Two characterizations of Sheffer-Dunkl sequences

Alejandro Gil Asensi, Judit Mínguez Ceniceros · 2025

Read Two characterizations of Sheffer-Dunkl sequences by Alejandro Gil Asensi, Judit Mínguez Ceniceros (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-2501-01364.txt.

Introduction

Alejandro Gil Asensi and Judit Minguez Ceniceros pin down a family of special polynomials in the Dunkl setting in a 2025 paper. Sheffer sequences, a broad clan that includes many classic ones, can be told apart by Stieltjes integrals. The Dunkl twist swaps the plain derivative for a deformed operator with a reflection built in.

Two new integral tests are proven to mark exactly which sequences count as Sheffer-Dunkl. The tests mirror the classic theory while respecting the twist. Such rules let researchers spot members of the family on sight. The Dunkl branch of the subject gains tidy membership rules.

Catalog record

CREATORAlejandro Gil Asensi, Judit Mínguez Ceniceros
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE26 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-01364

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Two characterizations of Sheffer-Dunkl sequences Alejandro Gil Asensi, Judit Mínguez Ceniceros Abstract Sheffer polynomials can be characterized using different Stieltjes integrals. These families of polynomials have been recently extended to the Dunkl context. In this way some classical operators as the derivative operator or the difference operator are replaced as analogous operators in the Dunkl universe. In this paper we establish two Stieltjes integrals that help us to characterize the Sheffer-Dunkl polynomials. MSC 2020 Classification]11B83,44A60, 11B68 keywords moment problems, Sheffer-Dunkl sequences. pacs [ † † equal-contributors: These authors contributed equally to this work. 1 Introduction Let g(t) and f(t) be some formal power series given by g(t)=\sum_{n=0}^{\infty}\frac{b_{n}}{n!}t^{n},\,b_{0}\neq 0,\quad f(t)=\sum_{n=1}^{\infty}\frac{a_{n}}{n!}t^{n},\,a_{1}\neq 0. (1.1) Let \mathrm{L}_{f}\colon\mathcal{P}\to\mathcal{P} (where \mathcal{P} is the space of polynomials) be the linear operator given by \mathrm{L}_{f}p(x):=\sum_{n=0}^{\infty}\frac{a_{n}}{n!}\frac{d^{n}}{dx^{n}}p(x). (1.2) A Sheffer sequence \{s_{n}(x)\}_{n=0}^{\infty} for the pair (g(t),f(t)) , where g(t) and f(t) are as in ( 1.1 ), is defined using any of the two equivalent definitions: • by a Taylor generating expansion \frac{1}{g(\overline{f}(t))}e^{x\overline{f}(t)}=\sum_{n=0}^{\infty}s_{n}(x)\frac{t^{n}}{n!}, (1.3) where \overline{f}(t) denotes the compositional inverse of f(t) ; • by the linear operator \mathrm{L}_{f}s_{n}(x)=ns_{n-1}(x),\quad n\geq 1. (1.4) Sheffer sequences were studied in [ 1 , 15 , 16 , 17 , 18 , 19 ] using Umbral Calculus. Particular cases of Sheffer polynomials are Appell polynomials , which are obtained taking f(t)=t , (in this case ( 1.4 ) is reduced to s_{n}^{\prime}(x)=ns_{n-1}(x) and \overline{f}(t)=t in ( 1.3 )) and associated polynomials for f(t) , which are obtained taking g(t)=1 . Other characterizations of this kind of polynomials can be found in [ 21 ] and [ 23 ] using Stieltjes integrals. Thorne in [ 23 ] gives a characterization for Appell polynomials that can be easily extended to Sheffer polynomials as points out Sheffer in [ 21 ] . In particular, he proves that: Thorne Theorem . A sequence of polynomials \{A_{n}(x)\}_{n=0}^{\infty} is an Appell sequence ( f(t)=t ) if and only if there exists a function \alpha(x) of bounded variation on (0,\infty) with the following properties: i) the moment integrals \mu_{n}=\int_{0}^{\infty}x^{n}\,d\alpha(x) all exist and \mu_{0}\neq 0 ; ii) if A_{n}^{(r)}(x) denotes the r -derivative of A_{n}(x) , \int_{0}^{\infty}A^{(r)}_{n}(x)\,d\alpha(x)=n!