Atomic decomposition for an affine Weyl group of type G 2 G_{2}
Bárbara Muniz, David Plaza, Claudia Rojas-Andías · 2025
Read Atomic decomposition for an affine Weyl group of type G 2 G_{2} by Bárbara Muniz, David Plaza, Claudia Rojas-Andías (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-2512-02559.txt.
Introduction
Bárbara Muniz, David Plaza, and Claudia Rojas-Andías prove a structural fact about a small but rich algebra. The algebra is the spherical Hecke algebra of type G2, tied to an affine Weyl group. Its Kazhdan-Lusztig basis is a special basis used to study symmetry and representation theory.
The authors show each basis element splits into atomic pieces, an atomic decomposition. As a side benefit, they give a new algorithm for generalized Kostka-Foulkes polynomials in this type. Concrete decompositions like this make hard algebraic objects easier to compute with and compare, both by hand and by machine. Small cases done with care often crack the big ones open in time, and this one is no odd case out.
Catalog record
| CREATOR | Bárbara Muniz, David Plaza, Claudia Rojas-Andías |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-02559
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Atomic decomposition for an affine Weyl group of type G 2 G_{2} Bárbara Muniz, David Plaza, Claudia Rojas-Andías Abstract . We show that the elements of the Kazhdan–Lusztig basis of the spherical Hecke algebra of type G 2 G_{2} have an atomic decomposition. As a by-product, we obtain a new algorithm to compute generalized Kostka–Foulkes polynomials in type G 2 G_{2} . 1. Introduction Kazhdan–Lusztig polynomials arise as the transition coefficients between the canonical (Kazhdan–Lusztig) basis and the standard basis of the Hecke algebra attached to a Coxeter system [ 4 ] . They are fundamental objects in representation theory and geometry because they encode important structural information in both settings. In this paper we work with an affine Weyl group of type G_{2} and focus on a distinguished family of elements indexed by dominant weights, which we call spherical elements . They are obtained by translating the longest element of the finite Weyl group by a dominant weight (see Section 2 for the precise definition). For spherical elements, Kazhdan–Lusztig polynomials coincide with Kostka–Foulkes polynomials or q -weight multiplicities [ 8 , 3 ] , and in particular their coefficients are known to be non-negative. In type A, Lascoux discovered a further positivity phenomenon for spherical elements [ 5 ] . He introduced a basis now called the atomic basis and proved that the canonical basis elements expand in this basis with coefficients in \mathbb{N}[q] . We refer to this phenomenon by saying that Kazhdan–Lusztig basis elements have an atomic decomposition. Lecouvey and Lenart [ 6 ] later studied atomic decompositions in other types using the theory of crystal. Their work provides evidence that atomicity extends beyond type A and naturally raises the problem of determining for which affine types, and for which dominant weights, the spherical elements admit atomic decompositions. Beyond type A , atomicity may fail, and a precise characterization of the weights and affine types for which it holds is still out of reach. In this paper we solve this question for an affine Weyl group of type G_{2} . For a dominant weight \lambda , we write \underline{\mathbf{H}}_{\lambda} for the spherical Kazhdan–Lusztig basis element and \mathbf{N}_{\lambda} for the corresponding element of the atomic basis. Our main result is the following. Theorem A . For every dominant weight \lambda we have an expansion \underline{\mathbf{H}}_{\lambda}=\sum_{\mu\leq\lambda}a_{\lambda,\mu}(q)\mathbf{N}_{\mu}, where \leq denotes the dominance order on weights and the polynomials a_{\lambda,\mu}(q) have non-negative coefficients. In other words, in type G_{2} , all the Kazhdan–Lusztig basis elements associated to spherical elements admits an atomic decomposition. We give two proofs of this result. The first proof relies on the step-by-step decompositions of the pre-canonical bases introduced in [ 7 ] . These decompositions do not immediately give atomicity, since one of the intermediate transitions is not positive. It is therefore necessary to verify that the negative terms that arise during the process cancel out, yielding the desired positive expansion. The second proof is obtained by modifying the definition of the pre-canonical bases so that the step-by-step decompositions become positive at each stage, along the line of what was done in [ 9 ] for type C_{2} . This approach is particularly well-suited for the construction of crystalline realizations of the atomic decomposition, an ongoing line of work by the first author. Such concrete realizations are extremely valuable, as they can be used to combinatorially describe Kostka–Foulkes polynomials as q -weight multiplicities — a long-standing open problem beyond type A . We expect that similar modifications of the pre-canonical bases may recover step-by-step positivity in other types as well. As an application of our results, we provide an explicit algorithm for computing generalized Kostka–Foulkes polynomials in type G_{2} . The paper is organized as follows. Section 2 reviews the necessary background on the affine Weyl group of type G_{2} , the spherical Hecke algebra, the spherical elements, and the pre-canonical bases. Section 3 contains the first proof of A via step-by-step decompositions. Section 4 develops the adjusted bases and presents the second proof. Section 5 describes the computation of generalized Kostka–Foulkes polynomials. Finally, we add an appendix with a SageMath implementation of our algorithms to compute the atomic decomposition, the pre-canonical bases and the Kostka–Foulkes polynomials. 