pith. machine review for the scientific record. sign in

arxiv: 2605.13124 · v1 · submitted 2026-05-13 · 🧮 math.NA · cs.NA

Recognition: unknown

Polynomial interpolation--regression on the sphere

Francesco Dell'Accio , Federico Nudo , Teresa E. P\'erez , Miguel A. Pi\~nar
Authors on Pith no claims yet

Pith reviewed 2026-05-14 18:42 UTC · model grok-4.3

classification 🧮 math.NA cs.NA MSC 65D05
keywords spherical polynomialsinterpolationleast squaresspherical harmonicsantipodal symmetryspherical designsVandermonde matricesKKT conditions
0
0 comments X

The pith

A spherical polynomial of degree r interpolates data at a chosen subset of nodes while minimizing the least-squares residual at the remaining nodes.

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

The paper defines a hybrid approximation method on the unit sphere that builds a polynomial of fixed degree r from scattered data values. Part of the data is matched exactly at designated interpolation nodes, while the rest of the nodes are used only to reduce the total squared error. When the associated Vandermonde matrices satisfy basic rank conditions, the resulting polynomial is the unique solution to this mixed problem and satisfies a discrete orthogonality relation between the residual and the full polynomial space. The same construction separates into independent even and odd pieces when the node set is antipodally symmetric, and the linear system becomes especially simple when the nodes form a spherical design.

Core claim

The interpolation-regression operator produces a unique spherical polynomial of degree r that interpolates the given data on a prescribed subset of nodes and minimizes the discrete least-squares residual on the remaining nodes. This approximant is characterized by the condition that the residual is orthogonal to the space of spherical polynomials of degree at most r with respect to the discrete inner product induced by the full sampling set.

What carries the argument

The interpolation-regression operator obtained by solving the constrained least-squares problem whose solution satisfies both exact interpolation at selected nodes and orthogonality of the residual to the polynomial space on the full node set.

If this is right

  • When the rank conditions hold, existence and uniqueness of the approximant follow directly from the KKT system.
  • For antipodally symmetric nodes the even and odd spherical-harmonic components decouple and can be solved independently.
  • The operator is equivariant under any orthogonal transformation that maps both the interpolation and sampling sets to themselves.
  • When the nodes form a spherical design the normal matrix is a scalar multiple of the identity, so the spectral condition number of the KKT matrix is known in closed form.

Where Pith is reading between the lines

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

  • The same mixed interpolation-regression idea could be tested on other compact manifolds where discrete polynomial spaces are well understood.
  • Choosing spherical-design nodes may yield more stable linear systems in floating-point arithmetic than generic point sets.
  • The explicit decoupling for antipodal data suggests similar block-diagonal structures might appear under other discrete symmetry groups.

Load-bearing premise

The Vandermonde matrices associated with the interpolation subset and the full sampling set must have full column rank.

What would settle it

A concrete node configuration in which the interpolation Vandermonde matrix has deficient rank, together with data values for which two distinct degree-r polynomials both interpolate the subset and achieve the same minimal residual on the remainder.

read the original abstract

We introduce an interpolation--regression operator for polynomial approximation on the unit sphere $\mathbb{S}^2$ from discrete samples. The approximant is a spherical polynomial of degree $r$ which interpolates the data on a prescribed subset of nodes and uses the remaining sampling nodes to minimize the residual in a least squares sense. Under natural rank assumptions on the associated Vandermonde matrices, the approximant is unique and is characterized by an orthogonality condition with respect to the discrete inner product on the sampling set. We then focus on the case in which the sampling and interpolation nodes are antipodally symmetric. In this setting, when the polynomial is expressed in real spherical harmonics, the constrained problem can be decomposed into independent even and odd components. In the same framework, we prove equivariance under the antipodal map and, more generally, under orthogonal transformations preserving the node sets. We also consider spherical designs. In this case, the normal matrix is a scalar matrix. Consequently, the spectral condition number of the associated KKT matrix can be written explicitly. Numerical experiments in both antipodal and non-antipodal settings illustrate the effectiveness of the proposed method.

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

2 major / 2 minor

Summary. The manuscript introduces a hybrid interpolation-regression operator for approximating functions on the unit sphere S^2 by spherical polynomials of fixed degree r. The approximant interpolates the data exactly on a prescribed subset of nodes while minimizing the discrete least-squares residual on the remaining nodes. Under rank assumptions on the associated Vandermonde matrices, the approximant is shown to be unique and to satisfy an orthogonality condition with respect to the discrete inner product induced by the full sampling set. The analysis specializes to antipodally symmetric node sets, where the problem decouples into independent even and odd components when expressed in the real spherical-harmonic basis; equivariance under the antipodal map and under orthogonal transformations preserving the node sets is established. For spherical designs the normal matrix reduces to a scalar multiple of the identity, yielding an explicit formula for the spectral condition number of the KKT matrix. Numerical experiments in both antipodal and non-antipodal regimes are presented to illustrate practical behavior.

