MAT 2125B Elementary Real Analysis

2024 Midterm Examination Solutions

Tanya Schmah
 MAT 2125B Winter 2024 Midterm Exam 2

[8pts] 1. Label the following statements as true (T) or false (F). No justification is needed.
(1 mark for each correct answer, 0 marks for an incorrect answer or no answer.)

T Every bounded monotone sequence of real numbers converges.

Solution. This is the Monotone Convergence Theorem.

T Every subsequence of a convergent sequence in Rd is Cauchy.

Solution. Every subsequence of a convergent sequence is convergent, hence Cauchy.

F Every subsequence of a convergent sequence in R is monotone.

Solution. For example, consider an := (−1)n /n for example.

F If (an )∞ a sequence such that an ≥ 0 for all n and an → 0 then by the Alternating
n=1 is P
Series Test, ∞ n=1 (−1)
n+1
an must converge.
Solution. We also need the condition that (an ) is decreasing. Without P∞ that,n+1 here is
aPcounter-example: if a2k = 1/k and am = 0 when m is odd, then n=1 (−1) an =
∞
k=1 1/k, which diverges.
P P∞
F If 0 ≤ an ≤ bn for all n ∈ N and ∞n=1 an converges then n=1 bn converges.
P P
Solution. The comparison test is the other way round: if bn converges then an
converges. The statement in the question is not always true. E.g. if an = 0 for all n, then
P
the statement given here would imply that every bn converges!

T If A ⊆ Rd is open and B ⊆ Rd is closed, then A \ B is open.

Solution. Rd \ B is open, so A \ B = A ∩ (Rd \ B) is also open.

T Every integer is a boundary point of Q in R.

Solution. In fact, every real number is a boundary point of Q in R, since any open
interval contains both rational and irrational numbers.

T The union of two compact subsets of Rd is always compact.

Solution. If two sets are both closed and bounded, then their union is both closed and
bounded.
 MAT 2125B Winter 2024 Midterm Exam 3

[8pts] 2. Let S be a subset of R.

(a) Define “M is an upper bound for S”. (1)

Solution. x ≤ M for all x ∈ S

(b) Define the supremum of S, denoted sup S. (2)

Solution. sup S = M if M is an upper bound for S and M ≤ L for every upper bound
L for S.
Acceptable alternative definition: sup S = M if x ≤ M for all x ∈ S and for every ε > 0
there is an element x ∈ S such that M − ε < x.

(c) Suppose sup S = M and T = {x + t : x ∈ S}. Show that sup T = M + t. (3)

Solution. For every y ∈ T ,we know y = x + t for some x ∈ S. Since x ≤ M , it follows
that y ≤ M + t. Hence M + t is an upper bound for T .
Let K be an upper bound for T . Then x + t ≤ K for all x ∈ S, which implies that
x ≤ K − t for all x ∈ S, i.e. K − t is an upper bound for S. Since sup S = M , we have
M ≤ K − t, so M + t ≤ K. It follows that sup T = M + t.

(d) Is it true that every non-empty set of real numbers that is bounded above has a supremum?
Justify your answer briefly. (2)
Solution. Yes, this is the Completeness Axiom of the real numbers.
 MAT 2125B Winter 2024 Midterm Exam 4

[8pts] 3. Let (an )∞

n=1 be a sequence of real numbers.

(a) Define what it means for (an )∞

n=1 to converge to a number L ∈ R. (3)
Solution.
∀ε > 0, ∃N ∈ N such that n ≥ N =⇒ |an − L| < ε.
or (equivalently):

∀ε > 0, ∃N ∈ N such that, for all n ≥ N, |an − L| < ε.

or (equivalently):

∀ε > 0, ∃N ∈ N such that |an − L| < ε for all n ≥ N.

(b) Prove the Squeeze Theorem: if (bn )∞ ∞

n=1 , (cn )n=1 are also sequences of real numbers and

an ≤ b n ≤ c n for all n ∈ N,

and limn→∞ an = limn→∞ cn = L, then limn→∞ bn = L. (5)

Solution. Let ε > 0 be given. Since an → L, there exists n0,a , such that for all
n ≥ n0,a ,
L − ε < an < L + ε.
Since cn → L, there exists n0,c , such that for all n ≥ n0,c ,

L − ε < cn < L + ε.

