On 1-factors of point determining graphs

A graph G is said to be point determining if and only if distinct points of G have distinct neighborhoods. For such a graph G, the nucleus is defined to be the set G″ consisting of all points ν of G for which G-ν is a point determining graph.In [4], Summer exhibited several families of graphs H such that if G⁰ = H, for some point determining graph G, then G has a 1-factor. In this paper, we extend this class of graphs.     

An upper bound on the size of a largest clique in a graph

A graph is point determining if distinct vertices have distinct neighborhoods. The nucleus of a point-determining graph is the set G^O of all vertices, v, such that G–v is point determining. In this paper we show that the size, ω(G), of a maximum clique in G satisfies ω(G) ? 2|π (G)^O|, where π(G) (the point determinant of G) is obtained from G by identifying vertices which have the same neighborhood.     

The edge nucleus of a point-determining graph

In this paper we investigate the edge nucleus E⁰(G) of a point-determining graph G. We observe several relationships between E⁰(G) and the nucleus G⁰ = {v ∈ V(G)∣ G ? v is point determining} and use these relationships to prove several properties of E⁰(G). In particular, we show that there are only a finite number of graphs with a given edge nucleus and we determine those graphs G for which |E⁰(G)| ≤ 2. We also show that an n-clique of a point-determining graph G contains at least n?2 edges of E⁰(G) and if G is totally point determining, then every odd cycle of G meets E⁰(G).     

On realizations of point determining graphs, and obstructions to full homomorphisms

A graph is point determining if distinct vertices have distinct neighbourhoods. A realization of a point determining graph H is a point determining graph G such that each vertex-removed subgraph G-x which is point determining, is isomorphic to H. We study the fine structure of point determining graphs, and conclude that every point determining graph has at most two realizations.A full homomorphism of a graph G to a graph H is a vertex mapping f such that for distinct vertices u and v of G, we have uv an edge of G if and only if f(u)f(v) is an edge of H. For a fixed graph H, a full H-colouring of G is a full homomorphism of G to H. A minimal H-obstruction is a graph G which does not admit a full H-colouring, such that each proper induced subgraph of G admits a full H-colouring. We analyse minimal H-obstructions using our results on point determining graphs. We connect the two problems by proving that if H has k vertices, then a graph with k+1 vertices is a minimal H-obstruction if and only if it is a realization of H. We conclude that every minimal H-obstruction has at most k+1 vertices, and there are at most two minimal H-obstructions with k+1 vertices.We also consider full homomorphisms to graphs H in which loops are allowed. If H has ? loops and k vertices without loops, then every minimal H-obstruction has at most (k+1)(?+1) vertices, and, when both k and ? are positive, there is at most one minimal H-obstruction with (k+1)(?+1) vertices.In particular, this yields a finite forbidden subgraph characterization of full H-colourability, for any graph H with loops allowed.     

Nuclei for totally point determining graphs

The nucleus (edge nucleus) of a point determining graph is defined by Geoffroy and Sumner to be the set of all points (edges) whose removal leaves the graph point determining. It is the purpose of this paper to develop the analogous concepts for totally point determining graphs, that is, graphs in which distinct points have distinct neighborhoods and closed neighborhoods.     

A linear algorithm for a core of a tree

A core of a graph G is a path P in G that is central with respect to the property of minimizing d(P) = Σ_υ?V(G)d(υ, P), where d(υ, P) is the distance from vertex υ to path P. We present a linear algorithm for finding a core of a tree. Since the core of a graph is not necessarily unique, we also output a list of all the vertices which are in some core.     

Strictly Deza line graphs

For a given graph G, its line graph L(G) is defined as the graph with vertex set equal to the edge set of G in which two vertices are adjacent if and only if the corresponding edges of G have exactly one common vertex. A k-regular graph of diameter 2 on υ vertices is called a strictly Deza graph with parameters (υ, k, b, a) if it is not strongly regular and any two vertices have a or b common neighbors. We give a classification of strictly Deza line graphs.     

Two applications of semi-stable graphs

A Michigan graph G on a vertex set V is called semi-stable if for some υ?V, Γ(G_υ) = Γ(G)_υ. It can be shown that all regular graphs are semi-stable and this fact is used to show (i) that if Γ(G) is doubly transitive then G = K_n or K?_n, and (ii) that Γ(G) can be recovered from Γ(G_υ). The second result is extended to the case of stable graphs.     

Unicyclic nonintegral sum graphs

A graph G = (V,E) is an integral sum graph if there exists a labeling S(G) ? Z such that V = S(G) and every two distinct vertices u, υ ∈ V are adjacent if and only if u + υ ∈ V. A connected graph G = (V,E) is called unicyclic if |V| = |E|. In this paper two infinite series are constructed of unicyclic graphs that are not integral sum graphs.     

The non-isolated vertices in the generating graph of a direct powers of simple groups

For a finite group G let Γ(G) denote the graph defined on the non-identity elements of G in such a way that two distinct vertices are connected by an edge if and only if they generate G. Many deep results on the generation of the finite simple groups G can be equivalently stated as theorems that ensure that Γ(G) is a rich graph, with several good properties. In this paper we want to consider Γ(G ^δ ) where G is a finite non-abelian simple group and G ^δ is the largest 2-generated power of G, with the aim to investigate whether the good generation properties of G still affect the behaviour of Γ(G ^δ ). In particular we prove that the graph obtained from Γ(G ^δ ) by removing the isolated vertices is 1-arc transitive and connected and we investigate the diameter of this graph. Moreover, some intriguing open questions will be introduced and their solutions will be exemplified for $G=\operatorname{Alt}(5)$ .     

