Answers are grounded in this paper and cite their sources.
We consider the iterative solution of symmetric saddle-point matrices with a singular leading block. We develop a new ideal positive definite block-diagonal preconditioner that yields a preconditioned operator with four distinct eigenvalues. We offer a few techniques for making the preconditioner practical and illustrate the effectiveness of our approach with numerical experiments. The novelty of the paper lies in the generality of the assumptions made: as long as the saddle-point matrix is nonsingular, there is no assumption on the specific rank of the leading block. Current ideal preconditioners typically rely either on invertibility or a high nullity of the leading block, and the new technique aims to bridge this gap. A spectral analysis is offered, accompanied by numerical experiments.
No takes yet. Share an insight, caveat, or question.
Bradley et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: