Limit of Function

Yibin Ren
December 12, 2020

1 Definition of Function limit

1.1 The cases x → +∞
Suppose f (x) is defined on (a, +∞). Roughly speaking, the limit of f (x) as x → +∞ is the number where
the function will “tend to”, denoted by

lim f (x) (1.1)

x→+∞

For example,
1 1
lim = lim = 0, lim 1 = 1. (1.2)
x→+∞ x x→+∞ x2 x→+∞

There are many functions having no limit as x → +∞. For example, sin x is a periodic function with value
in [−1, 1] and it oscillates between −1 and 1 as x → +∞; thus, sin x has no limit as x → +∞. As we known,
∞ is not a number and thus a function has no limit if it goes to ∞, such as
√
lim x = lim x = +∞ (1.3)
x→+∞ x→+∞

As sequence limit, we say

• f (x) is convergent as x → +∞ if it has a limit as x → +∞
• f (x) is divergent as x → +∞ if it has no limit as x → +∞
From the geometric viewpoint, the limit is the horizontal line where the function value will approach to.

y=A
x

Figure 1: Geometric Interpertation of lim f (x) = A

x→+∞

Theorem 1.1 (Heine). lim f (x) = A if and only if for any sequence {xn } with limn→∞ xn = +∞, we
x→+∞
have lim f (xn ) = A.
n→∞

Heine theorem reveals the relation of function limit and sequence limit. It is useful to find the divergence
of a function.
Corollary 1.2. For a function f (x),

1
 • if ∃{xn } with lim xn = +∞ and {f (xn )} is divergent as n → +∞, then f (x) is divergent as x → +∞;
n→∞

• if ∃{xn } and {yn } with lim xn = lim yn = +∞, such that lim f (xn ) 6= lim f (yn ), then f (x) is
n→∞ n→∞ n→∞ n→∞
divergent as x → +∞;

For example, let’s consider the function y = cos x. Set xn = nπ. Obviously,

lim xn = +∞ and lim cos xn = lim (−1)n divergent (1.4)

n→∞ n→∞ n→∞

Hence by Heine’s theorem, cos x is divergent as x → +∞.

Proposition 1.3. If a function is convergent as x → +∞, then its limit is unique.
Similarly, we can define the following limits

lim f (x) and lim f (x) (1.5)

x→−∞ x→∞

Note that x → ∞ includes the two direction x → +∞ and x → −∞. For example
(
1 1, x>0
lim = 0 and lim f (x) divergent where f (x) = (1.6)
x→∞ x x→∞ −1, x ≤ 0

1.2 The case x → x0

Given x0 , the limit of f (x) as x → x0 (or at x = x0 ) is the number where the function will “tend to”,
denoted by

lim f (x) (1.7)

x→x0

For example,

x2 − 1
lim x + 1 = 2 and lim =2 (1.8)
x→1 x→1 x − 1

This example shows that lim f (x) does not require x0 ∈ Dom(f ). We have similar properties of x → x0 as
x→x0
x → ∞. Its geometric meaning of limx→x0 f (x) = A is that the point (x, f (x)) will approach to (x0 , A) as
x → x0 .
• If f (x) has a limit as x → x0 , then it is called convergent as x → x0 .
• If f (x) has no limit as x → x0 , then it is called divergent as x → x0 .

Theorem 1.4 (Heine). lim f (x) = A if and only if for any sequence {xn } with limn→∞ xn = x0 , we have
x→x0
lim f (xn ) = A.
n→∞

Corollary 1.5. For a function f (x),

• if ∃{xn } with lim xn = x0 and {f (xn )} is divergent as n → +∞, then f (x) is divergent as x → x0 ;
n→∞

• if ∃{xn } and {yn } with lim xn = lim yn = x0 , such that lim f (xn ) 6= lim f (yn ), then f (x) is
n→∞ n→∞ n→∞ n→∞
divergent as x → x0 ;
1
Example 1.6. Consider lim cos . Choose xn = 1/nπ
x→0 x
The limit is also unique.

