CH 4 - Lecture Notes Prof. Mukwembi

The Sandwich Theorem & Trigonometric Limits
The Intermediate Value Theorem
MATH2001A: BASIC ANALYSIS
LIMITS AND CONTINUOUS FUNCTIONS

Professor S Mukwembi, University of The Witwatersrand
May 6, 2024
LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 1 / 100

LIMITS
The idea of a limit pervades all of mathematics. We intuitively
studied limits in first year.
HERE:
We will learn precise definitions and proofs.
REMARK
In the previous topics, we studied limits of sequences. Here,
we:
study limits of functions, which provide a foundation to
studying continuous functions.
NB
The functions have domains and ranges that are subsets of R.
Recall: Consider a function f defined near the point x = a,

but not necessarily at a itself.
In first year, we investigated the behaviour of f for values x in
the domain of f that are close to a.
Remark: Given a point x = a points that are close to, or
neighbourly to a, are important.
Question:
How do we precisely describe points that are close to a point
a?

Definition 1
Let a be a real number, i.e., a ∈ R.
An interval of the form (c, d) is called a neighbourhood
of a if c < a < d.
Let (c, d) be a neighbourhood of a. Then the set
(c, d)\{a}, i.e., (c, a) ∪ (a, d) is called a deleted
neighbourhood of a.

Example 1
The interval (1, 3) is a neighbourhood of 2.
The set (1, 2) ∪ (2, 3) is a deleted neighbourhood of 2.
The set (−1, 2) is not a neighbourhood of −9.
For ϵ > 0, the interval (a − ϵ, a + ϵ) is a neighbourhood
of a. We will sometimes call (a − ϵ, a + ϵ) the
ϵ-neighbourhood of a.
For ϵ > 0, the interval (a − 2ϵ , a + 5ϵ ) is a neighbourhood
of a.

Recall...
In first year, we looked at the intuitive meaning of:
As “x tends to a”, “f (x) tends to L”.

We now provide the precise definition of that:

Definition 2
Let f be a real function, a, L ∈ R. Assume that the domain
of f contains a deleted neighbourhood of a, i.e., f (x) is
defined for all x in a deleted neighbourhood of a.
Then we say that f (x) tends to L as x tends to a if:
For every ϵ > 0, we can find a δ > 0 such that whenever x is
in the deleted δ-neighbourhood of a, then f (x) is in the
ϵ-neighbourhood of L.
That is: ∀ϵ > 0, ∃δ > 0 such that
0 < |x − a| < δ =⇒ |f (x) − L| < ϵ.

Note:
If f (x) −→ L as x −→ a, we write
lim f (x) = L.
x−→a
The number L is called a limit of the function f at a.

We will show that if a limit exists, then it is unique.
We never use f (a) for L as these two may be different.
The number δ depends on ϵ, and also on a, if we vary a.

Example 2
Prove, from first principles, that x 2 −→ 4 as x −→ 2, i.e., if
f (x) = x 2 , then
lim f (x) = 4.
x−→2
Solution: Let ϵ > 0 be given. We need to prove that we can

find a δ > 0 such that:
0 < |x − a| < δ =⇒ |f (x) − L| < ϵ,
i.e.,
0 < |x − 2| < δ =⇒ |x 2 − 4| < ϵ.
For a suitable δ, we want to use the premise, 0 < |x − 2| < δ,
to get the conclusion, |x 2 − 4| < ϵ. We therefore need to
study the relationship between |x 2 − 4| and |x − 2|. Now,
|x 2 − 4| = |(x − 2)||(x + 2)|

= |(x − 2)||(x − 2) + 4|
≤ |x − 2|(|x − 2| + 4).
Recall that we are required to find δ > 0 such that whenever

0 < |x − 2| < δ, then |x 2 − 4| < ϵ. From the calculations
above, if the premise holds, then
|x 2 − 4| ≤ |x − 2|(|x − 2| + 4) < δ(δ + 4).
We therefore need to choose δ, so that δ(δ + 4) ≤ ϵ. Solving

≤ ϵ, we see that δ can be any positive
the inequality, δ(δ + 4) √
real number up to δ = 4 + ϵ − 2.

√
Hence given ϵ > 0, set δ = 4 + ϵ − 2. In that case, we can
conclude that whenever
√
0 < |x − 2| < δ = 4 + ϵ − 2,
then
√ √
|x 2 − 4| < δ(δ + 4) ≤ [ 4 + ϵ − 2][ 4 + ϵ − 2 + 4] = ϵ,
as desired.

We now wish to prove that the limit of a function at a point is

unique if it exists. First recall:
Lemma 1
Let L and M ∈ R be such that |L − M| < ϵ for all ϵ > 0.
Then L = M.
Theorem 1
If f (x) −→ L as x −→ a, then L is unique.
Proof: Suppose that f (x) −→ L and f (x) −→ M as x −→ a.

We must prove that L = M. We aim to use Lemma 1.

For this, let ϵ > 0 be given. We prove that |L − M| < ϵ so

that from Lemma 1, we get L = M. Now:
Since f (x) −→ L and ϵ > 0 (and so is 2ϵ ) is given, there
exists a δ1 > 0 such
ϵ
0 < |x − a| < δ1 =⇒ |f (x) − L| < .
2
Similarly, since f (x) −→ M and ϵ > 0 (and so is 2ϵ ) is
given, there exists a δ2 > 0 such
ϵ
0 < |x − a| < δ2 =⇒ |f (x) − M| < .
2

It follows that for δ = min{δ1 , δ2 } and 0 < |x − a| < δ, we

have
|L − M| = |(L − f (x)) − (f (x) − M)|

≤ |f (x) − L| + |f (x) − M|
ϵ ϵ
< +
2 2
= ϵ.
We conclude by Lemma 1 that L = M, as desired.

In first year, we learnt the idea of one sided limits. We now
give the precise definitions.

