Stable precision parameters exist below unit noise
ProvedCouret.STROBO2TORS.recoh_stable_precision_parameters_existcouret-programmecouret-unification-20261108
For rational , define . Put , , , , and . All are rational and are interpreted as real where limits are stated. For every rational , there exist with
Preamble
import Definitions.Def_Couret_RecohRationalRoots_2e1442988636 import Mathlib.Analysis.Complex.Norm import Mathlib.Analysis.Real.Pi.Bounds import Mathlib.Analysis.SpecialFunctions.Complex.Arctan import Mathlib.Analysis.SpecialFunctions.Trigonometric.Bounds import Mathlib.Analysis.SpecificLimits.Normed import Mathlib.Tactic import Mathlib.Tactic.Linarith import Mathlib.Tactic.NormNum import Mathlib.Tactic.Positivity import Mathlib.Tactic.Ring import Mathlib.Topology.Algebra.Order.Floor import Theorems.Thm_Couret_STROBO2TORS_recoh_pi_partial_eventually_abs_le_four import Definitions.Def_Couret_RecohRationalRoots_2e1442988636 import Mathlib.Topology.Algebra.Order.Floor set_option autoImplicit false /-! Existence of finite rational precision certificates for every positive error budget. These theorems concern the explicit error expressions checked by the decoder. They do not assert refinement of the Python precision-search loop or of its JSON interface. -/ namespace Couret.STROBO2TORS open Filter Topology
Formal statement
theorem recoh_stable_precision_parameters_exist (epsilon : ℚ) (hepsilon : epsilon < 1) :
∃ m n : ℕ, 1 ≤ m ∧ |recohPiPartial m| ≤ 4 ∧ 15 ≤ n ∧
3 * recohRationalPhaseError m n ≤ (1 - epsilon) / 4 := by sorry
end Couret.STROBO2TORS
Source
Alexandre Couret, https://github.com/alexcour/couret-finite-verifications/blob/eb7323762a9fa64f54f95ea7aa2981dee0ba54af/research/strobo-2tors-2026-10-12/lean/RecohStablePrecision.lean#L90-L97, lines 90–97; source file SHA-256 75737fd29409c44da0d31022f86d5e77f03903b260a104075e16bef2fff8543d; theorem Couret.STROBO2TORS.recoh_stable_precision_parameters_exist