A note on similar edges and edge-unique line graphs

If G is a connected graph having no vertices of degree 2 and L(G) is its line graph, two results are proven: if there exist distinct edges e and f with L(G) ? e ? L(G) ? f then there is an automorphism of L(G) mapping e to f; if $\text{G ? u ¦ G ? v}$ for any distinct vertices u, v, then $\text{L(G) ? e ¦ L(G) ? f}$ for any distinct edges e, f.     

Unoriented Laplacian maximizing graphs are degree maximal

A connected graph is said to be unoriented Laplacian maximizing if the spectral radius of its unoriented Laplacian matrix attains the maximum among all connected graphs with the same number of vertices and the same number of edges. A graph is said to be threshold (maximal) if its degree sequence is not majorized by the degree sequence of any other graph (and, in addition, the graph is connected). It is proved that an unoriented Laplacian maximizing graph is maximal and also that there are precisely two unoriented Laplacian maximizing graphs of a given order and with nullity 3. Our treatment depends on the following known characterization: a graph G is threshold (maximal) if and only if for every pair of vertices u,v of G, the sets N(u)?{v},N(v)?{u}, where N(u) denotes the neighbor set of u in G, are comparable with respect to the inclusion relation (and, in addition, the graph is connected). A conjecture about graphs that maximize the unoriented Laplacian matrix among all graphs with the same number of vertices and the same number of edges is also posed.     

Circuit preserving edge maps II

In Section 1 of this article we prove the following. Let f: G → G′ be a circuit surjection, i.e., a mapping of the edge set of G onto the edge set of G′ which maps circuits of G onto circuits of G′, where G, G′ are graphs without loops or multiple edges and G′ has no isolated vertices. We show that if G is assumed finite and 3-connected, then f is induced by a vertex isomorphism. If G is assumed 3-connected but not necessarily finite and G′ is assumed to not be a circuit, then f is induced by a vertex isomorphism. Examples of circuit surjections f: G → G′ where G′ is a circuit and G is an infinite graph of arbitrarily large connectivity are given. In general if we assume G two-connected and G′ not a circuit then any circuit surjection f: G → G′ may be written as the composite of three maps, f(G) = q(h(k(G))), where k is a 1-1 onto edge map which preserves circuits in both directions (the “2-isomorphism” of Whitney (Amer. J. Math. 55 (1933), 245–254) when G is finite), h is an onto edge map obtained by replacing “suspended chains” of k(G) with single edges, and G is a circuit injection (a 1-1 circuit surjection). Let f: G → M be a 1-1 onto mapping of the edges of G onto the cells of M which takes circuits of G onto circuits of M where G is a graph with no isolated vertices, M a matroid. If there exists a circuit C of M which is not the image of a circuit in G, we call f nontrivial, otherwise trivial. In Section 2 we show the following. Let G be a graph of even order. Then the statement “no trivial map f: G → M exists, where M is a binary matroid,” is equivalent to “G is Hamiltonian.” If G is a graph of odd order, then the statement “no nontrivial map f: G → M exists, where M is a binary matroid” is equivalent to “G is almost Hamiltonian,” where we define a graph G of order n to be almost Hamiltonian if every subset of vertices of order n − 1 is contained in some circuit of G.     

Representing Graphs in Steiner Triple Systems

Let G =  (V, E) be a simple graph and let T =  (P, B) be a Steiner triple system. Let φ be a one-to-one function from V to P. Any edge e =  {u, v} has its image {φ(u), φ(v)} in a unique block in B. We also denote this induced function from edges to blocks by φ. We say that T represents G if there exists a one-to-one function φ : V → P such that the induced function φ : E → B is also one-to-one; that is, if we can represent vertices of the graph by points of the triple system such that no two edges are represented by the same block. In this paper we examine when a graph can be represented by an STS. First, we find a bound which ensures that every graph of order n is represented in some STS of order f(n). Second, we find a bound which ensures that every graph of order n is represented in every STS of order g(n). Both of these answers are related to finding an independent set in an STS. Our question is a generalization of finding such independent sets. We next examine which graphs can be represented in STS’s of small orders. Finally, we give bounds on the orders of STS’s that are guaranteed to embed all graphs of a given maximum degree.     