\,\delta_{n,r}, where \delta_{n,r}=1 if n=r and \delta_{n,r}=0 if n\neq r . In this case, the function g(t) in ( 1.3 ) is given by g(t)=\sum_{n=0}^{\infty}\mu_{n}\frac{t^{n}}{n!}=\int_{0}^{\infty}e^{tx}\,d\alpha(x). On the other hand, Sheffer in [ 21 ] provides a characterization for the Sheffer sequences by raising a different integral. Sheffer Theorem . A sequence of polynomials \{s_{n}(x)\}_{n=0}^{\infty} is a Sheffer sequence for the pair (g(t),f(t)) if and only if there exists a function \beta(x) of bounded variation on (0,\infty) with the following properties: i) the moment integrals \omega_{n}=\int_{0}^{\infty}x^{n}\,d\beta(x) all exist and \omega_{0}\neq 0 ; ii) the polynomials s_{n}(x) can be expressed as s_{n}(x)=\int_{0}^{\infty}p_{n}(x+t)^{n}\,d\beta(t), (1.5) where p_{n}(x) are the associated polynomials for f(t) . In this case, the function g(t) is given by \frac{1}{g(t)}=\int_{0}^{\infty}e^{xt}\,d\beta(x)=\sum_{n=0}^{\infty}\omega_{n}\frac{t^{n}}{n!}. Note that if f(t)=t we have a characterization for Appell polynomials. In this case the associated polynomials in ( 1.5 ) are given by p_{n}(x)=x^{n} . The main complication when applying these theorems is to find, for a given sequence of moments \{c_{n}\}_{n=0}^{\infty} , the corresponding function w(x) such that c_{n}=\int_{-\infty}^{\infty}x^{n}\,dw(x). A moment problem is said to have a solution if w(x) is a positive measure. In [ 5 ] the author provides a technique to find the functions w(x) . In [ 11 ] , Sheffer polynomials are extended to Dunkl context. In this case, the derivative operator in ( 1.2 ) is replaced by the Dunkl operator \Lambda_{\nu}f(x)=\frac{d}{dx}f(x)+\frac{2\nu+1}{2}\left(\frac{f(x)-f(-x)}{x}\right), where \nu>-1 is a fixed parameter (see [ 4 , 20 ] ). Observe that the case \nu=-1/2 recovers the classical case \Lambda_{-1/2}=\frac{d}{dx} . The sequence of n! is replaced by \gamma_{n,\nu}=\begin{cases}2^{2k}k!\,(\nu+1)_{k},&\text{if $n=2k$},\\ 2^{2k+1}k!\,(\nu+1)_{k+1},&\text{if $n=2k+1$},\end{cases} (1.6) where (a)_{n} denotes the Pochhammer symbol (a)_{n}=a(a+1)(a+2)\cdots(a+n-1)=\frac{\Gamma(a+n)}{\Gamma(a)} (with a a non-negative integer); of course, \gamma_{n,-1/2}=n! . In order to define the Sheffer-Dunkl polynomials we take two formal power series in this way g(t)=\sum_{n=0}^{\infty}\frac{b_{n}}{\gamma_{n,\nu}}t^{n},\,b_{0}\neq 0,\quad f(t)=\sum_{n=1}^{\infty}\frac{a_{n}}{\gamma_{n,\nu}}t^{n},\,a_{1}\neq 0. (1.7) Then a Sheffer-Dunkl sequence \{s_{n,\nu}(x)\}_{n=0}^{\infty} for the pair (g(t),f(t)) as in ( 1.7 ) satisfies that \mathrm{L}_{f}s_{n,\nu}(x):=\sum_{n=1}^{\infty}\frac{a_{n}}{\gamma_{n,\nu}}\Lambda_{\nu}^{n}s_{n,\nu}(x)=\frac{\gamma_{n,\nu}}{\gamma_{n-1,\nu}}s_{n-1,\nu}(x), (1.8) where \Lambda_{\nu}^{0} is the identity operator and \Lambda_{\nu}^{n+1}=\Lambda_{\nu}(\Lambda_{\nu}^{n}) . Or, equivalently, \frac{1}{g(\overline{f}(t))}E_{\nu}(x\overline{f}(t))=\sum_{n=0}^{\infty}s_{n,\nu}(x)\frac{t^{n}}{\gamma_{n,\nu}}, (1.9) where \overline{f}(t) denotes the compositional inverse of f(t) and E_{\nu}(t) is called Dunkl exponential or Dunkl kernel, and it is defined by E_{\nu}(t)=\sum_{n=0}^{\infty}\frac{t^{n}}{\gamma_{n,\nu}}. This function plays the role of the exponential function in the classical case. It satisfies \Lambda_{\nu}(E_{\nu}(\lambda t))=\lambda E_{\nu}(\lambda t) and it can also be written as E_{\nu}(t)=\mathcal{I}_{\nu}(t)+\frac{1}{2(\nu+1)}G_{\nu}(t), where \mathcal{I}_{\nu}(t)=2^{\nu}\Gamma(\nu+1)\frac{J_{\nu}(it)}{(it)^{\nu}}=\sum_{n=0}^{\infty}\frac{t^{2n}}{\gamma_{2n,\nu}} and G_{\nu}(t)=t\mathcal{I}_{\nu+1}(t)=\sum_{n=0}^{\infty}\frac{t^{2n+1}}{\gamma_{2n,\nu}}=2(\nu+1)\sum_{n=0}^{\infty}\frac{t^{2n+1}}{\gamma_{2n+1,\nu}}. Here, J_{\nu}(t) is the Bessel function of order \nu (and hence, \mathcal{I}_{\nu}(t) is a small variation of the modified Bessel function of the second kind, I_{\nu}(t) ), see [ 24 ] or [ 14 ] . Therefore, \mathcal{I}_{\nu}(t) and G_{\nu}(t) play the role of the even part and the odd part of E_{\nu}(t) , respectively. We remark that, when \nu=-1/2 , we recover the classical case, i.e., \Lambda_{-1/2}=d/dt , E_{-1/2}(x)=e^{t} , \mathcal{I}_{-1/2}(t)=\cosh t and G_{-1/2}(t)=\sinh t . Finally, we are going to need the translation operator in this context. The Dunkl translation is defined as \tau_{y}f(x)=\sum_{n=0}^{\infty}\Lambda_{\nu}^{n}f(x)\frac{y^{n}}{\gamma_{n,\nu}},\qquad\nu>-1. (1.10) This is just a generalization of the classical translation, but considered as a Taylor expansion of a function f around a fixed point x , that is, f(x+y)=\sum_{n=0}^{\infty}f^{(n)}(x)\frac{y^{n}}{n!}. The Dunkl translation has a property than resembles the Newton binomial (x+y)^{n}=\sum_{k=0}^{n}\binom{n}{k}y^{k}x^{n-k} , which is \tau_{y}((\cdot)^{n})(x)=\sum_{k=0}^{n}\binom{n}{k}_{\!\nu}y^{k}x^{n-k}, where \binom{n}{j}_{\!\nu}=\frac{\gamma_{n,\nu}}{\gamma_{j,\nu}\gamma_{n-j,\nu}}. Of course, \binom{n}{j}_{\!\nu} becomes the ordinary binomial numbers in the case \nu=-1/2 . Some other properties of the Dunkl translation, including an integral expression that is more general than ( 1.10 ), can be found in [ 20 ] and [ 22 ] . In that setting, an Appell-Dunkl sequence \{A_{n,\nu}(x)\}_{n=0}^{\infty} for g(t) is a sequence of polynomials that satisfies \Lambda_{\nu}A_{n,\nu}(x)=\frac{\gamma_{n,\nu}}{\gamma_{n-1,\nu}}A_{n-1,\nu}(x) and its generating function is \frac{1}{g(t)}E_{\nu}(xt)=\sum_{n=0}^{\infty}A_{n,\nu}(x)\frac{t^{n}}{\gamma_{n,\nu}}. Also, the associated Dunkl polynomials p_{n,\nu}(x) for a function f(t) are defined by means of the generating function E_{\nu}(x\overline{f}(t))=\sum_{n=0}^{\infty}p_{n,\nu}(x)\frac{t^{n}}{\gamma_{n,\nu}}. Other important family of Sheffer-Dunkl polynomials are discrete Appell-Dunkl polynomials . They have been introduced and studied in [ 9 ] . The generating function of this kind of polynomials \{d_{n,\nu}(x)\}_{n=0}^{\infty} is given by \frac{1}{g(\overline{G_{\nu}}(t))}E_{\nu}(x\overline{G_{\nu}}(t))=\sum_{n=0}^{\infty}d_{n,\nu}(x)\frac{t^{n}}{\gamma_{n,\nu}}. The operator \mathrm{L}_{f} for the discrete Appell-Dunkl polynomials is \mathrm{L}_{G_{\nu}}=(\nu+1)(\tau_{1}-\tau_{-1}). (1.11) And it holds that \mathrm{L}_{G_{\nu}}d_{n,\nu}(x)=\frac{\gamma_{n,\nu}}{\gamma_{n-1,\nu}}d_{n-1,\nu}. If g(t)=1 we obtain the Dunkl factorial polynomials \{f_{n,\nu}(x)\}_{n=0}^{\infty} generated in the following way E_{\nu}(x\overline{G_{\nu}}(t))=\sum_{n=0}^{\infty}f_{n,\nu}(x)\frac{t^{n}}{\gamma_{n,\nu}}. (1.12) The paper is structured in this way. Section 2 is devoted to extend the Thorne Theorem to the Sheffer-Dunkl sequences of polynomials. In Section 3 we characterized these polynomials extending the Sheffer Theorem to the Dunkl context. Finally, in Section 4 we show several examples of Sheffer-Dunkl polynomials where we contruct the corresponding measures and obtain properties of these families, some of them, unkown. 2 Thorne Theorem for the Sheffer-Dunkl polynomials In this Section we present the Thorne Theorem extended to Sheffer-Dunkl polynomials. Theorem 1 . Let g(t) , f(t) be two formal series as in ( 1.7 ). Then a sequence \{s_{n,\nu}(x)\}_{n=0}^{\infty} is the Sheffer-Dunkl sequence for the pair (g(t),f(t)) if and only if there exists a function \alpha_{\nu}(x) of bounded variation on (-\infty,\infty) with the following properties i) there exist the moment integrals \mu_{n,\nu}=\int_{-\infty}^{\infty}x^{n}\,d\alpha_{\nu}(x),\quad n=0,1,\ldots, and \mu_{0,\nu}\neq 0 ; ii) the polynomials s_{n,\nu}(x) satisfy \int_{-\infty}^{\infty}\mathrm{L}_{f}^{r}s_{n,\nu}(x)\,d\alpha_{\nu}(x)=\gamma_{n,\nu}\delta_{n,r}, (2.1) where \mathrm{L}^{r}_{f} is the linear operator defined in ( 1.8 ) applied r times.

Read the complete text