RUS  ENG
Full version
JOURNALS // Zhurnal Vychislitel'noi Matematiki i Matematicheskoi Fiziki // Archive

Zh. Vychisl. Mat. Mat. Fiz., 2024 Volume 64, Number 12, Pages 2262–2269 (Mi zvmmf11887)

This article is cited in 1 paper

General numerical methods

A nonsingular matrix with a well-conditioned cosquare: how to bring IT to diagonal form by a congruence transformation

Kh. D. Ikramova, A. M. Nazarib

a Faculty of Computational Mathematics and Cybernetics, Lomonosov Moscow State University, Moscow, Russia
b Arak University, Arak, Islamic Republic Iran

Abstract: There exist efficient programs for bringing a diagonalizable matrix to diagonal form by a similarity transformation. In theory of congruence transformations, unitoid matrices are analogs of diagonalizable matrices. However, excepting Hermitian and, more generally, normal matrices, there are no recognized programs for bringing a unitoid matrix to diagonal form by a congruence transformation. We propose an algorithm that is able to perform this task for a special class of unitoid matrices, namely, nonsingular matrices whose cosquares are well-conditioned with respect to the complete eigenproblem. Examples are presented to illustrate the performance of the algorithm.

Key words: *-congruence transformation, similarity transformation, unitoid, cosquare, canonical angles, diagonally dominant matrix, condition number.

UDC: 512.643

Received: 09.01.2024
Revised: 09.01.2024
Accepted: 31.05.2024

DOI: 10.31857/S0044466924120035


 English version:
Computational Mathematics and Mathematical Physics, 2024, 64:12, 2770–2779

Bibliographic databases:


© Steklov Math. Inst. of RAS, 2026