Graphs with 1-Hamiltonian-connected cubes

The cube G³ of a connected graph G is that graph having the same vertex set as G and in which two distinct vertices are adjacent if and only if their distance in G is at most three. A Hamiltonian-connected graph has the property that every two distinct vertices are joined by a Hamiltonian path. A graph G is 1-Hamiltonian-connected if, for every vertex w of G, the graphs G and G?w are Hamiltonian-connected. A characterization of graphs whose cubes are 1-Hamiltonian-connected is presented.     

Signed graph factors and degree sequences

For a signed graph G and function , a signed f‐factor of G is a spanning subgraph F such that sdeg_F(υ) = f(υ) for every vertex υ of G, where sdeg(υ) is the number of positive edges incident with v less the number of negative edges incident with υ, with loops counting twice in either case. For a given vertex‐function f, we provide necessary and sufficient conditions for a signed graph G to have a signed f‐factor. As a consequence of this result, an Erd?s‐Gallai‐type result is given for a sequence of integers to be the degree sequence of a signed r‐graph, the graph with at most r positive and r negative edges between a given pair of distinct vertices. We discuss how the theory can be altered when mixed edges (i.e., edges with one positive and one negative end) are allowed, and how it applies to bidirected graphs. © 2006 Wiley Periodicals, Inc. J Graph Theory 52: 27–36, 2006     

Characterization of unigraphic and unidigraphic integer-pair sequences

Given a graph (digraph) G with edge (arc) set E(G) = {(u₁}, υ₁), (u₂, υ₂),?,(u_q, υ_q, where q = |E(G)|, we can associate with it an integer-pair sequence S_G = ((a₁, b₁), (a₂, b₂),?, (a_q, b_q)) where a_i, b_i are the degrees (indegrees) of u_i, υ_i respectively. An integer- pair sequence S is said to be graphic (digraphic) if there exists a graph (digraph) G such that _SG = S. In this paper we characterize unigraphic and unidigraphic integer-pair sequences.     

On the 2-factor index of a graph

The 2-factor index of a graph G, denoted by f(G), is the smallest integer m such that the m-iterated line graph L^m(G) of G contains a 2-factor. In this paper, we provide a formula for f(G), and point out that there is a polynomial time algorithm to determine f(G).     

When is G 2 a König–Egerváry Graph?

The independence number of a graph G, denoted by α(G), is the cardinality of a maximum independent set, and μ(G) is the size of a maximum matching in G. If α(G) + μ(G) equals its order, then G is a König–Egerváry graph. The square of a graph G is the graph G ² with the same vertex set as in G, and an edge of G ² is joining two distinct vertices, whenever the distance between them in G is at most two. G is a square-stable graph if it enjoys the property α(G) = α(G ²). In this paper we show that G ² is a König–Egerváry graph if and only if G is a square-stable König–Egerváry graph.     

Color degree and alternating cycles in edge-colored graphs

Let G be an edge-colored graph. An alternating cycle of G is a cycle of G in which any two consecutive edges have distinct colors. Let d^c(v), the color degree of a vertex v, be defined as the maximum number of edges incident with v that have distinct colors. In this paper, we study color degree conditions for the existence of alternating cycles of prescribed length.     

Archimedean ϕ ‐tolerance graphs

Let ? be a symmetric binary function, positive valued on positive arguments. A graph G = (V,E) is a ?‐tolerance graph if each vertex υ ∈ V can be assigned a closed interval I_υ and a positive tolerance t_υ so that xy ∈ E ? | I_x ∩ I_y|≥ ? (t_x,t_y). An Archimedean function has the property of tending to infinity whenever one of its arguments tends to infinity. Generalizing a known result of [15] for trees, we prove that every graph in a large class (which includes all chordless suns and cacti and the complete bipartite graphs K_2,k) is a ?‐tolerance graph for all Archimedean functions ?. This property does not hold for most graphs. Next, we present the result that every graph G can be represented as a ?_G‐tolerance graph for some Archimedean polynomial ?_G. Finally, we prove that there is a ?universal”? Archimedean function ? ^* such that every graph G is a ?^*‐tolerance graph. © 2002 Wiley Periodicals, Inc. J Graph Theory 41: 179–194, 2002     

The heterochromatic cycles in edge-colored graphs

Given a graph and an edge coloring C of G, a heterochromatic cycle of G is a cycle in which any pair of edges have distinct colors. Let d ^c (v), named the color degree of a vertex v, be the maximum number of distinct colored edges incident with v. In this paper, we give several sufficient conditions for the existence of heterochromatic cycles in edge-colored graphs.     

Contraction of Measures on Graphs

Given a finitely supported probability measure μ on a connected graph G, we construct a family of probability measures interpolating the Dirac measure at some given point o ∈ G and μ. Inspired by Sturm-Lott-Villani theory of Ricci curvature bounds on measured length spaces, we then study the convexity of the entropy functional along such interpolations. Explicit results are given in three canonical cases, when the graph G is either ? ⁿ , a cube or a tree.