallel Algorithm for the Nonsymmetric Eigenvalue Problem
A parallel algorithm for computing the eigenvalues and eigenvectors of a nonsymmetric matrix is proposed. The approach uses some features of the divide-and-conquer algorithm for the symmetric case by the first author and \textit{D. C. Sorensen} [ibid. 8, S 139--S 154 (1987; Zbl 0627.65033)]. The original problem is divided into two smaller and independent subproblems by a rank-one modification of the matrix (it is assumed that the matrix is already reduced to Hessenberg form and that the rank-one modification removes a subdiagonal element). Once the eigensystems of the smaller subproblems are known, it is possible to compute those of the original matrix. For the nonsymmetric case the eigensystem of the subproblem is used only to construct initial guesses for an iteration process. An iteration approach based on Newton's method is presented. The deflation step required to overcome multiple convergence to a particular eigenvalue, the convergence of the new algorithm, the parallel algorithm and the different parallel implementations and numerical results are given. Extension to a generalized eigenvalue problem is described.
- scientific article; zbMATH DE number 556483
- scientific article; zbMATH DE number 52560
- scientific article; zbMATH DE number 1399479
- scientific article; zbMATH DE number 4155755
- scientific article; zbMATH DE number 176408
- A fast and stable parallel QR algorithm for symmetric tridiagonal matrices
- On the Orthogonality of Eigenvectors Computed by Divide-and-Conquer Techniques
- Towards a Divide and Conquer Algorithm for the Real Nonsymmetric Eigenvalue Problem
- Divide and conquer algorithms for the bandsymmetric eigenvalue problem
- The Homotopy Continuation Algorithm for the Real Nonsymmetric Eigenproblem: Further Development and Implementation
- Parallel subspace method for non-Hermitian eigenproblems on the Connection Machine (CM2)
- On parallelizable eigensolvers
- Parallel algorithm for solving some spectral problems of linear algebra
- A parallel Davidson-type algorithm for several eigenvalues
- A parallel continuation algorithm for nonlinear eigenvalue problems
- A parallel QR algorithm for the nonsymmetric eigenvalue problem
- An inverse free parallel spectral divide and conquer algorithm for nonsymmetric eigenproblems
- Parallel computation of polynomials with minimal uniform norm and its application to large eigenproblems
- Parallel algorithms for reduction of a general matrix to upper Hessenberg form on a shared memory multiprocessor
- Continuation methods for the computation of zeros of Szegő polynomials
- Parallel spectral division using the matrix sign function for the generalized eigenproblem
- scientific article; zbMATH DE number 434519 (Why is no real title available?)
- scientific article; zbMATH DE number 434522 (Why is no real title available?)
- scientific article; zbMATH DE number 4070183 (Why is no real title available?)
- scientific article; zbMATH DE number 1247167 (Why is no real title available?)
- scientific article; zbMATH DE number 556483 (Why is no real title available?)
- Towards a Divide and Conquer Algorithm for the Real Nonsymmetric Eigenvalue Problem
- A Parallelizable Eigensolver for Real Diagonalizable Matrices with Real Eigenvalues
- QR -like algorithms for the nonsymmetric eigenvalue problem
- The Spectral Decomposition of Nonsymmetric Matrices on Distributed Memory Parallel Computers
- scientific article; zbMATH DE number 1548072 (Why is no real title available?)
- On the Schur Decomposition of a Matrix for Parallel Computation
- A Parallel Implementation of the Nonsymmetric QR Algorithm for Distributed Memory Architectures
- scientific article; zbMATH DE number 2095809 (Why is no real title available?)
- Evaluating products of matrix pencils and collapsing matrix products
- scientific article; zbMATH DE number 819138 (Why is no real title available?)
- Parallelizing the QR Algorithm for the Unsymmetric Algebraic Eigenvalue Problem: Myths and Reality
- scientific article; zbMATH DE number 7468269 (Why is no real title available?)
- An efficient Jacobi-like algorithm for parallel eigenvalue computation
- Exponentially convergent parallel algorithm for nonlinear eigenvalue problems
- Algorithm 826
- Computational Science – ICCS 2005
- scientific article; zbMATH DE number 2234457 (Why is no real title available?)
- A parallel algorithm for the reduction of a nonsymmetric matrix to block upper-Hessenberg form
- A case against a divide and conquer approach to the nonsymmetric eigenvalue problem
- A quadratically convergent parallel Jacobi process for diagonally dominant matrices with nondistinct eigenvalues
- Finding eigenvalues and eigenvectors of unsymmetric matrices using a distributed-memory multiprocessor
- A parallel algorithm for the eigenvalues and eigenvectors of a general complex matrix
This page was built for publication: A Parallel Algorithm for the Nonsymmetric Eigenvalue Problem
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4693191)
