Documentation

Anderson4D.DetParametrix.Paper42_Moment.R324FullFullStep1NormalizedClosure

Paper-faithful incoming input for R-324 Step 1 #

In (4.17) the kernel on the incoming leg is the input denoted G₀ in the paper. Proper intervals lying to the left of the retained terminal block may already have been collapsed into this leg, so it is not in general the free Green kernel. The only property used for the frequency-independent Step-1 estimate is the output of (4.13): measurability and the bound |G₀(z)| ≤ A |z|⁻².

This file is the minimal generalization of R324PaperStep1.lean needed at that seam. The terminal primitive kernel and all of the y/cosine/inserted kernel argument are unchanged. The incoming x integral is paid in L¹, as in the paper, rather than being incorrectly identified with the Fourier multiplier of greenFn.

The corrected form of (4.17) #

noncomputable def Anderson4D.r324Step1IntegrandWithIncoming (G0 J : T4 → ℝ) (α β : Z4) (xa xm x y : T4) :

The integrand (4.17) with the actual incoming removal output G₀.

Equations
  • One or more equations did not get rendered due to their size.
Instances For
    noncomputable def Anderson4D.r324Step1IntegralWithIncoming (G0 J : T4 → ℝ) (α β : Z4) :

    The corrected four-fold Step-1 amplitude.

    Equations
    • One or more equations did not get rendered due to their size.
    Instances For
      noncomputable def Anderson4D.r324IncomingMode (G0 : T4 → ℝ) (α : Z4) :

      The untranslated Fourier coefficient of the incoming removal output.

      Equations
      Instances For
        theorem Anderson4D.integral_charT4_mul_incoming_shift (G0 : T4 → ℝ) (α : Z4) (xa : T4) :
        ∫ (x : T4), charT4 α x * ↑(G0 (x - xa)) ∂paperMeasure = charT4 α xa * r324IncomingMode G0 α

        Translating the incoming leg extracts only its endpoint character; no special Fourier formula for G₀ is used.

        The unchanged endpoint and symmetry calculation #

        theorem Anderson4D.r324Step1WithIncoming_integral_y (G0 J : T4 → ℝ) (α β : Z4) (xa xm x : T4) :
        ∫ (y : T4), r324Step1IntegrandWithIncoming G0 J α β xa xm x y ∂paperMeasure = charT4 α x * ↑(G0 (x - xa)) * ↑(J (xa - xm)) * (↑(paperSecondOrderModeDecay β) * (charT4 β xm - charT4 β xa))

        The y integral is exactly the one in the original Step-1 proof.

        theorem Anderson4D.r324Step1WithIncoming_integral_x (G0 J : T4 → ℝ) (α β : Z4) (xa xm : T4) :
        ∫ (x : T4), ∫ (y : T4), r324Step1IntegrandWithIncoming G0 J α β xa xm x y ∂paperMeasure ∂paperMeasure = charT4 α xa * r324IncomingMode G0 α * (↑(J (xa - xm)) * (↑(paperSecondOrderModeDecay β) * (charT4 β xm - charT4 β xa)))

        The x integral records the arbitrary incoming Fourier coefficient.

        theorem Anderson4D.r324Step1WithIncoming_integral_xm (G0 : T4 → ℝ) {J : T4 → ℝ} (hJ : MemEClassT4 J) (α β : Z4) (xa : T4) (hcos : MeasureTheory.Integrable (fun (u : T4) => J u * (r324CharacterCos β u - 1)) paperMeasure) (hsin : MeasureTheory.Integrable (fun (u : T4) => J u * r324CharacterSin β u) paperMeasure) :

        The terminal primitive kernel still supplies the identical E-symmetry cancellation.

        Paying the incoming removal output in L¹ #

        theorem Anderson4D.norm_r324IncomingMode_le {G0 : T4 → ℝ} (hG0meas : Measurable G0) {A0 : ℝ} (hA0 : 0 ≤ A0) (hG0 : ∀ (z : T4), z ≠ 0 → |G0 z| ≤ A0 * invSqKer z) (α : Z4) :

        The Fourier coefficient of any (4.13) input costs only its L¹ inverse-square majorant.

        theorem Anderson4D.r324Step1WithIncoming_norm_le {G0 J : T4 → ℝ} (hG0meas : Measurable G0) {A0 : ℝ} (hA0 : 0 ≤ A0) (hG0 : ∀ (z : T4), z ≠ 0 → |G0 z| ≤ A0 * invSqKer z) (hJ : MemEClassT4 J) (α β : Z4) (hcos : MeasureTheory.Integrable (fun (u : T4) => J u * (r324CharacterCos β u - 1)) paperMeasure) (hsin : MeasureTheory.Integrable (fun (u : T4) => J u * r324CharacterSin β u) paperMeasure) :

        The generalized Step-1 integral differs from the old one only by the incoming L¹ factor.

