introvoyz045's picture
Migrated from GitHub
4a56cd6 verified
Raw History Blame Contribute Delete
2.53 kB
import Mathlib.Data.Fintype.Powerset
import Mathlib.Basic.Real.Basic
import Mathlib.Algebra.BigOperators.Ring.Finset
import Mathlib.Combinatorics.SimpleGraph.Copy
import Mathlib.Analysis.SpecialFunctions.Log.Basic
namespace OAI
noncomputable section
universe u
namespace LeanBlast.SecondKahnKalai
def bernoulliWeight {X : Type u} [Fintype X] [DecidableEq X]
(p : ℝ) (W : Finset X) : ℝ :=
∏ x, if x ∈ W then p else 1 - p
def bernoulliExpectation {X : Type u} [Fintype X] [DecidableEq X]
(p : ℝ) (f : Finset X β†’ ℝ) : ℝ :=
βˆ‘ W, bernoulliWeight p W * f W
def bernoulliProbability {X : Type u} [Fintype X] [DecidableEq X]
(p : ℝ) (P : Finset X β†’ Prop) : ℝ := by
classical
exact βˆ‘ W, if P W then bernoulliWeight p W else 0
/-- The unordered non-loop edges of the complete graph on `n` vertices. -/
abbrev Edge (n : β„•) := {e : Sym2 (Fin n) // Β¬e.IsDiag}
/-- A set of edges, interpreted as an ordinary graph on all `n` vertices. -/
def graphOfEdges {n : β„•} (W : Finset (Edge n)) : SimpleGraph (Fin n) :=
SimpleGraph.fromEdgeSet (Subtype.val '' (W : Set (Edge n)))
/-- The number of unordered edges, each counted once. -/
def edgeCount {V : Type u} (H : SimpleGraph V) : β„• := Nat.card H.edgeSet
/-- Expected number of actual unlabelled non-induced copies in `G(n,p)`. -/
def expectedCopies {V : Type u} (n : β„•) (p : ℝ) (H : SimpleGraph V) : ℝ :=
bernoulliExpectation p (fun W : Finset (Edge n) =>
((graphOfEdges W).copyCount H : ℝ))
/-- Probability that `G(n,p)` contains at least one non-induced copy. -/
def containmentProbability {V : Type u} (n : β„•) (p : ℝ)
(H : SimpleGraph V) : ℝ :=
bernoulliProbability p (fun W : Finset (Edge n) =>
0 < (graphOfEdges W).copyCount H)
def criticalThreshold {V : Type u} (n : β„•) (H : SimpleGraph V) : ℝ :=
sInf {p : ℝ | p ∈ Set.Icc 0 1 ∧ (1 / 2 : ℝ) ≀ containmentProbability n p H}
def expectationThreshold {V : Type u} (n : β„•) (H : SimpleGraph V) : ℝ :=
sInf {p : ℝ | p ∈ Set.Icc 0 1 ∧
βˆ€ F : H.Subgraph, (1 / 2 : ℝ) ≀ expectedCopies n p F.coe}
def logTwo (x : ℝ) : ℝ := Real.log x / Real.log 2
def SecondKahnKalaiBounds {V : Type u} (n : β„•) (H : SimpleGraph V) : Prop :=
criticalThreshold n H ≀
min 1 (2048 * Real.exp 50 * expectationThreshold n H *
(1 + logTwo (edgeCount H : ℝ))) ∧
criticalThreshold n H ≀
6144 * Real.exp 50 * expectationThreshold n H * logTwo (n : ℝ)
end LeanBlast.SecondKahnKalai
end
end OAI