OAI/Geometry/HyperbolicGroups/RankSelection.lean

GitHub source file

4 declarations, 0 contain sorry, 0 declared as axioms

  1. lemma OAI.HyperbolicObstruction.FiniteLaw.bernoulliProduct_sum_expect line 12
  2. lemma OAI.HyperbolicObstruction.IncidenceTemplate.Occurrences.expect_join line 31
  3. lemma OAI.HyperbolicObstruction.IncidenceTemplate.Occurrences.rank_bad_join_bound line 47
  4. theorem OAI.HyperbolicObstruction.IncidenceTemplate.Occurrences.exists_tables_of_estimate line 60

    Source documentation

    Simultaneous selection with the genuine bounded-rank event, not independent new copies of the arrays used for sampling.

    theorem exists_tables_of_estimate {J : Type*} [Fintype J]
        (T : J → ConditionalTest) (M n : ℕ) {α β δ ε τ : ℝ}
        (ha : 0 ≤ α) (hb : 7*α ≤ β) (hd : 0 ≤ δ)
        (hsmall : ∀ j, ((T j).k+1:ℝ)*(β+90*δ) ≤ 1/2)
        (hN : ∀ j, (T j).threshold ≤ Fintype.card (SymmetricAlphabet n))
        (hε0 : 0 < ε) (hε1 : ε < 1) (hτ0 : 0 < τ) (hτ2 : τ ≤ 2)
        (hest : (∑ j, (T j).rate (Fintype.card (SymmetricAlphabet n))) +
          ε*(60*(Fintype.card (SymmetricAlphabet n):ℝ)^2*
            Real.exp (-τ^2*(Fintype.card (SymmetricAlphabet n):ℝ)^(1-δ)/4) +
            rankRate M α β δ (Fintype.card (SymmetricAlphabet n))) < ε) :
        ∃ (w : TableModel.Entry (A := SymmetricAlphabet n) rowSupport → Bool)
    …

    Mentions: OAI.HyperbolicObstruction.IncidenceTemplate.ConditionalTest, OAI.HyperbolicObstruction.IncidenceTemplate.SymmetricAlphabet, OAI.HyperbolicObstruction.IncidenceTemplate.Occurrences.rankRate, OAI.HyperbolicObstruction.TableModel.Entry, OAI.HyperbolicObstruction.IncidenceTemplate.rowSupport