Set n0 := max{n0,a , n0,c }. Then for n ≥ n0 ,

L − ε < an ≤ bn ≤ cn < L + ε,

so
|bn − L| < ε
as required.
 MAT 2125B Winter 2024 Midterm Exam 5

[8pts] 4. (a) Define what it means for a series to converge. (2)

Solution. P The sequence of partial sums (Sk )∞Pkconverges, where,
k=1
for a series ∞ a
n=1 n the partial sums are S k = n=1 an .

(b) Define what it means for a series to converge absolutely. (1)

P∞ P∞
Solution. n=1 a n converges absolutely if n=1 |an | converges.

(c) Give an example of a series that converges but does not converge absolutely.
(No justification required) (2)
P∞ (−1)n+1
Solution. n=1 n
is an example.
Justification (not required): It converges by the Alternating Series Test. It does not
convergePabsolutely, because when taking the series with the absolute values of the terms
∞ 1
we get n=1 n , which is the Harmonic Series, which does not converge.

(d) Prove that the following series converges absolutely:

X
∞
sin(2n ) + 1
n=1
2n

(3)
Solution.
sin(2n ) + 1 2
0≤ < .
2n 2n
P∞ P
Since the geometric series converges (since 21 < 1), its multiple ∞
1
n=1 2n
2
n=1 2n con-
P∞ sin(2n )+1 P∞ sin(2 n
verges too, so by the Comparison test, n=1 2n
converges. Hence n=1 2n)+1
converges absolutely.
 MAT 2125B Winter 2024 Midterm Exam 6

[8pts] 5. Let A be a subset of Rd .

(a) Define “A is bounded”. (1)

Solution. There exists M ∈ N such that ∥x∥ ≤ M for all x ∈ A.
Acceptable variants: M ∈ R+ or M > 0; also “< M ” instead of “≤ M ”.

(b) Define the closure of A, denoted A. (2)

There are multiple equivalent definitions. Any one of them is fine, except that if you use
the term “boundary” or “accumulation point” then you must define that too.
Solution. A is the set of all limits of convergent sequences in A.

(c) Prove that if A is bounded, then A is bounded. (5)

Solution. Suppose M ∈ N is such that ∥x∥ ≤ M for all x ∈ A. Let y ∈ A. Then y is
the limit of some sequence (an ) in A. By definition of the limit, taking ε = 1, there exists
an n ∈ N such that ∥an − y∥ < 1. Then by the triangle inequality,

∥y∥ ≤ ∥y − an ∥ + ∥an ∥ < 1 + M.

This argument works for all y ∈ A, so A is bounded (by M + 1).

 MAT 2125B Winter 2024 Midterm Exam 7

[5pts] 6. (Bonus)
Prove that if (an )∞ ∞
n=1 is a bounded sequence then there exists a subsequence (ank )k=1 that
converges to lim sup an .
n→∞

Solution. We use the definition

lim sup an = inf{eventual upper bounds}

n→∞
= inf{β ∈ R : ∃n0 such that an ≤ β ∀n ≥ n0 }.

Set L := lim sup an . Note that every number greater than L is an eventual upper bound, while
every number less than L is not an eventual upper bound.
We will construct a subsequence (ank )∞
k=1 such that |ank − L| ≤
1
k
for all k. By the Squeeze
Theorem, it will follow that ank → L.
For k = 1, we note that L + 1 is an eventual upper bound so there is some n0 such that
an ≤ L + 1 for all n ≥ n0 . We also note that L − 1 is not an eventual upper bound, so there
exists n1 ≥ n0 such that an1 ≥ L − 1. Since n1 ≥ n0 , we also have an1 ≤ L + 1, and thus
|an1 − L| ≤ 1.
Next for k = 2, L + 1/2 is an eventual upper bound, so there is some new n0 such that
an ≤ L + 1/2 for all n ≥ n0 . We also note that L − 1/2 is not an eventual upper bound, so
there exists n2 ≥ max{n0 , n1 } such that an2 ≥ L − 1/2.
Next for k = 3, L + 1/3 is an eventual upper bound, so there is some new n0 such that
an ≤ L + 1/3 for all n ≥ n0 . We also note that L − 1/3 is not an eventual upper bound, so
there exists n3 ≥ max{n0 , n2 } such that an3 ≥ L − 1/3.
Continuing in this way produces the desired subsequence.