On the Super Edge-Magic Deficiency of Graphs

A (p, q) graph G is edge-magic if there exists a bijective function f: V(G) ∪ E(G) → {1,2,…,p + q} such that f(u) + f(v) + f(uv) = k is a constant, called the valence of f, for any edge uv of G. Moreover, G is said to be super edge-magic if f(V(G)) = {1,2,…,p}. The question studied in this paper is for which graphs is it possible to add a finite number of isolated vertices so that the resulting graph is super edge-magic? If it is possible for a given graph G, then we say that the minimum such number of isolated vertices is the super edge-magic deficiency, μ_s(G) of G; otherwise we define it to be + ∞.     

A characterization of block graphs

A block graph is a graph whose blocks are cliques. For each edge e=uv of a graph G, let N_e(u) denote the set of all vertices in G which are closer to u than v. In this paper we prove that a graph G is a block graph if and only if it satisfies two conditions: (a) The shortest path between any two vertices of G is unique; and (b) For each edge e=uv∈E(G), if x∈N_e(u) and y∈N_e(v), then, and only then, the shortest path between x and y contains the edge e. This confirms a conjecture of Dobrynin and Gutman [A.A. Dobrynin, I. Gutman, On a graph invariant related to the sum of all distances in a graph, Publ. Inst. Math., Beograd. 56 (1994) 18-22].     

On induced Folkman numbers

In 1970, Folkman proved that for any graph G there exists a graph H with the same clique number as G. In addition, any r ‐coloring of the vertices of H yields a monochromatic copy of G. For a given graph G and a number of colors r let f(G, r) be the order of the smallest graph H with the above properties. In this paper, we give a relatively small upper bound on f(G, r) as a function of the order of G and its clique number. © 2012 Wiley Periodicals, Inc. Random Struct. Alg., 40, 493–500, 2012     

Sum choice numbers of some graphs

Let f be a function assigning list sizes to the vertices of a graph G. The sum choice number of G is the minimum ∑_v∈V(G)f(v) such that for every assignment of lists to the vertices of G, with list sizes given by f, there exists proper coloring of G from the lists. We answer a few questions raised in a paper of Berliner, Bostelmann, Brualdi, and Deaett. Namely, we determine the sum choice number of the Petersen graph, the cartesian product of paths , and the complete bipartite graph K_3,n.     

Some remarks on the odd hadwiger’s conjecture

We say that H has an odd complete minor of order at least l if there are l vertex disjoint trees in H such that every two of them are joined by an edge, and in addition, all the vertices of trees are two-colored in such a way that the edges within the trees are bichromatic, but the edges between trees are monochromatic. Gerards and Seymour conjectured that if a graph has no odd complete minor of order l, then it is (l ? 1)-colorable. This is substantially stronger than the well-known conjecture of Hadwiger. Recently, Geelen et al. proved that there exists a constant c such that any graph with no odd K _k -minor is ck√logk-colorable. However, it is not known if there exists an absolute constant c such that any graph with no odd K _k -minor is ck-colorable. Motivated by these facts, in this paper, we shall first prove that, for any k, there exists a constant f(k) such that every (496k + 13)-connected graph with at least f(k) vertices has either an odd complete minor of size at least k or a vertex set X of order at most 8k such that G–X is bipartite. Since any bipartite graph does not contain an odd complete minor of size at least three, the second condition is necessary. This is an analogous result of Böhme et al. We also prove that every graph G on n vertices has an odd complete minor of size at least n/2α(G) ? 1, where α(G) denotes the independence number of G. This is an analogous result of Duchet and Meyniel. We obtain a better result for the case α(G)= 3.     

