Roots in the substitution group and in the group of Riordan matrices with ones in the main diagonal

Jorge Calero-Sanz, L. Felipe Prieto-Martínez · 2025

Read Roots in the substitution group and in the group of Riordan matrices with ones in the main diagonal by Jorge Calero-Sanz, L. Felipe Prieto-Martínez (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-2509-25972.txt.

Introduction

Jorge Calero-Sanz and Luis Felipe Prieto-Martinez study when iterative roots exist and are unique inside the substitution group of formal power series, and in the Riordan matrix group. The frame is matrix based, grounded in Riordan arrays. Their work came out in twenty twenty-five.

The substitution group links tightly to a Lagrange subgroup of Riordan matrices with ones on the main diagonal. Through that link, classic questions on algebraic completeness and root uniqueness become tractable. Results follow for varied base rings. The substitution view also eases root checks inside the Riordan group itself, the authors show.

Catalog record

CREATORJorge Calero-Sanz, L. Felipe Prieto-Martínez
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE51 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-25972

Browse the full library · About the Tower of Babel

Full text

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

Roots in the substitution group and in the group of Riordan matrices with ones in the main diagonal Jorge Calero-Sanz, L. Felipe Prieto-Martínez Abstract We investigate the existence and uniqueness of iterative roots of order n n within the substitution group of formal power series 𝒥 ⁡ ( Z ) \mathcal{J}(Z) -with coefficients in a commutative ring with unity Z Z - employing a matrix-based framework grounded in the Riordan group. We analyse the relationship between the substitution group and the Lagrange subgroup - a group of Riordan matrices - and explore some classic questions concerning algebraic completeness and uniqueness of root extractions. This approach allows us to obtain various results about the roots in 𝒥 ⁡ ( Z ) \mathcal{J}(Z) for different choices of Z Z . Furthermore, the examination of the substitution group facilitates the analysis of roots within the Riordan matrix group. MSC 13F25 , 13J05 , 15A99 , 16W60 1 Introduction The theory of formal power series has been a subject of significant interest, particularly due to its connections with algebraic and combinatorial structures [ 13 ] . Among these, the substitution group of formal power series with coefficients in a commutative ring with unity Z , denoted as \mathcal{J}(Z) and endowed with the composition operation, plays a central role in modeling iterative transformations and related algebraic operations [ 1 , 6 ] . A fundamental question in this context is the study of iterative roots: determining when a formal power series admits iterative roots of order n and under what conditions such roots are unique. Something similar happens with some groups of Riordan matrices with respect to the multiplication. A Riordan matrix is an infinite matrix that we can identify with a pair of formal power series in such a way that the product of two Riordan matrices is related to the multiplication and composition of the corresponding formal power series used for the identification. Recently, the algebraic study of groups of Riordan matrices have received some attention [ 2 , 3 , 4 , 8 , 9 , 10 ] but, as far as we know, a complete study of Riordan matrices admitting roots has not been thoroughly investigated. This paper deals with some aspects of the algebraic study of the substitution group and groups of Riordan matrices concerning questions about the existence and uniqueness of roots. 