Functional Analysis Lecture 4

[[lecture-data]]

← Lecture 3 | Lecture 5 →
Lecture Notes: Rodriguez, page 18

1. Normed and Banach Spaces

Recall from last time

Baire Category Theorem

Let M be a {as||complete} metric space and let {Cn}n be a {as||collection of closed subsets} of M such that M=nNCnThen {ha||at least one} of the Cn {ha||contain an open ball}
{ha||B(x,r)={yM:d(x,y)<r}}
ie, {ha||at least one} Cn has {ha||an interior point}.

Open Mapping Theorem

Let B1,B2 be two Banach spaces and let TB(B1,B2) be surjective. Then T is an open map.

ie for all open subsets UB1, we have T(U)B2 is open.

Proof

Note

We will first prove that if B(0,1)={bB1:||b||<1}, then T(B(0,1)) contains an open ball in B2 centered at 0. Then, since T is linear, we can scale everything properly to show that for any b1UB1 open, there exists ϵ>0 such that B(T(b1),ϵ)T(U)

Showing that the closure of the image T(B(0,1)) contains an open ball B(0,r)

T surjective means everything in B2 gets mapped to by something in B1 ie

B2=nNT(B(0,n))

So by the Baire Category Theorem, there exists some n0N such that T(B(0,n0)) contains an open ball.

Now, since T is a linear operator, n0T(B(0,1)) contains an open ball. ie, T(B(0,1)) contains an open ball. Equivalently, there exists some v0B2 and r>0 such that B(v0,4r)T(B(0,1)).

ie, we can pick some u1B(0,1) such that v1=T(u1)T(B(0,1)) such that ||v0v1||<2r

Note

we choose 4 here to make some calculations easier later

Then

B(v1,2r)B(v0,4r)T(B(0,1))

If ||v||<r, then

12(2v+v1)12B(v1,2r)12T(B(0,1))=T(B(0,12))

And this means

v=T(u12)+12(2v+v1)T(u12)+T(B(0,12))=T(u12+B(0,12))B(0,1) since u1B(0,1)T(B(0,1))

ie, (in B2), we have B(0,r)T(B(0,1)). This implies that

B(0,2nr)=2nB(0,r)2nT(B(0,1))=T(B(0,2n))

for all nN

Now we prove that B(0,r2)T(B(0,1)).

Showing that the image T(B(0,1)) contains an open ball B(0,r2)

Let ||v||<r2. Then vT(B(0,12)) . This implies there exists some b1B(0,12) such that ||vTb1||<r4

ie, vTb1T(B(0,12)) there is some b2B(0,14) such that

||vTb1Tb2||=||vT(b1b2)||<r412=r8

continuing in this manner, we obtain a sequence {bk}kNB1 such that

  1. ||bk||<2k
  2. ||vk=1nTbk||<2n1r

The series k=1bk is absolutely summable (since the norm of the sums are bounded). And since B1 is Banach, this means there exists some bB1 such that b=k=1bk

Further, by the triangle inequality, we have

||b||=limn||k=1nbk||limnk=1||bk||

And we can bound this

=k=1||bk||<k=12k=1

And since T is a bounded linear operator (and thus continuous), we have

Tb=limnT(k=1nbk)=limnk=1Tbk=v

ie, vT(B(0,1)). Thus B(0,r2)T(B(0,1)) in B2.

Thus, if 0 was an interior point of U in the domain, then it is still an interior point in the image of U.

Generalizing to any open set

Suppose UB1 is open and b2=Tb1T(U). Then there exists some ϵ>0 such that b1+B(0,ϵ)=B(b1,ϵ)U. And, from above, there exists some δ>0 such that B(0,δ)T(B(0,1)). ie

B(b2,ϵδ)=b2+ϵB(0,δ)b2+ϵT(B(0,1))=T(b1)+ϵT(B(0,1))=T(b1+ϵB(0,1))=T(b1+B(0,ϵ))T(U)

Since b1+B(0,ϵ) is contained in U. Thus we've found a ball around any b2T(U).

