Injective seminorm on the tensor of a finite family of normed spaces. #
Let π be a nontrivially normed field and E be a family of normed π-vector spaces Eα΅’,
indexed by a finite type ΞΉ. We define a seminorm on β¨[π] i, Eα΅’, which we call the
"injective seminorm". It is chosen to satisfy the following property: for every
normed π-vector space F, the linear equivalence
MultilinearMap π E F ββ[π] (β¨[π] i, Eα΅’) ββ[π] F
expressing the universal property of the tensor product induces an isometric linear equivalence
ContinuousMultilinearMap π E F ββα΅’[π] (β¨[π] i, Eα΅’) βL[π] F.
The idea is the following: Every normed π-vector space F defines a linear map
from β¨[π] i, Eα΅’ to ContinuousMultilinearMap π E F ββ[π] F, which sends x to the map
f β¦ f.lift x. Thanks to PiTensorProduct.norm_eval_le_projectiveSeminorm, this map lands in
ContinuousMultilinearMap π E F βL[π] F. As this last space has a natural operator (semi)norm,
we get an induced seminorm on β¨[π] i, Eα΅’, which, by
PiTensorProduct.norm_eval_le_projectiveSeminorm, is bounded above by the projective seminorm
PiTensorProduct.projectiveSeminorm. We then take the sup of these seminorms as F varies;
as this family of seminorms is bounded, its sup has good properties.
In fact, we cannot take the sup over all normed spaces F because of set-theoretical issues,
so we only take spaces F in the same universe as β¨[π] i, Eα΅’. We prove in
norm_eval_le_injectiveSeminorm that this gives the same result, because every multilinear map
from E = Ξ α΅’ Eα΅’ to F factors though a normed vector space in the same universe as
β¨[π] i, Eα΅’.
We then prove the universal property and the functoriality of β¨[π] i, Eα΅’ as a normed vector
space.
Main definitions #
PiTensorProduct.toDualContinuousMultilinearMap: Theπ-linear map fromβ¨[π] i, Eα΅’toContinuousMultilinearMap π E F βL[π] Fsendingxto the mapf β¦ f x.PiTensorProduct.injectiveSeminorm: The injective seminorm onβ¨[π] i, Eα΅’.PiTensorProduct.liftEquiv: The bijection betweenContinuousMultilinearMap π E Fand(β¨[π] i, Eα΅’) βL[π] F, as a continuous linear equivalence.PiTensorProduct.liftIsometry: The bijection betweenContinuousMultilinearMap π E Fand(β¨[π] i, Eα΅’) βL[π] F, as an isometric linear equivalence.PiTensorProduct.tprodL: The canonical continuous multilinear map fromE = Ξ α΅’ Eα΅’toβ¨[π] i, Eα΅’.PiTensorProduct.mapL: The continuous linear map fromβ¨[π] i, Eα΅’toβ¨[π] i, E'α΅’induced by a family of continuous linear mapsEα΅’ βL[π] E'α΅’.PiTensorProduct.mapLMultilinear: The continuous multilinear map fromΞ α΅’ (Eα΅’ βL[π] E'α΅’)to(β¨[π] i, Eα΅’) βL[π] (β¨[π] i, E'α΅’)sending a familyftoPiTensorProduct.mapL f.
Main results #
PiTensorProduct.norm_eval_le_injectiveSeminorm: The main property of the injective seminorm onβ¨[π] i, Eα΅’: for everyxinβ¨[π] i, Eα΅’and every continuous multilinear mapffromE = Ξ α΅’ Eα΅’to a normed spaceF, we haveβf.lift xβ β€ βfβ * injectiveSeminorm x.PiTensorProduct.mapL_opNorm: Iffis a family of continuous linear mapsfα΅’ : Eα΅’ βL[π] Fα΅’, thenβPiTensorProduct.mapL fβ β€ β i, βfα΅’β.PiTensorProduct.mapLMultilinear_opNorm: IfFis a normed vecteor space, thenβmapLMultilinear π E Fβ β€ 1.
