Documentation

MIPRE.Background.LIDT.Co.SelfImprovement.Theorems.Results.AddInUStep12.Selected

Selected add-in-u Step 1/2 Cauchy--Schwarz bounds #

Selected-family contraction inputs and raw √(2δ) estimates for the first two add-in-u moves in the self-improvement chain: the counterpart of the vendored SelfImprovement/Theorems/Results/AddInUStep12/Selected.lean (under MIPRE/Background/LIDT/MIPStarRE/LDT/) in the port of planning/c6b-plan.md (milestone M10, section "Port conventions"). It is the selected-pair twin of Co AddInUStep12/Raw: the sums run over the selected pairs (o, h) ∈ S_u with an arbitrary outcome operator M^u_o, in place of all polynomials h with the sandwiched operator.

Contents #

Every scalar is strategy.state.ev of a joint operator; leftTensor A, rightTensor A and opTensor A B are strategy.state.L A, strategy.state.R A and strategy.state.opTensor A B, and Xᴴ is star X. The vendored addInU_selected_filtered_tensor_sum_le_one names no state, its opTensor being the Kronecker product on the carrier; here it takes the model (ψ : SymModel 𝔓 K) as an explicit first argument (S is the selection), and its proof is the keystone's opTensor_le_one after summing the fibre into M^u_{tot} ⊗ T_{tot}. The vendored Hermitian facts (Matrix.nonneg_iff_posSemidef.mp (…outcome_pos _)).isHermitian.eq and the Matrix.conjTranspose_sum/conjTranspose_opTensor computation of Kᴴ = K are (IsSelfAdjoint.of_nonneg …).star_eq. The vendored proofs pass strategy.isNormalized to closenessOfInnerProduct_right/_left (vendored lines 448 and 662); the ported Cauchy--Schwarz bounds (Co Preliminaries/SwitchSandwichPrep/InnerProduct) have no normalization hypothesis, so it is dropped, and the vendored rw [hCop_zero, Matrix.mul_zero, ev_zero] steps are mul_zero (zero_mul) with VecState.ev_zero. No statement of the vendored file carries a swap, density or normalization hypothesis.

The contraction proofs reuse Co Raw's addInU_sum_projection_sandwich_le_one, which is why this file imports Co AddInUStep12/Raw beside Algebra, where the vendored file imports Algebra alone (the vendored AddInUStep34AndTransfer/Factored imports both). The two raw bounds share the new addInU_selected_sum_fiber_collapse, the selected form of Raw's addInU_sum_fiber_collapse; the vendored proofs repeat it inline as hcollapse.

Not ported #

Every declaration of the vendored file has a counterpart here.

New here #

References #

In LionSR/MIPStarRE at commit 507e8122, not in this repository:

theorem MIPRE.LIDT.Co.SelfImprovement.addInU_selected_filtered_tensor_sum_le_one {𝔓 : Type u_1} [CStarAlgebra 𝔓] [PartialOrder 𝔓] [StarOrderedRing 𝔓] {K : Type u_2} [NormedAddCommGroup K] [InnerProductSpace ℂ K] [CompleteSpace K] {Outcome : Type u_3} [Fintype Outcome] (ψ : SymModel 𝔓 K) (params : MIPStarRE.LDT.Parameters) [MIPStarRE.LDT.FieldModel params.q] (M : IdxSubMeas (MIPStarRE.LDT.Point params) Outcome 𝔓) (T : SubMeas (MIPStarRE.LDT.Polynomial params) 𝔓) (S : MIPStarRE.LDT.SelfImprovement.AddInUSelection params Outcome) (u v : MIPStarRE.LDT.Point params) (a : MIPStarRE.LDT.Fq params) :
∑ ah ∈ MIPStarRE.LDT.SelfImprovement.addInUSelectionPairs params S u with ah.2.toFun v = a, ψ.opTensor ((M u).outcome ah.1) (T.outcome ah.2) ≤ 1

The selected fiber tensor mass is a contraction.

