Ramsey numbers for matroids, which mimic properties of Ramsey numbers for graphs, have been denned as follows. Let k and l be positive integers. Then n(k, l) is the least positive integer n such that every connected matroid with n elements contains either a circuit with at least k elements or a cocircuit with at least l elements. We determine the largest known value of these numbers in the sense of maximizing both k and l. We also find extremal matroids with small circuits and cocircuits. Results on matroid connectivity, geometry, and extremal matroid theory are used here.

The choice number of a graph G is the minimum integer k such that for every assignment of a set S(v) of k colors to every vertex v of G, there is a proper coloring of G that assigns to each vertex v a color from S(v). By applying probabilistic methods, it is shown that there are two positive constants c1 and c2 such that for all m ≥ 2 and r ≥ 2 the choice number of the complete r-partite graph with m vertices in each vertex class is between c1r log m and c2r log m. This supplies the solutions of two problems of Erdős, Rubin and Taylor, as it implies that the choice number of almost all the graphs on n vertices is o(n) and that there is an n vertex graph G such that the sum of the choice number of G with that of its complement is at most O(n1/2(log n)1/2).

We prove that almost every r-regular digraph is Hamiltonian for all fixed r ≥ 3.

In this paper we solve the following problem on the lattice graph L(m1,…,mn) and the Hamming graph H(m1,…,mn), generalizing a result of Felzenbaum-Holzman-Kleitman on the n-dimensional cube (all mi = 2): Characterize the vectors (s1.…,sn) such that there exists a maximum matching in L, respectively, H with exactly si edges in the ith direction.

We focus our attention on the class RMAX(2) of NP optimization problems. Owing to recent developments in interactive proof techniques, RMAX(2) was shown to be the lowest class of logical classification that contains problems hard to approximate. Namely, the RMAX(2)-complete problem MAX CLIQUE (of finding the size of the largest clique in a graph) is not approximable in polynomial time within any constant factor unless NP=P.

We are interested in problems inside RMAX(2) that are not known to be complete but are still hard to approximate. We point out that one such problem is MAXlog n, n, considered by Berman and Schnitger: given m conjunctions, each of them consisting of log m propositional variables or their negations, find the maximal number of simultaneously satisfiable conjunctions. We also obtain the approximation hardness results for some other problems in RMAX(2). Finally, we discuss the question of whether or not the problems under consideration are RMAX(2)-complete.

We consider a «Maker-Breaker’ version of the Ramsey Graph Game, RG(n), and present a winning strategy for Maker requiring at most (n − 3)2n−1 + n + 1 moves. This is the fastest winning strategy known so far. We also demonstrate how the ideas presented can be used to develop winning strategies for some related combinatorial games.

We show distributional results for the length of the longest matching consecutive subsequence between two independent sequences A1, A2, …, Am and B1, B2, …, Bn whose letters are taken from a finite alphabet. We assume that A1, A2, … are i.i.d. with distribution μ and B1, B2, … are i.i.d. with distribution ν. It is known that if μ and v are not too different, the Chen–Stein method for Poisson approximation can be used to establish distributional results. We extend these results beyond the region where the Chen–Stein method was previously successful. We use a combination of ‘matching by patterns’ results obtained by Arratia and Waterman [1], and the Chen–Stein method to show that the Poisson approximation can be extended. Our method explains how the matching is achieved. This provides an explanation for the formulas in Arratia and Waterman [1] and thus answers one of the questions posed in comment F19 in Aldous [2]. Furthermore, in the case where the alphabet consists of only two letters, the phase transition observed by Arratia and Waterman [1] for the strong law of large numbers extends to the distributional result. We conjecture that this phase transition on the distributional level holds for any finite alphabet.

The structure of non-Hamiltonian graphs is studied in terms of neighbourhood-unions. Numerous results, such as (A), (B) and (C) below, are obtained.

Let G be a graph with maximum degree Δ(G). In this paper we prove that if the girth g(G) of G is greater than 4 then its chromatic number, χ(G), satisfies

where o(l) goes to zero as Δ(G) goes to infinity. (Our logarithms are base e.) More generally, we prove the same bound for the list-chromatic (or choice) number:

provided g(G) < 4.

Let P10\e be the graph obtained by deleting an edge from the Petersen graph. We give a decomposition theorem for cubic graphs with no minor isomorphic to P10\e. The decomposition is used to show that graphs in this class are 3-edge-colourable. We also consider an application to a conjecture due to Grötzsch which states that a planar graph is 3-edge-colourable if and only if it is fractionally 3-edge-colourable.

