Course Code: MTS 323

Course Title: Real Analysis II

Number of Units: 3 Units
Course Duration: Three hours per week
COURSE DETAILS:
Course Coordinator: Dr. Mewomo Oluwatosin, B.Sc., M.Sc., PhD
Email: [email protected], [email protected]
Office Location: COLNAS Extension
Other Lecturers: NIl

COURSE CONTENT:

1. Differentiation in ℜn

(a) differentiation of real valued functions of real variable

(b) directional derivatives

(c) partial derivatives

(d) higher order derivatives

(e) Taylor’s theorem

(f) classification of stationary points

(g) local extrema with constraints (method of Langrange’s multipli-

ers)

2. Integration in ℜ

(a) Riemann integral review

(b) Riemann Stieljes integral

1
 3. Function of bounded variation.

4. Sequences of functions

(a) pointwise convergence

(b) uniform convergence

(c) uniform Cauchy criterion

COURSE REQUIREMENT:
This is a compulsory course for all students in the Departments of Mathe-
matics and Statistics. Some knowledge from MTS101, MTS102 and MTS223
are required.It is also expected that students participate in all the course ac-
tivities and have a minimum of 75 percent attendance to be able to qualify
for the final examination. The grading will be based on weekly homework
assignment (10 percent), an in class - mid term test (20 percent) and a final
examination (70 percent).
READING LIST: The following are recommended:

1. R.G. Bartle; The elements of real analysis, 2nd edition, Jossey-Bass,

(1976).

2. W.R. Wade; An introduction to analysis, 2nd edition, Prentice, (2000).

3. T.M. Apostol; Mathematical analysis.

4. W. Rudin; Mathematical analysis.

2
1 Differentiation in ℜn
We recall the definition of derivative of a function f with Domain(f ) ⊂ ℜ
and Range(f ) ⊂ ℜ.

Definition 1.1 Let x◦ be an interior point of Dom(f ). A real number l is

called the derivative of f at x◦ , if

f (x◦ + h) − f (x◦ )
limh→0 = l,
h

when this exist, we denote it by f ′ (x◦ ) and so we write

f (x◦ + h) − f (x◦ )
limh→0 = f ′ (x◦ ) (h = x − x◦ ) (1.1)
h

We may generalized this notion to a map f : A → ℜm where A ⊂ ℜn .

Definition 1.2 Let f be a function with Dom(f ) ⊂ ℜn and Range(f ) ⊂ ℜm .

Let a be an interior point of Dom(f ) and let u be any vector in ℜn , a vector
v ∈ ℜm is called the directional derivative of f at a along the line determined
by u if
f (a + tu) − f (a)
limt→0 = v.
t
When this limit exit, we denote it by Du f (a). That is, the directional deriva-
tive of f at a in the direction of u is given by

f (a + tu) − f (a)
Du f (a) = limt→0 (1.2)
t

Using ϵ, δ notation, Du f (a) may be written as follows: Given ϵ > 0, there

exists a δ, δ(ϵ) > 0, such that 0 < |t| < δ(ϵ), then

f (a + tu) − f (a)
|Du f (a) − |<ϵ
t
3
 Remark: If a = (a1 , a2 , ..., an ) and u = (u1 , u2 , ..., un ), then we define
Du f (a) as

f (a1 + tu1 , a2 + tu2 , ..., an + tun ) − f (a1 , a2 , ..., an )

Du f (a) = limt→0
t

Definition 1.3 Let f be a function with Dom(f ) ⊂ ℜn and Range(f ) ⊂ ℜm .

Let a be an interior point of Dom(f ). The directional derivative of f at a
in the directions of the special vectors e1 , e2 , ...en (basis of ℜn ) are called the
partial derivatives of f with respect to the 1st, 2nd,...nth variable respectively.
Thus if a = (a1 , a2 , ..., an ), then since e1 = (1, 0, 0, ...0),
e2 = (0, 1, 0, ...0), ...en = (0, 0, ..., 1), we have from the above remark that

f (a1 + t, a2 , ..., an ) − f (a1 , a2 , ..., an )

De1 f (a) = limt→0
t

f (a1 a2 + tu2 , ..., an ) − f (a1 , a2 , ..., an )

De2 f (a) = limt→0
t
and in general

f (a1 , a2 , ..., an + t) − f (a1 , a2 , ..., an )

Den f (a) = limt→0
t

The vectors De1 f (a), De2 f (a), ...Den f (a) are often written as D1 f (a), D2 f (a), ..., Dn f (a).
Remark:
If m = 1, then f is a real-valued function, and D1 f (a), D2 f (a), ..., Dn f (a)
∂f (a) ∂f (a)
are usually written as ∂x1
, ∂x2 , ... ∂f∂x(a)
n
.

