May2018paper PDF

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Honours Analysis

MATH10068 Solutions and comments May 2018

Comment: This is the May 2018 Hons Analysis exam.


1.
(a) Let E ⊆ R, and forP each n ∈ N let fn : E → R be a function. Define what it means for the

series of functions P n=1 fn to converge pointwise on E. Define what it means for the

series of functions n=1 fn to converge uniformly on E. [6 marks]
(b) State the Weierstrass M -test for uniform convergence of a series of functions defined on E.
[5 marks]
(c) In each case find whether the given series of functions converges pointwise on the given
interval(s). Determine whether or not the convergence is uniform, giving reasons for your
answer.

X (−1)n x2n+1
(i) on I = [−1, 1] and I = R. [5 marks]
n=0
(2n + 1)!

X
(ii) (1 − x)xn on I = [0, a) for a < 1 and I = [0, 1]. [5 marks]
n=0

X xn −nx
(iii) e on I = [0, ∞) [4 marks]
n=0
1 + xn

Solution:
P∞
(a) Series n=1 fn (x) converges pointwise to a function s : E → R on E iff for every x ∈ E and
Pk
 > 0 there is an N ∈ N such that for k ≥ N we have | n=1 fn (x) − s(x)| < .
P∞
Series n=1 fn (x) converges uniformly to s : E → R on E iff for every  > 0 there is an N ∈ N
Pk
such that for k ≥ N and all x ∈ E we have | n=1 fn (x) − s(x)| < . Alternatively, iff for
Pk
every  > 0 there is an N ∈ N such that for k ≥ N we have supx∈E | n=1 fn (x) − s(x)| < .
Comment: Standard definition.

(b) Suppose there exists a sequence (Mn )n∈N such that



X
|fn (x)| ≤ Mn , ∀x ∈ E and n ∈ N and Mn < ∞.
n=1
P∞
Then the series n=1 fn (x) converges uniformly on E. Comment: Standard bookwork
(c) (i) The series converges pointwise on R (By the ratio test the radius of convergence of this
power series is R = ∞). The convergence is uniform on [−1, 1] (by the M -test) we have
(−1)n x2n+1

1
(2n + 1)! ≤ (2n + 1)!

P∞ 1
on [−1, 1] and n=0 (2n+1)! < ∞. The convergence is not uniform on R as

(−1)n x2n+1

sup
= ∞.
x∈R (2n + 1)!

(ii) By the M -test the series converges uniformly on any I = [0, a) for a < 1, since
X
|(1 − x)xn | ≤ an , and an < ∞.
n

1
Honours Analysis
MATH10068 Solutions and comments May 2018

From this the series converges (pointwise) for any 0 ≤ x < 1. We can check x = 1 by
hand and hence we have pointwise convergence on I = [0, 1]. The convergence fails to
be uniform there. To see this consider the partial sums of the series
n
X
sn (x) = (1 − x)xn = 1 − xn+1 .
k=0

Clearly, all sn are continuous function on [0, 1] but they converge to a function s(x) = 1
on [0, 1) but s(x) = 0 when x = 1. Hence s is not continuous and therefore the
convergence is not uniform.
(iii) We consider first I1 = [0, 1/2] where we have the estimate
xn −nx

≤ 2−n ,
1 + xn e

from which the uniform convergence follows by the M -test. On I2 = [1/2, ∞) we have
xn −nx

≤ e−n/2 ,
1 + xn e

which again gives uniform convergence of the series by the M -test as n e−n/2 < ∞.
P
From this the convergence is uniform on the union of these intervals and hence on the
whole [0, ∞).
Comment: Similar to seen examples.

2.
(a) Define, using step functions, what it means for a function f : R → R to be Riemann-
integrable. Show that a Riemann-integrable function must be bounded, and must be zero
outside a bounded interval. [8 marks]
(b) Suppose that M is a nonnegative integer. Use your definition from part (a) to show that the
function defined by f (x) = x for 0 ≤ x ≤ M and f (x) = 0 otherwise, is Riemann-integrable
and calculate its integral.
[8 marks]
(c) Is it true that for every f such that |f | is Riemann-integrable, f is also Riemann-integrable?
If so, prove it, if not, give a counterexample (with justification). [4 marks]
(d) Show that the function given by g(x) = 1 if xR = 1 − n−1 for some n ∈ N and g(x) = 0
otherwise is Riemann-integrable and calculate g. [5 marks]

Solution:
R R
(a) If φn and ψn are sequences of step functions satisfying φn ≤ fR ≤ ψn and
R ψn − φn → 0
as n → ∞, thenR f is Riemann-integrable, and the sequences φn and ψn both converge
to the value of f .
The second part follows from the definition of a step function. Since every step function is
bounded and vanishing outside a bounded interval, the fact that φn ≤ f ≤ ψn implies the
same for f .
Comment: Standard result.