        The corrected normalized reduction interface #

        def Anderson4D.R324Step1ReductionWithIncoming (ρ : SmoothCutoff) (lam ε : ℝ) (m : ℕ) (α β : Z4) (Cred : ℝ) (P : ℂ) :

        Minimal paper-faithful replacement for R324Step1Reduction: the incoming output of the proper-prefix removal is explicit and carries its own scale.

        Equations
        • One or more equations did not get rendered due to their size.
        Instances For

          The genuine terminal trace: the inserted-majorant estimate #

          Removing the genuinely empty post-terminal carrier #

          The carrier after the retained terminal head is empty, hence the apparently retained post-head Haar integral has mass one. This is the normalization fact deliberately left literal by the earlier Fubini bridge.

          At the singleton stop the complete active-scale carrier is exactly the incoming predecessor, the terminal internal chain, and the one outgoing Green slot. This is the local product decomposition needed to compare the actual Step-1 charge with the global budget invariant.

          Absolute integrability of the corrected four-point Step-1 kernel #

          The corrected four-point integrand is absolutely integrable whenever the incoming removal output and the retained primitive kernel are in L¹. This is the Fubini license used below; no estimate is taken here.

          The nested order used in the paper's display (4.17) is the grouped (z,w)-then-(x,y) product integral.

          Actual terminal inputs are in L¹ #

          theorem Anderson4D.R324WithinHalfResidualPrefix.R324FullPairingBudgetTerminalAdapter.lamEps_pow_integral_terminalRawLocal_eq_step1Section {ρ : SmoothCutoff} {C lam ε K A : ℝ} {q : ℕ} {κ : PartialPairing (Fin (2 * q))} {budget : R324FullPairingBudgetStopTrace κ} (data : R324FullPairingBudgetTerminalAdapter budget) (hε : 0 < ε) (hε1 : ε ≤ 1) (x y : T4) (hstop : MeasureTheory.Integrable (fun (v : data.geometry.trace.stopPrefix.SurvivingCoordinate → T4) => ↑(data.geometry.trace.stopPrefix.residualIntegrand ρ ε x y (data.geometry.trace.stopPrefix.reconstruct v))) (MeasureTheory.Measure.pi fun (x : data.geometry.trace.stopPrefix.SurvivingCoordinate) => paperMeasure)) :

          The terminal raw-local block, after the honest standard-block change of coordinates, is exactly the four physical kernels of (4.17). In particular the incoming factor is the actual predecessor edge of the reached trace.

          Exact paper-(4.17) representation of the complete terminal Fourier integral. This is the missing normalization seam: the incoming kernel is the reached predecessor edge, not an unjustified fresh copy of greenFn. All changes of integration order are licensed by the stopped-residual and four-point L¹ theorems.

          Proposition 4.1 applied to the normalized actual terminal edges makes the heterogeneous retained primitive kernel genuinely integrable.

          theorem Anderson4D.R324WithinHalfResidualPrefix.R324FullPairingBudgetTerminalAdapter.paperDecay_mul_norm_integral_terminalPrimitiveKernelDiff_le_scaled_inserted {ρ : SmoothCutoff} {C lam ε K A : ℝ} {q : ℕ} {κ : PartialPairing (Fin (2 * q))} {budget : R324FullPairingBudgetStopTrace κ} (data : R324FullPairingBudgetTerminalAdapter budget) (hε : 0 < ε) (supportConstant primitiveConstant : ℝ) (hprimitive : 0 ≤ primitiveConstant) (hlam : 0 ≤ lam) (hprop : ∀ (H : Fin (2 * residualBlockOrder data.geometry.terminalData.terminal.2 - 1) → T4 → ℝ), IsAdmissiblePrimitiveInput (residualBlockOrder data.geometry.terminalData.terminal.2) H → MemEClassT4 (primitiveKernelDiff ρ lam ε (residualBlockOrder data.geometry.terminalData.terminal.2) ⋯ H) ∧ MemEClassT4 (primitiveKernelInsertedDiff ρ lam ε (residualBlockOrder data.geometry.terminalData.terminal.2) ⋯ H) ∧ PrimitiveKernelBounds ρ lam ε (residualBlockOrder data.geometry.terminalData.terminal.2) ⋯ H supportConstant primitiveConstant) (β : Z4) :

          The terminal primitive phase with all genuine internal edge scales, paid by the inserted majorant of (4.4). This is the frequency-independent analogue of norm_integral_terminalPrimitiveKernelDiff_mul_..._le_scaled; the latter deliberately uses the Step-4 bound |cos-1| ≤ 2, whereas Step 1 keeps the quadratic cosine gain.

