pith. machine review for the scientific record. sign in

arxiv: 2603.00419 · v2 · submitted 2026-02-28 · 🧮 math.NA · cs.NA

Recognition: no theorem link

Inexact versions of several block-splitting preconditioners for indefinite least squares problems

Mohaddese Kaveh Shaldehi , Davod Khojasteh Salkuyeh
Authors on Pith no claims yet

Pith reviewed 2026-05-15 18:51 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords block-splitting preconditionersindefinite least squaresGMRES convergenceeigenvalue clusteringstationary iterative methodsthree-by-three block systemspreconditioned matrices
0
0 comments X

The pith

Inexact block-splitting preconditioners confine all eigenvalues of the preconditioned matrix to the unit disk centered at 1 for indefinite least squares systems.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops inexact versions of block-splitting preconditioners to solve three-by-three block linear systems that arise from a special class of indefinite least squares problems. It first derives convergence conditions for the associated stationary iterative methods. Under those conditions the eigenvalues of each preconditioned matrix lie inside the circle of radius one centered at the point (1,0) in the complex plane. This clustering property guarantees rapid convergence of the GMRES method, and the authors supply a detailed eigenpair analysis together with an explicit upper bound on the number of GMRES iterations required.

Core claim

The paper shows that, whenever the convergence conditions for the stationary iterative methods hold, every eigenvalue of the preconditioned matrix produced by the inexact block-splitting preconditioners lies inside the disk of radius 1 centered at (1,0). This spectral containment directly accelerates GMRES, and the accompanying eigenpair analysis yields a concrete iteration bound for the preconditioned systems.

What carries the argument

Inexact block-splitting preconditioners applied to the three-by-three block coefficient matrix, which force all eigenvalues of the preconditioned operator inside the unit disk centered at (1,0).

If this is right

  • The preconditioners accelerate GMRES convergence on the target three-by-three block systems.
  • An explicit upper bound on the number of GMRES iterations follows from the eigenpair analysis.
  • Convergence of the underlying stationary iterative methods is guaranteed once the derived conditions are met.
  • Numerical tests on representative problems confirm practical effectiveness of the preconditioners.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same block-splitting construction could be applied to other structured indefinite linear systems that admit a comparable three-by-three partitioning.
  • The radius-1 disk bound might be refined or extended to related saddle-point or least-squares formulations outside the special class considered here.
  • Practical performance could be further improved by choosing different inner solvers for the inexact steps, an aspect left open by the theoretical analysis.

Load-bearing premise

The indefinite least squares problems belong to the special class for which the stationary iterative methods are proven to converge.

What would settle it

Take any small concrete instance satisfying the stated convergence conditions, form the preconditioned matrix explicitly, and check whether any of its eigenvalues lie outside the circle of radius 1 centered at (1,0).

read the original abstract

This paper introduces inexact versions of several block-splitting preconditioners for solving the three-by-three block linear systems arising from a special class of indefinite least squares problems. We first establish the convergence conditions for the corresponding stationary iterative methods. Then, it follows that under these conditions, all eigenvalues of the preconditioned matrices are contained within a circle centered at $(1,0)$ with radius $1$. This property implies that these preconditioners are effective in accelerating the convergence of the GMRES method. Furthermore, we analyze the eigenpairs of the preconditioned matrices in detail and derive a theoretical upper bound on the number of GMRES iterations for solving the preconditioned systems. Ultimately, numerical experiments reveal the efficacy of the proposed preconditioners.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 0 minor

Summary. The paper introduces inexact versions of several block-splitting preconditioners for three-by-three block linear systems arising from a special class of indefinite least squares problems. It establishes convergence conditions for the associated stationary iterative methods, proves that all eigenvalues of the preconditioned matrices lie inside the disk centered at (1,0) with radius 1, performs a detailed eigenpair analysis, derives an explicit upper bound on GMRES iterations, and reports numerical experiments confirming practical effectiveness.

Significance. If the stated convergence conditions and eigenvalue bounds hold, the work supplies theoretically supported preconditioners that guarantee spectral containment in the unit disk around 1, thereby accelerating GMRES for these indefinite systems. The explicit iteration bound and the combination of stationary-iteration analysis with eigenpair study constitute a clear contribution to numerical linear algebra for block-structured least-squares problems.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive assessment of our manuscript on inexact block-splitting preconditioners for indefinite least squares problems and for recommending minor revision. The report correctly identifies the main contributions: convergence conditions for the stationary iterations, the eigenvalue containment result inside the unit disk centered at 1, the detailed eigenpair analysis, the explicit GMRES iteration bound, and the supporting numerical experiments.

Circularity Check

0 steps flagged

No significant circularity; standard linear-algebra implications only

full rationale

The central chain proceeds by deriving convergence conditions (spectral radius <1) for the stationary iteration matrices associated with the inexact block-splitting preconditioners, then invoking the textbook implication that this places all eigenvalues of the preconditioned operator inside the disk |λ−1|<1. This implication is a direct algebraic fact (eigenvalues of I−M^{-1}A lie inside the unit disk centered at 1 whenever ρ(M^{-1}A−I)<1) and does not reduce to any fitted parameter, self-definition, or self-citation. Subsequent eigenpair analysis and the explicit GMRES iteration bound are obtained by direct computation on the 3×3 block structure. No load-bearing step collapses to a prior result by the same authors or to an ansatz smuggled via citation. Numerical experiments supply independent empirical support.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claims rest on standard domain assumptions from numerical linear algebra regarding matrix properties and convergence of stationary iterations for block systems; no free parameters or invented entities are introduced in the abstract.

