Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

Rational minimax approximation of matrix-valued functions

T0 review · 3 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper claims that matrix-valued rational minimax approximation can be solved by maximizing a dual problem over the probability simplex, and that a checkable strong-duality condition certifies when the recovered rational function is the

desk verdict A genuine and mostly correct block extension of the d-Lawson framework, but the rational-case convergence theory does not cover the β=1 used in experiments. read the letter →

arxiv 2508.06378 v2 pith:WV4TZVS4 submitted 2025-08-08 math.NA cs.NA

Lei-Hong Zhang , Ya-Nan Zhang , Chenkun Zhang , Shanheng Han This is my paper · ORCID
classification math.NAcs.NA MSC 41A5041A2065D1590C46
keywords matrix-valuedrationalapproximationminimaxLawsoniterationLagrangiandualitystrongFrobeniusnormerrordiscretedualproblem
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tackles a discrete worst-case approximation problem: given samples of a matrix-valued function $F:\mathbb{C}\to\mathbb{C}^{s\times t}$, find a matrix rational function $R(x)=P(x)/q(x)$ with a common scalar denominator that minimizes the largest squared Frobenius error over the sample nodes. The authors show that this min-max problem can be linearized into a standard constrained minimization, and that its Lagrange dual is a maximization over the probability simplex. They prove weak duality always holds, and give a sufficient condition under which strong duality holds: if at the computed dual maximizer the recovered rational function has maximal squared error exactly equal to the dual value, then that rational function solves the original minimax problem. In the polynomial case (denominator degree zero), the condition is unnecessary and strong duality holds unconditionally. The numerical method m-d-Lawson solves the dual by iterated weighted least-squares-type updates, with convergence established for both polynomial and rational settings.

What carries the argument

The key object is the dual function $d(w)$, defined for each weight vector $w$ on the simplex as the minimum, subject to $\sum_\ell w_\ell |q(x_\ell)|^2=1$, of the weighted sum of squared Frobenius deviations. For fixed $w$, this minimum is the smallest eigenvalue of a Hermitian positive semidefinite pencil $(A_w,B_w)$, and after Vandermonde-with-Arnoldi orthogonalization its square root is the smallest singular value of a structured tall matrix. The Lawson update $w_\ell \leftarrow w_\ell \|F(x_\ell)-R(x_\ell)\|_F^\beta / \sum_i w_i \|F(x_i)-R(x_i)\|_F^\beta$ is a projected gradient ascent step for $d$, so each iteration only requires one thin SVD and one closed-form weight update. The pape

What would settle it

Take a discrete rational approximation of type $(n,d)$ with $d\ge1$ and a known exact minimax error (for instance a scalar rational function with a known equioscillation point distribution). Run m-d-Lawson to convergence and compute the duality gap (4.5), $|e(R)-d(w)|/e(R)$. If the gap stabilizes at a strictly positive value, the sufficient condition (4.1) is violated for that instance, so no global optimality certificate is obtained; collecting such instances would reveal when the paper's central guarantee actually applies.

Watch

Extended reading notes

Core claim

The central claim is that the matrix-valued rational minimax problem (1.2) can be solved through the dual maximization (2.19) over the simplex, and that m-d-Lawson is a reliable solver for that dual. The linearized primal (2.6) admits a Lagrange dual function $d(w)$ defined by a constrained minimum that, for fixed weights, reduces to the smallest eigenvalue of a Hermitian positive semidefinite pencil—equivalently, to the smallest singular value of a tall matrix built from weighted Vandermonde bases. Theorem 4.1 is the paper's hinge: if $w^*$ maximizes the dual and the rational function $R^*$ recovered from the minimizer of $d(w^*)$ satisfies $e(R^*) = \max_{x_\ell \in X} \|F(x_\ell)-R^*(x_\e

Load-bearing premise

The load-bearing premise is that, for rational approximations with denominator degree $d\ge1$, the computed dual maximizer produces a rational function whose largest squared Frobenius error exactly equals the dual value; this equality is not guaranteed in advance and is only verified after the computation through the duality gap (4.5).

