22.4. Continuous linear maps
If E and F are two topological vector spaces:
variable (𝕜 E F : Type*) [NormedField 𝕜]
[AddCommGroup E] [Module 𝕜 E] [TopologicalSpace E]
[ContinuousAdd E] [ContinuousSMul 𝕜 E]
[AddCommGroup F] [Module 𝕜 F] [TopologicalSpace F]
[ContinuousAdd F] [ContinuousSMul 𝕜 F]
we can speak of the continuous linear maps between E and F. The relevant definition is
ContinuousLinearMap and it has special notation as follows:
#check E →L[𝕜] F
22.4.1. Topologies on continuous linear maps
The space of continuous linear maps is naturally equipped with topology of bounded convergence. This topology coincides with the operator norm topology in case linear maps on Banach spaces.
variable (𝕜 E F : Type*) [NontriviallyNormedField 𝕜]
[NormedAddCommGroup E] [NormedSpace 𝕜 E]
[NormedAddCommGroup F] [NormedSpace 𝕜 F]
(f : E →L[𝕜] F)
example (x : E) : ‖f x‖ ≤ ‖f‖ * ‖x‖ := f.le_opNorm x
example {M : ℝ} (hMp : 0 ≤ M) (hM : ∀ x, ‖f x‖ ≤ M * ‖x‖) :
‖f‖ ≤ M := f.opNorm_le_bound hMp hM
Mathlib also knows about the following topologies on type synonyms of continuous linear maps:
-
Pointwise convergence (or strong operator topology):
PointwiseConvergenceCLM -
Compact convergence:
CompactConvergenceCLM -
Weak operator topology:
ContinuousLinearMapWOT
22.4.2. Topological dual
In the special case that F is the base field 𝕜, we have the abbreviation StrongDual 𝕜 E
for the E →L[𝕜] 𝕜. This space naturally carries the topology of bounded convergence, which
is also known as the strong topology. To obtain the weak dual use WeakDual.