🔢 mathematics

A Rodrigues Formula for Multiple Orthogonal Polynomials on the Simplex

This paper introduces a Rodrigues formula-based generalization of Jacobi–Piñeiro polynomials to the bivariate setting on the simplex and applies these new multiple orthogonal polynomials to solve the bivariate Hermite–Padé problem on the triangle.

Original authors: Lidia Fernández, Ana Foulquié-Moreno, Juan Antonio Villegas

Published 2026-08-05

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Lidia Fernández, Ana Foulquié-Moreno, Juan Antonio Villegas

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are a master baker trying to create the perfect cake. In the world of mathematics, "baking" often involves creating special shapes called polynomials. These aren't just random shapes; they are carefully crafted to fit specific rules, much like a key fitting into a lock.

For a long time, mathematicians have known how to bake these "keys" for simple, one-dimensional problems (like a straight line). They have a secret recipe called the Rodrigues formula. Think of this formula as a magical mixing bowl: you put in a specific weight (a flavor profile), stir it with a special tool (differentiation), and out pops a perfect polynomial key.

This paper is about taking that single-dimensional recipe and expanding it into a 3D world (specifically, a triangular slice of cake) and dealing with multiple flavor profiles at once.

Here is a breakdown of what the authors did, using simple analogies:

1. The Setting: The Triangular Cake

Most math problems happen on a straight line. But this paper happens on a triangle (specifically, a triangle where $x \ge 0$ , $y \ge 0$ , and $x+y \le 1$ ). Imagine a slice of pizza or a triangular piece of land. The authors are working with "weights" on this land. A weight is like a density map—some parts of the triangle are "heavier" or more important than others.

2. The Challenge: Multiple Flavors at Once

Usually, a polynomial is designed to be orthogonal (perfectly balanced) against just one weight.

The Old Way: You have one flavor (say, vanilla). You make a polynomial that balances perfectly against vanilla.
The New Way (Multiple Orthogonality): Imagine you have two different flavors (say, vanilla and chocolate) spread across the same triangle. You want to bake a single polynomial that balances perfectly against both flavors simultaneously.

This is tricky. In the old 1D world, there's usually only one unique way to do this. But in this 2D triangle world, there are many ways to arrange the ingredients. The authors had to figure out a specific recipe to get a result that works for both.

3. The Solution: The "Double-Action" Mixer

The authors created a new version of the Rodrigues formula (the magical mixing bowl).

In the old 1D world, you use one mixing tool.
In this new 2D world, they use two mixing tools (operators) in a row.

Think of it like this:

Tool 1 mixes the dough based on the "Vanilla" rules.
Tool 2 takes that result and mixes it again based on the "Chocolate" rules.

The authors proved that if you use these two tools in a specific order (and surprisingly, the order doesn't matter!), you get a special polynomial. This polynomial is a "Jack-of-all-trades" that satisfies the balance requirements for both flavors on the triangular land. They call these Jacobi-Piñeiro Multiple Orthogonal Polynomials.

4. The Application: The "Rational Approximation" Trick

Why do we care about these special cakes? The paper connects them to a problem called Hermite-Padé approximation.

Imagine you have a mysterious function (a black box) that spits out numbers, but it's too complicated to calculate directly. You want to approximate it with a simple fraction (a polynomial divided by another polynomial).

In 1D: The "denominator" of your fraction is the special polynomial you baked earlier.
In 2D (This Paper): The authors show that their new triangular polynomial can serve as the common denominator for approximating two different mysterious functions at the same time.

However, there is a twist. In the 1D world, the "numerator" (the top part of the fraction) is also a simple polynomial. In this 2D world, the numerator turns out to be a bit more complex—it involves hypergeometric functions (which are like infinite series of numbers). It's not a simple polynomial anymore, but the authors showed you can still calculate it exactly using their formulas.

5. The Proof: Did it Work?

The authors didn't just write down the theory; they actually baked the cake.

They chose specific "flavors" (mathematical parameters) for their triangle.
They calculated the polynomial using their new formula.
They used it to approximate two complex functions.
The Result: The approximation was very close to the real functions, especially when looking at numbers far away from the triangle (large values of $z$ and $w$ ). The error was tiny, proving their "recipe" works.

Summary

In short, this paper is a cookbook for a new type of mathematical ingredient.

They took a known 1D recipe for balancing multiple flavors.
They expanded it to work on a 2D triangle.
They proved that by using a "double-mixing" technique, you can create a polynomial that balances two different weight distributions at once.
They showed that this new polynomial is the perfect key to unlock better approximations for complex 2D functions, even though the "numerator" part of the solution is a bit more complicated than before.

It's a bridge between simple, one-dimensional math and the more complex, multi-dimensional world, showing us how to keep things balanced even when the rules get messy.