2
Example 1.7. Suppose
( (
x2 −1
x + 1, x 6= 1 x2 − 1 x−1 , x 6= 1
f (x) = x + 1, g(x) = , h(x) = , k(x) =
4, x=1 x−1 3, x=1

Then

lim f (x) = 2, lim g(x) = 2, lim h(x) = 2, lim k(x) = 2 (1.9)

x→1 x→1 x→1 x→1

The function limit at x = x0 is independent on the value of f (x0 ).

Similarly, we can define

lim f (x) and lim f (x) (1.10)

x→x−
0 x→x+
0

−
where Hence the direction x → x0 is the combination of the directions x → x+
0 and x → x0 .

x0 x
x→ x−
0 x+
0 ←x

Figure 2: Limit directions

Proposition 1.8.

lim f (x) = A ⇔ lim+ f (x) = lim− f (x) = A (1.11)

x→x0 x→x0 x→x0

Example 1.9. Solve lim f (x) and lim g(x) where

x→0 x→0
( (
x + 1, x>0 x − 1, x>0
f (x) = and g(x) = (1.12)
x2 + 1, x<0 x2 , x≤0

Solution. Easy to find that

lim f (x) = lim+ x + 1 = 1, lim f (x) = lim− x2 + 1 = 1 ⇒ lim f (x) = 1 (1.13)

x→0+ x→0 x→0− x→0 x→0

and

lim g(x) = lim x − 1 = −1, lim g(x) = lim x2 = 0 ⇒ lim g(x) divergent (1.14)
x→0+ x→0+ x→0− x→0− x→0

There are formal definitions for variant limits. For example,

Definition 1.10. Suppose f (x) is defined near x0 . The limit of f (x) at x0 is A if for ∀ > 0, ∃δ > 0 such
that

|f (x) − A| <  holds for 0 < |x − x0 | < δ. (1.15)

3
1.3 Properties
Locally, if a function is convergent at x0 , then it is bounded near x0 .
Proposition 1.11 (local boundedness). Suppose f (x) is convergent at x0 . Then there exist δ > 0 and
M > 0 such that

|f (x)| ≤ M for all 0 < |x − x0 | < δ (1.16)

Proposition 1.12 (local isotone property). Suppose lim f (x) = A > B. Then there exists δ > 0 such that
x→x0

f (x) > B for all 0 < |x − x0 | < δ (1.17)

In particular, when B = 0, then this proposition is that if the limit of a function at x0 is positive, then
the function itself must be positive near x0 . This proposition holds as well for the case A < B.
Theorem 1.13 (Squeeze theorem). Suppose that h(x) ≤ f (x) ≤ g(x) holds near x0 (except x0 ). If
lim h(x) = lim g(x) = A, then lim f (x) = A.
x→x0 x→x0 x→x0

Definition 1.14. Suppose f (x) is defined on an interval I.

• f (x) is called increasing if f (x1 ) ≤ f (x2 ) for x1 < x2
• f (x) is called decreasing if f (x1 ) ≥ f (x2 ) for x1 < x2
• f (x) is called bounded from above if ∃M such that f (x) ≤ M for all x ∈ I

• f (x) is called bounded from below if ∃M such that f (x) ≥ M for all x ∈ I
• f (x) is called bounded if it is bounded from above and below.
Theorem 1.15 (Monotone boundedness criterion). A function f (x) is convergent as x → x−
0 if one of the
following happens

• f (x) is increasing and bounded from above in some interval (a, x0 )

• f (x) is decreasing and bounded from below in some interval (a, x0 )
For the direction x → x+
0 , we have a similar result.