Example 1.4 Let f : ℜ2 → ℜ be defined by f (x, y) = x2 + y 2 . Find the

directional derivative of f at a = (2, 1) along a line determined by the vector
u = (3, −4).

4
 Solution To be provided in class.
Remark:
The existence of partial derivatives does not imply the existence of directional
derivatives as the following example shows.

Example 1.5 Define f : ℜ2 → ℜ by


 7 if x.y = 0
f (x, y) =
 5 if x.y ̸= 0

Assignment 1

1. If f : ℜ3 → ℜ is defined by f (x, y, z) = 2x2 − y + 6xy − z 3 + 3z.

Calculate the directional derivative of f at the origin in the direction
of the vectors

(a) u = (1, 2, 0)

(b) u = (2, 1, −3)

Find D1 f (0, 0, 0), D2 f (0, 0, 0) and D3 f (0, 0, 0).

2. Let f : ℜ2 → ℜ be defined by

 2xy 2
if (x, y) ̸= 0
x2 +y 4
f (x, y) =
 0 if (x, y) = 0

(a) Show that the partial derivative fx and fy exist at the origin and
at any point (w1 , w2 ) ̸= (0, 0).

(b) Show that f is discontinuous at the origin.

5
 Higher Derivatives:
Let f : ℜn → ℜk and let a be any point of ℜn and suppose that one of
the partial derivatives Dj f exist for all x in some open ball S(a, δ) ⊂ ℜn .
Define a new function g on S(a, δ) by g(x) = Dj f (x), then g is a function
with Dom(g) ⊂ ℜn and Range(g) ⊂ ℜk . In addition if a ∈ Dom(g), then
we may ask if Dk g(a) = Dk Dj f (a). If this partial derivative exist, we call it
the second partial derivative of f at a with respect to j variable and the k th
derivative. Putting x as (x1 , x2 , ..., xn ), then in a more familiar notation, we
write
∂ 2 f (a)
Dk Dj f (a) as
∂xk ∂xj

Example 1.6 Let f : ℜ2 → ℜ be defined by

( )
√ x
f (x, y) = xy + 1 + x2 sin √
1 + x2

Then
( ) ( )
∂f (x, y) x x 1 x
D1 f (x, y) = = y+ √ sin √ + cos √
∂x 1 + x2 1 + x2 1 + x2 1 + x2

From this, it follows that

D2 D1 f (x, y) = 1

Also,
∂f (x, y)
D2 f (x, y) = =x
∂y
D1 D2 f (x, y) = 1.

Thus D1 D2 f (x, y) = D2 D1 f (x, y).

6
Definition 1.7 Suppose f has a second partial derivatives on a neighbour-
hood of a which are continuous throughout the neighbourhood. We define
D2 f (a) as a function from ℜn × ℜn → ℜ by setting

∑
n ∑
n
2 2
D f (a)(u) = Di Dj f (a)ui uj ,
j=1 i=1

where u2 = (u, u) with u = (u1 , u2 , ..., un ) ∈ ℜn . The function D2 f (a) is

known as the second partial derivative of f at a, similarly D3 f (a) is the third
partial derivative of f at a.
∑
n ∑
n ∑
n
3 3
D f (a)(u) = Di Dj Dk f (a)ui uj uk ,
i=1 j=1 k=1

where u3 = (u, u, u) with u = (u1 , u2 , ..., un ) ∈ ℜn .

In general, the higher order derivatives take the following for:

For u = (h, k) and (x, y) in ℜ2 ,

D2 f (a)(u)2 = h2 fxx (a) + 2hkfxy (a) + k 2 fyy (a).

Once the partial derivatives are continuous, then fxy = fyx .

Also,

D3 f (a)(u)3 = h3 fxxx (a) + 2h2 fxxy (a) + 3hk 2 fxyy (a) + k 3 fyyy (a).