TODO #
If all
Eα΅’are separated and satisfySeparatingDual, then the seminorm onβ¨[π] i, Eα΅’is a norm. This uses the construction of a basis of thePiTensorProduct, hence depends on PR https://github.com/leanprover-community/mathlib4/pull/11156. It should probably go in a separate file.Adapt the remaining functoriality constructions/properties from
PiTensorProduct.
The linear map from β¨[π] i, Eα΅’ to ContinuousMultilinearMap π E F βL[π] F sending
x in β¨[π] i, Eα΅’ to the map f β¦ f.lift x.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The injective seminorm on β¨[π] i, Eα΅’. Morally, it sends x in β¨[π] i, Eα΅’ to the
sup of the operator norms of the PiTensorProduct.toDualContinuousMultilinearMap F x, for all
normed vector spaces F. In fact, we only take in the same universe as β¨[π] i, Eα΅’, and then
prove in PiTensorProduct.norm_eval_le_injectiveSeminorm that this gives the same result.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
- PiTensorProduct.instNormedSpace = { toModule := PiTensorProduct.instModule, norm_smul_le := β― }
The linear equivalence between ContinuousMultilinearMap π E F and (β¨[π] i, Eα΅’) βL[π] F
induced by PiTensorProduct.lift, for every normed space F.
Equations
- One or more equations did not get rendered due to their size.
Instances For
For a normed space F, we have constructed in PiTensorProduct.liftEquiv the canonical
linear equivalence between ContinuousMultilinearMap π E F and (β¨[π] i, Eα΅’) βL[π] F
(induced by PiTensorProduct.lift). Here we give the upgrade of this equivalence to
an isometric linear equivalence; in particular, it is a continuous linear equivalence.
Equations
- PiTensorProduct.liftIsometry π E F = { toLinearEquiv := PiTensorProduct.liftEquiv π E F, norm_map' := β― }
Instances For
The canonical continuous multilinear map from E = Ξ α΅’ Eα΅’ to β¨[π] i, Eα΅’.
Equations
- PiTensorProduct.tprodL π = (PiTensorProduct.liftIsometry π E (PiTensorProduct π fun (i : ΞΉ) => E i)).symm (ContinuousLinearMap.id π (PiTensorProduct π fun (i : ΞΉ) => E i))
Instances For
Let Eα΅’ and E'α΅’ be two families of normed π-vector spaces.
Let f be a family of continuous π-linear maps between Eα΅’ and E'α΅’, i.e.
f : Ξ α΅’ Eα΅’ βL[π] E'α΅’, then there is an induced continuous linear map
β¨α΅’ Eα΅’ β β¨α΅’ E'α΅’ by β¨ aα΅’ β¦ β¨ fα΅’ aα΅’.
Equations
- PiTensorProduct.mapL f = (PiTensorProduct.liftIsometry π E (PiTensorProduct π fun (i : ΞΉ) => E' i)) ((PiTensorProduct.tprodL π).compContinuousLinearMap f)
Instances For
Given submodules pα΅’ β Eα΅’, this is the natural map: β¨[π] i, pα΅’ β β¨[π] i, Eα΅’.
This is the continuous version of PiTensorProduct.mapIncl.
Equations
- PiTensorProduct.mapLIncl p = PiTensorProduct.mapL fun (i : ΞΉ) => (p i).subtypeL
Instances For
Upgrading PiTensorProduct.mapL to a MonoidHom when E = E'.
Equations
- PiTensorProduct.mapLMonoidHom = { toFun := PiTensorProduct.mapL, map_one' := β―, map_mul' := β― }
Instances For
The tensor of a family of linear maps from Eα΅’ to E'α΅’, as a continuous multilinear map of
the family.
Equations
- PiTensorProduct.mapLMultilinear π E E' = { toFun := PiTensorProduct.mapL, map_update_add' := β―, map_update_smul' := β― }.mkContinuous 1 β―