Pith. sign in

REVIEW 5 minor 18 references

Elliptic orthogonal polynomials and OPRL

T0 review · 0 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper establishes that on a rectangular torus, orthogonal elliptic a-polynomials with positive real weight have simple interlacing zeros, and that every real-line orthogonal polynomial family lifts to the torus.

desk verdict A sound, honest extension of OPRL to the torus: the interlacing and lifting theorems hold up, with only minor presentation issues. read the letter →

arxiv 2507.19656 v1 pith:GXIAENNF submitted 2025-07-25 math.CA

Victor Alves , Andrei Martinez-Finkelshtein This is my paper · ORCID
classification math.CA MSC 42C0514H5233E05
keywords ellipticorthogonalpolynomialsa-polynomialsonthereallineWeierstrassfunctionsnon-Hermitianorthogonalityzerointerlacingmultiplerationalmodificationofweights
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

The paper introduces elliptic orthogonal a-polynomials (a-EOPs): meromorphic functions on a genus-one torus with prescribed poles at 0 and an anchor point a, orthogonal against a weight on a contour $\Gamma$. It tries to show that the classical theory of orthogonal polynomials on the real line survives on the torus in a strong form: on a rectangular lattice, with a positive real weight on either horizontal contour $\gamma_1$ or shifted contour $\gamma_2$, the a-EOP $f_n$ has only simple zeros on the contour—$n$ of them on $\gamma_1$, and $n$ or $n+1$ on $\gamma_2$ according to parity—they interlace with $f_{n+1}$, and the zeros vary analytically with $a$. It then proves the converse direction of the same analogy: every OPRL family with positive weight $w$ on $[e_3,e_2]$ can be lifted to a-EOPs on a torus, with odd-degree members governed by the rationally modified weight $r_w(x)=(x-e_3)(e_2-x)/(e_1-x)w(x)$. The combination matters because it turns genus-one orthogonality into a two-way bridge with real-line orthogonality, yielding new interlacing statements for rationally deformed, including Jacobi, weights.

What carries the argument

The load-bearing object is the elliptic a-polynomial space $L(n\cdot 0+a)$: elliptic functions whose only poles are a pole of order at most $n$ at 0 and a simple pole at the anchor $a$. The construction runs on the basis $b_0=1$, $b_1=\zeta(z)-\zeta(z-a)-\zeta(a)=-(\wp'(z)+\wp'(a))/(2(\wp(z)-\wp(a)))$, $b_{2k}=\wp(z)^k$, $b_{2k+1}=-(1/2)\wp'(z)\wp(z)^{k-1}$. Orthogonality is defined through bi-moments $\mu_{i,j}=\int_\Gamma b_i b_j W$, with existence guaranteed by Andreeief's determinantal formula (3.3); in the real-positive setting the determinants $D_k$ are positive. The zero-location arguments use Abel's theorem on the torus, equating sums of zeros and poles modulo the lattice, to force the parity and interlacing of zeros. The OPRL bridge uses the Weierstrass parametrization $z\mapsto(\wp(z),-\wp'(z)/2)$, splitting the even and odd subsequences via evenness of $W$ at $a=\omega_1$, and the decomposition lemma (5.7) expressing any a-EOP as $p_{n,1}(\wp)+b_1p_{n,2}(\wp)+(\wp'(a)/2)p_{n,3}(\wp)$, whose components satisfy the multiple orthogonality relations (5.15)–(5.17).

What would settle it

For a rectangular torus with, say, $e_1=1$, $e_2=0$, $e_3=-1$ and $a=\omega_1/2$, take $W\equiv 1$ on $\gamma_2$ and compute the determinants $D_1,D_2,D_3$ by direct quadrature; if any $D_k\le 0$, or if the degree-2 a-EOP has a repeated zero or a zero outside $\gamma_2$, Theorem 4.2 fails. For Theorem 5.1, a direct check is to integrate the constructed $F_3$ against $b_0,b_1,b_2$ on $\gamma_2$ for the Jacobi weight of Example 5.2; a nonzero value would refute the lifting construction.