Taylor’s Theorem
For futher details on the notions of continuity, differentiability, Taylor’s
and Maclaurin’s theorems in 1-dimensional see MTS223 lecture note.
We are now in a position to state Taylor’s theorem for a real-valued
function f with Dom(f ) ⊂ ℜn .

7
Theorem 1.8 Let f be a real valued function with Dom(f ) ⊂ ℜn . Let a, b
be interior points of Dom(f ) and suppose that f has a continuous partial
derivatives of order m in an open set containing a, b and the line segement
joining a and b. Then there exists c on the line segement such that

Df (a)(b − a) D2 f (a)(b − a)2 Dm f (a)(b − a)m

f (b) = f (a) + + + ... +
1! 2! m!
If f (x, y) is a function of two variables, then Taylor’s theorem takes the
form
∂ ∂ 2
∂ ∂ (h ∂x + k ∂y ) f (x◦ , y◦ )
f (x◦ + h, y◦ + k) = f (x◦ , y◦ ) + (h + k )f (x◦ , y◦ ) +
∂x ∂y 2!
∂ ∂ 3
(h ∂x + k ∂y ) f (x◦ , y◦ )
+ + ...
3!
Note:
With h = x − x◦ , k = y − y◦ ,
n−1 (
∑ )m
∂ ∂
f (x, y) = h +k f (x◦ , y◦ ).
m=0
∂x ∂y

Example 1.9 Expand f (x, y) = x2 y + 3y − 2 in powers of x − 1 and y + 2.

Solution To be provided in class.

Example 1.10 Obtain the Maclaurin’s series of the function f (x, y = ex+y cosy
neglecting third and higher degrees.

Solution To be provided in class.

Extreme Value:

Definition 1.11 Let f be a real value function with Dom(f ) ⊂ ℜn . We say

that f has a local (or relative) maximum (minimum) at a point a ∈ Dom(f ),

8
if there is a neighbourhood U (a) of a such that f (x) ≤ f (a) (f (x) ≥ f (a)
for all x ∈ U (a) ∩ Dom(f ).

A function is said to have a local or relative extreme value at a if it either

have a relative maximum or relative minimum at a. A point a is called a
critical point if f ′ (a) = 0. A critical point that is not a local extreme is called
a saddle point.

Theorem 1.12 Let f be a real valued function with Dom(f ) ⊂ ℜn . If f is

differentiable at a and a is a local extremum, then f ′ (a) = 0.

Theorem 1.13 Let f be a real valued function with Dom(f ) ⊂ ℜn . If f is

differentiable at a and a is a local extremum, then

∂f (a) ∂f (a) ∂f (a)

= = ... = 0.
∂x1 ∂x2 ∂xn

Corollary 1.14 If f ′ (a) ̸= 0, then f cannot have an extreme value at a.

Remark:
We have not said that if f ′ (a) = 0, then f necessarily has an extremum at a.

Example 1.15 Let f : ℜ2 → ℜ be defined by

1. f (x, y) = x2 + y 2

2. f (x, y) = x2 − y 2

Describe the nature of the critical point of f in each case.

Solution To be provided in class.

9
Definition 1.16 1. A real valued function g defined on ℜn is called a
quadratic function if it has the form
n ∑
∑ n
g(h) = aij hi hj ,
j=1 i=1

h = (h1 , h2 , ..., hn ) ∈ ℜn and some marix (aij ), aij ∈ ℜ, i = 1, 2, ...n,

j = 1, 2, ...n.

2. A quadratic function g : ℜn → ℜ is said to be

(a) positive definite if g(h) ≥ 0 for every h ∈ ℜn and g(h) = 0 only

for h = 0

(b) negative definite if g(h) ≤ 0 for every h ∈ ℜn and g(h) = 0 only

for h = 0.

Theorem 1.17 If f : ℜ2 → ℜ and a is an interior point of Dom(f ). Suppose

f has continuous partial derivatives of order two on a neighbourhood of a and
that a is a critical point of f.

1. If D2 f (a)(u2 ) is positive definite, then f has a relative minimum at a.

2. If D2 f (a)(u2 ) is negative definite, then f has a relative maximum at a.

3. If D2 f (a)(u2 ) assume both positive definite and negative definite for

u ∈ ℜn , then f has a saddle point.