(see open mapping theorem)

Corollary

If B1,B2 are Banach spaces and T:B1B2 is a bijective bounded linear operator, then T1B(B2,B1) is also.

Proof

T1 is continuous for all UB1 open, (T1)1(U)=T(U) is open. And this is true by the open mapping theorem

see bijective bounded linear operators have bounded linear inverses

Corollary

If B1,B2 are Banach spaces, then B1×B2 with norm

||(b1,b2)||:=||b1||+||b2||

is a Banach space.

Proof

Exercise

  • just need to check all the definitions for the norm
  • check that Cauchy sequences in B1×B2 consist of Cauchy sequences in each of B1 and B2 (similar to completeness of R2)

see the cartesian product of banach spaces is banach

Closed Graph Theorem

If B1,B2 are Banach spaces and T:B1B2 is a linear operator, then

TB(B1,B2)Γ(T):={(u,Tu):uB1}B1×B2 is closed
Note

this may be easier to prove than proving something is a bounded linear operator. Normally, we need to show that sequences unu have TunTu

Proof

Proof ()

Suppose TB(B1,B2). Let {(un,Tun)}n be a sequence in Γ(T) such that unu and Tunv. Then

v=limnTun=T(limuun)=Tu

since T is continuous. Thus

(u,v)=(u,Tu)Γ(T)$$ie$Γ(T)$isclosed.
Proof ()

Define

  • π1:Γ(T)B1,π1(u,Tu)=u
  • π2:Γ(T)B2,π2(u,Tu)=Tu

Note that Γ(T) is a Banach space since Γ(T)B1×B2 is a closed subspace of B1×B2 (which is Banach by the corrolary above)

Further, we have π1B(Γ(T),B1) and π2B(Γ(T),B2)

||π2(u,v)||=||v||||u||+||v||=||(u,v)||

(we can see it is bounded by the norm of B1×B2 as we defined above, which must also be bounded)

π1:Γ(T)B1 is one-to-one and onto (bijective). Thus S:=π11:B1Γ(T) is also a bounded linear operator

And so T=π2S is the composition of two bounded linear operators, and is thus itself a bounded linear operator.

ie TB(B1,B2)

see closed graph theorem

Note

Hahn Banach Theorem

See Hahn-Banach theorem

This theorem tries to answer the question

Question

Given a general, nontrivial, normed space V, does VB(V,K)={0}?

(recall that we defined the dual space in the last lecture, and we define these specific duals to be "functionals")

Example

Spaces with nontrivial dual spaces

  • (p)=p where 1p+1p=1 for 1p<
  • (c0)=1

In general, we want to know if we have elements in the dual space.

framework / axioms from set theory

Partial Order

A partial order on a set E is a relation E such that

  1. for all eE, ee
  2. for all e,fE, if ef and fe then e=f
  3. for all e,f,gE, if ef and fg, then eg

see partial order

Upper Bound

An upper bound of a set DE is an element eE such that de for all dD.

see upper bound

Maximal Element

A maximal element of E is an element e such that if fE and efef

see maximal element

Similar definitions for a minimal element

Chain

If (E,) is a partially ordered set, a chain in E is a set CE if for all e,fC we have either ef or fe

ie we can compare all the elements in the chain C

see chain

Hamel basis

A Hamel basis HV is a linearly independent set such that every element of V is a finite linear combination of elements in H.

see Hamel basis

Note

We know from linear algebra that we can find a basis and calculate the dimension of the space. Next time, we'll see how to apply these to infinite vector spaces.

Zorn's Lemma

If every chain in a nonempty, partially ordered set E has an upper bound, then E contains a maximal element.

Without Proof

see Zorn's lemma

Note

Created 2025-06-05 Last Modified 2025-06-05