OAI/NumberTheory/EgyptianFractions/DensityAssembly.lean
12 declarations, 0 contain sorry, 0 declared as axioms
- def
OAI.Problem337.densityDivisorCountline 17Source documentation
The truncated divisor count appearing in the shifted moment hypothesis.
def densityDivisorCount (X : ℝ) (N : ℕ) : ℝ := …
- structure
OAI.Problem337.DensityBadFamilyline 23Source documentation
Explicit finite arithmetic data consumed by the density descent. No residue-distribution or divisor-moment estimate is asserted by this structure: these are the
many_residuesandmoment_boundfields that must be supplied.structure DensityBadFamily (S m ρ r c : ℝ) (d : ℕ) where …
- theorem
OAI.Problem337.card_le_cutoffline 54 - theorem
OAI.Problem337.DensityBadFamily.recurrenceline 62Source documentation
Pair counting and the actual shifted divisor moment give the normalized nonlinear recurrence, including inherited bad numerators and list multiplicity.
theorem DensityBadFamily.recurrence {S m ρ r c : ℝ} {d : ℕ} (F : DensityBadFamily S m ρ r c d) (hρ0 : 0 < ρ) (hρ1 : ρ ≤ 1) (hr : 1 < r) (j : ℕ) (hj : j < d) : ((F.bad j).card : ℝ) / F.cutoff j ≤ Real.exp (-c * m) + Real.exp (2 * S ^ (1 / 4 : ℝ)) * (((F.bad (j + 1)).card : ℝ) / F.cutoff (j + 1)) ^ (1 - 1 / r) := by …Mentions: OAI.Problem337.F, OAI.Problem337.DensityBadFamily
- theorem
OAI.Problem337.eventually_density_bad_card_le_eighthline 103Source documentation
The full counting/analytic density conclusion, conditional on explicit residue and shifted-divisor-moment data, uniformly in all finite families.
theorem eventually_density_bad_card_le_eighth (c K r : ℝ) (hc : 0 < c) (hK : 0 ≤ K) (hr : 2 ≤ r) (hKr : 8 * K ≤ r) : ∀ᶠ S : ℝ in atTop, ∀ (m ρ : ℝ) (d : ℕ) (F : DensityBadFamily S m ρ r c d), S / (2 * Real.log S) ≤ m → (d : ℝ) ≤ K * Real.log S → 0 < ρ → ρ ≤ 1 → ((F.bad 0).card : ℝ) ≤ F.cutoff 0 / 8 := by …Mentions: OAI.Problem337.F, OAI.Problem337.DensityBadFamily
- structure
OAI.Problem337.DensityResidueFamilyline 130Source documentation
Raw residue data before the good sets are constructed. The terminal expansions, small-residue supply and shifted divisor moments are explicit inputs; all finite propagation is proved below.
structure DensityResidueFamily (S m ρ r c : ℝ) (d : ℕ) where …
- def
OAI.Problem337.DensityResidueFamily.GoodSetsline 170Source documentation
The exact recursive good-set specification needed for both effects of the construction: closure counts bad residues, and decomposition propagates lists.
def DensityResidueFamily.GoodSets {S m ρ r c : ℝ} {d : ℕ} (F : DensityResidueFamily S m ρ r c d) (G : ℕ → Finset ℕ) : Prop := …Mentions: OAI.Problem337.F, OAI.Problem337.DensityResidueFamily
- def
OAI.Problem337.DensityResidueFamily.toBadFamilyline 182Source documentation
Every small successful index for a bad numerator lands at a bad numerator at the next level. In particular a zero residue cannot remain bad.
def DensityResidueFamily.toBadFamily {S m ρ r c : ℝ} {d : ℕ} (F : DensityResidueFamily S m ρ r c d) (G : ℕ → Finset ℕ) (hG : F.GoodSets G) : DensityBadFamily S m ρ r c d := by …Mentions: OAI.Problem337.F, OAI.Problem337.DensityResidueFamily, OAI.Problem337.DensityBadFamily
- theorem
OAI.Problem337.DensityResidueFamily.good_expansionsline 238Source documentation
The same good-set specification supplies genuine unit-fraction lists; this is not an additional expansion hypothesis at intermediate levels.