2 Methods
2.1 Rational Operation Rules
Theorem 2.1 (Rational operation rules). Suppose f (x) and g(x) are convergent as x → x0 . Then f (x) +
g(x), f (x) · g(x) and f (x)/g(x) are convergent as x → x0 . Moreover


lim f (x) + g(x) = lim f (x) + lim g(x) (2.1)
x→x0 x→x0 x→x0


lim f (x) · g(x) = lim f (x) · lim g(x) (2.2)
x→x0 x→x0 x→x0
lim f (x)
f (x) x→x0
lim = (requiring lim g(x) 6= 0) (2.3)
x→x0 g(x) lim g(x) x→x0
x→x0

Corollary 2.2. Suppose f (x) and g(x) are convergent x → x0 . Then for a, b ∈ R and n ∈ N,
 n
lim [af (x) + bg(x)] = a lim f (x) + b lim g(x) and lim [f (x)]n = lim f (x) (2.4)
x→x0 x→x0 x→x0 x→x0 x→x0

4
 The above theorem and corollary hold for other limit directions as well.
Example 2.3. Solve the following limits

x3 + 1 x3 − 1 x2 + 4
 
1 2
lim , lim , lim , lim − 2
x→1 2x + 1 x→1 x − 1 x→∞ 2x2 + x + 3 x→1 x−1 x −1

0 ∞
(The last three are indeterminate forms of type , , ∞ − ∞.)
0 ∞
Solution. For the first one, we can directly use the rational operation rules:

x3 + 1 lim x3 + 1 2
x→1
lim = = . (2.5)
x→1 2x + 1 lim 2x + 1 3
x→1

For the second one, it is type 00 . The factor x − 1 can be cancelled by using x3 − 1 = (x − 1)(x2 + x + 1).
Hence
x3 − 1 (x − 1)(x2 + x + 1)
lim = lim = lim x2 + x + 1 = 3. (2.6)
x→1 x − 1 x→1 x−1 x→1

∞
For the third one, it is type ∞ . We can use the method of cancelling main terms of numerator and
2
denominator which are both x . Hence

x2 + 4 x2 1 + x42 1
lim = lim · = . (2.7)
x→∞ 2x2 + x + 3 x→∞ x2 2 + 1 + 32 2
x x

For the fourth one, it is type ∞ − ∞. We can reduce to a common denominator and cancell the zero
factor in denominator. Hence
 
1 2 x+1 2 x−1 1 1
lim − 2 = lim 2 − 2 = lim = lim = . (2.8)
x→1 x − 1 x −1 x→1 x − 1 x − 1 x→1 x2 − 1 x→1 x + 1 2

Actually, we have similar results for quotient of polynomials:


an
n n−1 bm , n=m

an x + an−1 x + · · · + a1 x + a0 
lim = ∞, n>m (2.9)
x→∞ bm xm + bm−1 xm−1 + · · · + b1 x + b0 
0, n<m


Example 2.4. Determine the constants a and b such that

 2 
x +3
lim + ax + b = 0.
x→∞ x−2

Solution. Denote the left limit by I and reduce to a common denominator

 2
x2 + 3 + (ax + b)(x − 2) (1 + a)x2 + (b − 2a)x + 3 + 4b

x +3
I = lim + ax + b = lim = lim
x→∞ x−2 x→∞ x−2 x→∞ x−2
(2.10)

From (2.9), a = −1; otherwise, I = ∞ which leads a contradiction with the assumption. By (2.9) again,
I = b − 2a = 0 and then b = 2a = −2.

5
2.2 Composition
Theorem 2.5 (Weak Version). Suppose y = f (u), u = g(x) and g(x) 6= u0 for x 6= x0 . Then

lim g(x) = u0 , lim f (u) = A =⇒ lim f ◦ g(x) = A = lim f (u)

x→x0 u→u0 x→x0 u→u0

which is also true for u0 = ∞ or x0 = ∞.