Example 1.18 Verify that the origin is a critical point of the function f (x, y) =
x2 − xy + y 2 and determine whether at this point f has a local mimimum,
local maximum or neither.

10
 Solution To be provided in class.

Example 1.19 Discuss the behaviour of the following function defined from
ℜ2 into ℜ defined by

f (x, y) = 2x5 y + 3xy 5 + xy.

Solution To be provided in class.

Remark:
In some situations, the determination of the sign of D2 f (a) may not be easy,
we give the following more practical test.

Theorem 1.20 Let f be a real-valued function with Dom(f ) ⊂ ℜ2 , have

continuous partial derivate of order three on an open set containing a critical
point a in ℜ2 , and let ∆ = fxx (a)fyy (a) − (fxy (a))2 .

1. If ∆ > 0 and if fxx (a) > 0, then f has a relative minimum at a.

2. If ∆ > 0 and if fxx (a) < 0, then f has a relative maximum at a.

3. If ∆ < 0, then the point a is a saddle point of f.

We next consider extreme problems with some copnstraints which can be

put in the form maximise or minimise P = f (x, y) subject to the condition
g(x, y) = 0.

Example 1.21 Suppose a manufacturer is producing cements at two loca-

tions A and B. Let us assume that the cost of paying for the inspection of the
work at both locations depends on the number of inspections x at A and y at

11
B according to the formula C(x, y) = 2x2 + xy + y 2 + 25. How many inspec-
tions should be made at each site to minimize his cost if the total number of
inspection must be 16?

Solution To be provided in class.

Example 1.22 Of all the rectangles having the same perimeter 10 meters.
Find the one having the greatest area.

Solution To be provided in class.

Remark:
In some cases, it is difficult to solve the resulting equation in a closed form for
some of the unknown in terms of the other. Thus, we result to the method
of Lagrange’s multiplier.
Lagrange’s Multiplier
In finding extreme values of a function f with Dom(f ) ⊂ ℜn , and
Range(f ) ⊂ ℜ, subject to the restraining condition g(x) = 0, we apply the
following procedure:

1. Define the function F on Dom(f ) ∩ Dom(g) into ℜ by

F (x) = f (x) + λg(x).

2. Solve the n + 1 equations for λ1 , λ2 , ...λn .

Fx1 = 0, Fx2 = 0, ...., Fxn = 0, g(x) = 0. The constant λ which is not
wanted in the end is called the Lagrange’s multipliers.

If there are two restraining conditions g(x) = 0 and h(x) = 0, we shall

consider the function

F (x) = f (x) + λg(x) + βh(x)

12
and solve the equations Fx1 = 0, Fx2 = 0, ...., Fxn = 0, g(x) = 0, h(x) = 0.
We shall illustrate this method with some examples.

Example 1.23 Find the shortest distance from the point (1,0) to the parabola
y 2 = 4x.

Solution To be provided in class.

Example 1.24 Find the point on the surface x2 + y 2 + z 2 = 1 at which

f (x, y, z) = xyz is stationary.

Solution To be provided in class.

Example 1.25 Find the points on the sphere x2 + y 2 + z 2 = 36 that are

closest and farthest from the point (1,2,2).

Solution To be provided in class

13
 Assignment 3:

1. Obtain the Taylor’s series of tan−1 ( xy ) about the point (1,1) up to the
second degree.

2. Determine the nature of the turning points of the following

(a) f (x, y) = x3 y + 3x + y

(b) f (x, y) = sinx + siny + sin(x + y) for 0leqx ≤ 2π, 0 ≤ y ≤ 2π.

(c) f (x, y) = 16 + 4x + 7y − 2x2 − y 2

3. Find the shortest distance from the origin to the hyperbola

x2 + 8xy + 7y 2 − 225 = 0.

4. Find the point on the plane 2x − 3y − 4z = 25 which is nearest to the

point (3,2,1).

5. Find the maximum and minimum value of x2 + y 2 + z 2 subject to the

x2 y2 z2
condition 4
+ 5
+ 25
and z = x + y.

6. At what point on the curve x2 + y 2 = 1 does the product xy have a

maximum?

2 Integration in ℜ
Riemann Integral and its basic properties were studied in MTS223, for quick
review and details on Riemann Integral, see MTS 223 lecture note.
In this section, we shall study Riemann-Stieljes Integral (R-S integral)
which is a generalization of Riemann integral.

