Year: 2024
Type: article
Abstract: High Resolution Image Download MS PowerPoint Slide Given a sparse matrix A, we define and discuss a matrix B that approximates A –1, has exactly the same sparsity structure as A, and varies contravariantly with A . Unlike many other sparse inverse approximations, the iso-inverse B is not constructed by minimizing the residual ∥ AB – I ∥ but, instead, by stipulating that AB = I for a well-defined subset of the elements of I . The iso-inverse is potentially useful in electronic structure calculations with nonorthogonal localized molecular orbitals.
Authors: Peter M. W. Gill, Martin Mrovec
Institution: The University of Sydney
Language: English
Topic: Matrix Theory and Algorithms
Subfield: Computational Theory and Mathematics
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Funder: University of Sydney
OpenAlex ID: W4404863001