Lecture 2 Basics II

1 Sequences and Limits in R

In this section, we discuss real sequences and their limits. We begin with the defini-
tion of a linearly ordered set, the supremum, and the infimum.

Definition 1.1. Suppose R is a binary relation on a set X. X is called a linearly

ordered set if R is complete, transitive, and antisymmetric.

Loosely speaking, when X (equipped with R) is a linearly ordered set, we can

arrange the elements of X on an axis according to the corresponding R. One example
is the real line R with ≥.

Definition 1.2. Let X be a linearly ordered set and S ⊆ X. We say that

(i) S is bounded above if ∃ b ∈ X such that ∀x ∈ S, we have bRx. We call b an
upper bound of S.
(ii) S is bounded below if ∃ b ∈ X such that ∀x ∈ S, we have xRb. We call b a
lower bound of S.
(iii) S is bounded if it is bounded above and bounded below.

1
Definition 1.3. Let X be a linearly ordered set and S ⊆ X. We say that
(i) a ∈ X is the supremum/least upper bound of S if (1) a is an upper bound of
S, and (2) b ∈ X is an upper bound of S implies that bRa. We use sup S to denote
the supremum of S.
(ii) a ∈ X is the infimum/greatest lower bound of S if (1) a is a lower bound of
S, and (2) b ∈ X is a lower bound of S implies that aRb. We use inf S to denote
the infimum of S.

Example 1.1. Let X = R and S = [a, b). Then a is the infimum of S, and b is the
supremum of S. In this case the infimum of S is in S but the supremum of S is not
in S.

From here on, we will assume that X = R and R =≥. For any S ⊆ R, let
−S = {x ∈ R : −x ∈ S}. Verify that inf S = − sup −S and sup S = − inf −S.

1.1 Real Sequences and Limits

There are many way to construct the set of real numbers, R. Here, we adopt the
completeness axiom, i.e., we assume that the completeness axiom is true without the
need of proving it.

The Completeness Axiom. ∀ nonempty S ⊆ R that is bounded above, sup S ∈ R.

√ √
Remark 1.1. Let S = {x ∈ Q : x ≤ 2}. Then sup S = 2 ∈ R\Q. This means
that Q does not satisfy the completeness axiom: S ⊆ Q is nonempty and bounded
√
above (by any rational number larger than 2), but sup S ̸∈ Q.

2
Definition 1.4. A real sequence (an )∞
n=1 is a function a : N+ → R s.t. a(n) = an

for any n ∈ N+ .

Often we write (an ) instead of (an )∞

n=1 for simplicity.

Definition 1.5. A real sequence (an ) converges to a limit α ∈ R if ∀ε > 0, ∃N ∈ N+

s.t. ∀n ≥ N , we have |an − α| < ε. We write limn→∞ an = α or an → α.

Remark 1.2. If an → α, then ∀ε > 0, only finitely many an ’s are not in (α−ε, α+ε).

Exercise 1.1. ∀ε > 0, there are only finitely many an ’s that are not in (α − ε, α + ε).
Show that an → α.

Definition 1.6. A real sequence that does not converge is said to diverge.

Remark 1.3. ¬(∃α ∈ R, ∀ε > 0, ∃N ∈ N+ s.t. ∀n ≥ N, we have |an − α| < ε) ≡

(∀α ∈ R, ∃ε > 0, ∀N ∈ N+ , ∃n ≥ N s.t. |an − α| ≥ ε).

Definition 1.7. A real sequence (an ) is bounded if a(N+ ) is bounded, and (an ) is
monotone if it is a monotone (increasing or decreasing) function.

Theorem 1.1. Every monotone bounded real sequence converges (to some α ∈ R).

Proof. Since a(N+ ) is bounded above and below in R, sup a(N+ ) and inf a(N+ ) exists.
Let (an ) be increasing and α = sup a(N+ ). The case in which (an ) is decreasing is
similar. We want to show an → α. For any ε > 0, since α − ε is not an upper bound
of a(N+ ), there exists aN > α − ε. For any n ≥ N , an ≥ aN > α − ε. Since α is an
upper bound of a(N+ ), we know that an ≤ α < α + ε. So (an ) converges to α.

3
Definition 1.8. A subsequence (ank )∞ ∞
k=1 of (an )n=1 is a sequence in which 1 ≤ n1 <

n2 < n3 < · · · .

Theorem 1.2. Every real sequence has a monotone subsequence.

Proof. Call m ∈ N+ nice if ∀n > m, an ≤ am . Suppose {m ∈ N+ : m is nice}

contains infinite elements. Pick n1 ≤ n2 ≤ n3 ≤ · · · from the set of nice points.
Then by construction, we have an1 ≥ an2 ≥ an3 ≥ · · · . So (ank )∞
k=1 is a decreasing

subsequence and we are done.