For fixed points u, v and a value a, the selected sum over pairs (o,h) ∈ S_u with h(v)=a is bounded by the full product M^u_{\mathrm{tot}} ⊗ T_{\mathrm{tot}}, hence by the identity.

theorem MIPRE.LIDT.Co.SelfImprovement.addInU_selected_step1_C_contraction {𝔓 : Type u_1} [CStarAlgebra 𝔓] [PartialOrder 𝔓] [StarOrderedRing 𝔓] {K : Type u_2} [NormedAddCommGroup K] [InnerProductSpace ℂ K] [CompleteSpace K] {Outcome : Type u_3} [Fintype Outcome] (params : MIPStarRE.LDT.Parameters) [MIPStarRE.LDT.FieldModel params.q] (strategy : SymStrat params 𝔓 K) (M : IdxSubMeas (MIPStarRE.LDT.Point params) Outcome 𝔓) (T : SubMeas (MIPStarRE.LDT.Polynomial params) 𝔓) (S : MIPStarRE.LDT.SelfImprovement.AddInUSelection params Outcome) (uv : MIPStarRE.LDT.Point params × MIPStarRE.LDT.Point params) :
∑ a : MIPStarRE.LDT.Fq params, star (∑ ah ∈ MIPStarRE.LDT.SelfImprovement.addInUSelectionPairs params S uv.1 with ah.2.toFun uv.2 = a, strategy.state.opTensor ((M uv.1).outcome ah.1) (T.outcome ah.2) * strategy.state.R ((strategy.pointMeasurement uv.2).outcome a)) * ∑ ah ∈ MIPStarRE.LDT.SelfImprovement.addInUSelectionPairs params S uv.1 with ah.2.toFun uv.2 = a, strategy.state.opTensor ((M uv.1).outcome ah.1) (T.outcome ah.2) * strategy.state.R ((strategy.pointMeasurement uv.2).outcome a) ≤ 1

Cauchy--Schwarz contraction side condition for selected Step 1.

For a fixed (u, v), the selected fiber sum K_a = ∑_{(o,h) ∈ S_u, h(v)=a} M^u_o ⊗ T_h is a contraction. Sandwiching by the right-register point projector A^v_a and summing over a is therefore bounded by the identity.

theorem MIPRE.LIDT.Co.SelfImprovement.addInU_selected_step2_C_contraction {𝔓 : Type u_1} [CStarAlgebra 𝔓] [PartialOrder 𝔓] [StarOrderedRing 𝔓] {K : Type u_2} [NormedAddCommGroup K] [InnerProductSpace ℂ K] [CompleteSpace K] {Outcome : Type u_3} [Fintype Outcome] (params : MIPStarRE.LDT.Parameters) [MIPStarRE.LDT.FieldModel params.q] (strategy : SymStrat params 𝔓 K) (M : IdxSubMeas (MIPStarRE.LDT.Point params) Outcome 𝔓) (T : SubMeas (MIPStarRE.LDT.Polynomial params) 𝔓) (S : MIPStarRE.LDT.SelfImprovement.AddInUSelection params Outcome) (uv : MIPStarRE.LDT.Point params × MIPStarRE.LDT.Point params) :
∑ a : MIPStarRE.LDT.Fq params, (∑ ah ∈ MIPStarRE.LDT.SelfImprovement.addInUSelectionPairs params S uv.1 with ah.2.toFun uv.2 = a, strategy.state.L ((strategy.pointMeasurement uv.2).outcome a) * strategy.state.opTensor ((M uv.1).outcome ah.1) (T.outcome ah.2)) * star (∑ ah ∈ MIPStarRE.LDT.SelfImprovement.addInUSelectionPairs params S uv.1 with ah.2.toFun uv.2 = a, strategy.state.L ((strategy.pointMeasurement uv.2).outcome a) * strategy.state.opTensor ((M uv.1).outcome ah.1) (T.outcome ah.2)) ≤ 1

Cauchy--Schwarz contraction side condition for selected Step 2.

