Numerical Analysis Module 2

Fundamentals of Vector Spaces

Sachin C. Patwardhan
Dept. of Chemical Engineering,
Indian Institute of Technology, Bombay
Powai, Mumbai, 400 076, Inda.
Email: sachinp@iitb.ac.in

Contents
1 Introduction

2 Vector Spaces

3 Normed Linear Spaces and Banach Spaces

10

4 Inner Product Spaces and Hilbert Spaces

16

5 Gram-Schmidt Process and Orthogonal Polynomials

22

6 Induced Matrix Norms

6.1 Computation of 2-norm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2 Other Matrix Norms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

27
28
29

7 Summary

30

8 Exercise

30

Introduction

When we begin to use the concept of vectors for formulating mathematical models for physical systems, we start with the concept of a vector in the three dimensional coordinate space.
From the mathematical viewpoint, the three dimensional space can be looked upon as a
set of objects, called vectors, which satisfy certain generic properties. While working with
mathematical modeling we need to deal with variety of such sets containing dierent types
of objects. It is possible to distill essential properties satised by all the vectors in the three
dimensional vector space and develop a more general concept of a vector space, which is a set
of objects that satisfy these generic properties. Such a generalization can provide a unied
view of problem formulations and the solution techniques. Generalization of the concept of
the vector and the three dimensional vector space to any general set is not su cient. To
work with these sets of generalized vectors, we also need to generalize various algebraic and
geometric concepts, such as magnitude of a vector, convergence of a sequence of vectors,
limit, angle between two vectors, orthogonality etc., on these sets. Understanding the fundamentals of vector spaces helps in developing a unied view of many seemingly dierent
numerical schemes. In this module, fundamentals of vector spaces are briey introduced. A
more detailed treatment of these topics can be found in Luenberger [2] and Kreyzig [1].
A word of advice before we begin to study these grand generalizations. While dealing with
the generalization of geometric notions in three dimensions to more general vector spaces,
it is di cult to visualize vectors and surfaces as we can do in the three dimensional vector
space. However, if you understand the geometrical concepts in the three dimensional space
well, then you can develop an understanding of the corresponding concept in any general
vector space. In short, it is enough to know your school geometry well. We are only building
qualitatively similar structures on the other sets.

Vector Spaces

The concept of a vector space will now be formally introduced. This requires the concept of
closure and eld.
Denition 1 (Closure) A set is said to be closed under an operation when any two elements
of the set subject to the operation yields a third element belonging to the same set.
Example 2 The set of integers is closed under addition, multiplication and subtraction.
However, this set is not closed under division.

Example 3 The set of real numbers (R) and the set of complex numbers (C) are closed
under addition, subtraction, multiplication and division.
Denition 4 (Field) A eld is a set of elements closed under addition, subtraction, multiplication and division.
Example 5 The set of real numbers (R) and the set of complex numbers (C) are scalar
elds. However, the set of integers is not a eld.
A vector space is a set of elements, which is closed under addition and scalar multiplication. Thus, associated with every vector space is a set of scalars F (also called as scalar eld
or coe cient eld) used to dene scalar multiplication on the space. In functional analysis,
the scalars will be always taken to be the set of real numbers (R) or complex numbers (C).
Denition 6 (Vector Space): A vector space X is a set of elements called vectors and
scalar eld F together with two operations. The rst operation is called addition which
associates with any two vectors x; y 2 X a vector x + y 2 X , the sum of x and y. The
second operation is called scalar multiplication, which associates with any vector x 2 X and
any scalar a vector x (a scalar multiple of x by ):
Thus, when X is a linear vector space, given any vectors x; y 2 X and any scalars
; 2 R; the element x + y 2 X. This implies that the well known parallelogram law
in three dimensions also holds true in any vector space. Thus, given a vector space X and
scalar eld F , the fallowing properties hold for any x; y; z 2 X and any scalars ; 2 F :
1. Commutative law: x + y = y + x
2. Associative law: x + (y + z) = (x + y) + z
3. There exists a null vector 0 such that
4. Distributive laws:
5.

x = 0 when

x + 0 = x for all x 2 X

(x + y) = x + y, ( + ) x = x + x and

= 0 and x = x when

(x) =

( x)

= 1:

6. For convenience 1x is dened as x and called as negative of a vector. We have

x + ( x) = 0; where 0 represents zero vector in X:
Example 7 (X Rn ; F R) : n
x 2 X can be expressed as

x=

dimensional real coordinate space. A typical element

h

x1 x2 ::::: xn

where xi denotes the ith element of the vector.

3

iT

Example 8 (X

C n; F

C) : n

dimensional complex coordinate space.

