arXiv++ Combinatorics

We find a new approach to computing the remainder of a polynomial modulo $x^n-1$; such a computation is called modular enumeration. Given a polynomial with coefficients from a commutative $\mathbb{Q}$-algebra, our first main result constructs the remainder simply from the coefficients of residues of the polynomial modulo $Φ_d(x)$ for each $d\mid n$. Since such residues can often be found to have nice values, this simplifies a number of modular enumeration problems; indeed in some cases, such residues are already known while the related modular enumeration problem has remained unsolved. We list six such cases which our technique makes easy to solve. Our second main result is a formula for the unique polynomial $a$ such that $a \equiv f \mod Φ_n(x)$ and $a\equiv 0 \mod x^d-1$ for each proper divisor $d$ of $n$. We find a formula for remainders of $q$-multinomial coefficients and for remainders of $q$-Catalan numbers modulo $q^n-1$, reducing each problem to a finite number of cases for any fixed $n$. In the prior case, we solve an open problem posed by Hartke and Radcliffe. In considering $q$-Catalan numbers modulo $q^n-1$, we discover a cyclic group operation on certain lattice paths which behaves predictably with regard to major index. We also make progress on a problem in modular enumeration on subset sums posed by Kitchloo and Pachter.

Consider a system of \(n\) players in which each initially starts on a different team. At each time step, we select an individual winner and an individual loser randomly and the loser joins the winner's team. The resulting Markov chain and stochastic matrix clearly have one absorbing state, in which all players are on the same team, but the combinatorics along the way are surprisingly elegant. The expected number of time steps until each team is eliminated is a ratio of binomial coefficients. When a team is eliminated, the probabilities that the players are configured in various partitions of \(n\) into \(t\) teams are given by multinomial coefficients. The expected value of the time to absorbtion is \((n-1)^2\) steps. The results depend on elementary combinatorics, linear algebra, and the theory of Markov chains.

We present a bijective algorithm with which an arbitrary permutation decomposes canonically into elementary blocks which we call families, which are sets with a specified number of ascents and descents. We show that families, arranged in an arbitrary order in a sequence, are in bijection with permutations. The permutation decomposes canonically, by inserting parentheses, into a tree having as nodes a class of permutations which we call primitive. Primitive permutations can be assembled from very simple data. The data for the trees into which a permutation decomposes can be written in a form similar to the decimal classification of a library. We axiomatize that data. It has a structure very different from the permutation which it encodes, with shuffles and pairings instead of reorderings. These structures are similar to the fundamental processes in quantum field theory. Our main bijective structure algorithm gives explicit, additive multinomial formulae for the number of permutations with given sets of elements under and over the diagonal, or with given ascent and descent values. The multinomial expressions obtained this way give a new class of bicolored set statistics, between set partitions and set compositions, called shifted multinomials. These provide for the first time additive multinomial expressions for Eulerian numbers and derangements, as part of a sequence of new combinatorial objects. These multinomial expressions satisfy inductive relations involving only immediate neighbors, similar to the relations satisfied by the Eulerian numbers.

We prove an inverse relation and a family of convolution formulas involving partial Bell polynomials. Known and some presumably new combinatorial identities of convolution type are discussed. Our approach relies on an interesting multinomial formula for the binomial coefficients. The inverse relation is deduced from a parametrization of suitable identities that facilitate dealing with compositions of Bell polynomials.

This note examines linear combinations of multi-indexed sequences and derives the multivariate generating function of such a linear combination in terms of the original sequence's m.g.f. Applications include finding distributions and moments of non-negative discrete random variables conditioned on non-negative linear combinations of the original variables. Examples include independent Poisson r.v.'s and a $d$-variate multinomial distribution.

In this note we give a numerical expression for the bandwidth $bw(P_{n}^{d})$ of the $d$-product of a path with $n$ edges, $P_{n}^{d}$. We prove that this bandwidth is given by the sum of certain multinomial coefficients. We also show that $bw(P_{n}^{d})$ is bounded above and below by the largest coefficient in the expansion of $(1+x+...+x^{n})^{k}$, with $k\in{d,d+1}$. Moreover, we compare the asymptotic behavior of $bw(P_{n}^{d})$ with the bandwidth of the labeling obtained by ordering the vertices of $P_{n}^{d}$ in lexicographic order.

We give a proof of a conjecture of A. Lacasse in his doctoral thesis which has applications in machine learning algorithms. The proof relies on some interesting binomial sums identities introduced by Abel (1839), and on their generalization to the multinomial case by Hurwitz (1902).