Suppose {m ∈ N+ : m is nice} contains finitely many elements (may contain
none). Pick n1 ∈ N+ such that ∀ nice m, n1 > m. By construction n1 is not nice.
Thus, ∃n2 > m1 s.t. an2 > an1 . Also, n2 > n1 cannot be nice. Then ∃n3 > n2
s.t. an3 > an2 , and so on. Then we have constructed a subsequence (ank )∞
k=1 that is

increasing.

Theorem 1.3. Every bounded real sequence has a convergent subsequence.

Proof. Every sequence has a monotone subsequence. Since the sequence is bounded,
this monotone subsequence is bounded and hence converges.

Theorem 1.4. Suppose an → α, bn → β. Then

(i) (an + bn ) → α + β.
(ii) (an bn ) → αβ.
(iii) If β ̸= 0 and bn ̸= 0 ∀n ∈ N+ , then (an /bn ) → α/β.

Theorem 1.5. Suppose an → α, bn → β. Then

(i) If an ≥ bn ∀n ∈ N+ , then α ≥ β.
(ii) If an ≥ cn ≥ bn ∀n ∈ N+ and α = β, then cn → α = β.

4
 1 1
Does an > bn ∀n ∈ N+ imply α > β? The answer is no. Let an = n
and bn = n+1
.
an > bn but both sequences converge to 0.

Definition 1.9. A real sequence is a Cauchy sequence if ∀ε > 0, ∃N ∈ N+ s.t.

∀n, m ≥ N , we have |an − am | < ε.

Theorem 1.6. Every convergent real sequence is Cauchy.

ε
Proof. For any ε > 0, take 2
> 0. Since (an ) converges to α, there exists N ∈ N+ s.t.
ε
for any n, m ≥ N , we have |an − α| < 2
and |am − α| < 2ε . By triangle inequality,

|an − am | = |an − α + α − am | = |(an − α) − (am − α)|

ε ε
≤ |an − α| + |am − α| < + = ε.
2 2

Thus, we have shown that for any ε > 0, there exists N ∈ N+ such that for any
n, m ≥ N , |an − am | < ε.

Theorem 1.7. Every real Cauchy sequence is bounded.

Proof. For any ε > 0, there exists N ∈ N+ s.t. for any n ≥ N , we have |aN − an | < ε.
Then, for any n ≥ N , aN −ε < an < aN +ε. For any n ∈ N+ , min{a1 , . . . , aN −1 , aN −
ε} ≤ an ≤ max{a1 , . . . , aN −1 , aN + ε}.

Theorem 1.8. Every real Cauchy sequence converges (to some α ∈ R).

Sketch of the proof: (i) Every Cauchy sequence is bounded. (i) There exists a
convergent subsequence, say the limit is α. (ii) Show that an → α directly.

5
 P∞
Definition 1.10. For a real sequence (an ), we call n=1 an = a1 + a2 + · · · an

infinite series. Let bn = ni=1 ai . If limn→∞ bn exists, we say that ∞

P P
n=1 an converges

and write ∞
P P∞
n=1 an = limn→∞ bn ; otherwise, we say that n=1 an diverges.

P∞ 1
P∞ α
Example 1.2. n=1 n diverges. n=1 αβ n−1 = 1−β
if |β| < 1.

2 Real Functions

Now we discuss real functions.

Definition 2.1. A real function is a function f : A → R such that A ⊆ R. We write

(i) f = a if f (x) = a ∀x ∈ A.
(ii) f ≥ g if f (x) ≥ g(x) ∀x ∈ A.
(iii) f > g if f ≥ g and g ≱ f .
(iv) f ≫ g if f (x) > g(x) ∀x ∈ A.

Definition 2.2. Let f : A → R and g : A → R be real functions and let α ∈ R. We

define
(i) f + g be the function s.t. (f + g)(x) = f (x) + g(x) ∀x ∈ A.
(ii) αf be the function s.t. (αf )(x) = α · f (x).
(iii) f · g be the function s.t. (f · g)(x) = f (x) · g(x).
 
(iv) fg be the function s.t. fg (x) = fg(x)(x)
, if g(x) ̸= 0 for any x ∈ A.

Definition 2.3. A real function f : A → R is (weakly) increasing if x > x′ ⇒

f (x) ≥ f (x′ ). It is strictly increasing if x > x′ ⇒ f (x) > f (x′ ).

6
Definition 2.4. A real function f : A → R converges to a limit q ∈ R as x ap-
proaches p ∈ R if ∀ε > 0, ∃δ > 0 s.t. ∀x ∈ A s.t. 0 < |x − p| < δ, we have
|f (x) − q| < ε. We write limx→p f (x) = q.