Definition 3
of f contains an interval (a, d) with d > a, i.e., f (x) is defined
for all x in (a, d).
Then “f (x) −→ L as x −→ a+ ” is defined to mean:
∀ϵ > 0, ∃δ > 0
such that
a < x < a + δ =⇒ |f (x) − L| < ϵ.
If f (x) −→ L as x −→ a+ , we write
lim f (x) = L.
x−→a+

Definition 4
of f contains an interval (c, a) with c < a, i.e., f (x) is defined
for all x in (c, a).
Then “f (x) −→ L as x −→ a− ” is defined to mean:
∀ϵ > 0, ∃δ > 0
such that
a − δ < x < a =⇒ |f (x) − L| < ϵ.
If f (x) −→ L as x −→ a− , we write
lim f (x) = L.
x−→a−

Theorem 2
Assume that f (x) is defined in a deleted neighbourhood of a.
Then
lim f (x) = L ⇐⇒ lim f (x) = L = lim + f (x).

x−→a x−→a− x−→a
Proof: (=⇒) Assume that lim f (x) = L. We show that

x−→a
lim f (x) = L = lim + f (x).

x−→a− x−→a
Let ϵ > 0 be given. Since lim f (x) = L, ∃δ > 0 such that

x−→a
0 < |x − a| < δ =⇒ |f (x) − L| < ϵ,
i.e.,
∃δ > 0 such that
x ∈ (a − δ, a) ∪ (a, a + δ) =⇒ |f (x) − L| < ϵ,
i.e.,
∃δ > 0 such that
x ∈ (a−δ, a) =⇒ |f (x)−L| < ϵ and x ∈ (a, a+δ) =⇒ |f (x)−L| < ϵ.
We conclude that
lim f (x) = L and lim + f (x) = L,

as desired.

(⇐=) Conversely, assume that
lim f (x) = L = lim + f (x).

We show that lim f (x) = L. For this, let ϵ > 0 be given.

x−→a
Since lim − f (x) = L, there exists a δ1 such that
x−→a
x ∈ (a − δ1 , a) =⇒ |f (x) − L| < ϵ.
Similarly, since lim + f (x) = L, there exists a δ2 such that

x−→a
x ∈ (a, a + δ2 ) =⇒ |f (x) − L| < ϵ.

Set δ = min{δ1 , δ2 }. Then
x ∈ (a − δ, a + δ)\{a} =⇒ x ∈ (a − δ1 , a) or x ∈ (a, a + δ2 )
=⇒ |f (x) − L| < ϵ.
Therefore,
lim f (x) = L,
x−→a
as desired.

Example 3
x
Let f (x) = for x ∈ R\{0}.
|x|
1 Prove, from first principles, that lim + f (x) = 1.
x−→0
2 Prove, from first principles, that lim − f (x) = −1.
x−→0
3 Does f (x) have a limit at x = 0?
Proof: 1.) Let ϵ > 0 be given. We find a δ > 0 such that
a < x < a + δ =⇒ |f (x) − L| < ϵ,
i.e.,
0 < x < δ =⇒ |f (x) − 1| < ϵ.

Now for x > 0,

x
|f (x) − 1| = − 1 = 0,
|x|
and setting δ to be any positive number yields that
0 < x < δ =⇒ |f (x) − 1| = 0 < ϵ,
as desired.
2.) Exercise.
3.) Since
lim f (x) ̸= lim + f (x),
x−→0− x−→0
the limit does not exist by Theorem 2.

Remark:
When studying x −→ a, the deleted δ-neighbourhood,
(a − δ, a + δ)\{a}, of a is considered.
Question:
What do we consider when we study x −→ ∞?
A: We use (K , ∞), i.e., x > K . Formally,
Definition 5
Let f be a function defined on a set containing an interval of
the form (c, ∞). Then “f (x) −→ L as x −→ ∞” is defined to
mean:
∀ϵ > 0 ∃K > 0 such that
x > K =⇒ |f (x) − L| < ϵ.
If f (x) −→ L as x −→ ∞, then we write lim f (x) = L.
x−→∞
Example 4
1
Let f (x) = . Find
x
lim f (x)
x−→∞
and prove your result.

Solution: From first year,
1
lim f (x) = lim
= 0.
x−→∞ x x−→∞
We now justify. For this, let ϵ > 0 be given. We find K > 0

such that
x > K =⇒ |f (x) − L| < ϵ,
i.e.,
1
x > K =⇒ − 0 < ϵ.
x
Note that
1 1 1
− 0 < ϵ ⇐⇒ < ϵ ⇐⇒ |x| > .
x |x| ϵ
Hence, set K = 1ϵ . Thus if x > K = 1ϵ , then 1

x
< ϵ, and so
1 1 1
−0 = = < ϵ,
x |x| x
as required.

Definition 6
Let f be a function defined on a set containing an interval of
the form (−∞, c). Then “f (x) −→ L as x −→ −∞” is
defined to mean:
∀ϵ > 0 ∃K < 0 such that
x < K =⇒ |f (x) − L| < ϵ.

If f (x) −→ L as x −→ −∞, then we write lim f (x) = L.
x−→−∞

Example 5
1
Let f (x) = . Find
x
lim f (x)
x−→−∞
and prove your result.
Solution: Exercise.

We now turn to infinite limits. They are treated similarly:

Definition 7
Let f be a function whose domain includes a deleted
neighbourhood of the number a.
Then “f (x) −→ ∞ as x −→ a” is defined to mean:
∀K > 0 ∃δ > 0 such that
0 < |x − a| < δ =⇒ f (x) > K .

If f (x) −→ ∞ as x −→ a, then we write lim f (x) = ∞.
x−→a

Example 6
2
Prove that −→ ∞ as x −→ 0.
x2
Solution: Let K > 0 be given. We need to find a δ > 0 such
that
2
0 < |x − 0| < δ =⇒ 2 > K .
r x
2
Now set δ = . Then
K
r
2 2 2
0 < |x − 0| < δ = =⇒ |x|2 < =⇒ 2 > K .
K K x
Thus,
2
lim 2 = ∞,
x−→0 x
as required.
Definition 8
Let f be a function whose domain includes a deleted
neighbourhood of the number a.
Then “f (x) −→ −∞ as x −→ a” is defined to mean:
∀K < 0 ∃δ > 0 such that
0 < |x − a| < δ =⇒ f (x) < K .