In this paper we obtained an original integer sequence based on the properties of the multinomial coefficient. We investigated a property of the sequence that shows connection with a primality testing. For any prime n the n-th term in the sequence is less by 1 than the number of partitions of n. We hypothesize the existence of an asymptotic algorithm of primality testing.

Recently, a novel method based on coding partitions [1]-[4] has been used to derive power series expansions to previously intractable problems. In this method the coefficients at $k$ are determined by summing the contributions made by each partition whose elements sum to $k$. These contributions are found by assigning values to each element and multiplying by an appropriate multinomial factor. This work presents a theoretical framework for the partition method for a power series expansion. To overcome the complexity due to the contributions, a programming methodology is created allowing more general problems to be studied than envisaged originally. The methodology uses the bi-variate recursive central partition (BRCP) algorithm, which is based on a tree-diagram approach to scanning partitions. Its main advantage is that partitions are generated in the multiplicity representation. During the development of the theoretical framework, scanning over partitions was seen as a discrete operation with an operator $L_{P,k}[ \cdot]$, whose summand depends on the coefficients of the two series when the original function is written as a pseudo-composite function. Simple modifications result in programs for other operators of specific types of partitions such as: (1) only odd or even elements, (2) a fixed number of elements, (3) discrete elements, (4) specific elements and (5) those restricted by element size. Another modification generates conjugate partitions by transposing Ferrers diagrams. The operator approach is then applied to the generating functions for both discrete and standard partitions. The main generalisation introduces a parameter $ω$, whose powers give the number of elements in the partitions while the coefficients become polynomials in $ω$. Finally, power series expansions for more advanced infinite products are derived, culminating in Heine's multi-parameter product.

Lascoux, Leclerc and Thibon\cite{LLT} introduced a family of symmetric polynomials, called LLT polynomials. We prove a $q$-multinomial expansion of the coefficients of LLT polynomials in the case where $ \boldsymbolμ = \underbrace{(μ,...,μ)}_{n}$ and define a $q$-analog of a sum of the plethysm multiplicities.

In the work of Varchenko, Zagier, Thibon, and Reiner, Saliola, Welker, linear algebraic properties of the multiplication map on the group algebra of the group algebra element are studied, which is the sum over all permutations weighted by q^{inv}, q^{maj}, inv. Here q is a variable, and inv and maj are the classical statistics inversion and major index. We define a multinomial descent statistic des_X and a multinomial inversion statistic inv_X. These new defined statistics are the multinomial expressions of the classical statistics descent des and inversion. We determine the spectrum and the multiplicity of each element of the spectrum of the analogously defined multiplication map on the group algebra for both des_X and inv_X. As corollaries we deduce the spectrum and the multiplicity of each element of the spectrum of the defined multiplication map on the group algebra for the statistics des, maj and inv.

We derive explicit expressions for the generating series of the fundamental solutions of the $A_r$ quantum $Q$-system of Ref. [P. Di Francesco and R. Kedem, arXiv:1006.4774 [math-ph]], expressed in terms of any admissible initial data. These involve products of quantum multinomial coefficients, coded by the initial data structure.

A new algorithm for computing the multivariate Faà di Bruno's formula is provided. We use a symbolic approach based on the classical umbral calculus that turns the computation of the multivariate Faà di Bruno's formula into a suitable multinomial expansion. We propose a MAPLE procedure whose computational times are faster compared with the ones existing in the literature. Some illustrative applications are also provided.

We give a recursion for the multivariate Rogers-Szegö polynomials, along with another recursive functional equation, and apply them to compute special values. We also consider the sum of all $q$-multinomial coefficients of some fixed degree and length, and give a recursion for this sum which follows from the recursion of the multivariate Rogers-Szegö polynomials, and generalizes the recursion for the Galois numbers. The sum of all $q$-multinomial coefficients of degree $n$ and length $m$ is the number of flags of length $m-1$ of subspaces of an $n$-dimensional vector space over a field with $q$ elements. We give a combinatorial proof of the recursion for this sum of $q$-multinomial coefficients in terms of finite vector spaces.

In this expository article, we discuss the relation between the Gaussian binomial and multinomial coefficients and ordinary binomial and multinomial coefficients from a combinatorial viewpoint, based on expositions by Butler, Knuth and Stanley.