14
 We begining with the following concepts:
A partition of the closed and bounded interval [a,b] is a finite set of points
P = {x0 , x1 , ..., xn }, such that a = x0 < x1 < x2 < ... < xn = b. We denote
the set of all partitions of [a,b] by P[a,b] .
For any partition P of [a,b], we let

Mi = sup{f (x) : xi−1 ≤ x ≤ xi }

and
mi = inf {f (x) : xi−1 ≤ x ≤ xi }

Definition 2.1 Let α be a monotone increasing function on [a,b]. Corre-

sponding to each p ∈ P[a,b] , we write

0 ≤ ∆αi = α(xi ) − α(xi−1 ) i = 1, 2, ..., n.

For any bounded real valued function f on [a,b], we put

∑
n
U (p, f, α) = Mi ∆αi ,
i=1

∑
n
L(p, f, α) = mi ∆αi .
i=1

Also, we write
∫ b
f (x)dα(x) = inf {U (p, f, α) : p ∈ P[a,b] } (2.1).
a

This is called the upper R-S integral of f with respect to alpha.

∫ b
f (x)dα(x) = sup{L(p, f, α) : p ∈ P[a,b] } (2.2).
a

This is called the lower R-S integral of f with respect to alpha.

15
 If the LHS of (2.1) and LHS of (2.2) are equal, we denote their common
value by ∫ ∫
b b
f dα or f (x)dα(x) (2.3).
a a

If (2.3) exit, we say that f is Riemann-Stieljes (R-S) integrable with re-

spect to α on [a,b].

We recall that a partition Q is said to finer than a partition P ( or is

called a refinement of P) on the interval [a,b] if P and Q are partitions of
[a,b] and P ⊂ Q.

Theorem 2.2 If p∗ is a refinement of p, then

1. L(p, f, α) ≤ L(p∗ , f, α)

2. U (p, f, α) ≥ U (p∗ , f, α)

Proof To be provided in class.

The next result gives a necessary and sufficient condition for f to be R-S
integrable with respect to α.

Theorem 2.3 If f is a real valued function on [a,b], then f is R-S integrable

if and only if given epsilon > 0, there exists a partition p ∈ P[a,b] such that

U (p, f, α) − L(p, f, α) < ϵ (2.4)

Proof To be provided in class.

Theorem 2.4 If f is continuous on [a,b], then f is R-S integrable on [a,b].

Proof To be provided in class.

16
Theorem 2.5 If f is monotone on [a,b], and α is continuous on [a,b], then
f is R-S integrable on [a,b].

Proof To be provided in class.

Theorem 2.6 (Basic properties of R-S integral) Let f1 , f2 be R-S integrable

and λ an arbitrary constant. Then

1. f1 + f2 and λf1 are R-S integrable on [a,b]. Furthermore

∫ b ∫ b ∫ b
(f1 + f2 )dα = f1 dα + f2 dα
a a a

and ∫ ∫
b b
(λf1 )dα = λ f1 dα
a a

2. If f1 (x) ≤ f2 (x) for every x ∈ [a, b]. Then

∫ b ∫ b
f1 dα ≤ f2 dα
a a

3. If f is R-S integrable on [a,b] and a ≤ c ≤ b. Then f is R-S integrable

on [a,c] and on [c,b]. Furthermore,
∫ b ∫ c ∫ b
f dα = f dα + f dα
a a c

3 Function of Bounded Variation

Definition 3.1 Let f be a function defined on [a,b]. For each p ∈ P[a,b] , let
Wp (f ) be the real number given by

∑
n
Wp (f ) = |f (xi ) − f (xi−1 )|
i=1

17
If for each p ∈ P[a,b] , there is a constant M > 0, such that Wp (f ) ≤ M for
every p ∈ P[a,b] , then f is said to be of bounded variation on [a,b].

Since Wp (f ) is always less than or equal to M, p ∈ P[a,b] , then

V = {Wp (f ) : p ∈ P[a,b] }

is bounded above and non-empty. By completeness axiom, V has a supremum

since it is bounded above.

Definition 3.2 Let f be of bounded variation on [a,b] and for each p ∈ P[a,b] ,
the number
vf (a, b) = sup{Wp (f ) : p ∈ P[a,b] }

is called the total variation of f on [a,b].

