An explicit finite Gaussian-series enclosure for the dyadic theta function
ProvedMasulloRH.theta_gaussian_truncation_errorDefine the three-scale Gaussian
and the logarithmic theta function
For and , let
Then the finite symmetric truncation has the explicit error bound
The denominator is strictly positive for every allowed . The estimate includes , when the finite sum is empty. It bounds the actual theta function used by the mission's Gaussian approximation definition and needs no Riemann-hypothesis, Mertens, or density assumption.
This is a pointwise truncation estimate useful for rigorous theta evaluation. Additional control of quadrature and integration tails is needed to turn it into enclosures for the mission's weighted Gram integrals; that stronger numerical-certification procedure is not asserted here.
Formalization Note. The theta function and the finite sum are complex-valued with real summands; the left side is their complex norm. The finite sum is indexed in Lean by , with Gaussian argument .
import Definitions.Def_MasulloRH_GaussianThetaApproximation set_option autoImplicit false open scoped BigOperators open MasulloRH.GaussianApproximation
theorem MasulloRH.theta_gaussian_truncation_error (x : ℝ) (K : ℕ) :
‖theta x - 2 * ∑ n ∈ Finset.range K,
((Real.exp (-Real.pi * (((n : ℝ) + 1) * Real.exp x) ^ 2) -
3 * Real.exp (-Real.pi * (2 * (((n : ℝ) + 1) * Real.exp x)) ^ 2) +
2 * Real.exp (-Real.pi * (4 * (((n : ℝ) + 1) * Real.exp x)) ^ 2) : ℝ) : ℂ)‖ ≤
12 * Real.exp (-Real.pi * (Real.exp x) ^ 2 * ((K : ℝ) + 1) ^ 2) /
(1 - Real.exp (-Real.pi * (Real.exp x) ^ 2 * (2 * (K : ℝ) + 3))) := by sorry