The Four-in-a-Tree Problem in Triangle-Free Graphs

The three-in-a-tree algorithm of Chudnovsky and Seymour decides in time O(n ⁴) whether three given vertices of a graph belong to an induced tree. Here, we study four-in- a-tree for triangle-free graphs. We give a structural answer to the following question: what does a triangle-free graph look like if no induced tree covers four given vertices? Our main result says that any such graph must have the “same structure”, in a sense to be defined precisely, as a square or a cube. We provide an O(nm)-time algorithm that given a triangle-free graph G together with four vertices outputs either an induced tree that contains them or a partition of V(G) certifying that no such tree exists. We prove that the problem of deciding whether there exists a tree T covering the four vertices such that at most one vertex of T has degree at least 3 is NP-complete.     

On the reduced signless Laplacian spectrum of a degree maximal graph

For a (simple) graph G, the signless Laplacian of G is the matrix A(G)+D(G), where A(G) is the adjacency matrix and D(G) is the diagonal matrix of vertex degrees of G; the reduced signless Laplacian of G is the matrix Δ(G)+B(G), where B(G) is the reduced adjacency matrix of G and Δ(G) is the diagonal matrix whose diagonal entries are the common degrees for vertices belonging to the same neighborhood equivalence class of G. A graph is said to be (degree) maximal if it is connected and its degree sequence is not majorized by the degree sequence of any other connected graph. For a maximal graph, we obtain a formula for the characteristic polynomial of its reduced signless Laplacian and use the formula to derive a localization result for its reduced signless Laplacian eigenvalues, and to compare the signless Laplacian spectral radii of two well-known maximal graphs. We also obtain a necessary condition for a maximal graph to have maximal signless Laplacian spectral radius among all connected graphs with given numbers of vertices and edges.     

Finite common coverings of pairs of regular graphs

The graph G is a covering of the graph H if there exists a (projection) map p from the vertex set of G to the vertex set of H which induces a one-to-one correspondence between the vertices adjacent to v in G and the vertices adjacent to p(v) in H, for every vertex v of G. We show that for any two finite regular graphs G and H of the same degree, there exists a finite graph K that is simultaneously a covering both of G and H. The proof uses only Hall's theorem on 1-factors in regular bipartite graphs.     

Minimal 2-matching-covered graphs

A perfect 2-matching M of a graph G is a spanning subgraph of G such that each component of M is either an edge or a cycle. A graph G is said to be 2-matching-covered if every edge of G lies in some perfect 2-matching of G. A 2-matching-covered graph is equivalent to a “regularizable” graph, which was introduced and studied by Berge. A Tutte-type characterization for 2-matching-covered graph was given by Berge. A 2-matching-covered graph is minimal if G−e is not 2-matching-covered for all edges e of G. We use Berge’s theorem to prove that the minimum degree of a minimal 2-matching-covered graph other than K₂ and K₄ is 2 and to prove that a minimal 2-matching-covered graph other than K₄ cannot contain a complete subgraph with at least 4 vertices.     

On defect-d matchings in graphs

A defect-d matching in a graph G is a matching which covers all but d vertices of G. G is d-covered if each edge of G belongs to a defect-d matching. Here we characterise d-covered graphs and d-covered connected bipartite graphs. We show that a regular graph G of degree r which is (r ? 1)-edge-connected is 0-covered or 1-covered depending on whether G has an even or odd number of vertices, but, given any non-negative integers r and d, there exists a graph regular of degree r with connectivity and edge-connectivity r ? 2 which does not even have a defect-d matching. Finally, we prove that a vertex-transitive graph is 0-covered or 1-covered depending on whether it has an even or odd number of vertices.     

Every minor-closed property of sparse graphs is testable

Suppose G is a graph of bounded degree d, and one needs to remove ?n of its edges in order to make it planar. We show that in this case the statistics of local neighborhoods around vertices of G is far from the statistics of local neighborhoods around vertices of any planar graph G^′. In fact, a similar result is proved for any minor-closed property of bounded degree graphs.The main motivation of the above result comes from theoretical computer-science. Using our main result we infer that for any minor-closed property P, there is a constant time algorithm for detecting if a graph is “far” from satisfying P. This, in particular, answers an open problem of Goldreich and Ron [STOC 1997] [20], who asked if such an algorithm exists when P is the graph property of being planar. The proof combines results from the theory of graph minors with results on convergent sequences of sparse graphs, which rely on martingale arguments.     