If f (x) −→ −∞ as x −→ a, then we write
lim f (x) = −∞.

x−→a

Definition 9
Let f be a real function defined on a set containing an interval
of the form (c, ∞).
Then “f (x) −→ ∞ as x −→ ∞” is defined to mean:
∀A > 0 ∃K > 0 such that
x > K =⇒ f (x) > A.

If f (x) −→ ∞ as x −→ ∞, then we write
lim f (x) = ∞.
x−→∞

Remark:
Similar definitions hold with all other combinations of limits
involving ∞ and −∞, mutatis mutandis for one sided limits.

Example 7
1
Let f (x) = . Find lim + f (x), lim − f (x), lim f (x).
x x−→0 x−→0 x−→0
Solution: From first year, lim + f (x) = ∞. We prove this. Let

x−→0
K > 0 be given. We find δ > 0 such that
0 < x < δ =⇒ f (x) > K .

1
Set δ = K
. Then
1 1
0<x <δ= =⇒ K < = f (x).
K x
Hence, lim + f (x) = ∞.
x−→0

From first year, lim − f (x) = −∞. We prove this. Let K < 0
x−→0
be given. We find δ > 0 such that
−δ < x < 0 =⇒ f (x) < K .
Set δ = − K1 . Then
1 1
= −δ < x < 0 =⇒ = f (x) < K .
K x
Hence, lim − f (x) = −∞.
x−→0

Since
lim f (x) ̸= lim − f (x),
x−→0+ x−→0
lim f (x) does not exist.

x−→0
Question:
How do we calculate limits of very complicated functions?
A: We use limit laws.

We now state and prove some of the limit laws.
Theorem 3
Let a, c ∈ R and assume that the real functions f and g are
defined in a deleted neighbourhood of a and that
lim f (x) = L ∈ R and lim g (x) = M ∈ R both exist. Then:
x−→a x−→a

1 lim c = c.
x−→a
2 lim [f (x) + g (x)] = lim f (x) + lim g (x) = L + M.
x−→a x−→a x−→a
3 lim [f (x) − g (x)] = lim f (x) − lim g (x) = L − M.
4 lim [cf (x)] = c lim f (x) = cL.
x−→a x−→a
5 lim [f (x)g (x)] = [ lim f (x)][ lim g (x)] = LM.
f (x) lim f (x) L

6 If M ̸= 0, lim = x−→a = .
x−→a g (x) lim g (x) M
x−→a
f (x)
7 If L ̸= 0 and M = 0, lim does not exist.
x−→a g (x)
8 If n ∈ N, lim [f (x)n ] = [ lim f (x)]n = Ln .
x−→a x−→a
9 lim x = a.
x−→a
10 If n ∈ N, lim x n = an .
x−→a
√ √
11. If n ∈ N, lim n x = n a. If n is even, we assume that
x−→a
a > 0. p q √n
12. If n ∈ N, lim n f (x) = n lim f (x) = L.
x−→a x−→a
13. If lim |f (x)| = 0, then lim f (x) = 0.
x−→a x−→a
Proof: We will only prove (2), (5) and (6). The rest are left as
exercises.
(2) Let ϵ > 0 (and so is 2ϵ ) be given. Since lim f (x) = L,
x−→a
there exists δ1 > 0 such that
ϵ
0 < |x − a| < δ1 =⇒ |f (x) − L| < . (1)
2

Similarly, since lim g (x) = M, there exists δ2 > 0 such that

x−→a
ϵ
0 < |x − a| < δ2 =⇒ |g (x) − M| < . (2)
2
We set δ = min{δ1 , δ2 }. Then for
0 < |x − a| < δ,
we have 0 < |x − a| < δ1 and 0 < |x − a| < δ2 . It follows that
|[f (x) + g (x)] − [L + M]| = |[f (x) − L] + [g (x) − M]|

≤ |f (x) − L| + |g (x) − M|
ϵ ϵ
≤ + (by (1) and (2))
2 2
= ϵ.
Hence, f (x) + g (x) converges to L + M as x −→ a.

(5). Assume that lim f (x) = L and lim g (x) = M. We

x−→a x−→a
prove that
lim [f (x)g (x)] = [ lim f (x)][ lim g (x)] = LM.

We consider two cases separately.

Case A: L = M = 0.
Let ϵ > 0 be given.
Since lim f (x) = 0, for the number 1, there exists δ1 > 0
x−→a
such that
0 < |x − a| < δ1 =⇒ |f (x) − 0| < 1. (3)

Similarly, since lim g (x) = 0, there exists δ2 > 0 such that

x−→a
0 < |x − a| < δ2 =⇒ |g (x) − 0| < ϵ. (4)
We set δ = min{δ1 , δ2 }. Then for
0 < |x − a| < δ,
we have 0 < |x − a| < δ1 and 0 < |x − a| < δ2 . It follows that
|[f (x)g (x)] − [0 · 0]| = |f (x)g (x)|

≤ |f (x)||g (x)|
< 1 · ϵ (by (3) and (4))
= ϵ.
Hence, f (x)g (x) converges to LM = 0 as x −→ a, and Case

A is proven.
Case B: Now assume that L and M are arbitrary. Then, we

will use the equation:
f (x)g (x) = [f (x) − L][g (x) − M] + L[g (x) − M] + f (x)M. (5)
Note that
lim [f (x) − L] = lim f (x) − lim L (by (3))

= L − L (by (1))
= 0. (6)

Similarly,
lim [g (x) − M] = lim g (x) − lim M (by (3))

= M − M (by (1))
= 0. (7)

From (5), we get
lim f (x)g (x)

x−→a
= lim ([f (x) − L][g (x) − M] + L[g (x) − M] + f (x)M])
x−→a
= lim [f (x) − L][g (x) − M] + lim L[g (x) − M] +
x−→a x−→a
lim f (x)M (by (2))
x−→a
= lim [f (x) − L] lim [g (x) − M] +
x−→a x−→a
lim L lim [g (x) − M] +
x−→a x−→a
lim f (x) lim M (by (5))
x−→a x−→a
= 0 · 0 + L · 0 + LM (by (6) and (7))
= LM,
as desired.
(6). Assume that lim f (x) = L and lim g (x) = M ̸= 0. We

x−→a x−→a
prove that
f (x) lim f (x) L
lim = x−→a = .
x−→a g (x) lim g (x) M
x−→a
We consider two cases separately.

