OAI/Analysis/KadisonSimilarity/FreeProduct/WordPropagationTransfer.lean
5 declarations, 0 contain sorry, 0 declared as axioms
Transfer the multiplicity propagation estimate to the actual orthogonal word columns used by the finite Fourier test.
- def
OAI.KadisonSimilarity.repeatedWordInclusionline 24Source documentation
The concrete column at a fixed word degree, endpoint, and repeated letter.
def repeatedWordInclusion (n : ℕ) (R : K →ₗᵢ[ℂ] J n) (d : E) (r : F) (hd : ‖d‖ = 1) (hr : ‖r‖ = 1) : K →ₗᵢ[ℂ] H where … - theorem
OAI.KadisonSimilarity.repeatedWordInclusion_applyline 36 - theorem
OAI.KadisonSimilarity.norm_repeatedWordInclusion_error_leline 44Source documentation
A multiplicity error has exactly the same norm in a normalized word column. This supplies the full-output error needed in the Hankel test.
theorem norm_repeatedWordInclusion_error_le (n : ℕ) (R : K →ₗᵢ[ℂ] J n) (d : E) (r : F) (hd : ‖d‖ = 1) (hr : ‖r‖ = 1) (Y : H →L[ℂ] H) (B : K →L[ℂ] K) (Bn : J n →L[ℂ] J n) (hY : ∀ v, Y (U.inclusion n (hilbertTensorInsert v (endpointWord d (fun _ : Fin n => r)))) = U.inclusion n (hilbertTensorInsert (Bn v) (endpointWord d (fun _ : Fin n => r)))) {η : ℝ} (hη : 0 ≤ η) (hprop : ∀ x, ‖Bn (R x) - R (B x)‖ ≤ η * ‖x‖) : ‖Y.comp (repeatedWordInclusion U n R d r hd hr).toContinuousLinearMap - (repeatedWordInclusion U n R d r hd hr).toContinuousLinearMap.comp B‖ ≤ η := by …Mentions: OAI.KadisonSimilarity.hilbertTensorInsert, OAI.KadisonSimilarity.endpointWord, OAI.KadisonSimilarity.repeatedWordInclusion
- theorem
OAI.KadisonSimilarity.repeatedWordInclusion_orthogonalline 70Source documentation
Word columns at different degrees are orthogonal.
theorem repeatedWordInclusion_orthogonal (n m : ℕ) (hnm : n ≠ m) (Rn : K →ₗᵢ[ℂ] J n) (Rm : K →ₗᵢ[ℂ] J m) (d : E) (r : F) (hd : ‖d‖ = 1) (hr : ‖r‖ = 1) (x y : K) : inner ℂ (repeatedWordInclusion U n Rn d r hd hr x) (repeatedWordInclusion U m Rm d r hd hr y) = 0 := … - theorem
OAI.KadisonSimilarity.free_corner_from_multiplicity_propagationline 79Source documentation
The finite Fourier test consumes the derived multiplicity propagation bounds, with the orthogonality and full-column errors constructed here.
theorem free_corner_from_multiplicity_propagation {N : ℕ} (hN : 0 < N) (R : ∀ n, K →ₗᵢ[ℂ] J n) (d : E) (r : F) (hd : ‖d‖ = 1) (hr : ‖r‖ = 1) (Y T : H →L[ℂ] H) (B S : K →L[ℂ] K) (Bn : ∀ n, J n →L[ℂ] J n) (hY : ∀ n v, Y (U.inclusion n (hilbertTensorInsert v (endpointWord d (fun _ : Fin n => r)))) = U.inclusion n (hilbertTensorInsert (Bn n v) (endpointWord d (fun _ : Fin n => r)))) (hYstar : ∀ n v, Y.adjoint (U.inclusion n (hilbertTensorInsert v (endpointWord d (fun _ : Fin n => r)))) = U.inclusion n (hilbertTensorInsert ((Bn n).adjoint v) (endpointWord d (fun _ : Fin n => r)))) …Mentions: OAI.KadisonSimilarity.hilbertTensorInsert, OAI.KadisonSimilarity.endpointWord