Table of Contents
Fetching ...

Tilings in quasi-random $k$-partite hypergraphs

Shumin Sun

TL;DR

The paper establishes a sharp density threshold for embedding balanced $k$-partite $k$-graph factors in $(p, ext{μ})$-dense, quasi-random host hypergraphs under a vanishing partite minimum codegree condition, proving that $p>1/2$ suffices when each part has size $n$ and the factor $F$ has $m$ vertices per part. It introduces a lattice-based absorption framework with two key absorbing lemmas tailored to different density regimes, and combines absorption with almost-tiling arguments to obtain perfect tilings. The work proves optimality of the $p>1/2$ threshold via probabilistic constructions and shows that, for $p>1/2$, the partite minimum codegree can vanish while still guaranteeing all fixed $k$-partite $k$-graph factors, while also delivering an asymptotic $n/2$ codegree threshold without quasi-randomness. Methodologically, the paper leverages the weak hypergraph regularity lemma, cluster hypergraphs, and reachability/closure tools to bridge local abundance of $F$-copies to global tiling. Overall, it advances the understanding of tilings in multipartite hypergraphs under quasi-randomness by combining probabilistic constructions, absorption, and regularity techniques to achieve tight thresholds and robust tiling results.

Abstract

Given $k\ge 2$ and two $k$-graphs ($k$-uniform hypergraphs) $F$ and $H$, an $F$-factor in $H$ is a set of vertex disjoint copies of $F$ that together cover the vertex set of $H$. Lenz and Mubayi were first to study the $F$-factor problems in quasi-random $k$-graphs with a minimum degree condition. Recently, Ding, Han, Sun, Wang and Zhou gave the density threshold for having all $3$-partite $3$-graphs factors in quasi-random $3$-graphs with vanishing minimum codegree condition $Ω(n)$. In this paper, we consider embedding factors when the host $k$-graph is $k$-partite and quasi-random with partite minimum codegree condition. We prove that if $p>1/2$ and $F$ is a $k$-partite $k$-graph with each part having $m$ vertices, then for $n$ large enough and $m\mid n$, any $p$-dense $k$-partite $k$-graph with each part having $n$ vertices and partite minimum codegree condition $Ω(n)$ contains an $F$-factor. We also present a construction showing that $1/2$ is best possible. Furthermore, for $1\leq \ell \leq k-2$, by constructing a sequence of $p$-dense $k$-partite $k$-graphs with partite minimum $\ell$-degree $Ω(n^{k-\ell})$ having no $K_k(m)$-factor, we show that the partite minimum codegree constraint can not be replaced by other partite minimum degree conditions. On the other hand, we prove that $n/2$ is the asymptotic partite minimum codegree threshold for having all fixed $k$-partite $k$-graph factors in sufficiently large host $k$-partite $k$-graphs even without quasi-randomness.

Tilings in quasi-random $k$-partite hypergraphs

TL;DR

The paper establishes a sharp density threshold for embedding balanced -partite -graph factors in -dense, quasi-random host hypergraphs under a vanishing partite minimum codegree condition, proving that suffices when each part has size and the factor has vertices per part. It introduces a lattice-based absorption framework with two key absorbing lemmas tailored to different density regimes, and combines absorption with almost-tiling arguments to obtain perfect tilings. The work proves optimality of the threshold via probabilistic constructions and shows that, for , the partite minimum codegree can vanish while still guaranteeing all fixed -partite -graph factors, while also delivering an asymptotic codegree threshold without quasi-randomness. Methodologically, the paper leverages the weak hypergraph regularity lemma, cluster hypergraphs, and reachability/closure tools to bridge local abundance of -copies to global tiling. Overall, it advances the understanding of tilings in multipartite hypergraphs under quasi-randomness by combining probabilistic constructions, absorption, and regularity techniques to achieve tight thresholds and robust tiling results.

Abstract

Given and two -graphs (-uniform hypergraphs) and , an -factor in is a set of vertex disjoint copies of that together cover the vertex set of . Lenz and Mubayi were first to study the -factor problems in quasi-random -graphs with a minimum degree condition. Recently, Ding, Han, Sun, Wang and Zhou gave the density threshold for having all -partite -graphs factors in quasi-random -graphs with vanishing minimum codegree condition . In this paper, we consider embedding factors when the host -graph is -partite and quasi-random with partite minimum codegree condition. We prove that if and is a -partite -graph with each part having vertices, then for large enough and , any -dense -partite -graph with each part having vertices and partite minimum codegree condition contains an -factor. We also present a construction showing that is best possible. Furthermore, for , by constructing a sequence of -dense -partite -graphs with partite minimum -degree having no -factor, we show that the partite minimum codegree constraint can not be replaced by other partite minimum degree conditions. On the other hand, we prove that is the asymptotic partite minimum codegree threshold for having all fixed -partite -graph factors in sufficiently large host -partite -graphs even without quasi-randomness.
Paper Structure (13 sections, 19 theorems, 33 equations)

This paper contains 13 sections, 19 theorems, 33 equations.

Key Result

Theorem 1.2

Let $k\geq 3$ be an integer. Given $0<\varepsilon, \alpha <1$, and a $k$-partite $k$-graph $F$ with each part having $m$ vertices, there exist an $n_0$ and $\mu >0$ such that the following holds for $n\geq n_0$. If a $(\frac{1}{2}+\varepsilon, \mu)$-dense $k$-partite $k$-graph $H$ with $n$ vertices

Theorems & Definitions (36)

  • Definition 1.1: ($p,\mu$)-denseness
  • Theorem 1.2
  • Theorem 1.3
  • Theorem 1.4
  • Theorem 1.5
  • proof : Proof of Theorem \ref{['cons1']}
  • proof : Proof of Theorem \ref{['cons2']}
  • Lemma 3.1: Absorbing Lemma I
  • Lemma 3.2: Absorbing Lemma II
  • Lemma 3.3
  • ...and 26 more