Case A: f (x) = 1. We prove that
1 1
lim = .
x−→a g (x) M

|M|
Since lim g (x) = M, for 2
> 0, there exists δ1 > 0 such
x−→a
that
|M|
0 < |x − a| < δ1 =⇒ |g (x) − M| < . (8)
2
For this 0 < |x − a| < δ1 , we have
|M| |M|
|g (x)| = |M+[g (x)−M]| ≥ |M|−|g (x)−M| ≥ |M|− = ,
2 2
1 2
and so, |g (x)|
≤ |M|
. Therefore,
1 1 |M − g (x)| 2|M − g (x)| 2|M − g (x)|

− = ≤ = .
g (x) M |g (x)M| |M||M| |M|2
|M|2
Since lim g (x) = M, for 2
ϵ > 0, there exists δ2 > 0 such
x−→a
that
|M|2
0 < |x − a| < δ2 =⇒ |g (x) − M| < ϵ. (9)
2

Now set δ = min{δ1 , δ2 }. It follows for 0 < |x − a| < δ that
1 1 |M − g (x)| |M|2 1
− ≤2 < 2 ϵ = ϵ.
g (x) M |M|2 2 |M|2
Hence,
1 1
lim = .
x−→a g (x) M

Case B: f (x) is arbitrary. Then
f (x) 1
lim = lim f (x) ·
x−→a g (x) x−→a g (x)
1
= lim f (x) · lim (by (5))
x−→a x−→a g (x)
1
= L· (by Case A)
M
L
= ,
M
as desired.

Question:
How do we find limits of polynomials or rational functions?
Recall: Polynomials are of the form
f (x) = bn x n + bn−1 x n−1 + · · · + b2 x 2 + b1 x + b0
where bi ∈ R, i = 0, 1, 2, . . . , n and n, any non-negative

integer, whereas rational functions are of the form
p(x)
f (x) = ,
q(x)
where p(x) and q(x) are polynomials.

By Limit Laws, Theorem 3, we have

lim [f (x) + g (x)] = lim f (x) + lim g (x) = L + M.
lim [cf (x)] = c lim f (x) = cL.
x−→a x−→a
lim x = a.
x−→a
If n ∈ N, lim x n = an .
x−→a
Corollary 1
Let f be a polynomial or rational function. Then
lim f (x) = f (a).

x−→a

Proof: Let f (x) = bn x n + bn−1 x n−1 + · · · + b2 x 2 + b1 x + b0 .

Then
lim f (x) = lim [bn x n + bn−1 x n−1 + · · · + b2 x 2 + b1 x + b0 ]

x−→a x−→a
= lim [bn x n ] + lim [bn−1 x n−1 ] + · · · + lim [b2 x 2 ]
+ lim [b1 x] + lim [b0 ]
x−→a x−→a
= bn lim x n + bn−1 lim x n−1 + · · · + b2 lim x 2
+b1 lim x + lim b0
x−→a x−→a
n n−1
= bn a + bn−1 a + · · · + b2 a2 + b1 a + b0
= f (a),
as claimed.
Rational functions: Exercise.
Corollary 2
All the limit rules in Theorem 3 remain true if x −→ a is
replaced by any of the following:
x −→ a+ , x −→ a− , x −→ ∞, x −→ −∞.
Proof:
For x −→ a+ and x −→ a− , replace 0 < |x − a| < δ with
0 < x − a < δ and −δ < x − a < 0, respectively, in the
proof of each statement.
For x −→ ∞ and x −→ −∞, the proofs are analogous to
that of sequences.
Question:
What if f has infinite limits?

Theorem 4
Assume that lim f (x) = ∞, lim g (x) = ∞ and
x−→a x−→a
lim h(x) = c ∈ R. Then:
x−→a
1 f (x) + g (x) −→ ∞ as x −→ a,
2 f (x) + h(x) −→ ∞ as x −→ a,
3 f (x)g (x) −→ ∞ as x −→ a,

∞ if c > 0
4 f (x)h(x) −→ as x −→ a.
−∞ if c < 0

We prove (3), i.e., f (x)g (x) −→ ∞ as x −→ a, and leave the

rest as exercises.
Assume that f (x) −→ ∞ and g (x) −→ ∞ as x −→ a.
Let A > 0 be given. Since f (x) −→ ∞ as x −→ a, let δ1 > 0
be such that
0 < |x − a| < δ1 =⇒ f (x) > 1.
Similarly, since g (x) −→ ∞ as x −→ a, let δ2 > 0 be such

that
0 < |x − a| < δ2 =⇒ g (x) > A.
Let δ = min{δ1 , δ2 }. It follows that for all 0 < |x − a| < δ, we
have
f (x)g (x) > 1 · A = A,
as desired.
Question:
How do we compute the limits of functions that are bounded
(squeezed, trapped pinched or sandwiched) between two other
functions that have the same limit?
E.g.,
A: Two officers (and a drunk) theorem - if two police

officers are escorting a drunk prisoner between them, and both
officers go to a cell, then the prisoner must also end up in the
cell. LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 54 / 100
We now state and prove the Sandwich Theorem:

Theorem 5
Let a ∈ R ∪ {∞, −∞} and assume that f , g and h are real
functions defined in a deleted neighbourhood of a. If
f (x) ≤ g (x) ≤ h(x) for x in a deleted neighbourhood of a and
lim f (x) = L = lim h(x), then lim g (x) = L.

Proof: First assume that a, L ∈ R.

Since lim f (x) = L, let δ1 > 0 be such that
x−→a
0 < |x − a| < δ1 =⇒ |f (x) − L| < ϵ.

Similarly, since lim h(x) = L, let δ2 > 0 be such that
x−→a
0 < |x − a| < δ2 =⇒ |h(x) − L| < ϵ.