Remark 2.1. |x − p| > 0 rules out x = p, so f (p) can be outside the ε-interval.

Definition 2.5. A real function f : A → R converges to q ∈ R from the right (left) as

x approaches p ∈ R if ∀ε > 0, ∃δ > 0 s.t. ∀x ∈ A s.t. p < x < p + δ (p − δ < x < p),
we have |f (x) − q| < ε. We write limx→p+ f (x) = q (limx→p− f (x) = q).

Theorem 2.1. Suppose that f : A → R and g : A → R satisfy limx→p f (x) = q1 and

limx→p g(x) = q2 . Then
(i) limx→p (αf (x) + βg(x)) = αq1 + βq2 ;
(ii) limx→p f (x)g(x) = q1 q2 ;
f (x) q1
(iii) limx→p g(x)
= q2
if g(x) ̸= 0 for any x ∈ A and q2 ̸= 0.

Definition 2.6. A function f : A → R is continuous at p ∈ A if limx→p f (x) = f (p).

A function f : A → R is continuous if ∀p ∈ A, f is continuous at p.

Theorem 2.2. A function f : A → R is continuous at p ∈ A if and only if ∀

sequence (xn ) in A such that xn → p, we have f (xn ) → f (p).

Remark 2.2. (i) limx→p f (x) = q if and only if ∀ sequence (xn ) in A\{p} such that
xn → p, f (xn ) → q.
(ii) limx→p− f (x) = q if and only if ∀ increasing sequence (xn ) in A\{p} such that
xn → p, f (xn ) → q.
(iii) limx→p+ f (x) = q if and only if ∀ decreasing sequence (xn ) in A\{p} such
that xn → p, f (xn ) → q.

7
Exercise 2.1. Let f : R → R be given by

x sin 1 , if x ̸= 0


x
f (x) =

0, if x = 0.


Is f continuous at 0?
Yes, it indeed is continuous at 0. Now we prove it.

Proof. We want to show that limx→0 f (x) = f (0) = 0. By definition, it suffices to

show that ∀ε > 0, ∃δ > 0 such that if 0 < |x − 0| < δ then |f (x) − 0| < ε. Observe
that |f (x) − 0| = |f (x)| = |x|| sin x1 | ≤ |x|. Thus, ∀ε > 0, we can simply let δ = ε.
Then 0 < |x| < δ implies that |f (x)| ≤ |x| < δ = ε.

Example 2.1. There are several typical types of discontinuities.

(i) limx→p+ f (x) or limx→p− f (x) do not exist:


sin 1 , if x ̸= 0


x
f (x) = at x = 0.

0, if x = 0.


Here we show that limx→0+ f (x) does not exists.

Proof. We want to construct two decreasing sequence in A\{0}, (xn ) and (x′n ) such
that xn → 0, x′n → 0 but limn→∞ f (xn ) ̸= limn→∞ f (x′n ). Let xn = 1
2nπ+ π2
and
x′n = 1
2nπ− π2
for n ∈ N+ . Then for any n ∈ N+ , f (xn ) = 1 and f (x′n ) = −1. Thus
limn→∞ f (xn ) = 1 ̸= −1 = limn→∞ f (x′n ).

(ii) limx→p+ f (x) and limx→p− f (x) both exist but are distinct.

8
 (iii) limx→p+ f (x) = limx→p− f (x) ̸= f (p).

Definition 2.7. A real function f : A ⊆ R is uniformly continuous if ∀ε > 0,

∃δ > 0 s.t. ∀x1 , x2 ∈ A, if |x1 − x2 | < δ then |f (x1 ) − f (x2 )| < ε.

Exercise 2.2. Show that if a real function f : A ⊆ R is uniformly continuous,

then it is continuous. Can you find a function that is continuous but not uniformly
continuous?

2.1 Intermediate Value Theorems and Extreme Value The-

orem

In this section, we introduce some useful theorems for continuous functions.

Theorem 2.3 (Intermediate Value Theorem). Suppose f : [a, b] → R is continuous.

Then ∀y ∈ (min{f (a), f (b)}, max{f (a), f (b)}), ∃x ∈ (a, b) s.t. f (x) = y.

Remark 2.3. What if f is not continuous? How about f : [a, c] ∪ [d, b] → R? How
about f : (c, d) → R where c < a and b < d?

Theorem 2.4 (Extreme Value Theorem). Suppose f : [a, b] → R is continuous.

Then ∃x1 , x2 ∈ [a, b] s.t. f (x1 ) = supx∈[a,b] f (x) and f (x2 ) = inf x∈[a,b] f (x).

Remark 2.4. What if f is not continuous? How about f : [a, c] ∪ [d, b] → R? How
about f : (a, b) → R?