Before the strong version of composition rule which is more convenient to use, we first introduce the
notion of continuity.
Definition 2.6.
• A function is called continues at x = x0 if lim f (x) = f (x0 ).
x→x0

• A function is called continues in its domain if it is continues at each point in the domain.
By rational operation rules, all polynomials are continues.
Theorem 2.7 (Strong Version). Suppose y = f (u), u = g(x). Then

lim g(x) = u0 , lim f (u) = f (u0 ) =⇒ lim f ◦ g(x) = f (u0 )

x→x0 u→u0 x→x0
 
⇐⇒ lim f [g(x)] = f lim g(x)
x→x0 x→x0

In other words, if f (x) is continues, then we can exchange the order of f and lim . The case for x0 = ∞
x→x0
is also true.
Example 2.8. From the graph of y = ex :

9 y

1
x
−3 −2 −1 1 2 3

From the graph, we find lim ex = 1. Using this to prove

x→0

lim ex = ex0 , lim ax = ax0

x→x0 x→x0

Proof. For the first limit, set f (y) = ey and g(x) = x − x0 . Since lim ey = 1, then f (y) is continues at y = 0.
x→0
By strong version of composition rule, we have
 
lim x−x0
lim ex−x0 = lim f g(x) = f = e0 = 1


lim g(x) = ex→x0 (2.11)
x→x0 x→x0 x→x0

which leads that

lim ex = ex0 lim ex−x0 = ex0 (2.12)

x→x0 x→x0

6
Hence ex is continues.
For the second limit, since ex is continues, by strong version of composition rule, we have
lim x ln a
lim ax = lim exlna = ex→x0 = ex0 ln a = ax0 (2.13)
x→x0 x→x0

which shows that ax is continues as well.

Similarly, all logarithm functions are continues. By strong version of composition rule, we have (for
p ∈ R)
lim p ln x
lim xp = lim ep ln x = ex→x0 = ep ln x0 = xp0 (2.14)
x→x0 x→x0

which implies that xp is continues. Using strong version of composition rule again, we have
Proposition 2.9.
 p  p
lim f (x) = lim f (x) (2.15)
x→x0 x→x0

In particular, we known
p q
lim 1 + x2 = 1 + x20 (2.16)
x→x0

Actually, all basic elementary functions are continues. By composition rule and rational operation rules,
we find
Theorem 2.10. All elementary functions are continues.
For example,

lim sin x = sin x0 , lim cos x = cos x0 . (2.17)

x→x0 x→x0

3 Two important limits

sin x
The graph of the function y = x is as follows:

y
1

(
sin x
sin x x , x 6= 0
Theorem 3.1. lim = 1. Hence the function f (x) = is continues.
x→0 x 1, x=0
tan x 1 − cos x
Example 3.2. Solve lim , lim
x→0 x x→0 x2

7
Solution. For the first one, we have
tan x sin x 1
lim = lim · =1 (3.1)
x→0 x x→0 x cos x

For the second one, using double angle identity 1 − cos x = 2 sin2 x2 , we have
2
2 sin2 x
sin x2

1 − cos x 2 1 1
lim = lim = lim x = . (3.2)
x→0 x2 x→0 x2 x→0 2
2 2

1
Example 3.3. Solve lim x arcsin .
x→∞ x
Argument. Simplify the function u = x arcsin x1 . Denote y = 1
x and t = arcsiny. Hence

1 t
u= arcsin y = . (3.3)
y sin t

which implies that u = x arcsin x1 is composed by

t 1
u = f (t) = , t = g(y) = arcsin y, y = h(x) = (3.4)
sin t x
The former two are continues. Due to composition rule, we know
 

t
lim u = lim f g h(x) = lim f (t) = lim = 1. (3.5)
x→∞ x→∞ t→0 t→0 sin t

Only the following simplified solution is required in the notebook.