Whereas the cylindrical version of an Eden cluster in the plane has a surface roughness with a fractal dimension predicted by theory, the central version has hitherto seemed to conflict with theory. However, a fresh way of analysing computer simulations of the central version shows that this anomaly is more apparent than real, and the central version can thereby be reconciled with theory. As a by-product, we obtain statistical data on the properties of the central version in the plane. The macroscopic shape of a central cluster is not circular, and microscopic roughness depends weakly upon the angular direction of portions of the surface. Rather surprisingly, the edge method of construction gives a more nearly circular shape than the external and internal methods. For higher dimensions than the plane, the corresponding treatment is more difficult, and there the situation remains obscure. Higher dimensions and certain other clusters (e.g. Richardson clusters) are treated briefly in Section 6. The theory of surface roughness uses a spatial generalization of martingales, called a serial harness: this is also described in Section 6.

In this work we show that that many of the basic results concerning the theory of the characteristic polynomial of a graph can be derived as easy consequences of a determinantal identity due to Jacobi. As well as improving known results, we are also able to derive a number of new ones. A combinatorial interpretation of the Christoffel-Darboux identity from the theory of orthogonal polynomials is also presented. Finally, we extend some work of Tutte on the reconstructibility of graphs with irreducible characteristic polynomials.

Let G be a connected graph with n vertices and m edges (multiple edges allowed), and let k ≥ 2 be an integer. There is an algorithm with (optimal) running time of O(m) that finds

(i) a bipartite subgraph of G with ≥ m/2 + (n − 1)/4 edges,

(ii) a bipartite subgraph of G with ≥ m/2 + 3(n−1)/8 edges if G is triangle-free,

(iii) a k-colourable subgraph of G with ≥ m − m/k + (n−1)/k + (k − 3)/2 edges if k ≥ 3 and G is not k-colorable.

Infinite families of graphs show that each of those lower bounds on the worst-case performance are best possible (for every algorithm). Moreover, even if short cycles are excluded, the general lower bound of m − m/k cannot be replaced by m − m/k + εm for any fixed ε > 0; and it is NP-complete to decide whether a graph with m edges contains a k-colorable subgraph with more than m − m/k + εm edges, for any k ≥ 2 and ε> 0, ε < 1/k.

We consider special types of mod-λ flows, called odd and even mod-λ flows, for directed graphs, and prove that the numbers of such flows can be interpolated by polynomials in λ with the degree given by the cycle rank of the graph. The proofs involve computation of the number of integer solutions in a polyhedral region of Euclidean space using theorems due to Ehrhart. The resulting reciprocity properties of the interpolating polynomials for even flows are considered. The analogous properties of odd and even mod-λ potential differences and their associated potentials are also developed.

Let f be a de Morgan read-once function of n variables. Let fε be the random restriction obtained by independently assigning to each variable of f, the value 0 with probability (1 -ε)/2, the value 1 with the same probability, and leaving it unassigned with probability ε. We show that fε depends, on the average, on only O(εαn + εn1/α) variables, where . This result is asymptotically the tightest possible. It improves a similar result obtained recently by Håstad, Razborov and Yao.

A k×n array with entries from the q-letter alphabet {0, 1, …, q − 1} is said to be t-covering if each k × t submatrix has (at least one set of) qt distinct rows. We use the Lovász local lemma to obtain a general upper bound on the minimal number K = K(n, t, q) of rows for which a t-covering array exists; for t = 3 and q = 2, we are able to match the best-known such bound. Let Kλ = Kλ(n, t, q), (λ ≥ 2), denote the minimum number of rows that guarantees the existence of an array for which each set of t columns contains, amongst its rows, each of the qt possible ‘words’ of length t at least λ times. The Lovász lemma yields an upper bound on Kλ that reveals how substantially fewer rows are needed to accomplish subsequent t-coverings (beyond the first). Finally, given a random k × n array, the Stein–Chen method is employed to obtain a Poisson approximation for the number of sets of t columns that are deficient, i.e. missing at least one word.

Long regressive sequences in well-quasi-ordered sets contain ascendingsubsequences of length n. The complexity of the corresponding function H(n) is studied in the Grzegorczyk-Wainer hierarchy. An extension to regressive canonical colourings is indicated.

Suppose that each vertex of a graph independently chooses a colour uniformly from the set {1, …, k}; and let Si be the random set of vertices coloured i. Farr shows that the probability that each set Si is stable (so that the colouring is proper) is at most the product of the k probabilities that the sets Si separately are stable. We give here a simple proof of an extension of this result.