Watch

Extended reading notes

Core claim

On its own terms, the central claim is a pair of theorems. Theorem 4.2 says that when the lattice is rectangular (all $e_i$ real and distinct), the contour $\Gamma$ is $\gamma_1$ or $\gamma_2$, and the anchor $a$ lies on the other contour, the monic a-EOPs $f_n$ associated to a positive real weight have only simple zeros on the orthogonality contour: $n$ of them on $\gamma_1$ for $\Gamma=\gamma_1$, and $n$ or $n+1$ on $\gamma_2$ for $\Gamma=\gamma_2$ according to parity, with interlacing $f_{n+1}<f_n$ on $\gamma_1$ and $f_n<f_{n+1}$ or the reverse on $\gamma_2$; an extra zero on the other contour occurs exactly when $f_n$ has a pole at $a$. Theorem 5.1 states the lifting: for $e_3<e_2<e_1$ with $e_1=-e_2-e_3$, every positive weight $w$ on $[e_3,e_2]$ produces, through $\wp$, a positive even weight $W$ on $\gamma_2$ for which the even subsequence $F_{2j}=P_j(\wp(z);w)$ and the odd subsequence $F_{2j+1}=-(1/2)\wp'(z)/(\wp(z)-e_1)P_j(\wp(z);r_w)$ are exactly the monic a-EOPs with $a=\omega_1$, where $r_w$ is the rational modification of $w$ in (5.2). A decomposition theorem (5.8–5.12) extends the mechanism to non-symmetric weights and anchors, expressing a-EOPs through type II multiple orthogonality conditions on the real line.

Load-bearing premise

The load-bearing premise is that the orthogonality data stay in the real-positive rectangular regime—$e_1,e_2,e_3$ real and distinct, $\Gamma=\gamma_1$ or $\gamma_2$, $a$ off $\Gamma$ with $\wp(a)\in\mathbb{R}$, and $W>0$ on $\Gamma$—so that every Andreeief determinant $D_k$ is positive; the explicit lifting formulas of Section 5.3 additionally require $P_m(\wp(a),w)$ and $P_{m-1}(\wp(a),\widetilde{w})$ to be nonzero.

Editorial extensions

If this is right

  • For any positive real weight on $\gamma_1$ or $\gamma_2$ of a rectangular torus, the a-EOPs form an interlacing sequence $f_{n+1}<f_n$ on $\gamma_1$ (and the analogous parity statement on $\gamma_2$), so the torus reproduces the whole OPRL zero picture including analytic dependence of zeros on the anchor $a$.
  • Theorem 5.1 gives an explicit bijection between a-EOPs with even weight and anchor at a half-period, and pairs of OPRL sequences for $w$ and its rational modification $r_w$; consequently $P_n(\cdot;w)$ and $P_{n-1}(\cdot;r_w)$ interlace on $[e_3,e_2]$.
  • Specializing to the square lattice with a Jacobi weight yields explicit a-EOPs whose odd terms are built from Jacobi polynomials of parameters $\alpha+1,\beta+1$, and whose even terms for the Christoffel-deformed weight are built from Christoffel–Darboux kernels; the torus interlacing then gives the interlacing chain stated in Corollary 5.5.
  • For general non-symmetric weights and anchors, each a-EOP decomposes into two real-line polynomial components that satisfy type II multiple orthogonality relations (5.15)–(5.17), so the elliptic theory reduces to a vector of real-line orthogonal polynomials with shifted rational weights.

Reading between the lines

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

  • If the interlacing theorem extends past rectangular tori, where the Andreeief determinants may be complex, one would obtain a genuine higher-genus Sturm-type oscillation theory; a numerical check of the sign variation of $D_k$ for non-rectangular lattices would test this.
  • The limit $a\to 0$, noted in the paper as excluding degree-1 functions, should degenerate the a-EOPs to the no-anchor elliptic polynomials; checking whether the interlacing of Theorem 4.2 degenerates to the zero locations of those polynomials would clarify the connection to exceptional orthogonal polynomials.
  • Corollary 5.4 suggests a general principle: rational modifications of the form $(x-e_3)(e_2-x)/(e_1-x)$ preserve interlacing for arbitrary positive weights on $[e_3,e_2]$, not just Jacobi weights; a direct proof for a generic $w$ would be a testable strengthening.
  • The analyticity of zeros in $a$, combined with the explicit Cauchy-transform formulas for the lifting constants (5.23), gives a concrete route to compute zero velocities $\partial z_k/\partial a$, which could be compared with numerical differentiation.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

