Total Positivity of Analytic Bases through Symmetric Functions

P. Díaz, E. Mainar · 2026

Read Total Positivity of Analytic Bases through Symmetric Functions by P. Díaz, E. Mainar (2026) 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-2601-19998.txt.

Introduction

P. Díaz and E. Mainar study when families of functions form a totally positive system. Total positivity is a strong regularity property: certain matrices built from the functions have all minors of the right sign. Such systems are prized in approximation and design because they behave predictably. The authors examine analytic bases, families built from analytic functions.

Their route goes through bidiagonal factorisation of collocation matrices. They derive explicit formulas for the initial minors using Schur functions, classical objects from algebra. The structure of those formulas yields sufficient conditions for total positivity of generic analytic systems. Knowing a basis is totally positive guarantees stable interpolation. The work links symmetric function theory to practical approximation. The tests are strict by design. One bad sign and the whole set fails. That is why the checks mean so much here.

Catalog record

CREATORP. Díaz, E. Mainar
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-19998

Browse the full library · About the Tower of Babel

Full text

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

Total Positivity of Analytic Bases through Symmetric Functions Thanks: This work was partially supported by Spanish research grants PID2022-138569NB-I00 (MCI/AEI) and RED2022-134176-T (MCI/AEI) and by Gobierno de Aragón (E41 _ \_ 23R, S60 _ \_ 23R). P. Díaz, E. Mainar Abstract This paper studies the bidiagonal factorization of the collocation matrices of analytic bases using symmetric functions. Explicit formulas for their initial minors are derived in terms of Schur functions. The structure of these formulas permits establishing sufficient conditions for the total positivity of generic systems of analytic functions. In addition, they have been found to lead to generalizations of the Cauchy identity for certain families of functions. Mathematics Subject Classification: 15A23, 05E05, 65F40 ∎ 1 Introduction A class of matrices that has been extensively analyzed during the last decades are those with all their minors being nonnegative—these are known as totally positive (TP) matrices. The set of TP matrices possesses rich mathematical properties and numerous applications, attracting significant interest across various fields of mathematics, including approximation theory, combinatorics, computer-aided geometric design, and economics (cf. Ando ; FJ11 ; Gasca1 ; Gasca2 ; Gasca3 ). An important property of TP matrices is their closure under multiplication, meaning that the product of two TP matrices is also TP (see Theorem 3.1 of Ando ). This property opened the possibility of factorizing TP matrices into products of simpler TP factors, which is a topic widely studied in numerical linear algebra and computational mathematics, providing computational advantages, particularly in numerical algorithms where efficiency and stability are key concerns Marco3 ; dop1 ; dop2 ; Koev1 ; Koev2 ; Koev4 ; Demmel1 ; Demmel2 . As shown in Gasca2 ; Gasca3 , any nonsingular TP matrix A\in\mathbb{R}^{n\times n} admits a bidiagonal factorization of the form A=F_{n-1}\cdots F_{1}DG_{1}\cdots G_{n-1}, (1) where F_{i},G_{i}\in\mathbb{R}^{n\times n} are TP lower and upper triangular bidiagonal matrices, and D is a nonsingular TP diagonal matrix. This factorization enables the highly accurate solution of various numerical linear algebra problems Martinez2024 . When certain conditions are met, machine precision is achieved even for extremely ill-conditioned problems and large dimensions—a property known as High Relative Accuracy (HRA). It turns out that all the parameters of the factorization ( 1 ) can be explicitly determined from the initial minors of A Gasca1 ; Gasca2 , \det A[i-j+1,\dots,i\mid 1,\dots,j],\quad\det A^{T}[i-j+1,\dots,i\mid 1,\dots,j],\quad\quad 0<j\leq i\leq n. (2) Moreover, it is well known that a matrix A is TP if and only if all the initial minors in ( 2 ) are nonnegative Gasca1 ; Gasca3 . Collocation matrices encode the values of a given system of functions at a sequence of parameters within the domain (see ( 3 )). From an algebraic point of view, TP colocation matrices play a fundamental role in discretization, interpolation, and various other applications. A crucial observation is that any minor of a collocation matrix is an antisymmetric function of the subset of nodes involved. This property extends to their initial minors ( 2 ). Thus, for TP collocation matrices, all entries in the factorization ( 1 ) inherit a natural symmetry, a fact that lies at the core of an interesting connection between the