Editorial extensions

If this is right

  • If the sufficient condition (4.1) holds, the computed $R^*$ is a provable global solution of (1.2), giving users a numerical certificate of optimality via the duality gap (4.5).
  • For matrix-valued polynomial minimax approximation ($d=0$), strong duality is unconditional, so m-d-Lawson with $\beta\in(0,2)$ monotonically increases $d(w)$ and converges to the global minimax polynomial.
  • Complementary slackness implies the optimal weight vector is supported only on extreme points, so non-extreme nodes can be discarded to reduce cost without changing the solution.
  • At least $\min_{i,j}(n_{ij}+2)$ extreme points must exist at the optimum, a matrix analogue of the equioscillation property that can be used as a sanity check.
  • In the rational case ($d\ge1$), the framework inherits a Ruttan-type strong-duality behavior: the condition is frequently satisfied in practice, so the scalar success of d-Lawson extends to block/vector-valued functions.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper: if strong duality holds generically for matrix-valued rational problems, m-d-Lawson would provide a certified alternative to RMSE-based fitting in multiport microwave modeling and nonlinear eigenvalue problems, where worst-case error is the operative metric.
  • Beyond the paper: the duality gap (4.5) can be used as an adaptive certification switch—run Lawson until the gap is small, then declare the fit minimax; a persistent gap would signal a type or node-set change rather than more iterations.
  • Beyond the paper: the same dual formulation likely extends to other matrix norms, such as the spectral norm, by replacing the Frobenius trace terms with the dominant singular value of the error matrix and reweighting by the corresponding singular vectors.
  • Beyond the paper: because Example 2.1 shows the infimum in (1.2) can be unattainable, any practical solver must distinguish the certified regime (gap zero) from the merely approximate regime—a distinction the paper's framework makes explicit.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper develops a duality-based framework for discrete rational minimax approximation of matrix-valued functions with a common scalar denominator, minimizing the worst-case Frobenius-norm error over sampled nodes. It reformulates the original problem (1.2) as a linearized optimization (2.6), derives the Lagrange dual problem (2.19), and proves weak duality (Theorem 2.3). Strong duality is established conditionally in Theorem 4.1 via the checkable equality d(w*) = e(R*), and unconditionally for the polynomial case d=0 in Theorem 4.2. The proposed algorithm, m-d-Lawson (Algorithm 1), generalizes Lawson's iteration to the matrix-valued setting using the dual function. Section 7 gives convergence results: a monotonicity statement for the polynomial case for beta in [0,2], and for the rational case only a local monotonicity result for sufficiently small Lawson exponent beta (Theorem 7.2). Numerical experiments compare m-d-Lawson with v-AAA-Lawson, block-AAA, RKFIT, Loewner and VF on three test problems.

Significance. If the framework is fully validated, the paper would provide a principled optimization formulation and a practical algorithm for a problem class that is usually handled by heuristics; this is a useful extension of the authors' scalar d-Lawson work to matrix-valued functions with a common denominator. The block treatment in Sections 2 and 4 is algebraically coherent, and the sufficient-condition/checkable-gap perspective in Theorem 4.1 is genuinely useful. The paper also ships reproducible MATLAB code and gives extensive comparisons, which strengthens its practical value. However, the convergence guarantee for the rational case, which is the central computational claim, is considerably weaker than the narrative suggests: Theorem 7.2 does not cover the implemented beta=1 regime and does not prove convergence to a dual maximizer.