0 major / 5 minor

Summary. The paper develops a genus-one analogue of orthogonal polynomials, called elliptic orthogonal a-polynomials (a-EOPs): meromorphic functions on a torus with a pole of order at most n at 0 and at most a simple pole at a fixed anchor point a, orthogonal with respect to a weight W on a contour Γ. After setting up a monic basis (2.9)–(2.10), the authors prove a determinant/Andreeief formula, a five-term recurrence under multiplication by ℘, and Christoffel–Darboux formulae (Lemmas 3.1–3.2). In the real-positive setting on a rectangular torus with Γ=γ1 or γ2, Theorem 4.2 asserts that the a-EOPs have simple real zeros on the orthogonality contour and that consecutive polynomials interlace, with analytic dependence of zeros on a. Section 5 establishes a constructive correspondence with OPRL: for even weights and a=ω1, Theorem 5.1 identifies even and odd a-EOPs with OPRL for a weight w and its rational modification rw; Section 5.2 gives a multiple-orthogonality decomposition; Section 5.3 gives an explicit general lift for a∈γ1. The paper closes with Jacobi examples and interlacing consequences for rationally modified Jacobi weights.

Significance. The central contributions are the real-positive interlacing theorem and the explicit, parameter-free OPRL-to-a-EOP lifting theorem. The proofs are built from standard tools—Andreeief's formula, Abel's theorem, and the change of variable x=℘(z)—and are internally consistent; I checked the Section 5.3 denominator concern raised by the internal reader, and it does not land, because for a∈γ1 one has ℘(a)>e1>e2 while all zeros of the relevant OPRL lie in (e3,e2). The explicit Jacobi examples give concrete interlacing statements that appear new and are easily testable. The overlap with [3,4] is explicitly acknowledged and is limited to the pole-at-0 construction, while the anchor-point framework, the interlacing results, and the lifting correspondence are genuinely new.

minor comments (5)
  1. [Definition 4.1 and Corollary 5.4] The notation 'F < in γ' in Definition 4.1 is missing its second argument, and the assertion 'F_n < F_{n+2}' in Corollary 5.4 uses an interlacing notion for functions whose zero counts differ by two, which is not covered by Definition 4.1; the definition should be extended, or a cyclic convention for contours with identified endpoints should be stated.
  2. [Section 5.1, proof of Theorem 5.1] In the displayed integral for n=2j+1 and m=2l+1, the factor 1/4 and the sign are inconsistent with the definitions of F_n, W, and (2.4); the computation should give ∫_{e3}^{e2} P_j(x;rw)P_l(x;rw)(x-e2)(x-e3)/(x-e1) w(x) dx up to an irrelevant overall sign, so the displayed factor should be corrected.
  3. [Proposition 4.5] In the real-coefficient argument following the construction of Φ, the solution vector is denoted λ=(λ_1,...,λ_m) and λ∈R^m, but the system involves the m+1 coefficients λ_0,...,λ_m; the index range should be corrected to λ∈R^{m+1}.
  4. [Proposition 4.7, case (ii)] The interlacing argument for Γ=γ2 should explicitly handle the cyclic interval between the last and first zeros of f_n on the closed contour γ2; as written, the linear-interval argument only yields zeros between consecutive zeros in the linear order, and the remaining zero is accounted for only if a cyclic convention is adopted.
  5. [Introduction and Corollary 5.4] The citation 'J. Dunham [11]' should be 'D. Jackson', matching reference [11]; in the proof of Corollary 5.4, the interval '[e2,e3]' should be '[e3,e2]'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity found: the main constructions and theorems are derived from explicit inputs and external classical results, not from their own conclusions.