axioms (1)
  • domain assumption Convergence conditions for the stationary iterative methods hold under the problem structure
    Invoked to establish eigenvalue bounds for the preconditioned matrices

pith-pipeline@v0.9.0 · 5430 in / 1117 out tokens · 36100 ms · 2026-05-15T18:51:12.936223+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

23 extracted references · 23 canonical work pages

  1. [1]

    Axelsson, Iterative Solution Methods, Cambridge University Press, Cambridge, 1994

    O. Axelsson, Iterative Solution Methods, Cambridge University Press, Cambridge, 1994

    work page 1994

  2. [2]

    Benzi, G.H

    M. Benzi, G.H. Golub, A preconditioner for generalized saddle point problems, SIAM J. Matrix Anal. Appl. 26 (2004) 20–41

    work page 2004

  3. [3]

    Benzi, G.H

    M. Benzi, G.H. Golub, J. Liesen, Numerical solution of saddle point problems, Acta Numer. 26 (2004) 1-137

    work page 2004

  4. [4]

    Bojanczyk, N.J

    A. Bojanczyk, N.J. Higham, H. Patel, Solving the indefinite least squares problem by hyperbolic QR factorization, SIAM J. Matrix Anal. Appl. 24 (2003) 914–931

    work page 2003

  5. [5]

    Bojanczyk, Algorithms for indefinite linear least squares problems, Linear Algebra Appl

    A.W. Bojanczyk, Algorithms for indefinite linear least squares problems, Linear Algebra Appl. 623 (2021) 104–127

    work page 2021

  6. [6]

    Chandrasekaran, M

    S. Chandrasekaran, M. Gu, A.H. Sayed, A stable and efficient algorithm for the indefinite linear least-squares problem, SIAM J. Matrix Anal. Appl. 20 (1998) 354–362

    work page 1998

  7. [7]

    Golub, C.F

    G.H. Golub, C.F. Van Loan, An analysis of the total least squares problem, SIAM J. Numer. Anal. 17 (1980) 883–893

    work page 1980

  8. [8]

    Hassibi, A.H

    B. Hassibi, A.H. Sayed, T. Kailath, Recursive linear estimation in Krein spaces. I. Theory, in: Proceedings of 32nd IEEE Conference on Decision and Control, IEEE, 1993, pp. 3489–3494. 18M. Kaveh Shaldehi, D.K. Salkuyeh

    work page 1993

  9. [9]

    J. Li, K. Xin, and L. Meng, A block upper triangular splitting method for solving block three-by- three linear systems arising from the large indefinite least squares problem, Appl. Math. Comput. 505 (2025) 129546

    work page 2025

  10. [10]

    Liu, A.J

    Q.H. Liu, A.J. Liu, Block SOR methods for the solution of indefinite least squares problems, Calcolo 51 (2014) 367–379

    work page 2014

  11. [11]

    Meng, K.L

    L.S. Meng, K.L. Xin, J. Li, Variable parameter Uzawa method for solving the indefinite least squares problem, Numer. Algorithms Numer Algor 99 (2025) 1043–1061

    work page 2025

  12. [12]

    L. Meng, P. Li, K. Xin, Alternating direction implicit iterative method for solving indefinite least squares problem, Appl. Math. Lett. 171 (2025) 109677

    work page 2025

  13. [13]

    L. Meng, K. Xin, Y. Huang, New iterative method with fast convergence of the momentum for solving indefinite least squares problem, Japan J. Indust. Appl. Math. 42 (2025) 2021–2036

    work page 2025

  14. [14]

    Saad, A flexible inner-outer preconditioned GMRES algorithm, SIAM J

    Y. Saad, A flexible inner-outer preconditioned GMRES algorithm, SIAM J. Sci. Comput. 14 (1993) 461–469

    work page 1993

  15. [15]

    Saad, Iterative methods for sparse linear systems, Second Edition, SIAM, 2003

    Y. Saad, Iterative methods for sparse linear systems, Second Edition, SIAM, 2003

    work page 2003

  16. [16]

    Saad, M.H

    Y. Saad, M.H. Schultz,GMRES: A generalized minimal residual algorithm for solving nonsym- metric linear systems, SIAM J. Sci. Stat. Comput. 7 (1986) 856–869

    work page 1986

  17. [17]

    Sayed, B

    A.H. Sayed, B. Hassibi, T. Kailath, Inertia properties of indefinite quadratic forms, IEEE Signal Process. Lett. 3 (1996) 57–59

    work page 1996

  18. [18]

    Song, USSOR method for solving the indefinite least squares problem, Int

    J. Song, USSOR method for solving the indefinite least squares problem, Int. J. Comput. Math. 97 (2020) 1781–1791

    work page 2020

  19. [19]

    Van Huffel, J

    S. Van Huffel, J. Vandewalle, The Total Least Squares Problem: Computational Aspects and Analysis, SIAM, Philadelphia, 1991

    work page 1991

  20. [20]

    Xu, A backward stable hyperbolic QR factorization method for solving indefinite least squares problem, J

    H.G. Xu, A backward stable hyperbolic QR factorization method for solving indefinite least squares problem, J. Shanghai Univ. 8 (2004) 391–396

    work page 2004

  21. [21]

    K. Xin, L. Meng, Block-splitting preconditioners for indefinite least squares problem, Comput. Appl. Math. 44 (2025) 61

    work page 2025

  22. [22]

    K. Xin, L. Meng, J. Li, Y. Huang, A preconditioned shift-splitting iteration method for solving indefinite least squares problem, J. Appl. Math. Comput. 72 (2026) 49

    work page 2026

  23. [23]

    Zhang, H.Y

    Y.J. Zhang, H.Y. Li, Splitting-based randomized iterative methods for solving indefinite least squares problem, Appl. Math. Comput. 446 (2023) 127892

    work page 2023