Publications (96)59.38 Total impact
 [Show abstract] [Hide abstract]
ABSTRACT: Suppose that t 2 is an integer, and randomly label t graphs with the integers 1 : : : n. We give sufficient conditions for the number of edges common to all t of the labelings to be asymptotically Poisson as n ! 1. We show by example that our theorem is, in a sense, best possible. For Gn a sequence of graphs of bounded degree, each having at most n vertices, Tomescu [7] has shown that the number of spanning trees of Kn having k edges in common with Gn is asymptotically e Gamma2s=n (2s=n) k =k! Theta n nGamma2 , where s = s(n) is the number of edges in Gn . As an application of our Poissonintersection theorem, we extend this result to the case in which maximum degree is only restricted to be O(n log log n= log n). We give an inversion theorem for falling moments, which we use to prove our Poissonintersection theorem. AMSMOS Subject Classification (1990): 05C30; Secondary: 05A16, 05C05, 60C05 the electronic journal of combinatorics 6 (1999), #R36 2 1. Introduction and St...11/1999;  [Show abstract] [Hide abstract]
ABSTRACT: Suppose that t # 2 is an integer, and randomly label t graphs with the integers 1 ...n. We give sufficient conditions for the number of edges common to all t of the labelings to be asymptotically Poisson as n ##. We show by example that our theorem is, in a sense, best possible. For G n a sequence of graphs of bounded degree, each having at most n vertices, Tomescu [7] has shown that the number of spanning trees of K n having k edges in common with G n is asymptotically e 2s/n (2s/n) k /k! n n2 , where s=s(n) is the number of edges in G n .As an application of our Poissonintersection theorem, we extend this result to the case in which maximum degree is only restricted to be O(n log log n/ log n). We give an inversion theorem for falling moments, which we use to prove our Poissonintersection theorem.The electronic journal of combinatorics 10/1999; · 0.53 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: We prove various properties of C(n; q), the set of nvertex q edge labeled connected graphs. The domain of validity of the asymptotic formula of Erdos and R'enyi for jC(n; q)j is extended and the formula is seen to be the first term of an asymptotic expansion. The same is done for Wright's asymptotic formula. We study the number of edges in a random connected graph in the random edge model G n;p . For certain ranges of n and q, we determine the probability that a random edge (resp. vertex) of a random graph in C(n; q) is a bridge (resp. cut vertex). We also study the degrees of random vertices. Section 1. Introduction In this paper all graphs are labeled, and an (n; q) graph is one having n vertices and q edges. Let C(n; q) be the set of connected (n; q) graphs and let c(n; q) = jC(n; q)j. We will often speak of things chosen at random. Unless stated otherwise, this means uniformly at random. For example,"Let e be a random edge of C(n; q)" means that we choose a graph G 2 C(n; q) ...Random Structures and Algorithms 03/1999; · 1.05 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: Let D be a set of positive integers. Let m(n) be the number of n edged rooted maps on the sphere all of whose vertex degrees (or, dually, face degrees) lie in D. Using Brown's technique, we obtain the generating function for m(n) implicitly. We use it to prove that, when gcd(D) is even, m(n) C(D)n Gamma5=2 fl(D) n : It also yields known formulas for various special D. 2 Counting degree restricted maps Section 1: Introduction Let D be a set of positive integers containing some element exceeding 2, let M(x; y) = P i M i (x)y i be the generating function by edges and root face degree for rooted maps on the sphere such that each nonroot face degree lies in D and let m(n) be the number of n edged rooted maps all of whose face degrees lie in D. Define the coefficient operator with respect to y by [y k ] i P i0 f i (x)y i j = f k (x) and define [x k ] similarly. We will prove Theorem 1. There exist unique power series R 1 (x) and R 2 (x) such that R 1 = x 2 X i2D ...03/1999; 
Article: Multivariate Asymptotics for Products of Large Powers with Applications to Lagrange Inversion
[Show abstract] [Hide abstract]
