arXiv++ Combinatorics

We study forcing pairs for quasirandom graphs. Chung, Graham, and Wilson initiated the study of families $\\mathcal F$ of graphs with the property that if a large graph $G$ has approximately homomorphism density $p^{e(F)}$ for some fixed $p\in(0,1]$ for every $F\in \mathcal F$, then $G$ is quasirandom with density $p$. Such families $\mathcal F$ are said to be forcing. Several forcing families were found over the last three decades and characterising all bipartite graphs $F$ such that $(K_2,F)$ is a forcing pair is a well-known open problem in the area of quasirandom graphs, which is closely related to Sidorenko's conjecture. In fact, most of the known forcing families involve bipartite graphs only. We consider forcing pairs containing the triangle $K_3$. In particular, we show that if $(K_2,F)$ is a forcing pair, then so is $(K_3,F')$, where $F'$ is obtained from $F$ by replacing every edge of $F$ by a triangle (each of which introduces a new vertex). For the proof we first show that $(K_3,C'_4)$ is a forcing pair, which strengthens related results of Simonovits and Sós and of Conlon et al.

A well known result due to Andrásfai, Erdős, and Sós asserts that for $r\ge 2$ every $K_{r+1}$-free graph on $n$ vertices with $δ(G)>\frac{3r-4}{3r-1}n$ is $r$-partite. We study related questions in the context of weighted graphs, which are motivated by recent work on the Ramsey-Turán problem for cliques.

An important question in extremal graph theory raised by Vera T. Sós asks to determine for a given integer $t\ge 3$ and a given positive real number $δ$ the asymptotically supremal edge density $f_t(δ)$ that an $n$-vertex graph can have provided it contains neither a complete graph $K_t$ nor an independent set of size $δn$. Building upon recent work of Fox, Loh, and Zhao [The critical window for the classical Ramsey-Turán problem, Combinatorica 35 (2015), 435-476], we prove that if $δ$ is sufficiently small (in a sense depending on $t$), then \[ f_t(δ)= \begin{cases} \frac{3t-10}{3t-4}+δ-δ^2 & \text{ if $t$ is even,} \cr \frac{t-3}{t-1}+δ& \text{ if $t$ is odd.} \end{cases} \]

Given an integer $k \geq 2$ and a real number $γ\in [0, 1]$, which graphs of edge density $γ$ contain the largest number of $k$-edge stars? For $k=2$ Ahlswede and Katona proved that asymptotically there cannot be more such stars than in a clique or in the complement of a clique (depending on the value of $γ$). Here we extend their result to all integers $k\ge 2$.

We show that for an infinitely many natural numbers $k$ there are $k$-uniform hypergraphs which admit a `rescaling phenomenon' as described in [9]. More precisely, let $\mathcal{A}(k,I, n)$ denote the class of $k$-graphs on $n$ vertices in which the sizes of all pairwise intersections of edges belong to a set $I$. We show that if $k=rt^2$ for some $r\ge 1$ and $t\ge 2$, and~$I$ is chosen in some special way, the densest graphs in $\mathcal{A}(rt^2,I, n)$ are either dominated by stars of large degree, or basically, they are `$t$-thick' $rt^2$-graphs in which vertices are partitioned into groups of $t$ vertices each and every edge is a union of $tr$ such groups. It is easy to see that, unlike in stars, the maximum degree of $t$-thick graphs is of a lower order than the number of its edges. Thus, if we study the graphs from $\mathcal{A}(rt^2,I, n)$ with a prescribed number of edges $m$ which minimize the maximum degree, around the value of $m$ which is the number of edges of the largest $t$-thick graph, a rapid, discontinuous phase transition can be observed. Interestingly, these two types of $k$-graphs determine the structure of all hypergraphs in $\mathcal{A}(rt^2,I, n)$. Namely, we show that each such hypergraph can be decomposed into a $t$-thick graph $H_T$, a special collection $H_S$ of stars, and a sparse `left-over' graph $H_R$.

P. Erdős [On extremal problems of graphs and generalized graphs, Israel Journal of Mathematics 2 (1964), 183-190] characterised those hypergraphs $F$ that have to appear in any sufficiently large hypergraph $H$ of positive density. We study related questions for $3$-uniform hypergraphs with the additional assumption that $H$ has to be uniformly dense with respect to vertex sets. In particular, we characterise those hypergraphs $F$ that are guaranteed to appear in large uniformly dense hypergraphs $H$ of positive density. We also review the case when the density of the induced subhypergraphs of $H$ may depend on the proportion of the considered vertex sets.

We give a simple graph-theoretic proof of a classical result due to C. St. J. A. Nash-Williams on covering graphs by forests. Moreover we derive a slight generalisation of this statement where some edges are preassigned to distinct forests.