This is the left-register analogue of addInU_selected_step1_C_contraction: for each fixed (u,v), the operators C_a = A^v_a \otimes I · K_a, with K_a the selected fiber tensor mass, have ∑_a C_a C_a^* ≤ I.

theorem MIPRE.LIDT.Co.SelfImprovement.addInU_selected_sum_fiber_collapse {K : Type u_2} [NormedAddCommGroup K] [InnerProductSpace ℂ K] [CompleteSpace K] {Outcome : Type u_3} [Fintype Outcome] {params : MIPStarRE.LDT.Parameters} [MIPStarRE.LDT.FieldModel params.q] (V : VecState K) (v : MIPStarRE.LDT.Point params) (s : Finset (Outcome × MIPStarRE.LDT.Polynomial params)) (X : MIPStarRE.LDT.Fq params → Outcome × MIPStarRE.LDT.Polynomial params → K →L[ℂ] K) (hX : ∀ (a : MIPStarRE.LDT.Fq params), ∀ ah ∉ {ah ∈ s | ah.2.toFun v = a}, X a ah = 0) :
∑ a : MIPStarRE.LDT.Fq params, ∑ ah : Outcome × MIPStarRE.LDT.Polynomial params, V.ev (X a ah) = ∑ ah ∈ s, V.ev (X (ah.2.toFun v) ah)

Collapsing a double sum ∑ₐ ∑_{(o,h)} ev(X a (o,h)) onto the selected pairs with a = h(v), when X a (o,h) vanishes unless (o,h) ∈ s and h(v) = a: the step that turns the fibre-indexed Cauchy--Schwarz sums of the two selected raw bounds into the single sums over the selected pairs of the algebraic alignment.

theorem MIPRE.LIDT.Co.SelfImprovement.addInU_selected_cs_chain_step1_abs_le_sqrt_two_delta {𝔓 : Type u_1} [CStarAlgebra 𝔓] [PartialOrder 𝔓] [StarOrderedRing 𝔓] {K : Type u_2} [NormedAddCommGroup K] [InnerProductSpace ℂ K] [CompleteSpace K] {Outcome : Type u_3} [Fintype Outcome] (params : MIPStarRE.LDT.Parameters) [MIPStarRE.LDT.FieldModel params.q] (strategy : SymStrat params 𝔓 K) (M : IdxSubMeas (MIPStarRE.LDT.Point params) Outcome 𝔓) (T : SubMeas (MIPStarRE.LDT.Polynomial params) 𝔓) (S : MIPStarRE.LDT.SelfImprovement.AddInUSelection params Outcome) (delta : ℝ) (hssc : strategy.state.BipartiteSSCRel (MIPStarRE.LDT.uniformDistribution (MIPStarRE.LDT.Point params)) strategy.pointMeasurement.toIdxSubMeas delta) :
|addInUSelectedCSChainQ0 params strategy M T S - addInUSelectedCSChainQ1 params strategy M T S| ≤ √(2 * delta)

Raw selected |Q₀ - Q₁| ≤ √(2δ) bound for the first add-in-u move.

This is the selection-parametrized form of addInU_cs_chain_step1_abs_le_sqrt_two_delta. It applies the weighted Cauchy--Schwarz estimate to the selected pairs (o,h) ∈ S_u; the contraction side condition is addInU_selected_step1_C_contraction.

theorem MIPRE.LIDT.Co.SelfImprovement.addInU_selected_cs_chain_step2_abs_le_sqrt_two_delta {𝔓 : Type u_1} [CStarAlgebra 𝔓] [PartialOrder 𝔓] [StarOrderedRing 𝔓] {K : Type u_2} [NormedAddCommGroup K] [InnerProductSpace ℂ K] [CompleteSpace K] {Outcome : Type u_3} [Fintype Outcome] (params : MIPStarRE.LDT.Parameters) [MIPStarRE.LDT.FieldModel params.q] (strategy : SymStrat params 𝔓 K) (M : IdxSubMeas (MIPStarRE.LDT.Point params) Outcome 𝔓) (T : SubMeas (MIPStarRE.LDT.Polynomial params) 𝔓) (S : MIPStarRE.LDT.SelfImprovement.AddInUSelection params Outcome) (delta : ℝ) (hssc : strategy.state.BipartiteSSCRel (MIPStarRE.LDT.uniformDistribution (MIPStarRE.LDT.Point params)) strategy.pointMeasurement.toIdxSubMeas delta) :
|addInUSelectedCSChainQ1 params strategy M T S - addInUSelectedCSChainQ2 params strategy M T S| ≤ √(2 * delta)

Raw selected |Q₁ - Q₂| ≤ √(2δ) bound for the second add-in-u move.

This is the selection-parametrized form of addInU_cs_chain_step2_abs_le_sqrt_two_delta. It uses the left-action Cauchy--Schwarz estimate with the selected Step 2 contraction side condition.