ABSTRACT: An asymptotic estimate is given for the coe#cients of products of large powers of generating functions. This theorem and another local limit theorem which is useful for conditioning are applied to various combinatorial enumeration problems that involve multivariate Lagrange inversion. 1991 AMS Class. No. Primary: 41A63 Secondary: 05A16, 05C05, 41A60 ### ########### ###### ## ############# # ####### ### # 1. Introduction If f(0) #= 0 has a (possibly formal) power series expansion at 0, the equation w = xf(w) determines the power series w(x). Two forms of the Lagrange inversion formula are g n =[x n ] g(w)=[x n ] # g(x)f(x) n {(1  xf # (x)/f (x)} # (1) =(1/n)[x n ] # xg # (x)f(x) n # , (2) where [x n ] h(x) denotes the coe#cient of the monomial x n in the power series h(x). We obtained asymptotics for g n from (2) for some types of formal power series [6]. When f has a nonzero radius of convergence, various authors have studied the asymptotics of [x n ] g...The electronic journal of combinatorics 02/1999; · 0.53 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: An asymptotic estimate is given for the coefficients of products of large powers of generating functions. This theorem and another local limit theorem which is useful for conditioning are applied to various combinatorial enumeration problems that involve multivariate Lagrange inversion. 1991 AMS Class. No. Primary: 41A63 Secondary: 05A16, 05C05, 41A60 the electronic journal of combinatorics 6 (1999), #R8 2 1. Introduction If f(0) 6= 0 has a (possibly formal) power series expansion at 0, the equation w = xf(w) determines the power series w(x). Two forms of the Lagrange inversion formula are gn = [x n ] g(w) = [x n ] i g(x)f(x) n f(1 Gamma xf 0 (x)=f(x)g j (1) = (1=n)[x n ] Gamma xg 0 (x)f(x) n Delta ; (2) where [x n ] h(x) denotes the coefficient of the monomial x n in the power series h(x). We obtained asymptotics for gn from (2) for some types of formal power series [6]. When f has a nonzero radius of convergence, various authors have studied the asymp...02/1999;  [Show abstract] [Hide abstract]
ABSTRACT: Let M be a map on a surface S. The edgewidth of M is the length of a shortest noncontractible cycle of M . The facewidth (or, representativity) of M is the smallest number of intersections a noncontractible curve in S has with M . (The edgewidth and facewidth of a planar map may be defined to be infinity.) A map is an LEWembedding if its maximum face valency is less than its edgewidth. For several families of rooted maps on a given surface, we prove that there are positive constants c 1 and c 2 , depending on the family and the surface, such that 1. almost all maps with n edges have facewidth and edgewidth greater than c 1 log n and 2. the fraction of such maps which are LEWembeddings and the fraction which are not LEWembeddings both exceed n Gammac 2 . 2 1. Introduction We begin with some definitions: ffl A map is a connected graph G embedded in a surface S (a closed 2manifold) such that all components of S Gamma G are simply connected regions. These components ar...Journal of Graph Theory 02/1999; · 0.63 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: The determinant that is present in traditional formulations of multivariate Lagrange inversion causes difficulties when one attempts (d+1) dGamma1 terms in contrast to the d! terms of the determinantal form. Thus it is likely to prove useful only for asymptotic purposes. 1991 AMS Classification Number. Primary: 05A15 Secondary: 05C05, 40E99 the electronic journal of combinatorics 5 (1998), #Rxx 2 1. Introduction Many researchers have studied the Lagrange inversion formula, obtaining a variety of proofs and extensions. Gessel [4] has collected an extensive set of references. For more recent results see Haiman and Schmitt [6], Goulden and Kulkarni [5], and Section 3.1 of Bergeron, Labelle, and Leroux [3]. Let boldface letters denote vectors and let a vector to a vector power be the product of componentwise exponentiation as in x n = x n1 1 Delta Delta Delta x nd d . Let [x n ] h(x) denote the coefficient of x n in h(x). Let ka i;j k denote the determinant of the d ...02/1999;  [Show abstract] [Hide abstract]
ABSTRACT: this paper, all limits are understood to be taken through those values of nCombinatorics Probability and Computing 02/1999; · 0.61 Impact Factor 
Article: 01 Laws for Maps
[Show abstract] [Hide abstract]
