The Cayley transform #
This file constructs the bounded Cayley transform of a possibly unbounded self-adjoint operator.
The key point is that A + iI, regarded as a map from the domain of A, is bijective. Its inverse
can therefore be composed with A - iI, and the resulting everywhere-defined linear map is an
isometry, hence continuous.
theorem
add_I_surjective_of_isSelfAdjoint
{E : Type u_1}
[NormedAddCommGroup E]
[InnerProductSpace ℂ E]
[CompleteSpace E]
(A : E →ₗ.[ℂ] E)
(hA : IsSelfAdjoint A)
:
Function.Surjective fun (x : ↥A.domain) => ↑A x + Complex.I • ↑x
For a self-adjoint operator, A + iI maps its domain onto the ambient space.
theorem
sub_I_surjective_of_isSelfAdjoint
{E : Type u_1}
[NormedAddCommGroup E]
[InnerProductSpace ℂ E]
[CompleteSpace E]
(A : E →ₗ.[ℂ] E)
(hA : IsSelfAdjoint A)
:
Function.Surjective fun (x : ↥A.domain) => ↑A x - Complex.I • ↑x
For a self-adjoint operator, A - iI maps its domain onto the ambient space.
noncomputable def
cayleyLinearMap
{E : Type u_1}
[NormedAddCommGroup E]
[InnerProductSpace ℂ E]
[CompleteSpace E]
(A : E →ₗ.[ℂ] E)
(hA : IsSelfAdjoint A)
:
The linear map underlying the Cayley transform of a self-adjoint operator.
Equations
- cayleyLinearMap A hA = minusI A ∘ₗ ↑(LinearEquiv.ofBijective (plusI A) ⋯).symm
Instances For
noncomputable def
cayleyTransform
{E : Type u_1}
[NormedAddCommGroup E]
[InnerProductSpace ℂ E]
[CompleteSpace E]
(A : E →ₗ.[ℂ] E)
(hA : IsSelfAdjoint A)
:
The Cayley transform (A - iI)(A + iI)⁻¹ of a self-adjoint partial linear map.
Equations
- cayleyTransform A hA = (cayleyLinearMap A hA).mkContinuous 1 ⋯
Instances For
theorem
cayleyTransform_apply_plus
{E : Type u_1}
[NormedAddCommGroup E]
[InnerProductSpace ℂ E]
[CompleteSpace E]
(A : E →ₗ.[ℂ] E)
(hA : IsSelfAdjoint A)
(x : ↥A.domain)
:
The Cayley transform sends (A + iI)x to (A - iI)x.
theorem
norm_cayleyTransform
{E : Type u_1}
[NormedAddCommGroup E]
[InnerProductSpace ℂ E]
[CompleteSpace E]
(A : E →ₗ.[ℂ] E)
(hA : IsSelfAdjoint A)
(y : E)
:
The Cayley transform preserves the norm.