Final Exam (2011)
Uploaded by
danFinal Exam (2011)
Uploaded by
danQuestion 1 (10 + 5 + 5 = 20 marks)
(a) (i) Define the Lebesgue outer measure m∗ and the Lebesgue
inner measure m∗ .
(ii) Prove that an interval is Lebesgue measurable.
(iii) If A is a Lebesgue measurable subset of R and h > 0, prove
that m(A + h) = m(A), where A + h = {x + h, x ∈ A} and
m is Lebesgue measure.
(b) Let f : R → R be differentiable. Prove that f ′ is Lebesgue
measurable.
(c) Let X ⊆ R and f, g : X → R be Lebesgue measurable. Show
that f g is measurable.
OVER PAGE
1
2
Question 2 (6 + 4 + 3 + 7 = 20 marks)
n
X
(a) Let f = ai χAi be a measurable simple function.
i=1
(i) Define the Lebesgue integral of f.
(ii) Let f, g be integrable simple functions. Prove that
Z Z Z
(f + g) = f + g.
(b) Show that sin(x2 ) is not Lebesgue integrable on [0, ∞). (Hint:
Let x2 = u and compare the integral to an infinite sum).
(c) Let Q denote the rational numbers. Find the Lebesgue integral
over [0, 1] of the function
(
x2 , x ∈ [0, 1] − Q
f (x) =
2x, x ∈ [0, 1] ∩ Q.
R R R
(Hint: ifR A, B are disjoint, A∪B f = A f + B f. If m(E) = 0,
what is E f ?).
(d) Outline the proof of the following result. If f is Riemann inte-
grable on [a, b], then f is Lebesgue integrable and
Z b Z b
R f (x)dx = L f (x)dx.
a a
OVER PAGE
3
Question 3 (2 + 8 + 5 + 5 = 20 marks)
(a) State the monotone convergence theorem.
R1 2
(b) (i) Calculate limn→∞ 0
n2 x3 e−nx dx.
(ii) Show that
Z 1 Z 1
2 3 −nx2 2
lim nxe dx 6= lim n2 x3 e−nx dx.
n→∞ 0 0 n→∞
(iii) Why does this not contradict the Dominated Convergence
Theorem? (Hint: Treat the integrand as a function of n
and find out where its maximum value occurs).
(c) Let {fn }∞
n=1 be a sequence of continuous functions converging
uniformly to f on [a, b]. Use the Dominated Convergence The-
orem to give a proof of Riemann’s result that
Z b Z b
lim fn (x)dx = f (x)dx.
n→∞ a a
(Hint: If |fn − f | < ǫ, then fn ∈ (f − ǫ, f + ǫ)).
(d) Calculate the integral
Z ∞
sin(αx) sin(βx)
F (t) = e−xt dx, t > 0.
0 x
You do not have to prove existence of the integral first. You may
need the identities cos(A − B) − cos(A + B) = 2 sin(A) sin(B)
and Z ∞
t
e−xt cos(x)dx = .
0 1 + t2
OVER PAGE
4
Question 4 (9 + 5 + 6 = 20 marks)
The Fourier transform of f ∈ L1 (R) is defined by
Z ∞
b
f (y) = f (x)e−iyx dx.
−∞
(a) Using the Fourier transform, solve the partial differential equa-
tion
ut = 2uxx + u, x ∈ R,
u(x, 0) = f (x), f ∈ L1 (R).
State clearly any assumptions
R ∞+ic you have
p made on the solution
2
u. You may assume −∞+ic e−ax dx = π/a for c ∈ R.
R
1 ∞
b π
(b) (i) Prove that f (y) = 2 −∞ f (x) − f (x + y ) e−iyx dx. (Hint:
Let x = z + πy in the definition of the Fourier transform
and find an expression for 2fb(y)).
(ii) Prove that lim|y|→∞ fb(y) = 0.
(c) Calculate the Fourier transform of f (x) = e−a|x| . Hence sum the
series ∞
X 2
.
n=−∞
1 + n2
You do not need to prove existence of the integral. (Hint: Pois-
son Summation).
OVER PAGE
5
Question 5 (2 + 10 + 5 + 3 = 20 marks)
(a) Define a Hilbert space.
(b) (i) Prove that in a Hilbert space the norm satisfies
kx + yk2 + kx − yk2 = 2kxk2 + 2kyk2.
(ii) Show that C([a, b]), the space of continuous functions on
[a, b], is a Banach space under the norm kf k∞ = supx∈[a,b] |f (x)|.
(Hint: Convergence in this norm is uniform)
(iii) Prove that C([a, b]) with the norm of (ii) is not a Hilbert
space. (Hint: Is part (i) satisfied by the functions x(t) = 1,
t−a
y(t) = b−a ?)
(c) Show that in a Banach space B, the series
∞
X
vn , vn ∈ B
n=1
is convergent if
∞
X
kvn k < ∞.
n=1
(d) If f ∈ L2 ([a, b]) prove that f ∈ L1 ([a, b]).
OVER PAGE
6
Question 6 (6 + 4 + 7 + 3 = 20 marks)
(a) Suppose that f ∈ L1 ([−π, π]) and that the Fourier coefficients
Rπ
are fb(n) = 2π
1
−π
f (y)e−iny dy. Show that
XN Z π
1
fb(n)e =
inx
f (y)DN (x − y)dy,
n=−N
2π −π
sin([N + 12 ]x)
where DN (x) = sin( 12 x)
is Dirichlet’s kernel.
(b) Suppose that f is continuous on [−π, π], but
∞
X
f (x) 6= fb(n)einx .
n=−∞
How can we recover f from the Fourier coefficients?
(c) Prove that if f is a positive µ integrable function, then
Z
ν(E) = f dµ
E
is a measure.
(f) State the Riesz Representation Theorem for Cc (R), the com-
pactly supported functions on R.
