Searcharxiv⌕ Search

arXiv subjects

Eiichi Bannai

Publications and source records attributed to Eiichi Bannai.

34 records · Page 2Linked to original sources
arXiv2015

On relative $t$-designs in polynomial association schemes

Motivated by the similarities between the theory of spherical $t$-designs and that of $t$-designs in $Q$-polynomial association schemes, we study two versions of relative $t$-designs, the counterparts of Euclidean $t$-designs for $P$- and/or $Q$-polynomial association schemes. We develop the theory based on the Terwilliger algebra, which is a noncommutative associative semisimple $\mathbb{C}$-algebra associated with each vertex of an association scheme. We compute explicitly the Fisher type lower bounds on the sizes of relative $t$-designs, assuming that certain irreducible modules behave nicely. The two versions of relative $t$-designs turn out to be equivalent in the case of the Hamming schemes. From this point of view, we establish a new algebraic characterization of the Hamming schemes.

math.CO↗
arXiv2015

More on spherical designs of harmonic index $t$

A finite subset $Y$ on the unit sphere $S^{n-1} \subseteq \mathbb{R}^n$ is called a spherical design of harmonic index $t$, if the following condition is satisfied: $\sum_{\mathbf{x}\in Y}f(\mathbf{x})=0$ for all real homogeneous harmonic polynomials $f(x_1,\ldots,x_n)$ of degree $t$. Also, for a subset $T$ of $\mathbb{N} = \{1,2,\cdots \}$, a finite subset $Y\subset S^{n-1}$ is called a spherical design of harmonic index $T,$ if $\sum_{\mathbf{x}\in Y}f(\mathbf{x})=0$ is satisfied for all real homogeneous harmonic polynomials $f(x_1,\ldots,x_n)$ of degree $k$ with $k\in T$. In the present paper we first study Fisher type lower bounds for the sizes of spherical designs of harmonic index $t$ (or for harmonic index $T$). We also study 'tight' spherical designs of harmonic index $t$ or index $T$. Here 'tight' means that the size of $Y$ attains the lower bound for this Fisher type inequality. The classification problem of tight spherical designs of harmonic index $t$ was started by Bannai-Okuda-Tagami (2015), and the case $t = 4$ was completed by Okuda-Yu (2015+). In this paper we show the classification (non-existence) of tight spherical designs of harmonic index 6 and 8, as well as the asymptotic non-existence of tight spherical designs of harmonic index $2e$ for general $e\geq 3$. We also study the existence problem for tight spherical designs of harmonic index $T$ for some $T$, in particular, including index $T = \{8,4\}$. We use (i) the linear programming method by Delsarte, (ii) the detailed information on the locations of the zeros as well as the local minimum values of Gegenbauer polynomials, (iii) the generalization by Hiroshi Nozaki of the Larman-Rogers-Seidel theorem on $2$-distance sets to $s$-distance sets, (iv) the theory of elliptic diophantine equations, and (v) the semidefinite programming method of eliminating some $2$-angular line systems for small dimensions.

math.CO↗
arXiv2014

Mackey's theory of $τ$-conjugate representations for finite groups. APPENDIX: On Some Gelfand Pairs and Commutative Association Schemes

The aim of the present paper is to expose two contributions of Mackey, together with a more recent result of Kawanaka and Matsuyama, generalized by Bump and Ginzburg, on the representation theory of a finite group equipped with an involutory anti-automorphism (e.g. the anti-automorphism $g\mapsto g^{-1}$). Mackey's first contribution is a detailed version of the so-called Gelfand criterion for weakly symmetric Gelfand pairs. Mackey's second contribution is a characterization of simply reducible groups (a notion introduced by Wigner). The other result is a twisted version of the Frobenius-Schur theorem, where "twisted" refers to the above-mentioned involutory anti-automorphism. APPENDIX: We consider a special condition related to Gelfand pairs. Namely, we call a finite group $G$ and its automorphism $σ$ satisfy Condition ($\bigstar$) if the following condition is satisfied: if for $x,y\in G$, $x\cdot x^{-σ}$ and $y\cdot y^{-σ}$ are conjugate in $G$, then they are conjugate in $K=C_G(σ)$. We study the meanings of this condition, as well as showing many examples of $G$ and $σ$ which do (or do not) satisfy Condition ($\bigstar$).

math.RT↗
arXiv2013

Spherical designs of harmonic index t

Spherical $t$-design is a finite subset on sphere such that, for any polynomial of degree at most $t$, the average value of the integral on sphere can be replaced by the average value at the finite subset. It is well-known that an equivalent condition of spherical design is given in terms of harmonic polynomials. In this paper, we define a spherical design of harmonic index $t$ from the viewpoint of this equivalent condition, and we give its construction and a Fisher type lower bound on the cardinality. Also we investigate whether there is a spherical design of harmonic index attaining the bound.