Remark:

1. 0 ≤ vf (a, b) < +∞.

2. If f is of bounded variation on [a,b], then f is bounded on [a,].

3. vf (a, b) = 0 if and only if f is constant.

Theorem 3.3 If f is of bounded variation on [a,b] and a < c < b, then f is

of bounded variation on [a,c] and [c,b].

Proof To be provided in class.

Theorem 3.4 If f is of bounded variation on [a,b] and a < c < b, then

vf (a, b) = vf (a, c) + vf (c, b).

18
 Proof To be provided in class.

Theorem 3.5 If f ′ exist and is bounded on [a,b], then f is of bounded vari-

ation on [a,b].

Proof To be provided in class.

Remark:
Not every continuous function is of bounded variation. Not that in Theorem
3.5 f is continuous. The following example shows that a continuous function
does not need to be of bounded variation.

Example 3.6 Consider the function


 xsin π if 0 < x ≤ 2
x
f (x) =
 0 if x = 0

f is clearly continuous on [0,2], we show that f is not of bounded variation

on [0,2].

Solution To be provided in class.

We next discuss some basic properties of functions of bounded variation.

Theorem 3.7 If f and g are of bounded variation on [a,b], then f + g and

f g are of bounded variation on [a,b].

Proof To be provided in class.

Remark:
f − g is of bounded variation on [a,b] if f and g are of bounded variation on
[a,b].

19
Definition 3.8 Let f : [a, b] → ℜ be of bounded variation on [a,b]. Define
vf (x) = vf (a, x), (a ≤ x ≤ b). We call vf the total variation of f on [a,b].

Remark:
vf is monotonic increasing on [a,b] with vf (a) = 0.

Theorem 3.9 Let f : [a, b] → ℜ be of bounded variation on [a,b]. Then

1. vf (a, y) = vf (a, x) + vf (x, y) (a ≤ x ≤ y ≤ b).

2. If f is continuous on [a,b], so is vf .

Proof To be provided in class.

Theorem 3.10 Let f be of bounded variation on [a,b] and let v be defined

on [a,b] as follows v(x) = vf (a, x) if a < x ≤ b, v(a) = 0. Then

1. v is an increasing function on [a,b]

2. v − f is an increasing function on [a,b]

Proof Exercise.

Theorem 3.11 Let f be of bounded variation on [a,b].Then f can be ex-

pressed as the difference of two increasing functions.

Proof To be provided in class.

Theorem 3.12 Every monotone increasing function on [a,b] is of bounded

variation on [a,b].

Proof Exercise.

20
 Assignment 4:

1. Show that if f is of bounded variation on [a,b] and ϵ > 0, then there

exists p ∈ P[a,n] such that Wp (f ) > vf (a, b) − ϵ.

2. Is the function


 xcos π x if x ̸= 0
2
f (x) =
 0 if x = 0

of bounded variation on [0,1]?

4 Sequences of Functions:Pointwise and Uni-

form Convergence
In this section, we shall discuss two kinds of convergence of sequence of
real-valued functions defined on a subset D of ℜ- pointwise and uniform con-
vergence. Pointwise convergence is the natural extension of the convergence
of sequences and series of numbers, but it lacks many of important desirable
properties. The stronger notion of uniform convergence will be shown to
posses these properties. We begin with the following definition.

Definition 4.1 Let (fn ) be a sequnce of real-valued functions defined on a

subset D of ℜ. Then (fn ) is said to converge pointwise on D if for x ∈ D,
the sequence of number (fn (x)) converges. If (fn ) converges pointwise on D,
then we define f : D → ℜ by

f (x) = limn→∞ fn (x) f or every x ∈ D.

21
 The main interest in sequences of real-valued functions is the following:
If (fn ) converges to f, we would like the limit function f to enjoy some of
the properties of the individual functions, fn . For example, let (fn ) converges
pointwise to f. The following questions are of interest:

1. If fn is continuous for each n, is f necessarily continuous?

2. If fn is integrable for each n, is f necessarily integrable?

3. If fn is differentiable for each n, is f necessarily differentiable?

4. If fn is differentiable for each n, and f is also differentiable, does (fn′ )

converges to f ′ ?

