Saturation for Non-Symmetric Macdonald Polynomials
Milo Bechtloff Weising, Alexander E. Black · 2025
Read Saturation for Non-Symmetric Macdonald Polynomials by Milo Bechtloff Weising, Alexander E. Black (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-00336.txt.
Introduction
Milo Bechtloff Weising and Alexander Black settle a conjecture about the shape hidden inside a family of polynomials. Non-symmetric Macdonald polynomials are refined building blocks in algebra, each spreading over a set of exponent points. The question was whether that spread always forms an M-convex set, a discrete shape with no dents, which would give the polynomials a prized property about their Newton polytopes.
The authors prove the supports are indeed M-convex. That closes a 2019 conjecture and, as fallout, answers a 2022 question about affine Demazure characters, which inherit the same good shape. Certain polytopes tied to Bruhat order then turn out to be generalized permutahedra, a well-understood family. The key step is a new reading of a known formula for the polynomials, recast as colorings of Dyck graphs, which exposes the convexity directly.
Catalog record
| CREATOR | Milo Bechtloff Weising, Alexander E. Black |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-00336
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Saturation for Non-Symmetric Macdonald Polynomials Milo Bechtloff Weising, Alexander E. Black Abstract . We prove that supports of nonsymmetric Macdonald polynomials are M M -convex. As a consequence, we resolve a 2019 conjecture of Monical, Tokcan, and Yong that they have the saturated Newton polytope property. As a further consequence, we prove that the moment polytopes of affine Schubert varieties in the affine Grassmannian of type GL \mathrm{GL} are generalized permutahedra. 1. Introduction Given a polynomial defined implicitly, it is not necessarily clear how to determine which of its monomials has a nonzero coefficient. For some families of polynomials, there is a polyhedral characterization of the monomials that arise. Namely, given a polynomial f\in\mathbb{R}[x_{1},x_{2},\dots,x_{n}] with f(x)=\sum_{\alpha\in\mathbb{Z}_{\geq 0}^{n}}c_{\alpha}x^{\alpha} , one may compute its Newton polytope given by \mathrm{Newton}(f)=\text{conv}(\{\alpha\in\mathbb{Z}^{n}:c_{\alpha}\neq 0\}) . Since f is a polynomial, this set is finite by definition, and so the Newton polytope is indeed a convex polytope. Every monomial is a lattice point in that polytope. We say that f has a saturated Newton polytope if the converse is true: \mathrm{Newton}(f)\cap\mathbb{Z}^{n}=\{\alpha\in\mathbb{Z}^{n}:c_{\alpha}\neq 0\}. A first example of this phenomenon is the determinant. Its Newton polytope is the Birkhoff polytope, and the only integer points it contains are precisely its vertices. Hence, the determinant has saturated Newton polytope. Any square free polynomial is more generally saturated, but the phenomenon becomes rarer and harder to verify in higher degree. Saturation gained attention recently as it holds for Lorentzian polynomials [ 10 ] . These polynomials appear in connection to the Kahler package of Adiprasito, Huh, and Katz [ 1 ] used to resolve longstanding conjectures in matroid theory related to log-concavity, as well as in the theory of real stability of polynomials used to establish fast mixing for random walks on basis exchange graphs of matroids [ 3 , 4 , 5 , 6 ] . Lorentzian polynomials have a stronger property than saturation. Their supports are M-convex sets , meaning that they have saturated Newton polytopes, and their Newton polytope is a generalized permutahedron in the sense of Postnikov [ 32 ] . In fact, it was shown in [ 10 ] that a polynomial has M-convex support if and only if it has the same support set as some Lorentzian polynomial. These results motivate the recent study of M-convexity and saturation of supports of polynomials. In [ 28 ] , Monical, Tokcan, and Yong observed that many polynomials appearing in algebraic combinatorics have saturated Newton polytopes, either via proof or conjecturally after computations. They proved this, for example, for Schur polynomials, symmetric Macdonald polynomials, elementary symmetric functions, and resultants among several other examples. They also provided several nonexamples, such as discriminants and modified Macdonald polynomials. Based on computational evidence from small examples, they conjectured Schubert polynomials, double Schubert polynomials, key polynomials, Grothendieck polynomials, Kronecker products of Schur polynomials, Lascoux polynomials, Demazure atoms, and non-symmetric Macdonald polynomials were all saturated. Later, Fink, Mészáros, and St. Dizier proved that Schubert polynomials and key polynomials had the saturated Newton polytope property