Technical Summary: Jacobi-Piñeiro Multiple Orthogonal Polynomials on the Simplex

Problem Statement
The paper addresses a gap in the theory of Orthogonal Polynomials and Special Functions: the lack of explicit Rodrigues-type formulas for constructing multiple orthogonal polynomials in a multivariate setting. While univariate multiple orthogonality (specifically Jacobi-Piñeiro polynomials) and multivariate standard orthogonality (polynomials on the simplex) are well-established, their intersection—bivariate multiple orthogonality defined via Rodrigues formulas—had not been previously explored. Furthermore, the application of such polynomials to the multivariate Hermite-Padé approximation problem remained an open area of study.

Methodology
The authors develop a framework for bivariate multiple orthogonal polynomials on the triangular domain (simplex) $T = \{(x,y) \in \mathbb{R}^2 : x \ge 0, y \ge 0, x+y \le 1\}$ . The methodology proceeds in three stages:

Definition via Rodrigues Operators: The authors define a family of bivariate Jacobi-Piñeiro polynomials, $P_{(n_1,k_1),(n_2,k_2)}^{(\vec{\alpha},\vec{\beta},\gamma)}(x,y)$ , associated with $r$ weight functions of the form $W_j(x,y) = x^{\alpha_j}y^{\beta_j}(1-x-y)^\gamma$ . These polynomials are constructed by composing differential Rodrigues operators, $D_j$ , which generalize the univariate Rodrigues operators used for classical Jacobi-Piñeiro polynomials.
- The operator $D_j$ is defined as $D_j[f] = \frac{1}{W_j} \partial^{n_j}_{x^{n_j-k_j}y^{k_j}} (x^{n_j-k_j+\alpha_j}y^{k_j+\beta_j} f)$ .
- The polynomial is formed by the composition $D_r \circ \dots \circ D_1 [(1-x-y)^{\sum n_i}]$ .
Theoretical Verification: The paper rigorously proves that this construction yields:
- A polynomial of total degree $\sum n_i$ .
- A symmetric construction where the order of applying operators (and thus the assignment of measures) does not alter the resulting polynomial.
- Satisfaction of specific multiple orthogonality conditions with respect to the measures $d\mu_j = W_j dx dy$ . Specifically, the polynomial is orthogonal to monomials $x^{n-k}y^k$ under measure $j$ if $0 \le n-k < n_j - k_j$ or $0 \le k \le k_j$ .
Application to Hermite-Padé Approximation: The authors apply these polynomials to the bivariate Hermite-Padé approximation problem. They consider functions $E_j(z,w)$ defined as Stieltjes transforms of the measures on the triangle. The goal is to find a common denominator polynomial and numerators such that the approximation error vanishes to a specific order at infinity.
- The bivariate Jacobi-Piñeiro polynomial serves as the common denominator.
- The numerator is derived from an integral representation involving the difference of the polynomial evaluated at the approximation point and the integration variables.

Key Contributions and Results

Explicit Rodrigues Formula for Bivariate MOPs: The primary contribution is the provision of an explicit Rodrigues-type formula (Equation 15) for bivariate multiple orthogonal polynomials. This serves as a natural generalization of both the univariate Jacobi-Piñeiro polynomials and the standard bivariate orthogonal polynomials on the simplex.
Symmetry and Commutativity: The authors prove that the differential operators used in the construction commute (Proposition 3.5), ensuring the resulting polynomial is independent of the ordering of the measures.
Solution to the Bivariate Hermite-Padé Problem: The paper establishes that the constructed polynomials solve the bivariate Hermite-Padé approximation problem for functions defined by Stieltjes transforms on the simplex.
Analytic Structure of the Numerator: A significant finding is the structural difference between the univariate and bivariate cases regarding the numerator of the approximant.
- In the univariate case, the numerator is a polynomial.
- In the bivariate case, the numerator $\Phi^{(j)}(z,w)$ is not a polynomial. Instead, it is shown to be a computable expression involving finite sums and hypergeometric functions (specifically $_2F_1$ ), as derived in Equations (39)–(43).
Numerical Validation: The authors provide a numerical implementation for a specific case with two measures and parameters $\alpha_1=0, \alpha_2=3/2, \beta_1=1/2, \beta_2=4/3, \gamma=0$ . The results demonstrate that the rational approximants converge to the target functions $E_1$ and $E_2$ , with accuracy improving as $z$ and $w$ increase (moving away from the domain of integration).

Significance and Claims
The paper claims to provide the first example of bivariate multiple polynomials defined via a Rodrigues formula, effectively bridging the gap between multivariate standard orthogonality and univariate multiple orthogonality. The work extends the methodology of Hermite-Padé approximation to two variables, demonstrating that while the denominator remains a polynomial, the numerator requires a more complex analytic representation involving hypergeometric functions.

The authors modestly frame the work as a foundational step, noting that while uniqueness of the solution (up to a constant) holds in the univariate case, the bivariate case may admit multiple solutions, with the Rodrigues-based construction being one specific valid solution. The work is presented as a theoretical extension with a concrete numerical demonstration, laying the groundwork for future research in multivariate multiple orthogonality and its applications in rational approximation.