major comments (3)
  1. [Section 7.3, Theorem 7.2 and Section 6, Algorithm 1] The only convergence result for the rational case d >= 1 is local: for each fixed k, there exists beta0(k) > 0 such that d(w^(k+1)) >= d(w^(k)) for all beta in (0, beta0(k)). It establishes neither a uniform beta0, nor convergence of {d(w^(k))} to max_{w in S} d(w). Yet Algorithm 1 is implemented and evaluated with beta = 1 (Section 6, Step 5), a regime to which the theorem provides no guarantee even for a single step. Therefore the abstract and Section 1 claim that 'convergence analysis of m-d-Lawson is established' overstates the proved statement; the monotone decrease of the duality gap in Example 8.2 is empirical evidence, not a consequence of the stated theorem.
  2. [Section 7.3, Theorem 7.2 assumptions] Theorem 7.2 assumes d(w^(k)) is a simple eigenvalue of the pencil and, in the proof, that q(beta) is continuously differentiable near beta = 0. It also assumes (A1) and (A2). None of these conditions is verified or monitored in the numerical experiments of Section 8, and no experiment reports the behavior for beta other than 1. Thus the numerical protocol is not covered by the theorem's hypotheses either. The authors should either prove a theorem valid for beta = 1 under verifiable assumptions, or state explicitly that beta = 1 is a heuristic choice for rational problems and the convergence theory applies only to sufficiently small Lawson exponents.
  3. [Section 4, Theorem 4.1 and stopping criterion (6.1)] For rational d >= 1, the optimality of a computed R* rests on the a posteriori condition d(w*) = e(R*) in (4.1). The paper gives no result showing that m-d-Lawson, with either beta = 1 or the small beta of Theorem 7.2, drives the duality gap (4.5) to zero or that a sequence of iterates reaches a dual maximizer. Hence the central claim that m-d-Lawson solves the dual problem (2.19) and hence the original problem (1.2) is conditional in two layers: strong duality must hold for the instance, and the iteration must converge to a maximizer. This limitation should be stated prominently as a theorem-level restriction, not only as a numerical check.
minor comments (4)
  1. [Title and Abstract] The title spacing 'matrix-valued function s' and the abstract 'minima x approximation' should be corrected. Proofreading is needed.
  2. [Example 8.3, Section 8] The text says '401 (i.e., m = 400) equidistant sampled frequencies' but then lists m = 400; this is inconsistent. Also Example 8.1's 'interval [1,100]i' should be written as {i omega : 1 <= omega <= 100} for clarity.
  3. [Figure captions] Figure 8.1 caption says 'the up-right represents' and 'demonstrates' where 'upper-right' and 'shows' are intended. Please also check that Figure 8.5's subfigure labels are legible in print.
  4. [Throughout] Several proofs are stated as 'same as [45]' or 'adapted from [43]' with details omitted. This is acceptable for a generalization, but the paper would benefit from a short summary of which algebraic steps in the block setting are new and which are direct transcriptions of the scalar arguments.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: the matrix-valued extension is independently obtained; several convergence proofs import scalar-case results from the authors' prior papers, but these are prior independent results and are not used as an unproven equivalent of the present claim.

full rationale

The paper's central derivation is not circular. The dual problem (2.19) is obtained by a straightforward Lagrange-duality calculation from the linearized minimax problem (2.6); weak duality (2.15) follows from the definition of d(w), and the strong-duality sufficient condition (4.1) is proved directly from weak duality plus the definition of the infimum. No fitted parameter is relabeled as a prediction: the algorithm maximizes the dual function and then checks the duality gap (4.5), which is an a posteriori certificate rather than an input. The proofs do rely heavily on the same authors' prior scalar results: Theorem 2.1 is a 'parallel generalization' of [43, Theorem 2.1], Theorem 2.2 is 'adopted by a slight modification' of [43, Theorem 2.4], Theorem 2.3 is an 'extension' of [45, Theorem 2.2], and Theorem 7.2 'adapts [43, Theorem 5.2]'. These are self-citations, and they are numerous, but they cite separate prior works with their own proofs, and the present paper supplies the block-structured matrix-valued argument rather than simply renaming the scalar result. The cited lemmas are not the same as the present conclusion, and the matrix-valued reformulation (common denominator, block Vandermonde, generalized eigenvalue dual function) is derived here independently of those citations. The most serious limitation is that the rational-case convergence theorem (Theorem 7.2) only guarantees monotonicity for sufficiently small β, while the implemented algorithm uses β = 1; this is a correctness/completeness gap, not a circularity. Overall, there is no step in which an equation is defined in terms of the target quantity, no fitted input is called a prediction, and no uniqueness theorem from the authors' prior work is used to force the chosen approach. Score 2 reflects the heavy but non-load-bearing self-citation.