via an exact computation of their Newton polytopes as a Minkowski sum of matroid base polytopes [ 17 ] . Besson, Jeralds, and Kiers later extended the results for key polynomials with an alternative proof for more general Coxeter type [ 8 ] , and Castillo, Cid Ruiz, Mohammadi, and Montaño showed saturation for double Schubert polynomials using the theory of multidegrees from commutative algebra [ 12 ] . For Grothendieck polynomials and Kronecker products of Schur polynomials, several special cases of the conjecture are known to hold [ 18 , 11 , 27 , 15 , 31 ] . Beyond those conjectures, in connection to the theory of Lorentzian polynomials, several other families of polynomials from algebraic combinatorics have been shown to have saturated Newton polytopes [ 25 , 24 , 19 , 35 , 16 , 26 , 2 , 14 , 30 , 23 ] . For Lascoux polynomials, Demazure atoms, and non-symmetric Macdonald polynomials, to our knowledge, nothing was known prior to this work even in special cases. Here we resolve the final case equivalent to Conjecture 3.8 of [ 28 ] : Theorem ( 3.4 ) . The supports of non-symmetric Macdonald polynomials are M-convex. In particular, they have the saturated Newton polytope property. The non-symmetric Macdonald polynomials \{E_{\mu}(x_{1},\dots,x_{n})|\mu\in\mathbb{Z}^{n}\} form an exceptional basis for the space of Laurent polynomials \mathbb{Q}(q,t)[x_{1}^{\pm 1},\dots,x_{n}^{\pm 1}] occurring across modern algebraic combinatorics in the geometry of Hilbert schemes [ 22 ] , the combinatorics of LLT polynomials [ 9 , 20 , 21 ] , and the study of double affine Hecke algebras [ 13 ] . The Newton polytopes of non-symmetric Macdonald polynomials arise naturally in the study of affine Schubert calculus. Besson and Hong noted in [ 7 ] that the moment polytopes of affine Schubert varieties in affine Grassmannians are precisely the convex hulls of certain lower order ideals in the Bruhat order of the co-weight lattice. In this case, these moment polytopes are the Newton polytopes of non-symmetric Macdonald polynomials. As a consequence of Theorem 3.4 , we find the following. Corollary ( 3.10 ) . For all \mu\in\mathbb{Z}^{n}, the moment polytope \mathrm{MP}(\overline{X}_{\mu}) of the affine Schubert variety \overline{X}_{\mu}\subset\mathrm{Gr}_{n} is a generalized permutahedron. Our proof relies on the combinatorial formula for non-symmetric Macdonald polynomials of Haglund, Haiman, and Loehr [ 21 ] together with some polyhedral geometric observations. We will prove our result by induction by noting that the trivial non-symmetric Macdonald polynomial E_{(0,\dots,0)}=1 has M-convex support and any non-symmetric Macdonald polynomial may be built from smaller ones starting with E_{(0,\dots,0)}=1 by applying a sequence of applications of two operations that both preserve M-convexity of the support sets of the polynomials. 1.1. Acknowledgments The authors would like to thank Sean Griffin, Joseph Pappe, and Germain Poullot for useful discussions during the FPSAC 2025 conference at Hokkaido University. 2. Non-symmetric Macdonald polynomials and their combinatorics 2.1. Non-symmetric Macdonald polynomials The non-symmetric Macdonald polynomials E_{\mu} may be defined in multiple equivalent ways. We refer the reader to [ 21 ] and [ 13 ] for an overview of these various definitions. We use the explicit combinatorial description of non-symmetric Macdonald polynomials due to Haglund, Haiman, and Loehr (Theorem 3.5.1 [ 21 ] ) throughout this paper and will thus take their combinatorial formula as a definition. However, before we may define the non-symmetric Macdonald polynomials E_{\mu}(x_{1},\dots,x_{n};q,t) , we must first define some required combinatorial objects. Definition 2.1 . [ 21 ] For a (weak) composition \mu=(\mu_{1},\dots,\mu_{n})\in\mathbb{Z}_{\geq 0}^{n} , define the column diagram of \mu as \mathrm{dg}^{\prime}(\mu):=\{(i,j)\in\mathbb{Z}_{\geq 0}^{2}:1\leq i\leq n,1\leq j\leq\mu_{i}\}. This is represented by a collection of boxes in positions given by \mathrm{dg}^{\prime}(\mu) . The augmented diagram of \mu is given by \widehat{\mathrm{dg}}(\mu):=\mathrm{dg}^{\prime}(\mu)\cup\{(i,0):1\leq i\leq n\}. Visually, to get \widehat{\mathrm{dg}}(\mu) we are adding a bottom row of boxes on length n below the diagram \mathrm{dg}^{\prime}(\mu) which we refer to as the basement of the diagram. A filling of \mu is a function \sigma:\mathrm{dg}^{\prime}(\mu)\rightarrow\{1,...,n\} and given a filling there is an associated augmented filling \widehat{\sigma}:\widehat{\mathrm{dg}}(\mu)\rightarrow\{1,...,n\} extending \sigma with the additional bottom row boxes filled according to \widehat{\sigma}((j,0))=j for j=1,\dots,n . Distinct lattice squares u,v\in\mathbb{Z}_{\geq 0}^{2} are said to attack each other if one of the following is true: • u