          The proof takes the modulus only after the signed primitive-pairing sum has been integrated and the E-symmetry cancellation has been applied.

          theorem Anderson4D.R324WithinHalfResidualPrefix.R324FullPairingBudgetTerminalAdapter.fullPairingHalfTerminalEstimateInserted {ρ : SmoothCutoff} {C lam ε K A : ℝ} {q : ℕ} {κ : PartialPairing (Fin (2 * q))} {budget : R324FullPairingBudgetStopTrace κ} (data : R324FullPairingBudgetTerminalAdapter budget) (x z : T4) (hxz : x ≠ z) (hε : 0 < ε) (hε1 : ε ≤ 1) (supportConstant primitiveConstant : ℝ) (hprimitive : 0 ≤ primitiveConstant) (hlam : 0 ≤ lam) (hprop : ∀ (H : Fin (2 * residualBlockOrder data.geometry.terminalData.terminal.2 - 1) → T4 → ℝ), IsAdmissiblePrimitiveInput (residualBlockOrder data.geometry.terminalData.terminal.2) H → MemEClassT4 (primitiveKernelDiff ρ lam ε (residualBlockOrder data.geometry.terminalData.terminal.2) ⋯ H) ∧ MemEClassT4 (primitiveKernelInsertedDiff ρ lam ε (residualBlockOrder data.geometry.terminalData.terminal.2) ⋯ H) ∧ PrimitiveKernelBounds ρ lam ε (residualBlockOrder data.geometry.terminalData.terminal.2) ⋯ H supportConstant primitiveConstant) (β : Z4) :

          Fixed-endpoint Step 1 for the actual terminal trace. The predecessor is the paper's G₀; its certificate scale is charged exactly once. The terminal internal edge scales are charged by the preceding theorem, while the outgoing Green difference is Fourier-integrated without a norm.

          theorem Anderson4D.R324WithinHalfResidualPrefix.R324FullPairingBudgetTerminalAdapter.norm_terminalRawLocalFourierIntegral_le_insertedScaleLedger {ρ : SmoothCutoff} {C lam ε K A : ℝ} {q : ℕ} {κ : PartialPairing (Fin (2 * q))} {budget : R324FullPairingBudgetStopTrace κ} (data : R324FullPairingBudgetTerminalAdapter budget) (hε : 0 < ε) (hε1 : ε ≤ 1) (supportConstant primitiveConstant : ℝ) (hprimitive : 0 ≤ primitiveConstant) (hlam : 0 ≤ lam) (hprop : ∀ (H : Fin (2 * residualBlockOrder data.geometry.terminalData.terminal.2 - 1) → T4 → ℝ), IsAdmissiblePrimitiveInput (residualBlockOrder data.geometry.terminalData.terminal.2) H → MemEClassT4 (primitiveKernelDiff ρ lam ε (residualBlockOrder data.geometry.terminalData.terminal.2) ⋯ H) ∧ MemEClassT4 (primitiveKernelInsertedDiff ρ lam ε (residualBlockOrder data.geometry.terminalData.terminal.2) ⋯ H) ∧ PrimitiveKernelBounds ρ lam ε (residualBlockOrder data.geometry.terminalData.terminal.2) ⋯ H supportConstant primitiveConstant) (alpha beta : Z4) :

          Uniform Step-1 estimate for the complete terminal Fourier half, with the two scale charges exposed exactly as they occur in the trace ledger. This is (4.17) followed by the paper's quadratic cosine cancellation and inserted-kernel majorant; no absolute value is taken before the signed terminal primitive sum has been integrated.

          Closing the literal trace scale ledger #

          The predecessor and all genuine internal terminal scales are no larger than the complete active budget product. The missing outgoing factor is exactly the initial Green scale A ≥ 1.

          theorem Anderson4D.R324WithinHalfResidualPrefix.R324FullPairingBudgetTerminalAdapter.terminalBudgetMultiplier_le_finalEvenPower_reusable {q ell : ℕ} {A K D : ℝ} (hq : 1 ≤ q) (hell : ell ≤ q) (hA : 1 ≤ A) (hK : 0 < K) (hD : 0 ≤ D) :
          A ^ (2 * q + 1) * K ^ ell * D ≤ (A ^ 2 * max 1 K * (D + 1)) ^ (2 * q)

          Public scalar ledger for reusing the complete full-half budget with a different terminal Fourier calculation. The Step 4(A) endpoint proof has the same stop-scale and suffix multiplier as Step 1; only its terminal factor D changes.