Assumptions & free parameters 1 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new entities. Its theoretical results rest on the usual mathematical architecture of convex duality, generalized eigenvalue problems, and Lawson iteration, plus the specific domain assumptions of distinct nodes, common denominator, and the ad hoc convergence assumptions (A1)-(A2). The only hand-chosen number is the Lawson exponent beta, which is an algorithmic parameter rather than a data-fitted quantity.

free parameters (1)
  • Lawson exponent beta = 1 (chosen in experiments, Step 5 of Algorithm 1)
    A hand-chosen algorithmic parameter, not fitted to data. The rational-case convergence theorem (Theorem 7.2) only proves monotonicity for sufficiently small beta, while the reported experiments use beta = 1, creating a gap between theory and the default numerical setting.
assumptions (6)
  • domain assumption The sample nodes are distinct and m >= max_{i,j}(n_ij + d + 2), as stated in (1.4).
    Used in Theorem 2.1 to prove attainability of the linearized problem and in Theorem 2.2. If m is smaller, the problem reduces to interpolation and the theoretical results do not apply.
  • domain assumption The common-denominator representation R(x) = P(x)/q(x) is used, and q(x_l) is nonzero at all nodes for the solution; zeros are handled by cancellation (Section 2.1).
    The linearization (2.6) and the dual function (2.16) are only valid when q does not vanish at the sample nodes, or when common zeros are canceled.
  • domain assumption For Theorem 2.2, the sufficient condition 0 <= d <= min_{i,j} n_ij is assumed.
    This condition is needed to construct the interpolation polynomials that make the perturbation argument work. The paper notes it covers the polynomial case d=0 and the diagonal rational case (d,d).
  • ad hoc to paper Convergence analysis in Section 7 assumes (A1) q^(k)(x_l) != 0 for all l,k and (A2) |J_w^(k)| >= max{d+1, max n_ij+1}.
    These assumptions are introduced specifically for the m-d-Lawson convergence proof (Lemma 7.1 and Theorem 7.2) and are not proven to hold for all problem instances; they are generic but unverified.
  • standard math The trace-minimization principle for Hermitian positive semidefinite pencils [25] is used to justify that the dual function value is the smallest generalized eigenvalue.
    Invoked in Theorem 2.3 and Proposition 3.1. Standard result in matrix pencil theory.
  • standard math Slater's condition holds for the polynomial case (Theorem 4.2).
    The convex program (4.4) has a strictly feasible interior point (take eta large enough), so Slater's condition applies and strong duality holds.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Rational minimax approximation of matrix-valued functions." pith.science (2026). https://pith.science/paper/WV4TZVS4

@misc{pith2026250806378,
  author       = {Pith},
  title        = {Pith review of: Rational minimax approximation of matrix-valued functions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WV4TZVS4}},
  note         = {Machine review of arXiv:2508.06378}
}
abstract

