Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

MATH20101 Real Analysis: Part 1.1. Limits of Functions

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

MATH20101 Real Analysis

These notes contain the statements of all results and examples but not
necessarily the proofs. For these, you will have to attend the lectures.

Part 1.1. Limits of functions


1.1.1 Introduction

Example 1.1.1 What happens to f : R \ {1} → R, given by

x2 − 1
f (x) = , x 6= 1,
x−1
as x gets close to 1?

If we substitute x = 1 we get f (1) = 00 which is undefined. Instead we


might try to substitute numbers close to 1:
x2 −1 x2 −1
x x−1
x x−1
1.1 2.1 0.9 1.9
1.01 2.01 0.99 1.99
1.001 2.001 0.999 1.999
1.0001 2.0001 0.9999 1.9999

It appears that if x is very close to 1 then f (x) is very close to 2. In fact,

x2 − 1 (x − 1) (x + 1)
= = x + 1,
x−1 x−1
though, as an equation, this only holds for x 6= 1, since we are never allowed
to divide by zero.

1
Graphically we have:

where the point (1, 2) is omitted from the graph. It is obvious from the graph
that as x approaches 1 then f (x) approaches 2.
Example 1.1.2 The function
π 
g : R\ {0} → R, g (x) = sin
x
is defined for all x 6= 0. What happens as x gets close to 0?
We can calculate a few values:
!  ! 
x sin πx x sin πx
1 0 −1 0
0.1 0 −0.1 0
0.01 0 −0.01 0
0.001 0 −0.001 0
0.0001 0 −0.0001 0
!π
and, in fact, sin x = 0 when x = ±10−n for all n ≥ 1.
It is important not to derive the wrong conclusion from this little infor-
mation. For we also have
!  ! 
x sin πx x sin πx
2
5
1 − 23 1
2
9
1 − 27 1
2 2
13
1 − 11 1
2 2
17
1 − 15 1
! 
and, in fact, sin πx = 1 when x = 2/ (1 + 4n) for all n ∈ Z.
From this table we might draw a completely different conclusion to that
drawn from the first.

2
In fact the function oscillates wildly near the origin as is seen in

1.1.2 Limit of a function at a finite point.

Definition 1.1.3 Define the open intervals (a, b) = {x : a < x < b} and the
closed intervals [a, b] = {x : a ≤ x ≤ b} along with the obvious open-closed
and closed-open intervals.

Definition 1.1.4 • A neighbourhood of a is an open interval (a − δ, a + δ)


for some δ > 0.
(Often in lectures I will write nbhd instead of the longer word, neigh-
bourhood).
• A deleted neighbourhood of a is a neighbourhood of a with a deleted,
i.e. removed. So it is really the union of two open intervals, i.e.
(a − δ, a) ∪ (a, a + δ) = {x : 0 < |x − a| < δ}
for some δ > 0.

Assumption Given a function f : A → R and a ∈ R, which may or may


not lie in A, we will assume that there exists a deleted neighbourhood of a
lying inside A.

Example 1.1.5 If
x2 − 1
f (x) = , A = R \ {1} ,
x−1
and a = 1 then the assumption is satisfied (even though a ∈
/ A).
If
f (x) = x, A = [0, 1] ,
and a = 1 then the assumption is not satisfied (even though a ∈ A).

3
Definition 1.1.6 Let f : A → R be a function whose domain contains a
deleted neighbourhood of a ∈ R. The limit of the function f at a is L
if, and only if, for all ε > 0 there exists a δ > 0 such that if 0 < |x − a| < δ
then |f (x) − L| < ε. That is:

∀ε > 0, ∃δ > 0, ∀x, 0 < |x − a| < δ ⇒ |f (x) − L| < ε.

(Recall that in logic we read “p ⇒ q” as “if p holds then q will hold”.)


We write
lim f (x) = L.
x→a

But we may also write “f (x) → L as x → a”, which we read as “f (x)


tends to L as x tends to a”. (Often in lectures I will write Lx→a f (x) = L).

It is important that you learn and memorise this definition.


It is also important that you note that in the definition we have

0 < |x − a| , i.e. x 6= a.

That is, the limit of a function is calculated at a point a, by looking at the


value of the function at points x near but not equal to the point a.