full rationale

The paper's central claims do not reduce to their inputs by construction. Theorem 4.2 is proved from the positivity of the weight W, the Andréief determinant formula, Abel's theorem, and a Wronskian argument; the interlacing conclusion is not assumed in the hypotheses and is obtained by a genuine zero-counting and sign-change proof. Theorem 5.1 is a constructive lifting result: starting from an arbitrary positive OPRL weight w on [e3,e2], the paper defines an explicit elliptic weight W and verifies orthogonality of the proposed functions F_n by direct change-of-variable integrals using the standard orthogonality of P_j(·;w) and P_j(·;r̃w). No parameter is fitted to the quantity being predicted, and no 'prediction' is statistically forced by a prior fit. The paper also contains no load-bearing self-citation: references [1]–[4] are external works by other authors, and the acknowledged overlap with [3,4] concerns similar constructions rather than a circular dependence on them. The Section 5.3 formulas do involve denominators P_m(℘(a);w) and P_{m-1}(℘(a);p̃w), and the paper does not discuss possible degeneracies, but this is an omitted technical hypothesis or gap, not a circular step: the construction would fail rather than silently assume its conclusion. The internal consistency of Theorems 4.2 and 5.1, together with the absence of fitted inputs and self-citation chains, supports a non-circularity score of 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No data fitting or empirical constants are involved. The only antecedents are standard elliptic function theory and explicit assumptions on the weight and contour; the a-EOP objects are definitions, not empirical entities.

assumptions (4)
  • standard math Weierstrass elliptic function identities, including (2.4), (2.7), and Abel's theorem on the torus.
    Used throughout Sections 2 and 4 to parametrize the torus, compute with b_j, and count or interlace zeros via pole-zero balance.
  • domain assumption All principal minors D_k = det(mu_{i,j})_{i,j=0}^{k-1} are nonzero, so the full sequence of orthogonal a-polynomials exists.
    Assumed in Section 3 before Lemma 3.1 for the recurrence and Christoffel-Darboux formula; in the positive real setting it follows from Andreeief's formula.
  • domain assumption The lattice is rectangular with real invariants, discriminant positive, the weight is positive and real on Gamma, and a is not on Gamma.
    Assumed in Section 4, eqs. (4.1)-(4.2), to make the basis b_j real on the contours, powering the zero theorems.
  • domain assumption w is a positive finite-moment weight on [e3,e2] for the OPRL lifting.
    Used in Theorem 5.1 and Section 5.3 to define the even torus weight W and the rational modification rw.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Elliptic orthogonal polynomials and OPRL." pith.science (2026). https://pith.science/paper/GXIAENNF

@misc{pith2026250719656,
  author       = {Pith},
  title        = {Pith review of: Elliptic orthogonal polynomials and OPRL},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GXIAENNF}},
  note         = {Machine review of arXiv:2507.19656}
}
abstract

We explore a class of meromorphic functions on elliptic curves, termed \emph{elliptic orthogonal a-polynomials} ($a$-EOPs), which extend the classical notion of orthogonal polynomials to compact Riemann surfaces of genus one. Building on Bertola's construction of orthogonal sections, we study these functions via non-Hermitian orthogonality on the torus, establish their recurrence properties, and derive an analogue of the Christoffel--Darboux formula. We demonstrate that, under real-valued orthogonality conditions, $a$-EOPs exhibit interlacing and simplicity of zeros similar to orthogonal polynomials on the real line (OPRL). Furthermore, we construct a general correspondence between families of OPRL and elliptic orthogonal functions, including a decomposition into multiple orthogonality relations, and identify new interlacing phenomena induced by rational deformations of the orthogonality weight.

Figures

Figures reproduced from arXiv: 2507.19656 by the authors.

Figure 1
Figure 1. Contours γ1 and γ2 on Ω. One of the main features of orthogonal polynomials on the real line (OPRL) is the property of their zeros: they are real, simple, and interlace with polynomials of the lower degree. In the rest of this section, we discuss the corresponding analogues. When discussing the interlacing of zeros of elliptic functions (i.e., meromorphic functions on T ), we restrict our attention to the zeros in t… view at source ↗
Figure 2
Figure 2. Plots of ℘pzq P R for 2ω1 “ 1 and 2ω3 “ 3i{2 on the contours γ1 (left) and γ2 (right). Here e1 « 6.57974, e2 « ´3.29624, and e3 « ´3.2835. In a similar fashion, ℘ptω1q is strictly decreasing on t P p0, 1q, taking all values in the interval pe1, `8q. By symmetry, ℘ptω1q is strictly increasing on t P p1, 2q (see [PITH_FULL_IMAGE:figures/full_fig_p014_2.png] view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

18 extracted references · 18 canonical work pages

  1. [1]

    M. Bertola. Pad´ e approximants on Riemann surfaces and KP tau functions.Anal. Math. Phys., 11(4):Paper No. 149, 38, 2021

    work page 2021

  2. [2]

    M. Bertola. Nonlinear steepest descent approach to orthogonality on elliptic curves.J. Approx. Theory, 276:Paper No. 105717, 33, 2022

    work page 2022

  3. [3]