Solution.
1 y= x1 1 t=arcsin y t
lim x arcsin ===== lim arcsin y ======== lim = 1. (3.6)
x→∞ x y→0 y t→0 sin t

Note that the limit direction should simultaneously change when a substitution is applied.
π
Example 3.4. Solve lim 3n sin .
n→∞ 3n
sin x
Argument. Heine’s theorem give the bridge between sequence limits and function limits. Since lim = 1,
x→0 x
π
by choosing a sequence {xn = n } and Heine’s theorem, we have
3
π sin xn sin x
lim 3n sin = lim π · = lim π · = π. (3.7)
n→∞ 3n n→∞ xn x→0 x

Solution.
π sin x
lim 3n sin n
= lim π · = π. (3.8)
n→∞ 3 x→0 x

8
 y
3

y = (1 + x1 )x
1 y=e

x
20 40 60 80 100

Figure 3: Graph of y = (1 + x1 )x

The next important limit is

 x
1
lim 1+ (3.9)
x→+∞ x
The graph of y = (1 + x1 )x is in Figure 3 From the graph, the function y = (1 + x1 )x is increasing and
1
bounded. By monotone boundedness criterion, it is convergent as x → +∞ and denote lim (1 + )x = A.
x→+∞ x
As we learned in the previous section, limn→∞ (1 + n1 )n = e. Hence by Heine’s theorem, A = e.
 x
1
Theorem 3.5. lim 1 + = e.
x→+∞ x
Since
 −y
1 y=−x 1 y−1
lim (1 + )x ===== lim (1 − )−y = lim
x→−∞ x y→+∞ y y→+∞ y
 y  y  y−1  
y 1 1 1
= lim = lim 1 + = lim 1 + 1+ =e
y→+∞ y − 1 y→+∞ y−1 y→+∞ y−1 y−1
then
 x
1
lim 1 + = e. (3.10)
x→∞ x
Example 3.6. Solve
 x  x−1
1 2 x+1
lim (1 + x) , lim
x 1− , lim
x→0 x→∞ x x→∞ x−1
Solution. For the first one, we know
 y
1 y= 1
x 1
lim (1 + x) x ====
= lim 1+ = e. (3.11)
x→0 y→∞ y
For the second one, we have
 x 1 2
 −2y  y −2
2 y =− x 1 1
lim 1 − ====== lim 1 + = lim 1+ = e−2 (3.12)
x→∞ x y→∞ y y→∞ y
For the third one, we have
 x−1  x−1 1 2  2y
x+1 2 y = x−1 1
lim = lim 1 + ======= lim 1 + = e2 (3.13)
x→∞ x − 1 x→∞ x−1 x→∞ y

9
4 Homework
Exercise 1.3 (A)
2. Are the following conclusions correct? If it is not correct, give a counter example.
1
(1) If lim f ( ) = A, then limx→0+ f (x) = A;
n→∞ n
(2) If lim f (x) = A, then lim [f (x)]k = Ak where k ∈ N
x→x0 x→x0

(3) If lim f (x) and lim [f (x) + g(x)] both exist, then lim g(x) must exist.
x→x0 x→x0 x→x0

(4) If lim f (x) and lim [f (x) · g(x)] both exist, then lim g(x) must exist.
x→x0 x→x0 x→x0

(5) If f (x) > 0 holds in its domain and lim f (x) = A, then A > 0.
x→x0

5. Do the following limits exsit?

(
sin x
(5) lim f (x) where f (x) = x , x<0
1/x
x→0 (1 + x) , x>0

6. Find the following limits:

√ √ √ x
(3) lim x( 2 + x − x) (5) lim √ √
x→+∞ x→0 3 2 + x − 3 2 − x
√ √
 
1 3
(7) lim − 3
(8) lim sin( x + 1 − x)
x→1 1 − x 1−x x→+∞

7. Utilizing the two important limits given in the text to find the following limits:
 5n
x sin x 1 2
(1) lim x cot (3) lim (6) lim (1 − 2x) x (8) lim 1+
x→0 2 x→π x−π x→0 n→∞ 5n

8. Determine the constants a in the following equality:


x+a
1 , x<0


 2+e x
(2) lim f (x) exists where f (x) =
x→0
sin x tan x



1−cos 2x
2
, x>0

Exercise 1.3 (B)

1. Find the following limits:
 3x−1
3x − 1
(2) lim
x→∞ 3x + 1

10