Example 1.1.7 Use the ε − δ definition to prove that


(i)
x2 − 1
lim = 2,
x→1 x − 1

(ii)
lim x3 + x2 − 4x = 4,
! 
x→2

(iii)
x2 + 2x + 2
lim = 2.
x→2 x+3
Solutions of (i) and (iii) in lectures, (ii) for student to try. We will see (iii)
again later.

Example 1.1.8 If limx→a f (x) = L exists then the limit is unique.

Proof in lectures.

4
Example 1.1.9 (In feedback class) Use the ε − δ definition and proof by
contradiction to show that π 
lim sin
x→0 x
does not exist.

Hint Assume the limit exists, and call it L. Choose ε = 1/3 in the definition
of limit to find a δ > 0. By an earlier example find points in (0, δ) for which
sin (π/x) is 1 and is 0. The contradiction will come from the fact that both
0 and 1 cannot be within ε = 1/3 of L.

5
1.1.3 One sided limits.

Can we calculate limx→0 x? We cannot verify the definition because any
deleted neighbourhood of 0 must
√ be of the form (−δ, 0) ∪ (0, δ) and so will
√ x for which x is not defined. We must conclude, therefore,
contain negative
that limx→0 x is not defined. Yet, from the graph,


it seems that we could talk meaningfully of the limit of x as x approaches
0 from the right.
Definition 1.1.10 • Right hand limit: Suppose that f : A → R is defined
for x to the right of a ∈ R, i.e. in an interval (a, a + α) for some α > 0.
Then
lim+ f (x) = L
x→a
if, and only if, for any ε > 0 there exists δ > 0 such that if a < x < a + δ
then |f (x) − L| < ε. That is:
∀ε > 0, ∃δ > 0, ∀x, a < x < a + δ ⇒ |f (x) − L| < ε.
We sometimes write f (a+ ) for limx→a+ f (x).
Definition 1.1.11 • Left hand limit: Suppose that f : A → R is defined
for x to the left of a ∈ R, i.e. in an interval (a − β, a) for some β > 0. Then
lim f (x) = L
x→a−

if, and only if, for any ε > 0 there exists δ > 0 such that if a − δ < x < a
then |f (x) − L| < ε. That is:
∀ε > 0, ∃δ > 0, ∀x, a − δ < x < a ⇒ |f (x) − L| < ε.
We sometimes write f (a− ) for limx→a− f (x).

Note how in both definitions we exclude x = a as we did in the definition of


the (two-sided) limit.

6
Example 1.1.12 Using the ε−δ definition calculate the two one-sided limits
for x/ |x| (the signum function, also written as sgnx) at x = 0. The graph
for this is:

−1

where the value of x/ |x| is left undefined at 0.


Solution in lectures.


Example 1.1.13 Verify the definition to prove that limx→0+ x = 0.

Solution in lectures.

Theorem 1.1.14 If limx→a+ f (x) = L exists then the limit is unique, if


limx→a− f (x) = L exists then the limit is unique.

Proof left to students.

Theorem 1.1.15 We have

lim f (x) = L
x→a

if, and only if,

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


x→a− x→a+

So a (two-sided) limit exists if and only if both one-sided limits exist and are
equal.

Proof in lectures.

7
Example 1.1.16 The limit
x
lim
x→0 |x|

does not exist.

Solutions in lectures.

Note i) If x → a is replaced by either x → a+ or x → a− then the one-sided


limit is again unique.
ii) You cannot use Theorem 1.1.15 to prove that limx→0 x/|x| does not
exist because neither limx→0+ x/|x| nor limx→0− x/|x| exist either!

8
1.1.4 Limit of a function at infinity.
Definition 1.1.17 Assume f : A → R is defined for all sufficiently large
positive x. Then
lim f (x) = L
x→+∞
if, and only if, for all ε > 0 there exists an K ∈ R such that if x > K then
|f (x) − L| < ε.
Since we are looking at large positive x it suffices to look for a positive
K. (If you found a negative K which worked, it would work for |K| .) Hence
the definition becomes:
∀ε > 0, ∃k > 0 : x > K ⇒ |f (x) − L| < ε. (1)
Definition 1.1.18 Assume f : A → R is defined for all sufficiently large
negative x. Then
lim f (x) = L
x→−∞
if, and only if, for all ε > 0 there exists an G ∈ R such that if x < G then
|f (x) − L| < ε.
Since we are looking at large negative x it suffices to look for a negative
G. Hence the definition becomes:
∀ε > 0, ∃G < 0 : x < G ⇒ |f (x) − L| < ε.
Note The symbols +∞ (also simply known as ∞) and −∞ are not real
numbers. You cannot say
1 1
= ∞ or = 0.
0 ∞
The symbols are used in this course simply as a shorthand notation.
So x → +∞ should be read as “x takes arbitrarily large positive values”,
similarly x → −∞ is shorthand for “x takes arbitrarily large negative values”.