math.MG↗
arXiv2013

On the existence of tight relative 2-designs on binary Hamming association schemes

It is known that there is a close analogy between "Euclidean t-designs vs. spherical t-designs" and "Relative t-designs in binary Hamming association schemes vs. combinatorial t-designs". In this paper, we want to prove how much we can develop a similar theory in the latter situation, imitating the theory in the former one. We first prove that the weight function is constant on each shell for tight relative t-designs on p shells on a wide class of Q-polynomial association schemes, including Hamming association schemes. In the theory of Euclidean t-designs on 2 concentric spheres (shells), it is known that the structure of coherent configurations is naturally attached. However, it seems difficult to prove this claim in a general context. In the case of tight 2-designs in combinatorial 2-designs, there are great many tight 2-designs, i.e., symmetric 2-designs, while there are very few tight 2e-designs for e no less than 2. So, as a starting point, we concentrate our study to the existence problem of tight relative 2-designs, in particular on 2 shells, in binary Hamming association schemes H(n,2). We prove that every tight relative 2-designs on 2 shells in H(n,2) has the structure of coherent configuration. We determined all the possible parameters of coherent configurations attached to such tight relative 2-designs for n at most 30. Moreover for each of them we determined whether there exists such a tight relative 2-design or not, either by constructing them from symmetric 2-designs or Hadamard matrices, or theoretically showing the non-existence. In particular, we show that for n congruent to 6 (mod 8), there exist such tight relative 2-designs whose weight functions are not constant. These are the first examples of those with non-constant weight.

math.CO↗
arXiv2012

Maximal $m$-distance sets containing the representation of the Johnson graph $J(n, m)$

We classify the maximal $m$-distance sets in $\mathbb{R}^{n-1}$ which contain the representation of the Johnson graph $J(n, m)$ for $m = 2, 3$. Furthermore, we determine the necessary and sufficient condition for $n$ and $m$ such that the representation of the Johnson graph $J(n, m)$ is not maximal as an $m$-distance set. Also, we classify the maximal two-distance sets in $\mathbb{R}^{n-1}$ which contain the representation of $J(n - 1, 2)$.

math.CO↗
arXiv2011

On a property of 2-dimensional integral Euclidean lattices

Let $L$ be any integral lattice in the 2-dimensional Euclidean space. Generalizing the earlier works of Hiroshi Maehara and others, we prove that for every integer $n>0$, there is a circle in the plane $\mathbb{R}^{2}$ that passes through exactly $n$ points of $L$.

math.CO↗
arXiv2010

Tight 9-designs on two concentric spheres

The main purpose of this paper is to show the nonexistence of tight Euclidean 9-designs on 2 concentric spheres in $\mathbb R^n$ if $n\geq 3.$ This in turn implies the nonexistence of minimum cubature formulas of degree 9 (in the sense of Cools and Schmid) for any spherically symmetric integrals in $\mathbb R^n$ if $n\geq 3.$

math.CO↗
arXiv2010

Toy models for D. H. Lehmer's conjecture II

In the previous paper, we studied the "Toy models for D. H. Lehmer's conjecture". Namely, we showed that the m-th Fourier coefficient of the weighted theta series of the $\mathbb{Z}^2$-lattice and the $A_{2}$-lattice does not vanish, when the shell of norm $m$ of those lattices is not the empty set. In other words, the spherical 4 (resp. 6)-design does not exist among the nonempty shells in the $\mathbb{Z}^2$-lattice (resp. $A_{2}$-lattice). This paper is the sequel to the previous paper. We take 2-dimensional lattices associated to the algebraic integers of imaginary quadratic fields whose class number is either 1 or 2, except for $\mathbb{Q}(\sqrt{-1})$ and $\mathbb{Q}(\sqrt{-3})$, then, show that the $m$-th Fourier coefficient of the weighted theta series of those lattices does not vanish, when the shell of norm $m$ of those lattices is not the empty set. Equivalently, we show that the corresponding spherical 2-design does not exist among the nonempty shells in those lattices.