5. If fn is integrable for each n, is it always true that

∫ ∫ ∫
limn→∞ fn (x)dx = limn→∞ fn (x)dx = f dx.
D D D

We give examples (in the class) to show that the answer to any of the
above five questions is no, if (fn ) converges only pointwise to f. There is
another mode of convergence under which all the above questions have affir-
mative answer under some additional condition. This mode of convergence
is called uniform convergence.

Definition 4.2 Let (fn ) be a sequence of functions defined on subset D of

ℜ. Then (fn ) is said to converge uniformly on D to f If for every ϵ > 0,
there exist N ∈ N such that |fn (x) − f (x)| < ϵ for every x ∈ D, and for every
n ≥ N.

22
 We recall that (fn ) converges to f on D pointwise if for each x ∈ D,
given ϵ > 0, there exists N = N (x, ϵ) such that |fn (x) − f (x)| < ϵ for every
n ≥ N (x, ϵ).
Remark:
The main difference between pointwise and uniform convergence is that for
pointwise convergence, the N ∈ N depends on ϵ and x, but for uniform
convergence, it is possible to find one N ∈ N depending only on ϵ that will
work for all x ∈ D.
Before given more examples, we give a useful criterion for testing if a
given sequence of functions converges uniformly. We begin with the following
definition and theorem.

Definition 4.3 [?] A sequence (fn ) of functions is called uniformly cauchy

on a set D if for any given ϵ > 0, there exists N ∈ N such that |fn (x) −
fm (x)| < ϵ for every x ∈ D, and for every n, m ≥ N.

Theorem 4.4 (Uniformly Cauchy Criterion) Let (fn ) be a sequence of func-

tions defined on a subset D of ℜ. Then, there exists a function f such that
(fn ) converges uniformly to f on D if and only if the following condition is
satisfied: for every ϵ > 0, there exists N ∈ N such that |fn (x) − fm (x)| < ϵ
for every x ∈ D, and for every n, m ≥ N.

Proof To be provided in class.

Theorem 4.5 Suppose (fn ) is a sequence of functions defined on a subset D

of ℜ. Then (fn ) converges uniformly to f on D if and only if limn→∞ βn = 0,
where βn = supx∈D |fn (x) − f (x)|.

23
 Proof Exercise.
We now give more examples.

Example 4.6 For each n ∈ N, define fn : [0, 1] → ℜ by

x
fn (x) = f or every x ∈ [0, 1], n ≥ 1.
3 + nx
Show that (fn ) converges uniformly on [0,1].

Solution To be provided in class.

xn
Example 4.7 Let fn (x) = 2+x
for x ∈ [0, 4].

1. Find the set D ⊂ [0, 4] for which f (x) = limn→∞ fn (x) is defined as a
real-valued function.

2. Show that if 0 < a < 1, the convergence is uniform on [0,a].

3. Show that the convergence is not uniform on [0,1].

To be provided in class.

Solution To be provided in class.

x2
Example 4.8 Let fn (x) = x2 +n
for x ∈ [0, ∞). Show that

1. f (x) = limn→∞ fn (x) = 0, for all x ∈ [0, ∞)

2. the convergence is not uniform on[0, ∞)

3. the convergence is uniform on [0,a) a ∈ ℜ.

Solution To be provided in class.

We finally give some consequences of uniform convergence of sequences
of real-valued functions.

24
Theorem 4.9 Let (fn ) be a sequence of integrable functions on [a,b].
Suppose fn → f uniformly on [a,b]. Then

1. f is integrable on [a,b]
∫b ∫b
2. limn→∞ a
fn (x)dx = a
f (x)dx.

Proof Exercise.

Theorem 4.10 Let (fn ) be a sequence of continuous real-valued functions on

[a,b]. and let f be a real-valued continuous function on [a,b]. Suppose (fn )
is monotone increasing to f, as n → ∞, for each x ∈ [a, b]. Then fn → f
uniformly on [a,b].

Proof Exercise.
Assignment 5

1. For each n ∈ N, let

n2 x
fn (x) = (x ∈ [0, 1].
1 + n2 x

(a) Show that (fn ) converges pointwise on [0,1].

(b) Is the convergence uniform? Justify your claim.

2. For each integer n ≥ 1, let

5
fn (x) = (x ∈ [0, 1].
5 + xn

(a) Show that (fn ) converges pointwise on [0,1].

(b) Is the convergence uniform? Justify your claim.

25