    Recurrence relations and the Christoffel-Darboux formula for elliptic orthogonal polynomials

    H. Desiraju and S. Lahiry. Recurrence relations and the Christoffel-Darboux formula for elliptic orthogonal polynomials. Preprint arXiv:2506.09582, 2025

    work page Pith review arXiv 2025

  4. [4]

    Desiraju, T

    H. Desiraju, T. L. Latimer, and P. Roffelsen. On a class of elliptic orthogonal polynomials and their integrability. Constructive Approximation, 2024. 10.1007/s00365-024-09687-z

    work page doi:10.1007/s00365-024-09687-z 2024

  5. [5]

    D. K. Dimitrov, M. E. H. Ismail, and F. R. Rafaeli. Interlacing of zeros of orthogonal polynomials under modification of the measure. J. Approx. Theory, 175:64–76, 2013

    work page 2013

  6. [6]

    https://dlmf.nist.gov/, Release 1.2.4 of 2025-03-15

    DLMF: NIST Digital Library of Mathematical Functions. https://dlmf.nist.gov/, Release 1.2.4 of 2025-03-15. F. W. J. Olver, A. B. Olde Daalhuis, D. W. Lozier, B. I. Schneider, R. F. Boisvert, C. W. Clark, B. R. Miller, B. V. Saunders, H. S. Cohl, and M. A. McClain, eds

    work page 2025

  7. [7]

    Driver, K

    K. Driver, K. Jordaan, and N. Mbuyi. Interlacing of the zeros of Jacobi polynomials with different parameters. Numer. Algorithms, 49(1-4):143–152, 2008

    work page 2008

  8. [8]

    Fasondini, S

    M. Fasondini, S. Olver, and Y. Xu. Orthogonal polynomials on a class of planar algebraic curves. Stud. Appl. Math., 151(1):369–405, 2023

    work page 2023

Show all 18 references
  1. [9]

    Fasondini, S

    M. Fasondini, S. Olver, and Y. Xu. Orthogonal polynomials on planar cubic curves. Found. Comput. Math. , 23(1):1–31, 2023

    2023

  2. [10]

    M. E. H. Ismail. Classical and quantum orthogonal polynomials in one variable , volume 98 of Encyclopedia of Mathematics and its Applications . Cambridge University Press, Cambridge, 2009. 26 V. ALVES AND A. MART ´INEZ-FINKELSHTEIN

    2009

  3. [11]

    D. Jackson. Orthogonal polynomials on a plane curve. Duke Math. J. , 3(2):228–236, 1937

    1937

  4. [12]

    A. I. Markushevich. Theory of functions of a complex variable. Vol. III . Prentice-Hall, Inc., Englewood Cliffs, NJ, english edition, 1967

    1967

  5. [13]

    R. Miranda. Algebraic curves and Riemann surfaces , volume 5 of Graduate Studies in Mathematics . American Mathematical Society, Providence, RI, 1995

    1995

  6. [14]

    Olver and Y

    S. Olver and Y. Xu. Orthogonal structure on a wedge and on the boundary of a square. Found. Comput. Math., 19(3):561–589, 2019

    2019

  7. [15]

    Olver and Y

    S. Olver and Y. Xu. Orthogonal polynomials in and on a quadratic surface of revolution. Math. Comp. , 89(326):2847–2865, 2020

    2020

  8. [16]

    Olver and Y

    S. Olver and Y. Xu. Orthogonal structure on a quadratic curve. IMA J. Numer. Anal. , 41(1):206–246, 2021

    2021

  9. [17]

    P. E. Spicer, F. W. Nijhoff, and P. H. van der Kamp. Higher analogues of the discrete-time Toda equation and the quotient-difference algorithm. Nonlinearity, 24(8):2229–2263, 2011

    2011

  10. [18]

    Szeg˝ o.Orthogonal Polynomials, volume Vol

    G. Szeg˝ o.Orthogonal Polynomials, volume Vol. 23 of American Mathematical Society Colloquium Publications . American Mathematical Society, New York, 1939. (VA) Instituto de Ciˆencias Matem´aticas e de Computac ¸˜ao (ICMC), Universidade de S ˜ao Paulo (USP), S˜ao Paulo, Brazil...

    1939

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.