Significance. If the rank hypotheses hold for the node configurations considered, the construction supplies a flexible, symmetry-aware polynomial approximation scheme on the sphere that blends exact interpolation with regression-based stability. The even-odd decomposition and the explicit conditioning result for designs constitute concrete theoretical advantages that could facilitate efficient, structure-preserving implementations in applications such as geophysical modeling or computer graphics.

major comments (2)
  1. [Main uniqueness theorem and abstract] The uniqueness and orthogonality characterization (stated in the abstract and developed in the main theoretical section) rest entirely on 'natural rank assumptions' on the Vandermonde matrices associated with the sampling and interpolation nodes. No general criterion guaranteeing full rank for arbitrary node sets on S^2 is supplied, nor are explicit rank computations or proofs furnished for the concrete antipodally symmetric and spherical-design configurations used throughout the analysis and experiments. This renders the central claim conditional on an unverified hypothesis in the geometric settings of interest.
  2. [Theoretical development and numerical experiments] The manuscript provides no a priori error analysis, stability estimates, or convergence rates for the interpolation-regression approximant as the degree r or the number of nodes increases. Such quantitative bounds are load-bearing for any numerical-analysis claim about the method's effectiveness and are absent from both the theoretical development and the numerical section.
minor comments (2)
  1. [Abstract and introduction] The phrase 'natural rank assumptions' appears repeatedly without a concise definition or forward reference to its precise statement; a brief parenthetical clarification in the abstract and introduction would improve readability.
  2. [Notation and linear-algebraic sections] Notation for the full sampling set, the interpolation subset, and the associated Vandermonde matrices should be introduced once and used consistently; occasional shifts between matrix and operator notation obscure the linear-algebraic arguments.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the careful reading and constructive feedback on our manuscript. We address the two major comments point by point below, indicating the revisions we will make where appropriate.

read point-by-point responses

  1. Referee: [Main uniqueness theorem and abstract] The uniqueness and orthogonality characterization (stated in the abstract and developed in the main theoretical section) rest entirely on 'natural rank assumptions' on the Vandermonde matrices associated with the sampling and interpolation nodes. No general criterion guaranteeing full rank for arbitrary node sets on S^2 is supplied, nor are explicit rank computations or proofs furnished for the concrete antipodally symmetric and spherical-design configurations used throughout the analysis and experiments. This renders the central claim conditional on an unverified hypothesis in the geometric settings of interest.

    Authors: We agree that the uniqueness result is stated under rank assumptions on the Vandermonde matrices. A general criterion valid for arbitrary node sets on S^2 lies outside the paper's scope, as it would require results from algebraic geometry or potential theory. For the specific configurations analyzed (antipodally symmetric sets and spherical designs), the assumptions hold because the real spherical harmonics up to degree r remain linearly independent and spherical designs of sufficient strength produce full-rank Gram matrices. We will add a clarifying remark after the main theorem, supported by references to known properties of spherical designs, and note that the rank was verified numerically for the node sets in the experiments. revision: partial

  2. Referee: [Theoretical development and numerical experiments] The manuscript provides no a priori error analysis, stability estimates, or convergence rates for the interpolation-regression approximant as the degree r or the number of nodes increases. Such quantitative bounds are load-bearing for any numerical-analysis claim about the method's effectiveness and are absent from both the theoretical development and the numerical section.

    Authors: We acknowledge that the manuscript does not contain a priori error bounds or convergence rates. Deriving such estimates for the hybrid operator would require additional tools (e.g., sphere-specific Marcinkiewicz-Zygmund inequalities or bounds on the associated Lebesgue constants) that go beyond the present focus on well-posedness, symmetry, and equivariance. The numerical experiments illustrate practical behavior, but we will expand the conclusions to discuss the observed convergence trends and explicitly state that a full theoretical error analysis remains future work. revision: partial

standing simulated objections not resolved
  • Deriving rigorous a priori error analysis, stability estimates, and convergence rates for the approximant

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper's core derivation defines an interpolation-regression operator via a constrained least-squares problem on spherical polynomials, then states uniqueness under explicit rank assumptions on the associated Vandermonde matrices and derives the orthogonality characterization directly from the KKT conditions of that problem. These steps are standard linear-algebra consequences of the setup and do not reduce any claimed result to a fitted input, self-definition, or self-citation chain. The antipodal decomposition and spherical-design normal-matrix simplification follow immediately from symmetry properties of the node sets without importing unverified ansatzes or renaming prior results. No load-bearing step equates a prediction to its own input by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard domain assumptions about polynomial spaces and node sets; no free parameters or invented entities are introduced.