math.NT↗
arXiv2010

An elementary approach to toy models for D. H. Lehmer's conjecture

In 1947, Lehmer conjectured that the Ramanujan's tau function $τ(m)$ never vanishes for all positive integers $m$, where $τ(m)$ is the $m$-th Fourier coefficient of the cusp form $Δ_{24}$ of weight 12. The theory of spherical $t$-design is closely related to Lehmer's conjecture because it is shown, by Venkov, de la Harpe, and Pache, that $τ(m)=0$ is equivalent to the fact that the shell of norm $2m$ of the $E_{8}$-lattice is a spherical 8-design. So, Lehmer's conjecture is reformulated in terms of spherical $t$-design. Lehmer's conjecture is difficult to prove, and still remains open. However, Bannai-Miezaki showed that none of the nonempty shells of the integer lattice $\ZZ^2$ in $\RR^2$ is a spherical 4-design, and that none of the nonempty shells of the hexagonal lattice $A_2$ is a spherical 6-design. Moreover, none of the nonempty shells of the integer lattices associated to the algebraic integers of imaginary quadratic fields whose class number is either 1 or 2, except for $\QQ(\sqrt{-1})$ and $\QQ(\sqrt{-3})$ is a spherical 2-design. In the proof, the theory of modular forms played an important role. Recently, Yudin found an elementary proof for the case of $\ZZ^{2}$-lattice which does not use the theory of modular forms but uses the recent results of Calcut. In this paper, we give the elementary (i.e., modular form free) proof and discuss the relation between Calcut's results and the theory of imaginary quadratic fields.

math.CO↗
arXiv2009

Euclidean designs and coherent configurations

The concept of spherical $t$-design, which is a finite subset of the unit sphere, was introduced by Delsarte-Goethals-Seidel (1977). The concept of Euclidean $t$-design, which is a two step generalization of spherical design in the sense that it is a finite weighted subset of Euclidean space, by Neumaier-Seidel (1988). We first review these two concepts, as well as the concept of tight $t$-design, i.e., the one whose cardinality reaches the natural lower bound. We are interested in $t$-designs (spherical or Euclidean) which are either tight or close to tight. As is well known by Delsarte-Goethals-Seidel (1977), in the study of spherical $t$-designs and in particular of those which are either tight or close to tight, association schemes play important roles. The main purpose of this paper is to show that in the study of Euclidean $t$-designs and in particular of those which are either tight or close to tight, coherent configurations play important roles. Here, coherent configuration is a purely combinatorial concept defined by D. G. Higman, and is obtained by axiomatizing the properties of general, not necessarily transitive, permutation groups, in the same way as association scheme was obtained by axiomatizing the properties of transitive permutation groups. The main purpose of this paper is to prove that Euclidean $t$-designs satisfying certain conditions give the structure of coherent configurations. Moreover we study the classification problems of Euclidean 4-designs on two concentric spheres with certain additional conditions.

math.CO↗
arXiv2009

Toy models for D. H. Lehmer's conjecture

In 1947, Lehmer conjectured that the Ramanujan $τ$-function $τ(m)$ never vanishes for all positive integers $m$, where the $τ(m)$ are the Fourier coefficients of the cusp form $Δ_{24}$ of weight 12. Lehmer verified the conjecture in 1947 for $m<214928639999$. In 1973, Serre verified up to $m<10^{15}$, and in 1999, Jordan and Kelly for $m<22689242781695999$. The theory of spherical $t$-design, and in particular those which are the shells of Euclidean lattices, is closely related to the theory of modular forms, as first shown by Venkov in 1984. In particular, Ramanujan's $τ$-function gives the coefficients of a weighted theta series of the $E_{8}$-lattice. It is shown, by Venkov, de la Harpe, and Pache, that $τ(m)=0$ is equivalent to the fact that the shell of norm $2m$ of the $E_{8}$-lattice is an 8-design. So, Lehmer's conjecture is reformulated in terms of spherical $t$-design. Lehmer's conjecture is difficult to prove, and still remains open. In this paper, we consider toy models of Lehmer's conjecture. Namely, we show that the $m$-th Fourier coefficient of the weighted theta series of the $\mathbb{Z}^2$-lattice and the $A_{2}$-lattice does not vanish, when the shell of norm $m$ of those lattices is not the empty set. In other words, the spherical 5 (resp. 7)-design does not exist among the shells in the $\mathbb{Z}^2$-lattice (resp. $A_{2}$-lattice).

math.NT↗
arXiv2009

The mathematics of Donald Gordon Higman

This is about the mathematics and life of Donald Gordon Higman, 1928-2006. He did important work in representation theory of groups and algebras and in algebraic combinatorics. Charles C. Sims and Donald Higman discovered and constructed one of the sporadic simple groups.

math.HO↗
arXiv2008

Association schemes related to universally optimal configurations, Kerdock codes and extremal Euclidean line-sets

H. Cohn et. al. proposed an association scheme of 64 points in R^{14} which is conjectured to be a universally optimal code. We show that this scheme has a generalization in terms of Kerdock codes, as well as in terms of maximal real mutually unbiased bases. These schemes also related to extremal line-sets in Euclidean spaces and Barnes-Wall lattices. D. de Caen and E. R. van Dam constructed two infinite series of formally dual 3-class association schemes. We explain this formal duality by constructing two dual abelian schemes related to quaternary linear Kerdock and Preparata codes.

math.CO↗