In this paper, we present a rigorous framework for rational minimax approximation of matrix-valued functions that generalizes classical scalar approximation theory. Given sampled data $\{(x_\ell, {F}(x_\ell))\}_{\ell=1}^m$ where ${F}:\mathbb{C} \to \mathbb{C}^{s \times t}$ is a matrix-valued function, we study the problem of finding a matrix-valued rational approximant ${R}(x) = {P}(x)/q(x)$ (with ${P}:\mathbb{C} \to \mathbb{C}^{s \times t}$ a matrix-valued polynomial and $q(x)$ a nonzero scalar polynomial of prescribed degrees) that minimizes the worst-case Frobenius norm error over the given nodes: $$ \inf_{{R}(x) = {P}(x)/q(x)} \max_{1 \leq \ell \leq m} \|{F}(x_\ell) - {R}(x_\ell)\|_{\rm F}. $$ By reformulating this min-max optimization problem through Lagrangian duality, we derive a maximization dual problem over the probability simplex. We analyze weak and strong duality properties and establish a sufficient condition ensuring that the solution of the dual problem yields the minimax approximant $R(x)$. For numerical implementation, we propose an efficient method (\textsf{m-d-Lawson}) to solve the dual problem, generalizing Lawson's iteration to matrix-valued functions. Convergence analysis of \textsf{m-d-Lawson} is established. Numerical experiments are conducted and compared to state-of-the-art approaches, demonstrating its efficiency as a novel computational framework for matrix-valued rational approximation.

Figures

Figures reproduced from arXiv: 2508.06378 by the authors.

Figure 8.1
Figure 8.1. The errors of the approximation R(x) with type (6,6) computed by m-d-Lawson(20) for Example 8.2. The left two columns show the absolute errors eij (xℓ) = |fij (xℓ) − rij (xℓ)| of rij (x), the up-right represents the F-norm error kF(xℓ) − R(xℓ)kF at each xℓ and the bottom-right demonstrates that the duality gap p e(R(k)) − p d(w(k)) decreases as k increases. 0 2 4 6 8 10 0 0.2 0.4 0.6 0.8 1 10-5 m-d-Lawson(20) 0 2 4 … view at source ↗
Figure 8.2
Figure 8.2. The F-norm errors kF(xℓ) − R(xℓ)kF of the approximation R(x) with type (6,6) computed by different methods for Example 8.2 [PITH_FULL_IMAGE:figures/full_fig_p036_8_2.png] view at source ↗
Figure 8.3
Figure 8.3. The errors of the polynomial approximation P(x) with type (12,0) computed by m-d-Lawson(20) for Example 8.2. The left two columns show the absolute errors eij (xℓ) = |fij (xℓ)− pij (xℓ)| of pij (x), the up-right represents the F-norm error kF(xℓ) − P(xℓ)kF at each xℓ, and the bottom-right demonstrates that the duality gap p e(P(k)) − p d(w(k)) decreases as k increases. produces more accurate approximation than v-AAA… view at source ↗
Figures from the paper (2 more)
Figure 8.4
Figure 8.4. Figure 8.4: RMSE (left column) and p e(R) (right column) from v-AAA-Lawson(20) and m-d￾Lawson(20) based on rational approximation types (nij , 20) and (nij + 2, nij ) with nij = 2, . . . , 20 for Example 8.2. 37 [PITH_FULL_IMAGE:figures/full_fig_p037_8_4.png]
Figure 8.5
Figure 8.5. Figure 8.5: Synthesized yj (x) (thick line in the right column) for j = 1, 2, 3 and fitted responses (thin dashed line in the right column) from m-d-Lawson(10) (top) and v-AAA-Lawson(10) (bottom) with noise level ς = 10−8 . The left two columns plot the roots (solid points) give…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Rational Minimax Approximations for Matrix-Valued Functions: Existence, Optimality and Algorithms

    math.OC 2026-06 conditional novelty 6.0 of 10

    Matrix-valued rational minimax approximants with a common denominator are proved to exist on dense point sets, with Kolmogorov/Ruttan optimality certificates and an equivalence to the m-d-Lawson dual conditions.

