Documentation

Atlas.ProbabilisticMethodsInCombinatorics.code.GalvinTetali

structure GalvinTetali.LoopGraph (W : Type u_1) :
Type u_1

A LoopGraph on a vertex set $W$: a symmetric relation Adj on $W$, allowing self-loops (unlike SimpleGraph). Used as the target graph $H$ in graph homomorphism counts.

  • Adj : WWProp
  • symm (u v : W) : self.Adj u vself.Adj v u
Instances For
    def GalvinTetali.IsGraphHom {V : Type u_1} {W : Type u_2} (G : SimpleGraph V) (H : LoopGraph W) (φ : VW) :

    IsGraphHom G H φ says that $\varphi : V \to W$ is a graph homomorphism from the simple graph $G$ on $V$ to the loop graph $H$ on $W$: adjacent vertices map to adjacent vertices.

    Instances For
      noncomputable def GalvinTetali.homCount {V : Type u_1} {W : Type u_2} [Fintype V] [Fintype W] (G : SimpleGraph V) (H : LoopGraph W) :

      The number $\mathrm{hom}(G, H)$ of graph homomorphisms from a finite simple graph $G$ to a finite loop graph $H$.

      Instances For
        theorem GalvinTetali.galvin_tetali {V : Type u_1} {W : Type u_2} [Fintype V] [Fintype W] (G : SimpleGraph V) (H : LoopGraph W) (d : ) (hd : 0 < d) (hreg : G.IsRegularOfDegree d) (hbip : G.IsBipartite) :
        (homCount G H) (homCount (completeBipartiteGraph (Fin d) (Fin d)) H) ^ ((Fintype.card V) / (2 * d))

        Theorem 10.4.14 (Galvin–Tetali 2004). For every bipartite $d$-regular graph $G$ on $n$ vertices and any loop graph $H$, $\mathrm{hom}(G, H) \le \mathrm{hom}(K_{d,d}, H)^{n/(2d)}$.