Setting δ = min{δ1 , δ2 }, for
0 < |x −a| < δ =⇒ L−ϵ < f (x) < L+ϵ, L−ϵ < h(x) < L+ϵ,
i.e.,
0 < |x − a| < δ =⇒ L − ϵ < f (x) ≤ g (x) ≤ h(x) < L + ϵ, .
Hence,
0 < |x − a| < δ =⇒ |g (x) − L| < ϵ,
so that g (x) −→ L as x −→ a, as claimed.
If a, L ∈ {∞, −∞}, exercise.
Other names: The Sandwich Theorem is also called the
Squeeze Theorem.

Theorem 6
Let f be defined on an interval (a, b), where a = −∞ and
b = ∞ are allowed. If a = −∞, we write a+ = −∞, and if
b = ∞, we write b − = ∞.
1 If f is increasing, then
lim f (x) = sup{f (x) : x ∈ (a, b)} and

x−→b −
lim f (x) = inf{f (x) : x ∈ (a, b)}.

x−→a+
2 If f is decreasing, then
lim f (x) = inf{f (x) : x ∈ (a, b)} and

x−→b −
lim f (x) = sup{f (x) : x ∈ (a, b)}.

x−→a+

Proof: We prove (1) with b ∈ R; other cases are proved

similarly and are left as exercises.
Let L = sup{f (x) : x ∈ (a, b)}.
Case A: L ∈ R.
By the characterisation of supremum, there is a c ∈ (a, b)
such that L − ϵ < f (c).
Now set δ = b − c > 0.
Thus, if b − δ < x < b, i.e., x ∈ (c, b), then c < x, and since
f is increasing, f (c) ≤ f (x) =⇒ f (x) ≤ L for all
x ∈ (c, b) ⊂ (a, b), so that
L − ϵ < f (c) ≤ f (x) ≤ L < L + ϵ
for all x. Hence, by definition, we conclude that f (x) −→ L as

x −→ b − .
Case B: L = ∞.
Then {f (x) : x ∈ (a, b)} is not bounded above.
Therefore, for each A ∈ R, there is c ∈ (a, b) such that
f (c) > A.
Since f is increasing, we must have
x ∈ (c, b) =⇒ A < f (c) ≤ f (x).
Therefore, f (x) −→ ∞ as x −→ b − .

Remark:
Some graphs of functions have breaks at certain points.
Question:
How do we precisely describe the property that at a point a,
the graph of a function f has no breaks?
A: We say that f is continuous at point a and define it

precisely as follows:

Definition 10
Let f be a real function, a ∈ R and assume that the domain of
f contains a neigbourhood of a, i.e., f (x) is defined for all x in
a neighbourhood of a. We say that f is continuous at a if
for all ϵ > 0, there exists δ > 0
such that
|x − a| < δ =⇒ |f (x) − f (a)| < ϵ,
i.e.,
such that
x ∈ (a − δ, a + δ) =⇒ f (x) ∈ (f (a) − ϵ, f (a) + ϵ).

Recall:
We say that f (x) tends to L as x tends to a if:
such that
0 < |x − a| < δ =⇒ |f (x) − L| < ϵ.
Remark:
The definitions of a function being continuous at a point, and
that of a limit of the function at that point are similar, but
different since f (a) may be different from L.

We can use limits to describe continuity:

Theorem 7
Let f be a real function, a ∈ R and assume that the domain of
f contains a neigbourhood of a, i.e., f (x) is defined for all x in
a neighbourhood of a. Then f is continuous at a point a if
and only if the following three conditions hold:
1. f (a) is defined, i.e., a is in the domain of f ,
2. lim f (x) exists, i.e., lim − f (x) = lim + f (x), and
3. f (a) = lim f (x).
x−→a
Proof: (=⇒) Assume that f is continuous at a. Condition 1.,

i.e., f (a) exists, holds from the hypothesis.
We now prove that 2. and 3. are true, i.e.,
L = f (a) = lim f (x).
x−→a
For this, let ϵ > 0 be given.
Since f is continuous at a, let δ > 0 be such that
|x − a| < δ =⇒ |f (x) − f (a)| < ϵ.
For this δ,
0 < |x − a| < δ =⇒ |f (x) − L| < ϵ,
and so
L = f (a) = lim f (x), as claimed.
x−→a

(⇐=) Conversely, assume that the three conditions hold. We

prove that f is continuous at a.
Since lim f (x) = f (a), there exists a δ > 0 such that
x−→a
0 < |x − a| < δ =⇒ |f (x) − f (a)| < ϵ.
Note that for this δ,
|x − a| < δ =⇒ |f (x) − f (a)| < ϵ,
because if x = a, then
|f (x) − f (a)| = |f (a) − f (a)| = 0 < ϵ.

Example 8
Say whether f is continuous at 2 or not, where
2
x if x ̸= 2
f (x) =
2 if x = 2.
Solution: Note that
lim f (x) = 4,
x−→2
exists, but it is not equal to f (2) = 2; hence f is NOT

continuous at 2.

Example 9
Say whether
sin x
f (x) =
x
is continuous at 0 or not.
Solution: Note that
lim f (x) = 1,
x−→0
exists, but f is not defined at 0; hence f is NOT continuous at

0.

Example 10
Say whether f is continuous at 0 or not, where
 sin x
 x if x ̸= 0
f (x) =
1 if x = 0.

Solution: Note that
lim f (x) = 1,
x−→0
exists, and f (0) = 1; hence f is continuous at 0.

Theorem 8
If f and g are continuous at a and if c ∈ R, then
1 the sum f + g ,
2 the difference f − g ,
3 the product fg ,
4 the quotient gf , where g (a) ̸= 0, and
5 the scalar multiple cf
are functions that are also continuous at a.

Proof: The theorem follows from limit laws and from Theorem
7, e.g.,
3. Assume that f and g are continuous at a. Then by
Theorem 7,
lim f (x) = f (a) and lim g (x) = g (a).

x→a x→a
It follows that
lim (fg )(x) = lim f (x)g (x)

x→a x→a
= lim f (x) lim g (x)
x→a x→a
= f (a)g (a)
= (fg )(a),
so that by Theorem 7, fg is continuous at a.

