The source parity indices coincide with the actual recurrence carrier.
theorem
MathlibNt.SieveTheory.eventually_lemma144_caseII_odd_rawRoundedFinal
(S : BoundingSieve)
(H : SwitchingPrinciple.SuzukiLemma144KappaOne.Section13HatLayers)
{Dmin : ℕ}
{d Δ C K : ℝ}
(hH : SwitchingPrinciple.SuzukiLemma144KappaOne.Section13HatSourceContract H)
(hd1 : 1 < d)
(hΔ0 : 0 < Δ)
(hΔ1 : Δ < 1)
(hC3 : 3 ≤ C)
(hK : 2 ≤ K)
(hlocal : SwitchingPrinciple.HasDimensionOneLocalProductBound S K)
(hDmin : 2 ≤ Dmin)
:
∀ᶠ (D : ℕ) in Filter.atTop, ∀ (N : ℕ) (s : ℝ),
Odd N →
3 ≤ N →
1 < s →
s ≤ 3 →
SwitchingPrinciple.SuzukiLemma144Equation1410.GlobalDepthLemma144InductionHypothesis (suzukiActualT S)
(fun (p : ℕ) => SwitchingPrinciple.suzukiVProduct S ↑p)
(fun (n D' : ℕ) (x : ℝ) => SwitchingPrinciple.SuzukiLemma144KappaOne.errorEnvelope H n (↑D') d x) 2 C K
Δ (N - 1) Dmin →
have z := ⌈↑D ^ (1 / s)⌉₊;
∑ n ∈ sourceParityIndices N, suzukiSourceV S n D z ≤ caseIIOddClaim145Endpoint S d N D + SwitchingPrinciple.suzukiVProduct S ↑z * (SuzukiFiniteContinuousLayers.finiteSourceLayer 1 2 N s + C * Real.exp √K * SwitchingPrinciple.SuzukiLemma144KappaOne.errorEnvelope H N (↑D) d s * Real.log ↑D ^ (-Δ) * caseIIConcreteRoundedRelativeBracket N (↑D) d Δ
(SwitchingPrinciple.SuzukiLemma144KappaOne.sourceSigma (↑D) d) C K)
The genuine large-parameter raw rounded Case-II producer. Its cutoff is
chosen before D, N, and s; the only recursive input left at a particular
N is the global depth-N-1 induction hypothesis.