2
Honours Analysis
MATH10068 Solutions and comments May 2018

(b) Consider the sequences of step functions defined by


n  
X j−1
φn (x) = M χ[M (j−1)/n ,M j/n) (x)
j=1
n

and
n  
X j
ψn (x) = M χ[M (j−1)/n ,M j/n) (x).
j=1
n

Then φn ≤ f ≤ ψn and
n n n−1
M2 X M2 X
 
j−1
Z
MX
φn = M = 2 (j − 1) = 2 j
n j=1 n n j=1 n j=1

M 2 (n − 1)n M2
= 2
= (1 − 1/n)
n 2 2
while
n n
M2 X
Z  
MX j
ψn = M = 2 j
n j=1 n n j=1

M 2 n(n + 1) M2
= = (1 + 1/n).
n2 2 2

2 2 2
Now the sequences M2 (1 − 1/n) and M2 (1 + 1/n) both converge to M2 as n → ∞ and
2
so the sequences φn and ψn both converge to M2 . Hence f is Riemann-integrable and
R R
2
f = M2 .
R

Comment: Done f (x) = x on [0, 1] in lectures.

(c) No, let f (x) = 1 if 0 ≤ x ≤ 1 and x is rational, f (x) = −1 if 0 ≤ x ≤ 1 and x is irrational,


and f (x) = 0 if x < 0 or x > 1. Then |f | = χ[0,1] is integrable as it is a step function, but f
is not since for every step function φ with φ ≤ f we must have R φ ≤R −χ[0,1] while for every
step function ψ with ψ ≥ f we must have ψ ≥ χ[0,1] . Hence ψ − φ ≥ 2, and so f is not
Riemann-integrable.
Comment: Seen.
(d) Let φn = 0 and let

1,
 if x ∈ [1 − 1/n, 1],
ψn (x) = 1, if x = 1 − 1/m for some m = 1, 2, . . . , n − 1,

0, otherwise.

ψn = n−1 → 0. So
R R
Then φn and ψn are step functions, Rφn ≤ g ≤ ψn and φn = 0 while
by our criterion, g is integrable and g = 0.
Comment: Very similar to seen.

3.
(a) Define the notion of uniform continuity of a function on an interval I of R. [6 marks]
(b) Show that if there are sequences (xn ) and (yn ) in I with |xn − yn | → 0 as n → ∞ but
|f (xn ) − f (yn )| ≥ 1 for all n, then f is not uniformly continuous on I. [5 marks]

3
Honours Analysis
MATH10068 Solutions and comments May 2018

(c) Use the sequences xn = log n and yn = log(n + 1) to deduce that the function f (x) = ex is
not uniformly continuous on R. [7 marks]
(d) If f : R → R is uniformly continuous on the intervals [−1, 1] and [0, 2] show that it is
uniformly continuous on the interval [−1, 2].
[7 marks]

Solution:
(a) f : I → R is uniformly continuous if for all  > 0 there is a δ > 0 such that x, y ∈ I and
|x − y| < δ implies |f (x) − f (y)| < .
Comment: Standard definition covered in workshop.
(b) As follows from (a) the uniform continuity of f implies that if |xn − yn | → 0 then |f (xn ) −
f (yn )| → 0 as n → ∞. Hence having |f (xn ) − f (yn )| ≥ 1 disproves the uniform continuity.
Comment: Seen.
(c) Clearly using mean value theorem
1 1
|yn − xn | = | log(n + 1) − log n| = |n + 1 − n| = ,
c c
for some c ∈ (n, n + 1). From this
1
|yn − xn | ≤ → 0, as n → ∞.
n
On the other hand

f (yn ) − f (xn ) = elog(n+1) − elog n = n + 1 − n = 1.

Hence using (b) we see that f is not uniformly continuous on R.


Comment: Seen similar.
(d) We have to establish

∀ε > 0∃δ > 0 : |x − y| < δ, x, y ∈ [−1, 2] =⇒ |f (x) − f (y)| < ε.

We are allowed to use uniform continuity on [−1, 1] and [0, 2]. That is we know that

∀ε > 0∃δ1 > 0 : |x − y| < δ1 , x, y ∈ [−1, 1] =⇒ |f (x) − f (y)| < ε

and
∀ε > 0∃δ2 > 0 : |x − y| < δ2 , x, y ∈ [0, 2] =⇒ |f (x) − f (y)| < ε.
For a fixed ε > 0 consider δ = min{δ1 , δ2 , 1} with δ1 , δ2 defined as above.
Consider any x, y ∈ [−1, 2]. Without loss of generality let x < y. If |x − y| < δ ≤ 1 then
either x ∈ [−1, 0] and hence y ∈ [−1, 1] or x ∈ [0, 2] and hence y ∈ [0, 2]. Hence in each case
x and y belongs simultaneously to one of the intervals [−1, 1] or [0, 2]. We can therefore use
known uniform continuity of f on these subintervals (c.f. above) to conclude that

|f (x) − f (y)| < ε,

as desired.
Comment: Unseen.