ABSTRACT: A class of finite structures has a 01 law with respect to a logic if every property expressible in the logic has a probability approaching a limit of 0 or 1 as the structure size grows. To formulate 01 laws for maps (i.e., embeddings of graphs in a surface), it is necessary to represent maps as logical structures. Three such representations are given, the most general being the full cross representation based on Tutte's theory of combinatorial maps. The main result says that if a class of maps has two properties, richness and large representativity, then the corresponding class of full cross representations has a 01 law with respect to firstorder logic. As a corollary the following classes of maps on a surface of fixed type have a firstorder 01 law: all maps, smooth maps, 2connected maps, 3connected maps, triangular maps, 2connected triangular maps, and 3connected triangular maps. c fl ??? John Wiley & Sons, Inc. Keywords: 01 law, maps 1. INTRODUCTION. In probability...Random Structures and Algorithms 02/1999; · 1.05 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: Introduction Modern algorithms [5, 7] for factoring integers or for solving discrete logarithms problems work in two phases. In the first one, one collects a huge amount of data that help create a large matrix that is triangularized in the second phase. For instance, the largest nontrivial number ever factored as of today is (10 211 Gamma 1)=9, which involved finding dependencies in a 4; 820; 249 Theta 4; 895; 741 boolean matrix (see [2]). The easiest way to solve the problem is to find a computer with enough memory so that the matrix fits in core and Gaussian elimination can be used. If such a behemoth is not available, alternative methods have to be used. A method that is widely used relies on the fact that the matrix we are interested in is sparse. For instance the matrix referred to above has only 48:1 nonzero coefficients per row onJournal of Algorithms 01/1999; 31:271290. · 0.50 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: The determinant that is present in traditional formulations of multivariate Lagrange inversion causes di#culties when one attempts to obtain asymptotic information. We obtain an alternate formulation as a sum of terms, thereby avoiding this di# culty. 1991 AMS Classification Number. Primary: 05A15 Secondary: 05C05, 40E99 the electronic journal of combinatorics 5 (1998), #R33 2 1. Introduction Many researchers have studied the Lagrange inversion formula, obtaining a variety of proofs and extensions. Gessel [4] has collected an extensive set of references. For more recent results see Haiman and Schmitt [6], Goulden and Kulkarni [5], and Section 3.1 of Bergeron, Labelle, and Leroux [3]. Let boldface letters denote vectors and let a vector to a vector power be the product of componentwise exponentiation as in x n = x n1 1 x n d d .Let[x n ]h(x) denote the coe#cient of x n in h(x). Let #a i,j # denote the determinant of the d d matrix with entries a i,j . A traditional formul...The electronic journal of combinatorics 08/1998; · 0.53 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: This paper begins with the observation that half of all graphs containing no induced path of length 3 are disconnected. We generalize this in several directions. First, we give necessary and sufficient conditions (in terms of generating functions) for the probability of connectedness in a suitable class of graphs to tend to a limit strictly between zero and one. Next we give a general framework in which this and related questions can be posed, involving operations on classes of finite structures. Finally, we discuss briefly an algebra associated with such a class of structures, and give a conjecture about its structure. 1 1 Introduction The class of graphs containing no induced path of length 3 has many remarkable properties, stemming from the following wellknown observation. Recall that an induced subgraph of a graph consists of a subset S of the vertex set together with all edges contained in S. Proposition 1.1 Let G be a finite graph with more than one vertex, containin...Combinatorics Probability and Computing 06/1998; · 0.61 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: The weight of a vector in the finite vector space GF(q) is the number of nonzero components it contains. We show that for a certain range of parameters (n, j, k, w) the number of kdimensional subspaces having j(q  1) vectors of minimum weight w has asymptotically a Poisson distribution with parameter # = . As the Poisson parameter grows, the distribution becomes normal. AMSMOS Subject Classification (1990). Primary: 05A16 Secondary: 05A15, 11T99 1.The electronic journal of combinatorics 01/1998; · 0.53 Impact Factor 
Article: The Fraction of Subspaces of
[Show abstract] [Hide abstract]