1.1 Main groups of interest In the sections that follow, we will always assume that Z denotes a commutative ring with unity and \mathbb{K} will always denote a field. We denote by Z[[x]] (respectively \mathbb{K}[[x]] ) to the set of formal power series f=f_{0}+f_{1}x+f_{2}x^{2}+\ldots with coefficients in Z (resp. \mathbb{K} ). This paper concerns about two groups. The first one is the following: Definition 1 . Let Z be any commutative ring with unity. The substitution group \mathcal{J}(Z) is the set of formal power series of the type g=x+g_{2}x^{2}+g_{3}x^{3}+\ldots in Z[[x]] with the operation of composition. The reader may find more information about this object in [ 1 , 6 , 7 ] . In the special case Z=\mathbb{F}_{p} (where \mathbb{F}_{p} denotes the corresponding finite field) the substitution group is known as the Nottingham group . More information about this special case can be found in the chapter The Nottingham Group by R. Camina in [ 14 ] . On the other hand, we have the second object of interest: Riordan matrices. Such objects are infinite matrices with a structure that generalizes Pascal’s Arithmetical Triangle. We have decided to present the following definition, which is a generalization, for any commutative ring with unity Z , of the original one appearing in [ 17 ] . See the notational remark below to put the definition in context with respect to the notation in current use. Definition 2 . Let Z be a commutative ring with unity. A Riordan matrix is an invertible infinite matrix of the type \begin{bmatrix}1\\ d_{10}&1\\ d_{20}&d_{21}&1\\ d_{30}&d_{31}&d_{32}&1\\ \vdots&\vdots&\vdots&\vdots&\ddots\end{bmatrix}, such that there exists two formal power series f\in(Z[[x]]\setminus xZ[[x]]) and g\in\mathcal{J}(Z) satisfying that, for every j\in\mathbb{N} , the generating function of the column [0,\dots,0,1,d_{j+1,j},\ldots]^{T} is f\cdot g^{j} , that is, f\cdot g^{j}=x^{j}+d_{j+1,j}x^{j+1}+d_{j+2,j}x^{j+2}+\ldots In this case, we denote the corresponding matrix by R(f,g) . We define the Riordan group \mathcal{R}^{\prime}(Z) as the set of all Riordan matrices endowed with the multiplication of matrices. Notational remark. If Z is a field \mathbb{K} of characteristic 0, we do not need ones in the main diagonal for an infinite lower triangular matrix with entries in Z to have an inverse. This leads to a different definition of the concept of Riordan matrix, which is the one in current use [ 2 , 3 , 4 , 8 , 9 , 10 ] , and it is also included herein in Definition 5 . The group of these alternative Riordan matrices is denoted by \mathcal{R}(\mathbb{K}) . The first derivative subgroup of \mathcal{R}(\mathbb{K}) is what we have defined here as \mathcal{R}^{\prime}(\mathbb{K}) [ 8 ] . This has been the motivation for our notation. Let us also define \mathcal{R}_{m}^{\prime}(Z) as the matrix group consisting in those (m+1)\times(m+1) matrices that are a principal submatrix of a Riordan matrix. We use the notation R_{m}(f,g) for the matrix obtained as a principal submatrix of R(f,g) . Let us remark that one same element in \mathcal{R}_{m}^{\prime}(Z) can be obtained as a submatrix of infinitely many different Riordan matrices. According to the definition of a Riordan matrix, the initial row and column of a matrix or submatrix begin indexing at j=0 . 1.2 Roots in groups Let G be a group (we will use multiplicative notation) and let \omega,g\in G . For some positive integer n\in\mathbb{N} , we say that \omega is a root of order n of g if \omega^{n}=g . A root of order 2 of g is usually referred just as a root of g . If the operation in G is the composition, as is the case of the group \mathcal{J}(Z) , then the roots are sometimes called iterative roots or fractional iterates . In this context, the notation \omega^{[n]} is more common for the iterated composition than \omega^{n} . In Analysis, the study of iterative roots, in some groups of functions (typically continuous) endowed with the composition, is of great interest. This has also motivated the study of iterative roots in \mathcal{J}(Z) , usually for the cases Z=\mathbb{R},\mathbb{C} . The reader may find more information in the survey [ 5 ] and in the references therein. In order to characterise a group by the number of roots that an arbitrary element can have, we present the following definitions, which can be found in [ 1 ] : Definition 3 . We say that: G is a group with unique root extraction (URE-group) if, for any g\in G and for any n\in\mathbb{N} , g has at most one root , that is, the equation \omega^{n}=g has at most one solution \omega\in G , for each g\in G ; G is an algebraically complete group or divisible group if the equation \omega^{n}=g has solutions for every g\in G and for every n\in\mathbb{N} . For instance, the group \mathcal{J}(\mathbb{Z}) is an URE-group, whereas the groups \mathcal{J}(\mathbb{Q}) , \mathcal{J}(\mathbb{R}) are algebraically complete (proposition 5.3 in [

Read the complete text