Documentation

LeanPool.ErdosGinzburgZiv.EGZ.Decomposition.IterationColorCapacity

Combining the three event capacities #

noncomputable def EGZ.FlagDecomposition.Iteration.progressColor {p d : ℕ} [NeZero p] [Fact (Nat.Prime p)] {f : FpCoord p d → ℕ} {s : ℕ → State p d f} {ε : ℝ} {δ : ℕ → ℝ} {g : ℕ → ℕ} {N : ℕ} (P : (i : ℕ) → i < N → Progress (s i) (s (i + 1)) ε (δ i) g) (i : ℕ) :

Colors outside the finite run are arbitrary and never used.

Equations
Instances For
    theorem EGZ.FlagDecomposition.Iteration.progressColor_eq {p d : ℕ} [NeZero p] [Fact (Nat.Prime p)] {f : FpCoord p d → ℕ} {s : ℕ → State p d f} {ε : ℝ} {δ : ℕ → ℝ} {g : ℕ → ℕ} {N : ℕ} (P : (i : ℕ) → i < N → Progress (s i) (s (i + 1)) ε (δ i) g) (i : ℕ) (hi : i < N) :
    progressColor P i = (P i hi).event.color
    theorem EGZ.FlagDecomposition.Iteration.progressColor_lt {p d : ℕ} [NeZero p] [Fact (Nat.Prime p)] {f : FpCoord p d → ℕ} {s : ℕ → State p d f} {ε : ℝ} {δ : ℕ → ℝ} {g : ℕ → ℕ} {N : ℕ} (P : (i : ℕ) → i < N → Progress (s i) (s (i + 1)) ε (δ i) g) (i : ℕ) (hi : i < N) :
    progressColor P i < 2 * (d + 1) ^ 2 + 1
    theorem EGZ.FlagDecomposition.Iteration.hasIntervalCapacity_of_event_bounds {p d : ℕ} [NeZero p] [Fact (Nat.Prime p)] {f : FpCoord p d → ℕ} {s : ℕ → State p d f} {ε : ℝ} {δ : ℕ → ℝ} {g : ℕ → ℕ} {N : ℕ} (P : (i : ℕ) → i < N → Progress (s i) (s (i + 1)) ε (δ i) g) (hδ : Antitone δ) (hδnonneg : ∀ (i : ℕ), 0 ≤ δ i) (hcomplete : ∀ (a b L : ℕ), a ≤ b → b < N → L < (d + 1) ^ 2 → (∀ i ∈ Finset.Icc a b, 2 * L ≤ progressColor P i) → {i ∈ Finset.Icc a b | progressColor P i = 2 * L}.card ≤ 2 ^ a) (hface : ∀ (a b L : ℕ), a ≤ b → b < N → L < (d + 1) ^ 2 → (∀ i ∈ Finset.Icc a b, 2 * L + 1 ≤ progressColor P i) → {i ∈ Finset.Icc a b | progressColor P i = 2 * L + 1}.card ≤ intervalCapacity d ε a) :

    Complete and face event bounds, together with the checked gap bound, give precisely the interval-capacity hypothesis of the finite-color lemma.

    theorem EGZ.FlagDecomposition.Iteration.length_lt_stoppingBound_of_event_bounds {p d : ℕ} [NeZero p] [Fact (Nat.Prime p)] {f : FpCoord p d → ℕ} {s : ℕ → State p d f} {ε : ℝ} {δ : ℕ → ℝ} {g : ℕ → ℕ} {N : ℕ} (P : (i : ℕ) → i < N → Progress (s i) (s (i + 1)) ε (δ i) g) (hδ : Antitone δ) (hδnonneg : ∀ (i : ℕ), 0 ≤ δ i) (hcomplete : ∀ (a b L : ℕ), a ≤ b → b < N → L < (d + 1) ^ 2 → (∀ i ∈ Finset.Icc a b, 2 * L ≤ progressColor P i) → {i ∈ Finset.Icc a b | progressColor P i = 2 * L}.card ≤ 2 ^ a) (hface : ∀ (a b L : ℕ), a ≤ b → b < N → L < (d + 1) ^ 2 → (∀ i ∈ Finset.Icc a b, 2 * L + 1 ≤ progressColor P i) → {i ∈ Finset.Icc a b | progressColor P i = 2 * L + 1}.card ≤ intervalCapacity d ε a) :