Motivated by the recent work of Chu [Electron. J. Combin. 17 (2010), #N24], we give simple proofs of Jensen's identity $$ \sum_{k=0}^{n}{x+kz\choose k}{y-kz\choose n-k} =\sum_{k=0}^{n}{x+y-k\choose n-k}z^k, $$ and Chu's and Mohanty-Handa's generalizations of Jensen's identity. We also give a quite simple proof of an equivalent form of Graham-Knuth-Patashnik's identity $$ \sum_{k\geq 0}{m+r\choose m-n-k}{n+k\choose n}x^{m-n-k}y^k =\sum_{k\geq 0}{-r\choose m-n-k}{n+k\choose n}(-x)^{m-n-k}(x+y)^k, $$ which was rediscovered, respectively, by Sun in 2003 and Munarini in 2005. Finally we give a multinomial coefficient generalization of this identity and raise two open problems.

Let $m$ and $n>0$ be integers. Suppose that $p$ is a prime dividing $m-4$ but not dividing $m$. We show that $ν_p(\sum_{k=0}^{n-1}\frac{\binom{2k}k}{m^k})$ and $ν_p(\sum_{k=0}^{n-1}\binom{n-1}{k}(-1)^k\frac{\binom{2k}k}{m^k})$ are at least $ν_p(n)$, where $ν_p(x)$ denotes the $p$-adic valuation of $x$. Furthermore, if $p>3$ then $$n^{-1}\sum_{k=0}^{n-1}\frac{\bi{2k}k}{m^k}=\frac{\binom{2n-1}{n-1}}{4^{n-1}} (mod p^{ν_p(m-4)})$$ and $$n^{-1}\sum_{k=0}^{n-1}\binom{n-1}{k}(-1)^k\frac{\binom{2k}k}{m^k}=\frac{C_{n-1}}{4^{n-1}} (mod p^{ν_p(m-4)}),$$ where $C_k$ denotes the Catalan number $\binom{2k}{k}/(k+1)$. This implies several conjectures of Guo and Zeng [GZ]. We also raise two conjectures, and prove that $n>1$ is a prime if and only if $$\sum_{k=0}^{n-1}multinomial{(n-1)k}{k,...,k}=0 (mod n),$$ where $multinomial{k_1+...+k_{n-1}}{k_1,...,k_{n-1}}$ denotes the multinomial coefficient $(k_1+...+k_{n-1})!/(k_1!... k_{n-1}!)$.

The Mahonian statistic is the number of inversions in a permutation of a multiset with $a_i$ elements of type $i$, $1\le i\le m$. The counting function for this statistic is the $q$ analog of the multinomial coefficient $\binom{a_1+...+a_m}{a_1,... a_m}$, and the probability generating function is the normalization of the latter. We give two proofs that the distribution is asymptotically normal. The first is {\it computer-assisted}, based on the method of moments. The Maple package {\tt MahonianStat}, available from the webpage of this article, can be used by the reader to perform experiments and calculations. Our second proof uses characteristic functions. We then take up the study of a local limit theorem to accompany our central limit theorem. Here our result is less general, and we must be content with a conjecture about further work. Our local limit theorem permits us to conclude that the coeffiecients of the $q$-multinomial are log-concave, provided one stays near the center (where the largest coefficients reside.)

Generalizing a classical problem in enumerative combinatorics, Mansour and Sun counted the number of subsets of $\Z_n$ without certain separations. Chen, Wang, and Zhang then studied the problem of partitioning $\Z_n$ into arithmetical progressions of a given type under some technical conditions. In this paper, we improve on their main theorems by applying a convolution formula for cyclic multinomial coefficients due to Raney-Mohanty.

Let $n=\prod_p p^{ν_p(n)}$ denote the canonical factorization of $n\in \N$. The binomial convolution of arithmetical functions $f$ and $g$ is defined as $(f\circ g)(n)=\sum_{d\mid n} (\prod_p \binom{ν_p(n)}{ν_p(d)}) f(d)g(n/d),$ where $\binom{a}{b}$ is the binomial coefficient. We provide properties of the binomial convolution. We study the $\C$-algebra $({\cal A},+,\circ,\C)$, characterizations of completely multiplicative functions, Selberg multiplicative functions, exponential Dirichlet series, exponential generating functions and a generalized binomial convolution leading to various Möbius-type inversion formulas. Throughout the paper we compare our results with those of the Dirichlet convolution *. Our main result is that $({\cal A},+,\circ,\C)$ is isomorphic to $({\cal A},+,*,\C)$. We also obtain a "multiplicative" version of the multinomial theorem.