## The shift-and-invert Arnoldi method has been popularly used for computing a number of eigenvalues close to a given shift and/or the associated eigenvectors of a large unsymmetric matrix pair, but there is no guarantee for the approximate eigenvectors, Ritz vectors, obtained by this method to co
A new method for accelerating Arnoldi algorithms for large scale eigenproblems
β Scribed by K. Dookhitram; R. Boojhawon; M. Bhuruth
- Publisher
- Elsevier Science
- Year
- 2009
- Tongue
- English
- Weight
- 276 KB
- Volume
- 80
- Category
- Article
- ISSN
- 0378-4754
No coin nor oath required. For personal study only.
π SIMILAR VOLUMES
The block Arnoldi method is one of the most commonly used techniques for large eigenproblems. In this paper, we exploit certain modified Ritz vectors to take the place of Ritz vectors in the thick-restarted block Arnoldi algorithm, and propose a modified thickrestarted block Arnoldi algorithm for la
When the matrix in question is uns)xnmetric, the approximate eigenvectors or Ritz vectors obtained by orthogonal projection methods including Arnoldi's method and the block Arnoldi method cannot be guaranteed to converge in theory even if the corresponding approximate eigenvalues or Ritz values do.
## A NEW METHOD FOR LARGE-SCALE CI CALCULATIONS A new method for obtaining the coefficients in a fv.rgc CI expansion is pmpnsed. Instead flf consfructing an intcrxtlon matrix, the expansion coefficients ;1~e obtained directly from the list of two-electron integrals by means of 3n iterative procedu
Population initialization is a crucial task in evolutionary algorithms because it can affect the convergence speed and also the quality of the final solution. If no information about the solution is available, then random initialization is the most commonly used method to generate candidate solution