OAI/Analysis/KadisonKastler/UniformGraphComparison.lean
3 declarations, 0 contain sorry, 0 declared as axioms
Uniform comparison of the two rescaled graph projections
Directional bounds proved at scales at least one extend to every positive scale by the reciprocal graph symmetry. The two directions then bound the norm of the difference of orthogonal projections.
- theorem
OAI.KadisonKastler.graph_directional_vector_le_of_normline 21Source documentation
The norm of a directional projection product bounds all defining graph vectors in that direction.
theorem graph_directional_vector_le_of_norm (M N : VonNeumannAlgebra H) (J : H ≃ₗᵢ⋆[ℂ] H) (Ω : H) (c B : ℝ) (hbound : ‖(1 - rescaledGraphProjection N J Ω c) * rescaledGraphProjection M J Ω c‖ ≤ B) (x : Algebra M) : ‖(1 - rescaledGraphProjection N J Ω c) (rescaledGraphVector J Ω c x.val)‖ ≤ B * ‖rescaledGraphVector J Ω c x.val‖ := by …Mentions: OAI.KadisonKastler.rescaledGraphProjection, OAI.KadisonKastler.Algebra, OAI.KadisonKastler.rescaledGraphVector
- theorem
OAI.KadisonKastler.graph_directional_all_positiveline 36Source documentation
A large-scale directional bound controls every positive scale.
theorem graph_directional_all_positive (M N : VonNeumannAlgebra H) (J : H ≃ₗᵢ⋆[ℂ] H) (hJ : Function.Involutive J) (Ω : H) (B : ℝ) (hlarge : ∀ c : ℝ, 1 ≤ c → ‖(1 - rescaledGraphProjection N J Ω c) * rescaledGraphProjection M J Ω c‖ ≤ B) {c : ℝ} (hc : 0 < c) (x : Algebra M) : ‖(1 - rescaledGraphProjection N J Ω c) (rescaledGraphVector J Ω c x.val)‖ ≤ B * ‖rescaledGraphVector J Ω c x.val‖ := by …Mentions: OAI.KadisonKastler.rescaledGraphProjection, OAI.KadisonKastler.Algebra, OAI.KadisonKastler.rescaledGraphVector
- theorem
OAI.KadisonKastler.uniform_rescaledGraphProjection_comparisonline 52Source documentation
Two large-scale directional estimates give a uniform norm estimate for the actual closed graph projections at every positive scale.
theorem uniform_rescaledGraphProjection_comparison (M N : VonNeumannAlgebra H) (J : H ≃ₗᵢ⋆[ℂ] H) (hJ : Function.Involutive J) (Ω : H) {B : ℝ} (hB : 0 ≤ B) (hMN : ∀ c : ℝ, 1 ≤ c → ‖(1 - rescaledGraphProjection N J Ω c) * rescaledGraphProjection M J Ω c‖ ≤ B) (hNM : ∀ c : ℝ, 1 ≤ c → ‖(1 - rescaledGraphProjection M J Ω c) * rescaledGraphProjection N J Ω c‖ ≤ B) {c : ℝ} (hc : 0 < c) : ‖rescaledGraphProjection M J Ω c - rescaledGraphProjection N J Ω c‖ ≤ 2 * B := …