Lean source theorem: rational_cubic_primitive_integral_tail
ProvedErdosProblems.Erdos243.PaperCompleteR11.rational_cubic_primitive_integral_tailerdos-243support-lemma
For a positive rational cubic profile κn(n+1)(n+2)+η and positive natural recurrence coordinates, if the disagreement set has no lower-density bound of 1/4, there are a tail N, positive constant gcd g, positive integer m and integer c with mg=6κ and cg=η, and m,c have a Bézout identity. On that tail the divided coordinates are positive and coprime, satisfy the same exact recurrence, and agreement with the original rational profile is equivalent at each original index to agreement of C(n)/g with m times the rising binomial coefficient plus c.
Preamble
import Definitions.Def_ErdosProblems_Erdos243_PaperCompleteR9_PolynomialCorrections import Definitions.Def_ErdosProblems_Erdos243_PaperCompleteR11_WindowIncidence import Definitions.Def_ErdosProblems_Erdos243_PaperCompleteR11_CubicIntegralNormalisation import Mathlib import Mathlib.Algebra.Order.BigOperators.GroupWithZero.Finset import Mathlib.Algebra.Ring.Basic import Mathlib.Analysis.SpecificLimits.Basic import Mathlib.Data.Fin.Pigeonhole import Mathlib.Data.Int.GCD import Mathlib.Data.Nat.ChineseRemainder import Mathlib.Data.Nat.Choose.Basic import Mathlib.Data.Nat.Find import Mathlib.Data.Nat.Prime.Basic import Mathlib.Data.Rat.Defs import Mathlib.Data.ZMod.Basic import Mathlib.Data.ZMod.QuotientRing import Mathlib.Order.Filter.AtTopBot.Basic import Mathlib.Tactic import Mathlib.Tactic.Ring /-! # Constructed primitive cubic tails below quarter density The stabilising divisor, the exact natural quotient recurrence, the integral profile coefficients, and their Bezout certificate are produced from the original orbit. The index is never reset: division by a fixed gcd does not replace n by n-N in the polynomial. This eliminates a normalisation supplier, but does NOT prove that the primitive constant is a unit. Integral primitive profiles with other constants remain in the universal-density problem. -/ open ErdosProblems.Erdos243.PaperCompleteR11
Formal statement
theorem ErdosProblems.Erdos243.PaperCompleteR11.rational_cubic_primitive_integral_tail
(a C D : ℕ → ℕ) (κ η : ℚ) (hκ : 0 < κ)
(hCpos : ∀ n, 0 < C n) (hDpos : ∀ n, 0 < D n)
(hC : ∀ n, C (n + 1) + D n = a n * C n)
(hD : ∀ n, D (n + 1) = a n * D n)
(hlow : ¬ LowerDensityAtLeast
{n : ℕ | (C n : ℚ) ≠ κ * (n : ℚ) * ((n : ℚ) + 1) * ((n : ℚ) + 2) + η}
(1 / 4)) :
∃ (N g : ℕ) (m c : ℤ), 0 < g ∧ 0 < m ∧
(m : ℚ) * (g : ℚ) = 6 * κ ∧ (c : ℚ) * (g : ℚ) = η ∧
(∃ x y : ℤ, m * x + c * y = 1) ∧
∀ n, N ≤ n →
Nat.gcd (C n) (D n) = g ∧
0 < C n / g ∧ 0 < D n / g ∧
Nat.Coprime (C n / g) (D n / g) ∧
C (n + 1) / g + D n / g = a n * (C n / g) ∧
D (n + 1) / g = a n * (D n / g) ∧
((C n : ℚ) = κ * (n : ℚ) * ((n : ℚ) + 1) * ((n : ℚ) + 2) + η ↔
((C n / g : ℕ) : ℤ) = m * risingBinomial n + c) := by sorry
Source
Lean source (Apache-2.0): https://github.com/wcook04/plectis-erdos-lean/blob/4fe59e0191169606942baae0949365070b7f419c/ErdosProblems/Erdos243/PaperCompleteR11/CubicPrimitiveNormalisation.lean#L98-L187
Related paper by Will Cook (CC-BY-4.0): https://github.com/wcook04/plectis-erdos/blob/6917e15ec4abc2623512254da93221e446eeb707/paper/243/erdos-243-reciprocal-tail-rigidity.tex#L1-L58
Paper's AI-assistance disclosure: https://github.com/wcook04/plectis-erdos/blob/6917e15ec4abc2623512254da93221e446eeb707/paper/paper-house-style.sty#L180-L188
Original Erdős problem and Koizumi prior work are distinguished in the paper: https://github.com/wcook04/plectis-erdos/blob/6917e15ec4abc2623512254da93221e446eeb707/paper/243/erdos-243-reciprocal-tail-rigidity.tex#L57-L110