Reference graph

Works this paper leans on

46 extracted references · 21 canonical work pages · cited by 1 Pith paper

  1. [45]

    Zhang, L

    L.-H. Zhang, L. Yang, W. H. Yang and Y.-N. Zhang, A convex dua l problem for the rational minimax approximation and Lawson’s iteration, Math. Comp. , 94 (2025), 2457–2494, DOI: https://doi.org/10.1090/mcom/4021

    doi:10.1090/mcom/4021 2025

  2. [43]

    Zhang and S

    L.-H. Zhang and S. Han, A convergence analysis of Lawson’s iter ation for computing polynomial and rational minimax approximations, SIAM J. Numer. Anal. , URL https: //arxiv.org/abs/2401.00778v3, to appear

    arXiv

  3. [1]

    A. C. Antoulas, S. Lefteriu and A. C. Ionita, Chapter 8: A Tutorial Introduction to the Loewner Framework for Model Reduction , 335–376, SIAM Publications, 2017, URL https: //epubs.siam.org/doi/abs/10.1137/1.9781611974829.ch8

    doi:10.1137/1.9781611974829.ch8 2017

  4. [2]

    Barrodale and J

    I. Barrodale and J. Mason, Two simple algorithms for discrete rat ional approximation, Math. Comp., 24 (1970), 877–891

    work page 1970

  5. [3]

    Berljafa and S

    M. Berljafa and S. G¨ uttel, Generalized rational Krylov decompo sitions with an application to rational approximation, SIAM J. Matrix Anal. Appl. , 36 (2015), 894–916, URL https: //doi.org/10.1137/140998081

    doi:10.1137/140998081 2015

  6. [4]

    Berljafa and S

    M. Berljafa and S. G¨ uttel, The RKFIT algorithm for nonlinear rat ional approximation, SIAM J. Sci. Comput. , 39 (2017), A2049–A2071, URL https://doi.org/10.1137/15M1025426

    doi:10.1137/15m1025426 2017

  7. [5]

    Boyd and L

    S. Boyd and L. Vandenberghe, Convex Optimization , Cambridge University Press, 2004. 39 -0.2 -0.15 -0.1 -0.05 0 -1 -0.5 0 0.5 1 original and computed zeros of pD computed original -0.1 -0.05 0 0.05 0.1 -1 -0.5 0 0.5 1 original and computed zeros of pN computed original -0.1 -0.08 -0.06 -0.04 -0.02 0 0.02 -1 -0.8 -0.6 -0.4 -0.2 0 original and computed zero...

    work page 2004

  8. [6]

    P. D. Brubeck, Y. Nakatsukasa and L. N. Trefethen, Vanderm onde with Arnoldi, SIAM Rev., 63 (2021), 405–415

    work page 2021

Show all 46 references
  1. [7]

    A. K. Cline, Rate of convergence of Lawson’s algorithm, Math. Comp. , 26 (1972), 167–176

    1972

  2. [8]

    Deschrijver, M

    D. Deschrijver, M. Mrozowski, T. Dhaene and D. de Zutter, Mac romodeling of multiport systems using a fast implementation of the vector fitting method, IEEE Microw. Wireless Compon. Lett., 18 (2008), 383–385. 40

    2008

  3. [9]

    Drmaˇ c, S

    Z. Drmaˇ c, S. Gugercin and C. Beattie, Quadrature-based vector fitting fo r discretized H2 approximation, SIAM J. Sci. Comput. , 37 (2015), A625–A652, URL https://doi.org/10.1 137/140961511

    2015

  4. [10]

    Ellacott and J

    S. Ellacott and J. Williams, Linear Chebyshev approximation in the c omplex plane using Lawson’s algorithm, Math. Comp. , 30 (1976), 35–44

    1976

  5. [11]

    G. H. Elliott, The construction of Chebyshev approximations in the comple x plane, PhD thesis, Faculty of Science (Mathematics), University of London, 1978

    1978

  6. [12]

    Filip, Y

    S.-I. Filip, Y. Nakatsukasa, L. N. Trefethen and B. Beckerman n, Rational minimax approxima- tion via adaptive barycentric representations, SIAM J. Sci. Comput. , 40 (2018), A2427–A2455