bases whose collocation matrices are TP and symmetric functions. The relationship above was first highlighted in YHZ in the particular context of Cauchy-Vandermonde matrices, where it was noted that the entries of ( 1 ) can be expressed in terms of Schur polynomials. Later, a systematic research line was initiated in DMR , where explicit formulas for the initial minors of collocation matrices of arbitrary polynomial bases were derived in terms of Schur functions. These results led to novel and practical criteria for determining the total positivity of polynomial bases. Subsequent applications of these findings were applied to Wronskian and Gramian matrices in DMR2 ; DMR3 , offering new insights and criteria for their total positivity. In this paper, we extend the approach developed in DMR ; DMR2 ; DMR3 to the more general framework of bases of analytic functions. From the theoretical point of view, the main result is presented in Theorem 3.1 , which establishes an expansion of the initial minors of the collocation matrix of a generic system of analytic functions in terms of Schur polynomials. From a practical perspective, Theorem 3.1 also provides effective criteria for determining total positivity when applied to specific families of functions. A particularly interesting case arises when considering any analytic function f and the system formed by the functions f_{i}(x)=f(a_{i}x) for distinct parameters a_{i} , i=1,\dots,n . In this setting, the achieved expansions yield an infinite family of identities including, as a remarkable particular case, the celebrated Cauchy identity ( 15 ). The Cauchy identity establishes fundamental links between determinants, symmetric polynomials, and partition functions, with deep connections to representation theory and algebraic combinatorics, inspiring various generalizations across mathematical structures Mac ; Stanley ; Bump ; ortho ; double ; Hall ; quantum . Therefore, the detailed derivation of this type of identity, along with illustrative examples, deserves a dedicated section in this article. The paper is organized as follows. In Section 2 , we introduce notation and key concepts on total positivity and symmetric functions, which play a fundamental role in the paper. Section 3 derives explicit formulas in terms of Schur polynomials for the initial minors ( 2 ) of collocation matrices of systems of analytic functions. As a consequence, a sufficient condition to guarantee the total positivity property of these bases is derived. Furthermore, in Section 4 , a family of Cauchy-type identities, extending the classical Cauchy identity, is derived. Finally, in Section 5 , we summarize our main findings and suggest directions for future research. 2 Total positivity and symmetric functions In the sequel, we shall use the following matrix notation. Given A\in\mathbb{R}^{n\times n} , the submatrix formed by rows i_{1},\ldots,i_{r} and columns j_{1},\ldots,j_{s} will be denoted by A[i_{1},\ldots,i_{r}\mid j_{1},\ldots,j_{s}]. For square submatrices, we shall use the shorthand notation A[i_{1},\ldots,i_{r}]:=A[i_{1},\ldots,i_{r}\mid i_{1},\ldots,i_{r}] . A symmetric function is a multivariate function that remains invariant under any permutation of its variables. Formally, a function f(x_{1},\ldots,x_{n}) is symmetric if f(x_{\sigma(1)},\ldots,x_{\sigma(n)})=f(x_{1},\ldots,x_{n}), for any permutation \sigma of the indices \{1,\ldots,n\} . Conversely, a function f is alternating if f(x_{\sigma(1)},\ldots,x_{\sigma(n)})=\text{sgn}(\sigma)f(x_{1},\ldots,x_{n}), where \text{sgn}(\sigma) denotes the signature of the permutation, taking values +1 or -1 depending on whether \sigma is even or odd. A matrix is called totally positive (TP) if all its minors are nonnegative and strictly totally positive (STP) if all its minors are strictly positive. Given a system \mathbf{f}=(f_{1},\ldots,f_{n}) of functions defined on I\subseteq\mathbb{R} and a sequence \mathbf{x}=\{x_{1},\ldots,x_{n}\} of parameters in I satisfying x_{1}<\cdots<x_{n} , the collocation matrix of \mathbf{f} at \mathbf{x} is defined as M_{\mathbf{f},\mathbf{x}}:=\left(f_{j}(x_{i})\right)_{i,j=1}^{n}. (3) The basis \mathbf{f} is called TP or STP if M_{\mathbf{f},\mathbf{x}} is TP or STP, respectively, for all choices of \mathbf{x} . When considering parameters \mathbf{x}=\{x_{1},\ldots,x_{n}\} and the collocation matrix A=M_{\mathbf{f},\mathbf{x}} in ( 3 ), the minors of A are alternant (or antisymmetric) functions of the parameters. Thus, although the collocation matrix M_{\mathbf{f},\mathbf{x}} itself does not inherently possess any symmetry in its nodes, its initial minors do. This fundamental observation underpins the intriguing relationship between symmetric functions and TP bases, a connection that has motivated the present study, as well as the preceding works DMR

Read the complete text