Scribd
Upload
141 views2 pages

Functional Analysis: Problem Sheet 3: Paul D. Mitchener

1. Shows that the dual space V* of a vector space V is never empty. 2. Proves that any vector space V can be written as the direct sum of a subspace Z and its annihilator W. 3. Applies the Stone-Weierstrass theorem to show that the set of finite linear combinations of cosines on the interval [0,π] is dense in the continuous functions on that interval.

Uploaded by

Zohaib Shabir
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
141 views2 pages

Functional Analysis: Problem Sheet 3: Paul D. Mitchener

1. Shows that the dual space V* of a vector space V is never empty. 2. Proves that any vector space V can be written as the direct sum of a subspace Z and its annihilator W. 3. Applies the Stone-Weierstrass theorem to show that the set of finite linear combinations of cosines on the interval [0,π] is dense in the continuous functions on that interval.

Uploaded by

Zohaib Shabir
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Functional Analysis: Problem Sheet 3

Paul D. Mitchener

1. (a) Let v ∈ V , v 6= 0. Define f : Span{v} → F by f (αv) = α. Then


f ∈ V ∗ . By the Hahn-Banach theorem, we have F ∈ V ∗ such that
F |Span(v) = f and kF k = kf k. In particular, F (v) = 1, so F 6= 0.
Thus V ∗ 6= {0}.
(b) Suppose x 6= y. If x and y are not linearly independent, then either
x = αy or y = αx for some α ∈ R, α 6= 1. Suppose that y = αx.
Then x 6= 0, and we can define f : Span{x, y} = Span{x} → F by
f (αx) = α.
If x and y are linearly independent, define f : Span{x, y} → F by
f (αx + βy) = α + β.
Since Span{x, y} is finite-dimensional in either of the above two cases,
f is a bounded linear map. By the Hahn-Banach theorem, we have
F ∈ V ∗ such that F |Span(x,y) = f and kF k = kf k. Note that
f (x) 6= f (y), so F (x) 6= F (y). The result now follows.
(c) Let x ∈ V . Define f : Span{x} → F by f (αx) = αkxk. Then kf k = 1,
|f (x)| = kxk. By the Hahn-Banach theorem, we have F ∈ V ∗ such
that F |Span(x) = f and kF k = kf k = 1. In particular, F (x) = kxk.
So by the above, if |g(x)| ≤ c whenever g ∈ V ∗ with kgk ≤ 1, then
in particular |F (x)| ≤ c. But F (x) = kxk, so kxk ≤ c are required.
2. Let P be the set of all subspaces U ⊆ V such that U ∩ W = {0}. Then P
is a partially ordered set, with relation ⊆.
Let Q be a totally ordered subset of P . Let U0 = ∪U ∈Q U . Certainly
0 ∈ U0 . Let x, y ∈ U0 , α, β ∈ F.
Then x ∈ U1 and y ∈ U2 for some U1 , U2 ∈ Q. Since Q is totally ordered,
either U1 ⊆ U2 or U2 ⊆ U1 . Suppose U1 ⊆ U2 .
Then x, y ∈ U2 . Since U2 is a subspace, we have that αx + βy ∈ U2 ⊆ U0 ,
so U0 is a subspace of V . Since U ∩ W = {0} for all U ∈ Q, we have that
U0 ∩ W = {0}. Hence U0 ∈ P , and U0 is an upper bound for Q.
By Zorn’s lemma, the set P therefore has a maximal element, Z. We claim
that Z + W = V .
Suppose Z + W 6= V . Then we have a vector v ∈ V such that v 6∈ Z + W .
Let Z 0 = {z + αv | z ∈ Z, α ∈ F}.
Then Z 0 is a subspace of V and Z 0 ∩ W = {0}. Further, Z ⊆ Z 0 and
Z 6= Z 0 . This contradicts Z 0 being the maximal element of P .

1
2

Hence Z + W = V and we are done.


3. By the formula
1
cos(mx) cos(nx) = (cos(m + n)x + cos(m − x)x)
2
we see that A is closed under multiplication. The set A is certainly a
vector space, so A is a subalgebra of CR [0, π]. It contains the constant
functions.
Let x, y ∈ [0, π] with x 6= y. Then cos x 6= cos y, so A separates points.
Hence by the Stone-Weierstrass theorem, the set A is dense in CR [0, π].

You might also like