Recall:
The composite function g ◦ f of functions g and f is defined
by (g ◦ f )(x) = g (f (x)).
Theorem 9
If f is continuous at a and g is continuous at f (a), then g ◦ f
is continuous at a.
Proof: Let ϵ > 0 be given. Since g is continuous at f (a),
there is a η > 0 such that
|y − f (a)| < η =⇒ |g (y ) − g (f (a))| < ϵ. (10)
For this η, since f is continuous at a, there is a δ > 0 such that

|x − a| < δ =⇒ |f (x) − f (a)| < η. (11)

For this δ, we have:
|x − a|< δ
=⇒
|f (x) − f (a)| < η (by 11)
=⇒
|y − f (a)| < η (setting y = f (x))
=⇒ |g (y ) − g (f (a))| < ϵ (by 10).
That is,
|x − a| < δ =⇒ |(g ◦ f )(x) = (g ◦ f )(a)| < ϵ;
hence g ◦ f is continuous at a.
We also have right and left continuity:

Definition 11
A function f is continuous from the right at a if
lim f (x) = f (a).

x−→a+
A function f is continuous from the left at a if
lim f (x) = f (a).

x−→a−

Example 11
Let
|x|+x

 2x
if x ̸= 0
f (x) =
0 if x = 0.

Determine the right and left continuity of f at x = 0.

Solution: Note first that f (0) = 0. Now for right continuity,
|x| + x
lim + f (x) = lim +
x−→0 x−→0 2x
x +x
= lim +
x−→0 2x
= 1
̸ = f (0);
and so f is not continuous from the right at 0.
Now for left continuity,
|x| + x
lim − f (x) = lim −
x−→0 x−→0 2x
−x + x
= lim −
x−→0 2x
= 0
= f (0);
and so f is continuous from the left at 0.

Remark:
It can be easily shown that f is continuous at a point a if
and only if f is continuous from the right and continuous
from the left at a.
If a ∈ Dom(f ) and if there is ϵ > 0 such that
Dom(f ) ∩ (a − ϵ, a + ϵ) = (a − ϵ, a], then we say that f is
continuous at a if lim − f (x) = f (a).
x−→a
If a ∈ Dom(f ) and if there is ϵ > 0 such that
Dom(f ) ∩ (a − ϵ, a + ϵ) = [a, a + ϵ), then we say that f is
continuous at a if lim + f (x) = f (a).
x−→a
The last two conventions will be learnt further in General
Topology.

Lemma 2
If f (x) −→ b as x −→ a and g is continuous at b, then
g (f (x)) −→ g (b) as x −→ a, which can be written as,
lim g (f (x)) = g ( lim f (x)).

x−→a x−→a
The above also holds if x −→ a+ or a− .

Proof: The function

 f (x) if x ∈ Dom(f ), x ̸= a
f (x) =
b if x = a,

is continuous both from the right and from the left at a. The
result follows by applying Theorem 9.
Definition 12
A function is continuous on a set X ⊂ R if f is continuous at
each x ∈ X . Here continuity is understood in the sense of the
above remark with X = Dom(f ). A function is said to be
continuous if it is continuous on its domain.

Example 12
√
Show that f (x) = x 2 − 4 is continuous.
Solution:
The domain of f is
Dom(f ) = {x ∈ R : |x| ≥ 2} = (−∞, −2] ∪ [2, ∞).
By Theorem 8, the function g (x) = x 2 − 4 is continuous
on R, and by law of limits,
the square root is continuous at each positive number.
Therefore, the composite function f is continuous on
(−∞, −2) ∪ (2, ∞).
The proof on limit laws can be adapted to show that the
square root is continuous from the right at 0.
Then it follows that f is continuous from the right at 2
and continuous from the left at −2.
Theorem 10
The following functions are continuous on their domains.
1. Polynomials
p(x) = an x n + an−1 x n−1 + · · · + a2 x 2 + a1 x + a0 , ai ∈ R,
n ∈ N.
p(x)
2. Rational functions , p and q ̸= 0 polynomials.
q(x)
3. Sums, differences, products and quotients of continuous
functions.
4. Root functions.
5. The trigonometric functions sin x, cos x, tan x, cosec x,
sec x and cot x.
6. The exponential function, e x .
7. The absolute value function |x|.
Proof: 1., 2. and 3. easily follow from previous theorems on

limits and continuity, as does 7.
However, 7 can be easily proved directly:
For each ϵ > 0 let δ = ϵ. Then, for |x − a| < δ we have
| |x| − |a| | ≤ |x − a| < δ = ϵ.
The continuity of sin and cos follows from the sum of angles
formulae and from the limits proved in Calculus I (the proofs
used the Sandwich Theorem, which now has been proved).
The continuity of the other trigonometric functions then
follows from part 3.
Finally, 6. (Tutorial).

Theorem 11
Let a ∈ R and let f be a real function which is defined in a
neighbourhood of a. Then f is continuous at a if and only if
for each sequence (xn ) in dom(f ) with lim xn = a, the
n→∞
sequence f (xn ) satisfies lim f (xn ) = f (a).
n→∞

Proof: (⇒) Let (xn ) be a sequence in dom(f ) with

lim xn = a. We must show that lim f (xn ) = f (a). Hence let
n→∞ n→∞
ϵ > 0. Since f is continuous at a, there is δ > 0 such that
|x − a| < δ ⇒ |f (x) − f (a)| < ϵ. (1)
Since lim xn = a, there is K ∈ R such that for n > K ,

n→∞
|xn − a| < δ. But then, by (1), |f (xn ) − f (a)| < ϵ for n > K .

(⇐) (indirect proof) Assume that f is not continuous at a.