We give a short survey of problems and results on (1) diameter graphs and hypergraphs, and (2) geometric Ramsey theory. We also make some modest contributions to both areas. Extending a well known theorem of Kahn and Kalai which disproved Borsuk's conjecture, we show that for any integer $r\ge 2$, there exist $\varepsilon=\varepsilon(r)>0$ and $d_0=d_0(r)$ with the following property. For every $d\ge d_0$, there is a finite point set $P\subset\mathbb{R}^d$ of diameter $1$ such that no matter how we color the elements of $P$ with fewer than $(1+\varepsilon)^{\sqrt{d}}$ colors, we can always find $r$ points of the same color, any two of which are at distance $1$.

We show that every 3-uniform hypergraph with $n$ vertices and minimum vertex degree at least $(5/9+o(1))\binom{n}2$ contains a tight Hamiltonian cycle. Known lower bound constructions show that this degree condition is asymptotically optimal.

Erdős conjectured that every $n$-vertex triangle-free graph contains a subset of $\lfloor n/2\rfloor$ vertices that spans at most $n^2/50$ edges. Extending a recent result of Norin and Yepremyan, we confirm this conjecture for graphs homomorphic to so-called Andrásfai graphs. As a consequence, Erdős' conjecture holds for every triangle-free graph $G$ with minimum degree $δ(G)>10n/29$ and if $χ(G)\leq 3$ the degree condition can be relaxed to $δ(G)> n/3$. In fact, we obtain a more general result for graphs of higher odd-girth.

For a $k$-uniform hypergraph $F$ let $\textrm{ex}(n,F)$ be the maximum number of edges of a $k$-uniform $n$-vertex hypergraph $H$ which contains no copy of $F$. Determining or estimating $\textrm{ex}(n,F)$ is a classical and central problem in extremal combinatorics. While for $k=2$ this problem is well understood, due to the work of Turán and of Erdős and Stone, only very little is known for $k$-uniform hypergraphs for $k>2$. We focus on the case when $F$ is a $k$-uniform hypergraph with three edges on $k+1$ vertices. Already this very innocent (and maybe somewhat particular looking) problem is still wide open even for $k=3$. We consider a variant of the problem where the large hypergraph $H$ enjoys additional hereditary density conditions. Questions of this type were suggested by Erd\H os and Sós about 30 years ago. We show that every $k$-uniform hypergraph $H$ with density $>2^{1-k}$ with respect to every large collections of $k$-cliques induced by sets of $(k-2)$-tuples contains a copy of $F$. The required density $2^{1-k}$ is best possible as higher order tournament constructions show. Our result can be viewed as a common generalisation of the first extremal result in graph theory due to Mantel (when $k=2$ and the hereditary density condition reduces to a normal density condition) and a recent result of Glebov, Král', and Volec (when $k=3$ and large subsets of vertices of $H$ induce a subhypergraph of density $>1/4$). Our proof for arbitrary $k\geq 2$ utilises the regularity method for hypergraphs.

We construct a Ramsey class whose objects are Steiner systems. In contrast to the situation with general $r$-uniform hypergraphs, it turns out that simply putting linear orders on their sets of vertices is not enough for this purpose: one also has to strengthen the notion of subobjects used from "induced subsystems" to something we call "strongly induced subsystems". Moreover we study the Ramsey properties of other classes of Steiner systems obtained from this class by either forgetting the order or by working with the usual notion of subsystems. This leads to a perhaps surprising induced Ramsey theorem in which designs get coloured.

We prove that for any given $\varepsilon>0$ and $d\in [0,1]$, every sufficiently large $(\varepsilon, d)$-dense graph $G$ contains for each odd integer $r$ at least $(d^r-\varepsilon)|V(G)|^r$ cycles of length $r$. Here, $G$ being $(\varepsilon, d)$-dense means that every set $X$ containing at least~$\varepsilon\,|V(G)|$ vertices spans at least $\tfrac d2\, |X|^2$ edges, and what we really count is the number of homomorphisms from an $r$-cycle into $G$. The result adresses a question of Y. Kohayakawa, B. Nagle, V. Rödl, and M. Schacht.

We prove that for any positive integers $n$ and $d$ there exists a collection consisting of $f=d\log n+O(1)$ subsets $A_1, A_2, \ldots, A_f$ of $[n]$ such that for any two distinct subsets $X$ and $Y$ of $[n]$ whose size is at most $d$ there is an index $i\in [f]$ for which $| A_i\cap X|$ and $|A_i\cap Y|$ have different parity. Here we think of $d$ as fixed whereas $n$ is thought of as tending to infinity, and the base of the logarithm is $2$. Translated into the language of combinatorial search theory, this tells us that \[ d \log n+O(1) \] queries suffice to identify up to $d$ marked items from a totality of $n$ items if the answers one gets are just whether an even or an odd number of marked elements has been queried, even if the search is performed non-adaptively. Since the entropy method easily yields a matching lower bound for the adaptive version of this problem, our result is asymptotically best possible. This answers a question posed by Dániel Gerbner and Balázs Patkós in Gyula O.H. Katona's Search Theory Seminar at the Rényi institute.