theorem DensityResidueFamily.good_expansions {S m ρ r c : ℝ} {d : ℕ} (F : DensityResidueFamily S m ρ r c d) (G : ℕ → Finset ℕ) (hG : F.GoodSets G) : ∀ j ≤ d, ∀ u ∈ G j, Descent.UnitListAtMost ((u : ℚ) / (F.denominator * F.factor j : ℕ)) (F.baseLength + (d - j)) := by …Mentions: OAI.Problem337.F, OAI.Problem337.DensityResidueFamily, OAI.Problem337.Descent.UnitListAtMost
- theorem
OAI.Problem337.eventually_residue_family_denseline 258Source documentation
Complete finite/analytic density construction from raw residue and moment supply. The intermediate good sets and their expansions are constructed here.
theorem eventually_residue_family_dense (c K r : ℝ) (hc : 0 < c) (hK : 0 ≤ K) (hr : 2 ≤ r) (hKr : 8 * K ≤ r) : ∀ᶠ S : ℝ in atTop, ∀ (m ρ : ℝ) (d : ℕ) (F : DensityResidueFamily S m ρ r c d), S / (2 * Real.log S) ≤ m → (d : ℝ) ≤ K * Real.log S → 0 < ρ → ρ ≤ 1 → ∃ G : Finset ℕ, G ⊆ Finset.Icc 1 ⌊F.cutoff 0⌋₊ ∧ ((Finset.Icc 1 ⌊F.cutoff 0⌋₊ \ G).card : ℝ) ≤ F.cutoff 0 / 8 ∧ ∀ u ∈ G, Descent.UnitListAtMost ((u : ℚ) / (F.denominator * F.factor 0 : ℕ)) (F.baseLength + d) := by …Mentions: OAI.Problem337.F, OAI.Problem337.DensityResidueFamily, OAI.Problem337.Descent.UnitListAtMost
- theorem
OAI.Problem337.Descent.UnitListAtMost.to_density_list_twoline 278Source documentation
Conversion to the list interface used by the elementary dense-family reduction, retaining the same length bound.
theorem Descent.UnitListAtMost.to_density_list_two {x : ℚ} {B : ℕ} (H : Descent.UnitListAtMost x B) (hx : x < 1) : ∃ l : List ℕ, (∀ n ∈ l, 2 ≤ n) ∧ l.length ≤ B ∧ (l.map (fun n : ℕ => (1 : ℚ) / (n : ℚ))).sum = x := by …Mentions: OAI.Problem337.Descent.UnitListAtMost
- theorem
OAI.Problem337.eventually_residue_family_dense_listsline 292Source documentation
A directly consumable list version of the density proposition. The only additional hypotheses are the supplied terminal-length budget and the fact that all represented fractions lie below one.
theorem eventually_residue_family_dense_lists (c K r : ℝ) (hc : 0 < c) (hK : 0 ≤ K) (hr : 2 ≤ r) (hKr : 8 * K ≤ r) : ∀ᶠ S : ℝ in atTop, ∀ (m ρ L : ℝ) (d : ℕ) (F : DensityResidueFamily S m ρ r c d), S / (2 * Real.log S) ≤ m → (d : ℝ) ≤ K * Real.log S → 0 < ρ → ρ ≤ 1 → ((F.baseLength + d : ℕ) : ℝ) ≤ L * Real.log S → F.cutoff 0 < ((F.denominator * F.factor 0 : ℕ) : ℝ) → ∃ G : Finset ℕ, ((Finset.Icc 1 ⌊F.cutoff 0⌋₊ \ G).card : ℝ) ≤ F.cutoff 0 / 8 ∧ ∀ u ∈ G, ∃ ds : List ℕ, (∀ n ∈ ds, 2 ≤ n) ∧ (ds.length : ℝ) ≤ L * Real.log S ∧ (ds.map (fun n : ℕ => (1 : ℚ) / (n : ℚ))).sum = …Mentions: OAI.Problem337.F, OAI.Problem337.DensityResidueFamily