    2018

  7. [13]

    G. H. Golub and C. F. Van Loan, Matrix Computations, 4th edition , Johns Hopkins University Press, Baltimore, Maryland, 2013

    2013

  8. [14]

    I. V. Gosea and S. G¨ uttel, Algorithms for the rational approx imation of matrix-valued func- tions, SIAM J. Sci. Comput. , 43 (2021), A3033–A3054, URL https://doi.org/10.1137/20 M1324727

    2021 doi

  9. [15]

    I. V. Gosea, C. Poussot-Vassal and A. C. Antoulas, Chapter 15 - Data-driven modeling and control of large-scale dynamical systems in the Loewner fram ework: Methodology and applications, in Numerical Control: Part A (eds. E. Tr´ elat and E. Zuazua), vol. 23 of Handbook of Nume...

    2022

  10. [16]

    Gustavsen, Improving the pole relocating properties of vec tor fitting, IEEE Trans

    B. Gustavsen, Improving the pole relocating properties of vec tor fitting, IEEE Trans. Power Deliv., 21 (2006), 1587–1592

    2006

  11. [17]

    Gustavsen, Matrix Fitting Toolbox, 2009, Https://www.sinte f.no/en/software/vector- fitting/downloads/matrix-fitting-toolbox/

    B. Gustavsen, Matrix Fitting Toolbox, 2009, Https://www.sinte f.no/en/software/vector- fitting/downloads/matrix-fitting-toolbox/

    2009

  12. [18]

    Gustavsen and A

    B. Gustavsen and A. Semlyen, Rational approximation of frequ ency domain responses by vector fitting, IEEE Trans. Power Deliv. , 14 (1999), 1052–1061

    1999

  13. [19]

    M. H. Gutknecht, On complex rational approximation. Part I: T he characterization problem, in Computational Aspects of Complex Analysis (H. Werneret at. ,eds.). Dordrecht : Reidel , 1983, 79–101

    1983

  14. [20]

    G¨ uttel and F

    S. G¨ uttel and F. Tisseur, The nonlinear eigenvalue problem, Acta Numer. , 26 (2017), 1–94

    2017

  15. [21]

    G¨ uttel, G

    S. G¨ uttel, G. M. Negri Porzio and F. Tisseur, Robust rational approximations of nonlinear eigenvalue problems, SIAM J. Sci. Comput. , 44 (2022), A2439–A2463, URL https://doi. org/10.1137/20M1380533

    2022 doi

  16. [22]

    N. J. Higham, G. M. N. Porzio and F. Tisseur, An updated set of nonlinear eigenvalue problems, Tech. report MIMS EPrint 2019.5, Manchester, United Kingdom, 2 019, URL http: //eprints.maths.manchester.ac.uk/, Http://eprints.maths.manchester.ac.uk/

    2019

  17. [23]

    J. M. Hokanson, Multivariate rational approximation using a sta bilized Sanathanan-Koerner iteration, 2020, URL arXiv:2009.10803v1

    2020 arXiv

  18. [24]

    C. L. Lawson, Contributions to the Theory of Linear Least Maximum Approxi mations, PhD thesis, UCLA, USA, 1961. 41

    1961

  19. [25]

    Liang, R.-C

    X. Liang, R.-C. Li and Z. Bai, Trace minimization principles for posit ive semi-definite pencils, Linear Algebra Appl. , 438 (2013), 3085–3106

    2013

  20. [26]

    Lietaert, K

    P. Lietaert, K. Meerbergen, J. P´ erez and B. Vandereycken , Automatic rational approximation and linearization of nonlinear eigenvalue problems, IMA J. Numer. Anal. , 42 (2022), 1087– 1115

    2022

  21. [27]

    A. J. Mayo and A. Antoulas, A framework for the solution of the generalized realization problem, Linear Algebra Appl. , 425 (2007), 634–662

    2007

  22. [28]