Then
∃ ϵ > 0 ∀ δ > 0 ∃ x ∈ dom(f ), |x−a| < δ and |f (x)−f (a)| ≥ ϵ.
In particular, for δ = n1 , n = 1, 2, . . . we find xn ∈ dom(f )

1
such that |xn − a| < and |f (xn ) − f (a)| ≥ ϵ. But then
n
lim xn = a, whereas (f (xn )) does not converge to f (a).
n→∞

Question:
Assume that f is a continuous function. Can its graph jump
from one side of a horizontal line y = k to the other without
intersecting the line at least once.
A:
Theorem 12 (Intermediate Value Theorem (IVT))
Suppose that f is continuous on the closed interval [a, b] with
f (a) ̸= f (b). Then for any number k between f (a) and f (b)
there exists a number c in the open interval (a, b) such that
f (c) = k.

Proof: Let
g (x) = f (x) − k, (x ∈ [a, b]).
Then g is continuous, and g (a) and g (b) have opposite signs:
g (a)g (b) < 0.
Let [a0 , b0 ] = [a, b] and use bisection to define intervals
[an , bn ] as follows: If [an , bn ] with g (an )g (bn ) < 0 has been
found, let d be the midpoint of the interval [an , bn ]. If
g (d) = 0, the result follows with c = d. If g (d) has the same
sign as g (bn ), then g (an ) and g (d) have opposite signs, and
putting an+1 = an , bn+1 = d, we have g (an+1 )g (bn+1 ) < 0.
Otherwise, if g (d) has the opposite sign to g (bn ), we put
an+1 = d, bn+1 = bn and get again g (an+1 )g (bn+1 ) < 0.

If this procedure does not stop, we obtain an increasing

sequence (an ) and a decreasing sequence (bn ), both of which
converge by a previous theorem. We observe that
1
bn = an + (bn−1 − an−1 ) = an + 2−n (b − a).
2
Then
c := lim bn = lim an + lim 2−n (b − a) = lim an .
n→∞ n→∞ n→∞ n→∞
Since a ≤ c ≤ b and g is continuous at c, it follows in view of

Theorem 11 that
g (c)2 = lim g (an ) lim g (bn ) = lim g (an )g (bn ) ≤ 0.
n→∞ n→∞ n→∞
Therefore g (c) = 0, which gives f (c) = k.

Since f (a) ̸= k and f (b) ̸= k, it follows that c ̸= a and c ̸= b,
so that a < c < b.
Note:
You have seen the definition of interval in first year and you
will recall that that the definition required several cases,
depending on whether the endpoints belong to the interval or
not and whether the interval is bounded (above, below),
namely (a, b), [a, b), (a, b], [a, b], (−∞, b), (−∞, b], (a, ∞),
[a, ∞), (−∞, ∞) where a, b ∈ R and a < b. However,
intervals can be characterized by one common property. For
this we need the following notion:
A subset S of R is called a singleton if the set S has exactly
one element.

Definition 13
1. A set S ⊂ R is called an interval if
(i) S ̸= ∅,
(ii) S is not a singleton,
(iii) if x, y ∈ S, x < y , then each z ∈ R with x < z < y
satisfies z ∈ S.
2. An interval of the form [a, b] with a < b is called a closed
bounded interval.

Note:
A subset S of R is an interval if and only if it contains at least
two elements and if all real numbers between any two
elements in S also belong to S.
Definition 14
For a function f : X → Y and A ⊂ X , the set
f (A) = {y ∈ Y : ∃ x ∈ A ∩ dom(f ), f (x) = y }

= {f (x) : x ∈ A ∩ dom(f )}
is called the image of A under f .

Corollary 3
Let I be an interval and let f be a continuous real function on
I . Then f (I ) is either an interval or a singleton.
Proof: Since I ̸= ∅, there is x ∈ I and so f (x) ∈ f (I ). Hence

f (I ) ̸= ∅. Hence we have to show that if f (I ) is not a
singleton, then it is an interval, that is, we must show that for
any x, y ∈ I with f (x) < f (y ) and any k ∈ R with
f (x) < k < f (y ) there is c ∈ I such that f (c) = k.
Indeed, from f (x) ̸= f (y ), we have x ̸= y . If x < y , then f is
continuous on [x, y ], and by the intermediate value theorem,
there is c ∈ (x, y ) with f (c) = k. A similar argument holds for
x > y.

Example 13
Let f (x) = x 2 . Then f ((−1, 2)) = [0, 4). Notice that
I = (−1, 2) is an open interval, while f (I ) is not.
Theorem 13
Let f be a real function which is continuous on [a, b], where
a < b. Then f is bounded on [a, b], i. e., f ([a, b]) is bounded.
Proof: Assume that f ([a, b]) is unbounded. Let d be the

midpoint of the interval [a, b]. Then at least one of the sets
f ([a, d]), f ([d, b]) would be unbounded, because otherwise
f ([a, b]) = f ([a, d]) ∪ f ([d, b]) would be bounded.

By induction, we find subintervals [an , bn ] of [a, b] such that

(an ) is increasing, (bn ) is decreasing, f ([an , bn ]) is unbounded,
and
1
bn = an + (bn−1 − an−1 ) = 2−n (b − a),
2
see the proof of Theorem 12, and we infer that both sequences
converge with
c := lim bn = lim an ∈ [a, b].

n→∞ n→∞
Since f is continuous at c, there is δ > 0 such that
f ((c − δ, c + δ) ∩ [a, b]) ⊂ (f (c) − 1, f (c) + 1).
Since c = lim bn , there is K ∈ N such that |bn − c| < δ for

n→∞
n > K.
Similarly, there is M ∈ N such that |an − c| < δ for n > M.

Let n > max{K , M}. Then
c − δ < an ≤ c ≤ bn < c + δ,
and
f ([an , bn ]) ⊂ f ((c − δ, c + δ) ∩ [a, b]) ⊂ (f (c) − 1, f (c) + 1)
would give the contradiction that f ([an , bn ]) would have to be

bounded as well as unbounded.

Theorem 14
A continuous function on a closed bounded interval achieves
its supremum and infimum.
Proof: Let [a, b] be a closed bounded interval and f be a

continuous function on [a, b]. We must show that there are
x1 , x2 ∈ [a, b] such that f (x1 ) = inf f ([a, b]) and
f (x2 ) = sup f ([a, b]).
We are going to show the latter; the proof of the first
statement is similar.