Example 1.1.19 Use the definition to calculate the limits at infinity, if they
exist, of the function
x2
f (x) = 2 .
x +1
Solutions for x → +∞ in lectures, whilst x → −∞ is left to students.

Theorem 1.1.20 If limx→+∞ f (x) exists, then the limit is unique. If limx→−∞ f (x)
exists, then the limit is unique.
Proof Left to students.

9
1.1.5 Divergence or “when the limit is infinite”.
In the theory of sequences we say that a sequence diverges if it doesn’t
converge. There are a number of ways it could diverge, e.g. it could oscillate,
i.e. an = (−1)n , n ≥ 1, or it could be unbounded, i.e. an = n, n ≥ 1.
For certain unbounded functions there is a type of limit that can still
be defined. The first definition below encapsulates the situation in which
given a function defined on a deleted neighbourhood of a ∈ R, and given
any real number K, which we might think of as positive and large, there is
some deleted neighbourhood of a on which the function is greater than K.
This can be repeated for each and every positive large K. Presumably the
larger the K the smaller the deleted neighbourhood. Then, if limx→a f (x)
is to be assigned a real value connected in some way with the values taken
by the function on deleted neighbourhoods of a, this value should be larger
than every positive large K. There is no such possible real value! Instead
we assign the symbol +∞ to limx→a f (x) .

Definition 1.1.21 • Let f : A → R be a function whose domain contains a


deleted neighbourhood of a ∈ R. We write

lim f (x) = +∞,


x→a

or say
“f (x) tends to + ∞ as x tends to a”
if, and only if, for all for all K > 0 there exists δ > 0 such that if
0 < |x − a| < δ then f (x) > K. That is:

∀K > 0, ∃δ > 0, ∀x, 0 < |x − a| < δ ⇒ f (x) > K. (2)

• Similarly, we write
lim f (x) = −∞,
x→a
or say
“f (x) tends to − ∞ as x tends to a”
if, and only if, for all for all G < 0 there exists δ > 0 such that if
0 < |x − a| < δ then f (x) < G. That is:

∀G < 0, ∃δ > 0, ∀x, 0 < |x − a| < δ ⇒ f (x) < G.

10
Note that here, in f (x) < G, both G and f (x) will be negative and so,
in fact, f (x) will be larger, in magnitude, than K!

Example 1.1.22 Verify the definition to show that


1
lim = +∞.
x→1 (x − 1)2

Solution Rough work. Given K > 0 we want


1 1
2 > K, which rearranges as |x − 1| <
√ .
(x − 1) K
End of rough Work.

Proof. Let K > 0 be given. Choose δ = 1/ K. Assume that
1
0 < |x − 1| < δ = √ .
K
This rearranges to give
1
> K.
(x − 1)2
Thus we have verified the definition that limx→1 1/ (x − 1)2 = +∞. 

It should not be hard for the student to supply definitions for

lim f (x) = +∞ or − ∞ and lim f (x) = +∞ or − ∞.


x→a+ x→a−

Further, the student should be able to also define

lim f (x) = +∞ or − ∞, and lim f (x) = +∞ or − ∞.


x→+∞ x→−∞

In fact you will be asked to do just this in a question on the Problem


Sheets.

Note In none of the cases above do we say that the limit exists. If we
say that lim f (x) exists we are implicitly assuming that it is finite. This
is because, as in the definitions of limits at infinity, the symbols +∞ and
−∞ are not real numbers. They are used simply as shorthand. To say
limx→a f (x) = +∞ is to say that f satisfies (3), and there is no use of the
symbol +∞ in that definition..

11

You might also like