axioms (1)
  • domain assumption Natural rank assumptions on the associated Vandermonde matrices
    Invoked to guarantee uniqueness of the approximant and the orthogonality characterization.

pith-pipeline@v0.9.0 · 5508 in / 1118 out tokens · 48427 ms · 2026-05-14T18:42:24.172673+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

21 extracted references · 1 canonical work pages

  1. [1]

    Dover Publications, New York (1972)

    Abramowitz, M., Stegun, I.A.: Handbook of Mathematical Functions with Formulas, Graphs, and Mathemat- ical Tables. Dover Publications, New York (1972)

    1972

  2. [2]

    Discrete Comput

    Aistleitner, C., Brauchart, J.S., Dick, J.: Point sets on the sphere with small spherical cap discrepancy. Discrete Comput. Geom.48, 990–1024 (2012)

    2012

  3. [3]

    Geometriae Dedicata43, 167–179 (1992)

    Bajnok, B.: Construction of sphericalt-designs. Geometriae Dedicata43, 167–179 (1992)

    1992

  4. [4]

    Bruni Bruno, L., Dell’Accio, F., Erb, W., Nudo, F.: Polynomial histopolation on mock-Chebyshev segments. J. Sci. Comput.104(2), 65 (2025)

    2025

  5. [5]

    ACM Trans

    Buss, S.R., Fillmore, J.P.: Spherical averages and applications to spherical splines and interpolation. ACM Trans. Graph.20, 95–126 (2001)

    2001

  6. [6]

    arXiv preprint arXiv:2601.18721 (2026)

    Cruz-Barroso, R., Fern´ andez, L., Marcell´ an, F.: A mixed interpolation-regression method for numerical in- tegration on the unit circle using zeros of para-orthogonal polynomials. arXiv preprint arXiv:2601.18721 (2026)

    work page arXiv 2026

  7. [7]

    Dolomites Res

    De Marchi, S., Dell’Accio, F., Nudo, F.: A mixed interpolation-regression approximation operator on the triangle. Dolomites Res. Notes Approx.17, 33–44 (2024)

    2024

  8. [8]

    De Marchi, S., Dell’Accio, F., Mazza, M.: On the constrained mock-Chebyshev least-squares. J. Comput. Appl. Math.280, 94–109 (2015)

    2015

  9. [9]

    Dell’Accio, F., Di Tommaso, F.: Interpolation of scattered data on the sphere by multinode Shepard operators. J. Sci. Comput.104, 1–21 (2025)

    2025

  10. [10]

    Dell’Accio, F., Di Tommaso, F., Nudo, F.: Generalizations of the constrained mock-Chebyshev least squares in two variables: Tensor product vs total degree polynomial interpolation. Appl. Math. Lett.125, 107732 (2022)

    2022

  11. [11]

    Dell’Accio, F., Marcell´ an, F., Nudo, F.: An interpolation–regression approach for function approximation on the disk and its application to cubature formulas. Adv. Comput. Math.51, 53 (2025)

    2025

  12. [12]

    Oxford University Press (1998)

    Freeden, W., Gervens, T., Schreiner, M.: Constructive approximation on the sphere: with applications to geomathematics. Oxford University Press (1998)

    1998

  13. [13]

    Golitschek, M., Light, W.: Interpolation by polynomials and radial basis functions on spheres. Constr. Approx. 17, 1–18 (2001)

    2001

  14. [14]

    Hannay, J., Nye, J.: Fibonacci numerical integration on a sphere. J. Phys. A: Math. Gen.37, 11591 (2004)

    2004

  15. [15]

    ACM Trans

    Keinert, B., Innmann, M., S¨ anger, M., Stamminger, M.: Spherical Fibonacci mapping. ACM Trans. Graph. 34, 1–7 (2015)

    2015

  16. [16]

    Makar, P., Karpik, S.: Basis-spline interpolation on the sphere: Applications to semi-Lagrangian advection. Mon. Weather Rev.124, 182 (1996)

    1996

  17. [17]

    Sloan, I.H., Womersley, R.S.: Constructive polynomial approximation on the sphere. J. Approx. Theory.103, 91–118 (2000)

    2000

  18. [18]

    Stoller, M.: Fourier interpolation from spheres. Trans. Am. Math. Soc.374(11), 8045–8079 (2021)

    2021

  19. [19]

    Xu, Y.: Polynomial interpolation on the unit sphere. SIAM J. Numer. Anal.41, 751–766 (2003)

    2003

  20. [20]

    Xu, Y.: Polynomial interpolation on the unit sphere and on the unit ball. Adv. Comput. Math.20, 247–260 (2004)

    2004

  21. [21]

    Zhou, Y., Chen, X.: Sphericalt ϵ-designs for approximations on the sphere. Math. Comput.87(314), 2831–2855 (2018)

    2018