By Theorem 13, S = f ([a, b]) is bounded. Let M = sup S. By

proof of contradiction, assume that f (x) ̸= M for all
x ∈ [a, b]. Define
1
g (x) = , x ∈ [a, b].
M − f (x)
By Theorem 10, g is continuous on [a, b], and by Theorem 13,

g ([a, b]) is bounded. So there is K ∈ R such that
0 < g (x) ≤ K for all x ∈ [a, b]. Hence for all x ∈ [a, b]:
1 1 1
≤ = M − f (x) ⇒ f (x) ≤ M − ,
K g (x) K
1
so that the number M − < M would be an upper bound of
K
f ([a, b]). This contradicts the fact that M = sup f ([a, b]).
Corollary 4
If f is continuous on [a, b], a < b, then either f ([a, b]) is a
singleton or f ([a, b]) = [c, d] with c < d.
Proof: From Corollary 3 and Theorems 13 and 14 it follows

that f ([a, b]) is either a singleton or a bounded interval which
contains both its infimum and supremum. But such an interval
is of the form [c, d] with c < d.

Theorem 15
Let I be an interval and f : I → R be a strictly monotonic
continuous function. Then f (I ) is an interval, and the inverse
function f −1 : f (I ) → R is continuous.
Proof: By Corollary 3, f (I ) is an interval. Assume that f is

strictly increasing. Then also f −1 is strictly increasing. Let
b ∈ f (I ), i. e., b = f (a) for some a ∈ I . If a is not the left
endpoint of I , then b is not the left endpoint of f (I ), and for
y ∈ f (I ) with y < b = f (a) we have f −1 (y ) < f −1 (f (a)) = a.

Therefore, f −1 is bounded above and increasing on

f (I ) ∩ (−∞, b), and thus
α = lim− f −1 (y )
y →b
exists and α ≤ f −1 (b) = a, from a previous theorem. Let

x0 ∈ I ∩ (−∞, a). Then f (x0 ) < f (a) = b and hence
x0 = f −1 (f (x0 )) ≤ α ≤ a.
Since x0 and a belong to the interval I , also α ∈ I . Since f is
continuous, it follows from a previous lemma that

f (α) = f lim− f (y ) = lim− f (f −1 (y )) = lim− y = b
−1
y →b y →b y →b
which gives
lim f −1 (y ) = α = f −1 (b).
y →b −

Therefore f −1 is continuous from the left. Similarly, one can

show that f −1 is continuous from the right. Therefore f −1 is
continuous. The case f strictly decreasing is similar.

CH 4 - Lecture Notes Prof. Mukwembi

Uploaded by

Copyright:

Available Formats

CH 4 - Lecture Notes Prof. Mukwembi

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CH 4 - Lecture Notes Prof. Mukwembi

Uploaded by

Copyright:

Available Formats

The Sandwich Theorem & Trigonometric Limits

The Intermediate Value Theorem

MATH2001A: BASIC ANALYSIS

LIMITS AND CONTINUOUS FUNCTIONS

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 1 / 100

Recall: Consider a function f defined near the point x = a,

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 3 / 100

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 4 / 100

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 5 / 100

In first year, we looked at the intuitive meaning of:

As “x tends to a”, “f (x) tends to L”.

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 6 / 100

0 < |x − a| < δ =⇒ |f (x) − L| < ϵ.

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 7 / 100

The number L is called a limit of the function f at a.

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 8 / 100

Solution: Let ϵ > 0 be given. We need to prove that we can

0 < |x − a| < δ =⇒ |f (x) − L| < ϵ,

|x 2 − 4| = |(x − 2)||(x + 2)|

Recall that we are required to find δ > 0 such that whenever

|x 2 − 4| ≤ |x − 2|(|x − 2| + 4) < δ(δ + 4).

We therefore need to choose δ, so that δ(δ + 4) ≤ ϵ. Solving

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 10 / 100

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 11 / 100

We now wish to prove that the limit of a function at a point is

Proof: Suppose that f (x) −→ L and f (x) −→ M as x −→ a.

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 12 / 100

For this, let ϵ > 0 be given. We prove that |L − M| < ϵ so

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 13 / 100

It follows that for δ = min{δ1 , δ2 } and 0 < |x − a| < δ, we

|L − M| = |(L − f (x)) − (f (x) − M)|

We conclude by Lemma 1 that L = M, as desired.

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 14 / 100

a < x < a + δ =⇒ |f (x) − L| < ϵ.

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 15 / 100

a − δ < x < a =⇒ |f (x) − L| < ϵ.

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 16 / 100

lim f (x) = L ⇐⇒ lim f (x) = L = lim + f (x).

Proof: (=⇒) Assume that lim f (x) = L. We show that

lim f (x) = L = lim + f (x).

Let ϵ > 0 be given. Since lim f (x) = L, ∃δ > 0 such that

0 < |x − a| < δ =⇒ |f (x) − L| < ϵ,

∃δ > 0 such that

x ∈ (a − δ, a) ∪ (a, a + δ) =⇒ |f (x) − L| < ϵ,

x ∈ (a−δ, a) =⇒ |f (x)−L| < ϵ and x ∈ (a, a+δ) =⇒ |f (x)−L| < ϵ.

lim f (x) = L and lim + f (x) = L,

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 18 / 100

(⇐=) Conversely, assume that

lim f (x) = L = lim + f (x).

We show that lim f (x) = L. For this, let ϵ > 0 be given.

Similarly, since lim + f (x) = L, there exists a δ2 such that

x ∈ (a, a + δ2 ) =⇒ |f (x) − L| < ϵ.

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 19 / 100

Set δ = min{δ1 , δ2 }. Then

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 20 / 100

Proof: 1.) Let ϵ > 0 be given. We find a δ > 0 such that

a < x < a + δ =⇒ |f (x) − L| < ϵ,

LIMITS AND CONTINUOUS FUNCTIONS MATH2001A: BASIC ANALYSIS 21 / 100

Now for x > 0,