A Generalization of Opsut’s Lower Bounds for the Competition Number of a Graph

The notion of a competition graph was introduced by Cohen in 1968. The competition graph C(D) of a digraph D is a (simple undirected) graph which has the same vertex set as D and has an edge between two distinct vertices x and y if and only if there exists a vertex v in D such that (x, v) and (y, v) are arcs of D. For any graph G, G together with sufficiently many isolated vertices is the competition graph of some acyclic digraph. In 1978, Roberts defined the competition number k(G) of a graph G as the minimum number of such isolated vertices. In general, it is hard to compute the competition number k(G) for a graph G and it has been one of the important research problems in the study of competition graphs to characterize a graph by its competition number. In 1982, Opsut gave two lower bounds for the competition number of a graph. In this paper, we give a generalization of these two lower bounds for the competition number of a graph.     

Multiple optima in local search

We consider the application of local search methods to the maximum independent set problem. These methods employ a relation R on the power set of the graph's vertices that identifies a set of vertices, U, with a collection of subsets of vertices, R(U), called the neighbors of U. If each set U has only polynomially many neighbors then we say that R is polynomially bounded. Further, given a graph, G, we call a permutation of G, φ(G), to be any graph that arises from G by relabeling the vertices. Our main result is slightly stronger than the following: we construct a graph G such that, for all polynomially bounded relations, R, most permutations φ(G) of the graph contain exponentially many strict local optima that are not global optima. That is, a single graph (up to relabelling of the vertices) exists that soundly defeats all polynomially bounded relations R. Corollaries follow for 0–1 integer programming and other hard optimization problems.     

Dimension and embedding theorems for geometric lattices

Let G be an n-dimensional geometric lattice. Suppose that 1 ? e, f ? n ? 1, e + f ? n, but e and f are not both n ? 1. Then, in general, there are E, F? G with dim E = e, dim F = f, E ? F = 1, and dim E ∧ F = e + f ? n ? 1; any exception can be embedded in an n-dimensional modular geometric lattice M in such a way that joins and dimensions agree in G and M, as do intersections of modular pairs, while each point and line of M is the intersection (in M) of the elements of G containing it.     

A characterization of maximal non-k-factor-critical graphs

A graph G of order p is k-factor-critical,where p and k are positive integers with the same parity, if the deletion of any set of k vertices results in a graph with a perfect matching. G is called maximal non-k-factor-critical if G is not k-factor-critical but G+e is k-factor-critical for every missing edge e∉E(G). A connected graph G with a perfect matching on 2n vertices is k-extendable, for 1?k?n-1, if for every matching M of size k in G there is a perfect matching in G containing all edges of M. G is called maximal non-k-extendable if G is not k-extendable but G+e is k-extendable for every missing edge e∉E(G) . A connected bipartite graph G with a bipartitioning set (X,Y) such that |X|=|Y|=n is maximal non-k-extendable bipartite if G is not k-extendable but G+xy is k-extendable for any edge xy∉E(G) with x∈X and y∈Y. A complete characterization of maximal non-k-factor-critical graphs, maximal non-k-extendable graphs and maximal non-k-extendable bipartite graphs is given.     

The square of every two-connected graph is Hamiltonian

A graph G is a line-critical block if κ(G) = 2 and if for any line e of G the graph G ? e has κ(G ? e) = 1.If G is a line-critical block, then G is either a DT-block (i.e., G is a two-connected graph in which every line is incident to a point of degree two), or G contains a specific two-connected subgraph which is a DT-block (Theorem 1). Using this result and results of the preceding paper on DT-graphs, a simple proof of the conjecture that the square of every two-connected graph is Hamiltonian is given.