    Nakatsukasa, O

    Y. Nakatsukasa, O. S` ete and L. N. Trefethen, The AAA algor ithm for rational approximation, SIAM J. Sci. Comput. , 40 (2018), A1494–A1522

    2018

  23. [29]

    Nakatsukasa and L

    Y. Nakatsukasa and L. N. Trefethen, An algorithm for real an d complex rational minimax approximation, SIAM J. Sci. Comput. , 42 (2020), A3157–A3179

    2020

  24. [30]

    Nocedal and S

    J. Nocedal and S. Wright, Numerical Optimization , 2nd edition, Springer, New York, 2006

    2006

  25. [31]

    J. R. Rice, The approximation of functions , Addison-Wesley publishing Company, 1969

    1969

  26. [32]

    Ruttan, A characterization of best complex rational appro ximants in a fundamental case, Constr

    A. Ruttan, A characterization of best complex rational appro ximants in a fundamental case, Constr. Approx. , 1 (1985), 287–296

    1985

  27. [33]

    Y. Saad, M. El-Guide and A. Mi¸ edlar, A rational approximation method for the nonlinear eigenvalue problem, Technical report, 2020, URL https://arxiv.org/abs/1901.01188v2

    2020 arXiv

  28. [34]

    E. B. Saff and R. S. Varga, Nonuniqueness of best approximatin g complex rational functions, Bull. Amer. Math. Soc. , 83 (1977), 375–377

    1977

  29. [35]

    Thiran and M.-P

    J.-P. Thiran and M.-P. Istace, Optimality and uniqueness conditio ns in complex rational Chebyshev approximation with examples, Constr. Approx. , 9 (1993), 83–103

    1993

  30. [36]

    Traina, Cauchy Method Modeling: Applied to Computer-Aided Tuning o f Microwave Filters and Duplexers , VDM Verlag Dr

    D. Traina, Cauchy Method Modeling: Applied to Computer-Aided Tuning o f Microwave Filters and Duplexers , VDM Verlag Dr. M¨ uller, 2010

    2010

  31. [37]

    Traina, G

    D. Traina, G. Macchiarella and T. K. Sarkar, Robust formulation s of the Cauchy method suitable for microwave duplexers modeling, IEEE Trans. Microw. Theory Tech. , 55 (2007), 974–982

    2007

  32. [38]

    L. N. Trefethen, Approximation Theory and Approximation Practice, Extende d Edition, SIAM, 2019

    2019

  33. [39]

    Williams, Numerical Chebyshev approximation in the complex plan e, SIAM J

    J. Williams, Numerical Chebyshev approximation in the complex plan e, SIAM J. Numer. Anal., 9 (1972), 638–649

    1972

  34. [40]

    Williams, Characterization and computation of rational Cheby shev approximations in the complex plane, SIAM J

    J. Williams, Characterization and computation of rational Cheby shev approximations in the complex plane, SIAM J. Numer. Anal. , 16 (1979), 819–827

    1979

  35. [41]

    D. E. Wulbert, On the characterization of complex rational app roximations, Illinois J. Math. , 24 (1980), 140–155

    1980

  36. [42]

    Yang, L.-H

    L. Yang, L.-H. Zhang and Y. Zhang, The Lq-weighted dual prog ramming of the linear Chebyshev approximation and an interior-point method, Adv. Comput. Math. , 50:80 (2024). 42

    2024

  37. [44]

    Zhang, Y

    L.-H. Zhang, Y. Su and R.-C. Li, Accurate polynomial fitting and e valuation via Arnoldi, Numer. Algebr. Control. , 14 (2024), 526–546

    2024

  38. [46]

    Zhang, Y.-N

    L.-H. Zhang, Y.-N. Zhang, L. Yang and R.-W. Xiao, A AAA-type alg orithm for the mi- crowave duplexer filtering, Numer. Algorithms , Https://doi.org/10.1007/s11075-025-02058-0, to appear. 43

    doi

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.