4. Let (X, d) be a metric space.

4
Honours Analysis
MATH10068 Solutions and comments May 2018

(a) Define what it means for a subset A ⊂ X to be connected. [5 marks]


(b) State without proof which are the connected subsets of R (in the usual metric of R).
[4 marks]
(c) Define what it means for a set F ⊂ X to be closed. Show carefully that {x ∈ X : d(x, a) ≤
r} is a closed set. [7 marks]
(d) Let E ⊂ X. Define the closure E of E. Let U = {x ∈ X : d(x, a) < r}. Is it always true
that U = {x ∈ X : d(x, a) ≤ r} if
(i) X = Rn with the usual metric?
(ii) X is an arbitrary metric space?
In each case justify your answer briefly. [9 marks]

Solution:

(a) A subset A ⊂ X is connected if there do not exist open and disjoint sets U, V ⊂ X such that

A ∩ U 6= ∅, B ∩ V 6= ∅, and A ⊂ U ∪ V.

Comment: Standard definition.


(b) A ⊂ R is connected iff A is any type of interval (open, closed or semi-open).
Comment: Bookwork.
(c) A set F ⊂ X is closed iff the complement X \ F is open. That is for any x ∈ X \ F there is
r > 0 such that B(x, r) ⊂ X \ F .
For F = {x : d(x, a) ≤ r}, clearly, X \ F = {x : d(x, a) > r}. Pick any y ∈ {x : d(x, a) > r}.
Since d(y, a) > r we have that r0 = d(y, a) − r > 0. Then for any

x ∈ B(y, r0 ) we have that d(x, a) ≥ d(y, a) − d(x, y) > d(y, a) − r0 = r.

From this B(y, r0 ) ⊂ {x : d(x, a) > r} which shows that {x : d(x, a) > r} is open and hence
{x : d(x, a) ≤ r} is closed.
Comment: Standard.
(d) For E ⊂ X the closure E is the smallest closed set in X that contains E (It always exists as
it is an intersection of all closed subsets of X that contains E).
To answer the second part, it’s more useful to use that

E = {x ∈ X : x is a limit point of E}.

(i) When X = Rn the answer is yes. Any point of the set {x : d(x, a) ≤ r} is a limit point
of U , to see this for x ∈ {x : d(x, a) ≤ r} consider a sequence
n
xn = a + (x − a) ∈ U, clearly xn → x.
n+1

(ii) No. Consider a discrete metric on X with U = {x : d(x, a) < 1} = {a} and hence
U = {a} but {x : d(x, a) ≤ 1} = X.

Comment: Similar question discussed.

5.

5
Honours Analysis
MATH10068 Solutions and comments May 2018

(a) Let (X, d) be a metric space. Define what is meant by a contraction on X. State and prove
Banach’s Contraction Mapping Theorem. [10 marks]
(b) Let g : [0, 1] → R be an arbitrary continuous function. Show that there is a unique continuous
function f : [0, 1] → R such that for all s ∈ [0, 1]
Z s
f (s) = f (t)(s − t)dt + g(s).
0

[15 marks]

Solution:
(a) A map f : X → X is a contraction if there exists an α < 1 such that for all x, y ∈ X we have
d(f (x), f (y)) ≤ αd(x, y). Banach’s theorem states that if X is complete and f : X → X is
a contraction, then there is a unique point x ∈ X such that f (x) = x.

Pick x0 ∈ X and let x1 = f (x0 ), x2 = f (x1 ) etc. so that xn+1 = f (xn ). Consider
d(xn+1 , xn ) = d(f (xn ), f (xn−1 )) ≤ αd(xn , xn−1 )) by hypothesis. Repeating, we have
d(xn+1 , xn ) ≤ αn d(x1 , x0 )
so that, when m ≥ n
d(xm , xn ) ≤ d(xm , xm−1 ) + · · · + d(xn+1 , xn ) ≤ (αm−1 + · · · + αn )d(x1 , x0 )
αn
≤ d(x1 , x0 )
1−α
since α < 1. This shows that (xn ) is a Cauchy sequence in X and hence, by completeness of
X, there is an x ∈ X to which it converges.

Now a contraction map is continuous, so continuity of f at x shows that f (x) = f (limn xn ) =


limn f (xn ) = limn xn+1 = x, so that indeed f (x) = x.

Finally, if there are x, y ∈ X with f (x) = x and f (y) = y, we have d(x, y) = d(f (x), f (y)) ≤
αd(x, y) which since α < 1 forces d(x, y) = 0, i.e. x = y.
Comment: Standard.
(b) On the complete metric space C([0, 1]) with the metric d∞ (f, g) = supt∈[0,1] |f (t) − g(t)|
consider the mapping T : C([0, 1]) → C([0, 1]) given by
Z s
T (f )(s) = f (t)(s − t)dt + g(s).
0

Now for any 0 ≤ s ≤ 1 we have


Z s
T (f )(s) − T (h)(s) = (f (t) − h(t))(s − t)dt
0

so that Z s
|T (f )(s) − T (h)(s)| ≤ |f (t) − h(t)|(s − t)dt
0
s
s2
Z
1
≤ sup |f (t) − h(t)| (s − t)dt = d∞ (f, h) ≤ d∞ (f, h)
0≤t≤1 0 2 2
so that
1
d∞ (T f, T h) ≤
d∞ (f, h).
2
Since α = 1/2 < 1, T is a contraction and therefore there is a unique fixed point f ∈ C([0, 1])
of T .

You might also like