          Uniform complete-half producer #

          theorem Anderson4D.R324WithinHalfResidualPrefix.R324FullPairingBudgetTerminalAdapter.exists_fullPairing_half_step1_bound (ρ : SmoothCutoff) :
          ∃ (Cdet : ℝ), 0 < Cdet ∧ ∀ (lam ε : ℝ) (q : ℕ) (κ : PartialPairing (Fin (2 * q))), 0 < lam → 0 < ε → ε ≤ 1 → 1 ≤ |Real.log ε| → q ≤ truncOrder ε → 1 ≤ q → κ.IsFull → ∀ (alpha beta : Z4), ‖↑(lamEps lam ε) ^ (2 * q) * ∑ τ : ReductionEndpointFiberAt κ, deterministicFullHalfIntegral ρ ε (2 * q) alpha beta ↑τ‖ ≤ (Cdet * lam) ^ (2 * q) / |Real.log ε|

          Paper Step 1, complete full-pairing half.

          For the canonical Proposition-4.1 trace, the complete endpoint-signature half has the paper's uniform (C λ)^(2q) / |log ε| bound. This theorem contains the finite-fibre producer, every proper-prefix collapse, the corrected incoming G₀ form of (4.17), the terminal signed cancellation, and the final scale/order absorption.

          Direct closure of the full/full refined fibre #

          theorem Anderson4D.exists_fullFull_refined_bound_of_normalizedStep1 (ρ : SmoothCutoff) :
          ∃ (primitiveConstant : ℝ), 0 < primitiveConstant ∧ ∀ (lam ε : ℝ) (m : ℕ) (alpha beta : Z4), 0 ≤ lam → 0 < ε → ε ≤ 1 → 1 ≤ |Real.log ε| → 1 ≤ m → m ≤ truncOrder ε → ∀ {s r : Finset (Fin (2 * m)) × Finset (Fin (2 * m))}, ∀ e0 ∈ momentRefinedContractionFiber m s r, e0.fst.IsFull → e0.snd.fst.IsFull → |lamEps lam ε| ^ (2 * m) * ‖momentRefinedDeterministicTermSum ρ ε m alpha beta s r‖ ≤ ∫ (z : T4), primitiveInsertedMajorant primitiveConstant lam ε 1 m z ∂paperMeasure

          Paper Step 1, full/full refined fibre.

          The exact finite-fibre product factorization is combined with the two unconditional normalized half estimates above. Fullness supplies the even-order witness m = 2q; the paper range 1 ≤ m excludes the empty q = 0 trace. The product has two logarithmic gains, one of which is discarded using 1 ≤ |log ε|, and the remaining gain is exactly the integrated inserted-majorant lower bound.

          The statement includes the degenerate coupling lam = 0, so downstream full/full routing needs no separate analytic interface.

          theorem Anderson4D.exists_fullFull_refined_bound_of_normalizedStep1_at_support (ρ : SmoothCutoff) {supportConstant : ℝ} (hsupport : 0 < supportConstant) :
          ∃ (primitiveConstant : ℝ), 0 < primitiveConstant ∧ ∀ (lam ε : ℝ) (m : ℕ) (alpha beta : Z4), 0 ≤ lam → 0 < ε → ε ≤ 1 → 1 ≤ |Real.log ε| → 1 ≤ m → m ≤ truncOrder ε → ∀ {s r : Finset (Fin (2 * m)) × Finset (Fin (2 * m))}, ∀ e0 ∈ momentRefinedContractionFiber m s r, e0.fst.IsFull → e0.snd.fst.IsFull → |lamEps lam ε| ^ (2 * m) * ‖momentRefinedDeterministicTermSum ρ ε m alpha beta s r‖ ≤ ∫ (z : T4), primitiveInsertedMajorant primitiveConstant lam ε supportConstant m z ∂paperMeasure

          Paper Step 1, full/full refined fibre at an arbitrary positive support constant.

          The loss supportConstant² / min supportConstant 1⁴ in the quantitative inserted-majorant lower bound is independent of the order. Enlarging the primitive constant by that loss absorbs it uniformly because m ≥ 1. No new reduction interface is introduced.