By a finite type-graph we mean a graph whose set of vertices is the set of all $k$-subsets of $[n]=\{1,2,\ldots, n\}$ for some integers $n\ge k\ge 1$, and in which two such sets are adjacent if and only if they realise a certain order type specified in advance. Examples of such graphs have been investigated in a great variety of contexts in the literature with particular attention being paid to their chromatic number. In recent joint work with Tomasz Łuczak, two of the authors embarked on a systematic study of the chromatic numbers of such type-graphs, formulated a general conjecture determining this number up to a multiplicative factor, and proved various results of this kind. In this article we fully prove this conjecture.

We investigate extremal problems for quasirandom hypergraphs. We say that a $3$-uniform hypergraph $H=(V,E)$ is $(d,η)$-quasirandom if for any subset $X\subseteq V$ and every set of pairs $P\subseteq V\times V$ the number of pairs $(x,(y,z))\in X\times P$ with $\{x,y,z\}$ being a hyperedge of $H$ is in the interval $d|X||P|\pmη|V|^3$. We show that for any $\varepsilon>0$ there exists $η>0$ such that every sufficiently large $(1/2+\varepsilon,η)$-quasirandom hypergraph contains a tetrahedron, i.e., four vertices spanning all four hyperedges. A known random construction shows that the density $1/2$ is best possible. This result is closely related to a question of Erdős, whether every weakly quasirandom $3$-uniform hypergraph $H$ with density bigger than $1/2$, i.e., every large subset of vertices induces a hypergraph with density bigger than $1/2$, contains a tetrahedron.

Extremal problems for $3$-uniform hypergraphs are known to be very difficult and despite considerable effort the progress has been slow. We suggest a more systematic study of extremal problems in the context of quasirandom hypergraphs. We say that a $3$-uniform hypergraph $H=(V,E)$ is weakly $(d,η)$-quasirandom if for any subset $U\subseteq V$ the number of hyperedges of $H$ contained in $U$ is in the interval $d\binom{|U|}{3}\pmη|V|^3$. We show that for any $\varepsilon>0$ there exists $η>0$ such that every sufficiently large weakly $(1/4+\varepsilon,η)$-quasirandom hypergraph contains four vertices spanning at least three hyperedges. This was conjectured by Erdős and Sós and it is known that the density $1/4$ is best possible. Recently, a computer assisted proof of this result based on the flag-algebra method was established by Glebov, Král', and Volec. In contrast to their work our proof presented here is based on the regularity method of hypergraphs and requires no heavy computations. In addition we obtain an ordered version of this result. The method of our proof allows us to study extremal problems of this type in a more systematic way and we discuss a few extensions and open problems here.

We investigate extremal problems for hypergraphs satisfying the following density condition. A $3$-uniform hypergraph $H=(V, E)$ is $(d, η,P_2)$-dense if for any two subsets of pairs $P$, $Q\subseteq V\times V$ the number of pairs $((x,y),(x,z))\in P\times Q$ with $\{x,y,z\}\in E$ is at least $d|\mathcal{K}_{P_2}(P,Q)|-η|V|^3,$ where $\mathcal{K}_{P_2}(P,Q)$ denotes the set of pairs in $P\times Q$ of the form $((x,y),(x,z))$. For a given $3$-uniform hypergraph $F$ we are interested in the infimum $d\geq 0$ such that for sufficiently small $η$ every sufficiently large $(d, η,P_2)$-dense hypergraph $H$ contains a copy of $F$ and this infimum will be denoted by $π_{P_2}(F)$. We present a few results for the case when $F=K_k^{(3)}$ is a complete three uniform hypergraph on $k$ vertices. It will be shown that $π_{P_2}(K_{2^r}^{(3)})\leq \frac{r-2}{r-1}$, which is sharp for $r=2,3,4$, where the lower bound for $r=4$ is based on a result of Chung and Graham [Edge-colored complete graphs with precisely colored subgraphs, Combinatorica 3 (3-4), 315-324].

We show that, given an infinite cardinal $μ$, a graph has colouring number at most $μ$ if and only if it contains neither of two types of subgraph. We also show that every graph with infinite colouring number has a well-ordering of its vertices that simultaneously witnesses its colouring number and its cardinality.

Turán's theorem is a cornerstone of extremal graph theory. It asserts that for any integer $r \geq 2$ every graph on $n$ vertices with more than ${\tfrac{r-2}{2(r-1)}\cdot n^2}$ edges contains a clique of size $r$, i.e., $r$ mutually adjacent vertices. The corresponding extremal graphs are balanced $(r-1)$-partite graphs. The question as to how many such $r$-cliques appear at least in any $n$-vertex graph with $γn^2$ edges has been intensively studied in the literature. In particular, Lovász and Simonovits conjectured in the 1970s that asymptotically the best possible lower bound is given by the complete multipartite graph with $γn^2$ edges in which all but one vertex class is of the same size while the remaining one may be smaller. Their conjecture was recently resolved for $r=3$ by Razborov and for $r=4$ by Nikiforov. In this article, we prove the conjecture for all values of $r$.