ABSTRACT: The weight of a vector in the finite vector space GF(q) n is the number of nonzero components it contains. We show that for a certain range of parameters (n; j; k; w) the number of kdimensional subspaces having j(q Gamma 1) vectors of minimum weight w has asymptotically a Poisson distribution with parameter = Gamma n w Delta (q Gamma 1) wGamma1 q kGamman . As the Poisson parameter grows, the distribution becomes normal. AMSMOS Subject Classification (1990). Primary: 05A16 Secondary: 05A15, 11T99 the electronic journal of combinatorics 4 (1997), #R3 2 1. Introduction Almost all the familiar concepts of linear algebra, such as dimension and linear independence, are valid without regard to the characteristic of the underlying field. An example of a characteristicdependent result is that a nonzero vector cannot be orthogonal to itself; researchers accustomed to real vector spaces must modify their "intuition" on this point when entering the realm of finite fields. L...01/1998;  [Show abstract] [Hide abstract]
ABSTRACT: Letd(n,q) be the number of labeled graphs withnvertices, q\le N={n \choose 2} edges, and no isolated vertices. Letx=q/nandk=2q−n. We determine functionswk∼1,a(x), andϕ(x) such that d(n,\ q)\sim w_k{N \choose q}\ e^{n\varphi (x)+a(x)} uniformly for allnandq>n/2.Journal of Combinatorial Theory Series A 10/1997; 80:124150. · 0.77 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: Let d(n; q) be the number of labeled graphs with n vertices, q N = Gamma n 2 Delta edges, and no isolated vertices. Let x = q=n and k = 2q Gamma n. We determine functions w k ¸ 1, a(x), and '(x) such that d(n; q) ¸ w k Gamma N q Delta e n'(x)+a(x) uniformly for all n and q ? n=2.09/1996;  [Show abstract] [Hide abstract]
ABSTRACT: LetAndenote thenthcycle index polynomial, in the variablesXj, for the symmetric group onnletters. We show that if the variablesXjare assigned nonnegative real values which are logconcave, then the resulting quantitiesAnsatisfy the two inequalitiesAn−1An+1⩽A2n⩽((n+1)/n)An−1An+1. This implies that the coefficients of the formal power series exp(g(u)) are logconcave whenever those ofg(u) satisfy a condition slightly weaker than logconcavity. The latter includes many familiar combinatorial sequences, only some of which were previously known to be logconcave. To prove the first inequality we show that in fact the differenceA2n−An−1An+1can be written as a polynomial with positive coefficients in the expressionsXjandXjXk−Xj−1Xk+1,j⩽k. The second inequality is proven combinatorially, by working with the notion of amarkedpermutation, which we introduce in this paper. The latter is a permutation each of whose cycles is assigned a subset of available markers {Mi, j}. Each marker has aweight, wt(Mi, j)=xj, and we relate the second inequality to properties of theweight enumerator polynomials. Finally, using asymptotic analysis, we show that the same inequalities hold fornsufficiently large when theXjare fixed with only finite many nonzero values, with no additional assumption on theXj.Journal of Combinatorial Theory Series A 04/1996; 74(1):57–70. · 0.77 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: In this paper we obtain asymptotics for the number of rooted 3connected maps on an arbitrary surface and use them to prove that almost all rooted 3connected maps on any fixed surface have large edgewidth and large facewidth. It then follows from the result of Roberston and Vitray [10] that almost all rooted 3connected maps on any fixed surface are minimum genus embeddings and their underlying graphs are uniquely embeddable on the surface.Journal of the Australian Mathematical Society 01/1996; 60(01):31  41. · 0.45 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: Let a(n;k) denote the number of combinatorial structures of size n with k compo nents. One often has P n;ka(n;k)x n yk=n !=e xp ' yC(x) " ,w hereC(x )i s frequently the exponential generating function for connected structures. How does a(n;k) behave as a function of k when n is large and C(x) is entire or has large singular ities on its circle of convergence? The FlajoletOdlyzko singularity analysis does not directly apply in such cases. We extend some of Hayman's work on admissi ble functions of a single variable to functions of several variables. As applications, we obtain asymptotics and local limit theorems for several set partition problems, decomposition of vector spaces, tagged permutations, and various complete graph covering problems.The electronic journal of combinatorics 01/1996; 3. · 0.53 Impact Factor
Publication Stats
1k  Citations  
59.38  Total Impact Points  
Top Journals
Institutions

1978–2006

University of California, San Diego
 Department of Mathematics
San Diego, CA, United States


1983–1993

University of Waterloo
 Department of Combinatorics & Optimization
Waterloo, Ontario, Canada


1978–1991

University of Georgia
 Department of Computer Science
Атина, Georgia, United States


1985

University of Pennsylvania
 Department of Mathematics
Philadelphia, PA, United States 
University of Newcastle
Newcastle, New South Wales, Australia