2. Preliminaries Throughout this paper we denote by \mathbb{N} the set of non-negative integers. 2.1. The affine Weyl group of type G_{2} In this paper we only work with a root system \Phi of type G_{2} . Therefore we fix an explicit realization of \Phi inside V=\mathbb{R}^{2} with the usual inner product denoted by \langle-,-\rangle . We take \Delta=\{\alpha_{1},\alpha_{2}\} as the set of simple roots, where (2.1) \displaystyle\alpha_{1}=(1,0)\quad\mbox{ and }\quad\alpha_{2}=\left(-\tfrac{3}{2},\,\tfrac{\sqrt{3}}{2}\right). Then the root system \Phi is given by (2.2) \Phi=\{\pm\alpha_{1},\pm\alpha_{2},\pm(\alpha_{1}+\alpha_{2}),\pm(2\alpha_{1}+\alpha_{2}),\pm(3\alpha_{1}+\alpha_{2}),\pm(3\alpha_{1}+2\alpha_{2})\}. The root system \Phi is depicted in Figure 1 , where simple roots are given in red and the positive roots that are not simple are drawn in blue. We denote by \Phi^{\vee}=\{\alpha^{\vee}\mid\alpha\in\Phi\} the dual root system, where \displaystyle\alpha^{\vee}=\frac{2\alpha}{\langle\alpha,\alpha\rangle} . 3\alpha_{1}+2\alpha_{2} 3\alpha_{1}+\alpha_{2} \alpha_{2} -3\alpha_{1}-2\alpha_{2} -3\alpha_{1}-\alpha_{2} -\alpha_{2} \alpha_{1} -\alpha_{1} 2\alpha_{1}+\alpha_{2} \alpha_{1}+\alpha_{2} -2\alpha_{1}-\alpha_{2} -\alpha_{1}-\alpha_{2} Figure 1 . A root system of type G_{2} Because the definition of precanonical bases involves the height of positive roots, we include in Table 1 the height of each positive root. Root \alpha_{1} \alpha_{2} \alpha_{1}+\alpha_{2} 2\alpha_{1}+\alpha_{2} 3\alpha_{1}+\alpha_{2} 3\alpha_{1}+2\alpha_{2} Height 1 1 2 3 4 5 {(2,-1)_{\varpi}} {(-3,2)_{\varpi}} {(-1,1)_{\varpi}} {(1,0)_{\varpi}} {(3,-1)_{\varpi}} {(0,1)_{\varpi}} . Table 1 . Heights of the six positive roots in the G_{2} root system. The fundamental weights \{\varpi_{1},\varpi_{2}\} are determined by the conditions \langle\varpi_{i},\alpha_{j}^{\vee}\rangle=\delta_{i,j} for i,j\in\{1,2\} . Concretely, we have \varpi_{1}=2\alpha_{1}+\alpha_{2} and \varpi_{2}=3\alpha_{1}+2\alpha_{2} . Let X=\mathbb{Z}\varpi_{1}\oplus\mathbb{Z}\varpi_{2} be the weight lattice and X^{+}=\mathbb{N}\varpi_{1}\oplus\mathbb{N}\varpi_{2} the set of dominant weights. To avoid confusion with coordinates in V , for \lambda=a_{1}\varpi_{1}+a_{2}\varpi_{2}\in X we use the shorthand {(a_{1},a_{2})_{\varpi}} to denote \lambda . We will frequently use the coefficients of the positive roots when written in the basis of fundamental weights, and we collect these coefficients in Table 1 . We also define \rho\in X^{+} as the half-sum of positive roots. In formulas, (2.3) \rho=\frac{1}{2}\sum_{\alpha\in\Phi^{+}}\alpha=5\alpha_{1}+3\alpha_{2}={(1,1)_{\varpi}}. For \alpha\in\Phi and m\in\mathbb{Z} we define the hyperplane H_{\alpha,m} by (2.4) H_{\alpha,m}=\{\lambda\in V\mid\langle\lambda,\alpha^{\vee}\rangle=m\} and the reflection s_{\alpha,m} along the hyperplane H_{\alpha,m} . In formulas, we have s_{\alpha,m}(\lambda)=\lambda-(\langle\lambda,\alpha^{\vee}\rangle-m)\alpha . The finite Weyl group W_{f} attached to \Phi is the group generated by the reflections \{s_{\alpha,0}\mid\alpha\in\Phi\} . Let s_{1}=s_{\alpha_{1},0} and s_{2}=s_{\alpha_{2},0} . It is well-known that W_{f} is generated by \{s_{1},s_{2}\} . Indeed W_{f} is a Coxeter system with presentation (2.5) W_{f}=\langle s_{1},s_{2}\mid s_{1}^{2}=s_{2}^{2}=1,\quad(s_{1}s_{2})^{6}=1\rangle. So that W_{f} is isomorphic to the group of symmetries of a regular hexagon. It longest element is w_{0}=(s_{1}s_{2})^{3}=(s_{2}s_{1})^{3} . The affine Weyl group W_{f} attached to \Phi is the group generated by the reflections \{s_{\alpha,m}\mid\alpha\in\Phi,\,m\in\mathbb{Z}\} . The connected components of the complement of hyperplanes (2.6) V\setminus\bigcup_{\alpha,m}H_{\alpha,m} are called alcoves. In our G_{2} -setting , alcoves correspond to triangles (see Figure 2 ) We denote by \mathcal{A} the set of alcoves. We call (2.7) A_{0}=\{\lambda\in V\mid-1<\langle\lambda,\alpha^{\vee}\rangle<0,\mbox{ for all }\alpha\in\Phi^{+}\} the fundamental alcove. There is a bijection W_{a}\rightarrow\mathcal{A} given by w\mapsto w(A_{0}) . This bijection allows us to uniquely identify elements in W_{a} with alcoves. The walls of A_{0} are supported in the hyperplanes H_{\alpha_{1},0} , H_{\alpha_{2},0} and H_{\beta,-1} , where \beta is the highest short root, so that in our setting we have