math / data /lean /OAI /Probability /ClassicalON /FamilyKernel.lean
introvoyz045's picture
Migrated from GitHub
4a56cd6 verified
Raw History Blame Contribute Delete
4.63 kB
import OAI.Probability.ClassicalON.ObservableBounds
namespace OAI
universe uE uI uV uΞ©
noncomputable section
open MeasureTheory
open scoped BigOperators InnerProductSpace ComplexConjugate
namespace ClassicalON
variable {I : Type uI} {E : Type uE} [Fintype I] [Fintype E]
def familyKernel (u : I β†’ E β†’ β„‚) (e l : E) : β„‚ := βˆ‘ i, u i e * conj (u i l)
def rowSquare (u : I β†’ E β†’ β„‚) (e : E) : ℝ := βˆ‘ l, β€–familyKernel u e lβ€–^2
def familyHSSquare (u : I β†’ E β†’ β„‚) : ℝ := βˆ‘ e, rowSquare u e
omit [Fintype E] in
@[simp] theorem familyKernel_conj (u : I β†’ E β†’ β„‚) (e l : E) :
familyKernel (fun i e => conj (u i e)) e l = conj (familyKernel u e l) := by
simp [familyKernel]
theorem rowSquare_nonneg (u : I β†’ E β†’ β„‚) (e : E) : 0 ≀ rowSquare u e :=
Finset.sum_nonneg fun _ _ => sq_nonneg _
theorem familyHSSquare_nonneg (u : I β†’ E β†’ β„‚) : 0 ≀ familyHSSquare u :=
Finset.sum_nonneg fun e _ => rowSquare_nonneg u e
theorem kernel_diag_sq_le_row (u : I β†’ E β†’ β„‚) (e : E) :
β€–familyKernel u e eβ€–^2 ≀ rowSquare u e :=
Finset.single_le_sum (fun l _ => sq_nonneg β€–familyKernel u e lβ€–) (Finset.mem_univ e)
theorem kernel_diag_norm_le_sqrt (u : I β†’ E β†’ β„‚) (e : E) :
β€–familyKernel u e eβ€– ≀ Real.sqrt (rowSquare u e) :=
Real.le_sqrt_of_sq_le (kernel_diag_sq_le_row u e)
namespace SpinSystem
variable {n : β„•} {V : Type uV} [Fintype V]
omit [Fintype V] [Fintype I] in
def coefficientObservable (S : SpinSystem n V E) (M : SpinOperator n) (e : E) :
C((V β†’ Spin n), β„‚) := ⟨fun Οƒ => (S.localCoefficient M Οƒ e : β„‚), by
unfold localCoefficient
fun_prop⟩
omit [Fintype V] [Fintype I] in
def currentObservable (S : SpinSystem n V E) (M : SpinOperator n) (u : E β†’ β„‚) :
C((V β†’ Spin n), β„‚) := ⟨_, S.continuous_complexEnergy M u⟩
omit [Fintype V] [Fintype I] [Fintype E] in
def conjugateObservable {Ξ© : Type uΞ©} [TopologicalSpace Ξ©] (f : C(Ξ©, β„‚)) : C(Ξ©, β„‚) :=
⟨fun Οƒ => conj (f Οƒ), Complex.continuous_conj.comp f.continuous⟩
omit [Fintype V] [Fintype I] [Fintype E] in
@[simp] theorem conjugateObservable_apply {Ξ© : Type uΞ©} [TopologicalSpace Ξ©]
(f : C(Ξ©, β„‚)) (Οƒ : Ξ©) : conjugateObservable f Οƒ = conj (f Οƒ) := rfl
omit [Fintype V] [Fintype I] [Fintype E] in
@[simp] theorem normObservable_conjugate {Ξ© : Type uΞ©} [TopologicalSpace Ξ©]
(f : C(Ξ©, β„‚)) : normObservable (conjugateObservable f) = normObservable f := by
ext Οƒ
exact Complex.norm_conj (f Οƒ)
omit [Fintype V] [Fintype I] in
@[simp] theorem currentObservable_conj (S : SpinSystem n V E) (M : SpinOperator n)
(u : E β†’ β„‚) : S.currentObservable M (fun e => conj (u e)) =
conjugateObservable (S.currentObservable M u) := by
ext Οƒ
exact S.complexEnergy_conj M u Οƒ
omit [Fintype V] [Fintype I] [Fintype E] in
@[simp] theorem coefficientObservable_conj (S : SpinSystem n V E) (M : SpinOperator n)
(e : E) : conjugateObservable (S.coefficientObservable M e) =
S.coefficientObservable M e := by
ext Οƒ
exact Complex.conj_ofReal _
omit [Fintype V] [Fintype E] in
theorem familyKernel_differential (S : SpinSystem n V E) (f : I β†’ V β†’ β„‚) (e : E) :
familyKernel (fun i => S.differential (f i)) e =
S.differential (fun x => βˆ‘ i, S.differential (f i) e * conj (f i x)) := by
funext l
simp only [familyKernel, differential, map_sub, mul_sub, Finset.sum_sub_distrib]
theorem row_current_moment [NeZero n] (S : SpinSystem n V E) (Ξ² : ℝ)
(hb : βˆ€ e, 0 ≀ S.strength e ∧ S.strength e ≀ Ξ²)
(M : SpinOperator n) (hM : M ∈ skewAdjoint (SpinOperator n)) (hMn : β€–Mβ€– ≀ 1)
(f : I β†’ V β†’ β„‚) (hP : βˆ€ i x s, S.pin x = some s β†’ f i x = 0) (e : E) :
S.averageLinear (normObservable (S.currentObservable M
(familyKernel (fun i => S.differential (f i)) e)) *
normObservable (S.currentObservable M
(familyKernel (fun i => S.differential (f i)) e))) ≀
Ξ² * rowSquare (fun i => S.differential (f i)) e := by
have hp : βˆ€ x s, S.pin x = some s β†’
(βˆ‘ i, S.differential (f i) e * conj (f i x)) = 0 := by
intro x s hs
simp [hP _ x s hs]
have h := S.complex_gradient_current_bound Ξ² hb M hM hMn
(fun x => βˆ‘ i, S.differential (f i) e * conj (f i x)) hp
rw [← S.familyKernel_differential f e] at h
change S.average (fun Οƒ =>
β€–S.complexEnergy M (familyKernel (fun i => S.differential (f i)) e) Οƒβ€– *
β€–S.complexEnergy M (familyKernel (fun i => S.differential (f i)) e) Οƒβ€–) ≀ _
simpa only [rowSquare, pow_two] using h
end SpinSystem
end ClassicalON
end
end OAI