Example 9 (X Rn ; F C) : This combination of set X and scalar eld F does not form
a vector space. For any x 2 X and any 2 C the vector x 2X:
=
Example 10 (X l1 ; F R) : Set of all innite sequence of real numbers. A typical vector
x of this space has form x = ( 1 ; 2 ; :::::::::; k ; ::::::::):
Example 11 (X C[a; b]; F R) : Set of all continuous functions over an interval [a; b]
forms a vector space. We write x = y if x(t) = y(t) for all t 2 [a; b] The null vector 0 in
this space is a function which is zero every where on [a; b] ;i.e.
f (t) = 0 for all t 2 [a; b]
If x(t) and y(t) are vectors from this space and is real scalar, then (x + y)(t) = x(t) +
y(t) and ( x)(t) = x(t) are also elements of C[a; b]:
Example 12 X C (n) [a; b]; F R : Set of all continuous and n times dierentiable functions over an interval [a; b] forms a vector space.
Example 13 X set of all real valued polynomial functions dened on interval [a; b] together with F R forms a vector space.
Example 14 The set of all functions ff (t) : t 2 [a; b]g for which
Zb

jf (t)jp dt < 1

holds is a linear space Lp :

Example 15 (X Rm Rn ; F R) : Here we consider the set of all m n matrices with
real elements. It is easy to see that, if A; B 2 X; then A + B 2 X and X is a vector
space. Note that a vector in this space is a m n matrix and the null vector corresponds to
m n null matrix.
In three dimension, we often have to work with a line or a plane passing through the
origin, which form a subspace of the three dimensional space. The concept of a sub-space
can be generalized as follows.
Denition 16 (Subspace): A non-empty subset M of a vector space X is called subspace
of X if every vector x + y is in M wherever x and y are both in M: Every subspace always
contains the null vector, I.e. the origin of the space x:
4

Thus, the fundamental property of objects (elements) in a vector space is that they can
be constructed by simply adding other elements in the space. This property is formally
dened as follows.
(m)
Denition 17 (Linear Combination): A linear combination of vectors x(1) ; x(2)
; ; :::::::x
in a vector space is of the form 1 x(1) + 2 x(2) + :::::::::::::: + m x(m) where ( 1 ; ::: m ) are
scalars.

Note that we are dealing with set of vectors

x(k) : k = 1; 2; ::::::m:

(1)

The individual elements in the set are indexed using superscript (k). Now, if X = Rn and
x(k) 2 Rn represents kth vector in the set, then it is a vector with n components which are
represented as follows
iT
h
(k)
(k)
(2)
x(k) = x(k)
::::
x
x
n
2
1

Similarly, if X = l1 and x(k) 2 l1 represents kth vector in the set, then x(k) represents an
innite sequence with elements denoted as follows
h
iT
(k)
(k)
(k)
(3)
x = x1 :::: xi
::::::

Denition 18 (Span of Set of Vectors): Let S be a subset of vector space X. The set
generated by all possible linear combinations of elements of S is called as span of S and
denoted as [S]. Span of S is a subspace of X.
Denition 19 (Linear Dependence): A vector x is said to be linearly dependent upon a
set S of vectors if x can be expressed as a linear combination of vectors from S. Alternatively,
x is linearly dependent upon S if x belongs to the span of S; i.e. x 2 [S]. A vector is said to
be linearly independent of set S, if it not linearly dependent on S . A necessary and su cient
condition for the set of vectors x(1) ; x(2) ; :::::x(m) to be linearly independent is that expression
m
X

(i)
ix

=0

(4)

i=1

implies that

= 0 for all i = 1; 2::::::m:

Denition 20 (Basis): A nite set S of linearly independent vectors is said to be basis

for space X if S generates X i.e. X = [S]:
Example 21 Basis, Span and Sub-spaces
5

1. Two dimensional plane passing through origin of R3 . For example, consider the set S
of collection of all vectors
x = x(1) + x(2)
where ;

2 R are arbitrary scalars and

2
3
2 3
2
4
6
7
6 7
(1)
(2)
x = 4 1 5 and x = 4 0 5
0
1

i.e. S = span x(1) ; x(2) :This set denes a plane passing through origin in R3 : Note
that a plane which does not pass through the origin is not a sub-space. The origin must
be included in the set for it to qualify as a sub-space.
h
iT
2. Let S = fvg where v = 1 2 3 4 5
and let us dene span of S as [S] = v
where 2 R represents a scalar. Here, [S] is one dimensional vector space and subspace
of R5
3. Let S = v(1) ; v(2) where

v(1)

6
6
6
=6
6
6
4

1
2
3
4
5

3
7
7
7
7
7
7
5

; v(2)

6
6
6
=6
6
6
4

5
4
3
2
1

3
7
7
7
7
7
7
5

(5)

Here span of S (i.e. [S]) is two dimensional subspace of R5 .

4. Consider set of nth order polynomials on interval [0; 1]. A possible basis for this space
is
p(1) (z) = 1; p(2) (z) = z; p(3) (z) = z 2 ; ::::; p(n+1) (z) = z n
(6)
Any vector p(t) from this space can be expressed as
p(z) =
=

0p
0

(1)

(z) +
1z

1p

(2)

(z) + ::::::::: +

+ :::::::::: +

nz

np

(n+1)

(z)

(7)

Note that [S] in this case is (n + 1) dimensional subspace of C[a; b].

5. Consider set of continuous functions over interval, i.e. C[
for this space is

; ]: A well known basis

x(0) (z) = 1; x(1) (z) = cos(z); x(2) (z) = sin(z);

(8)

x(3) (z) = cos(2z); x(4) (z) = sin(2z); ::::::::

(9)

It can be shown that C[

; ] is an innite dimensional vector space.

6. The set of all symmetric real valued n n matrices is a subspace of the set of all real
valued n n matrices. This follows from the fact that matrix A + B is a real values
symmetric matrix for arbitrary scalars ; 2 R when AT = A and BT = B:
Example 22 Show that functions 1, exp(t), exp(2t), exp(3t) are linearly independent over
any interval [a,b].
Let us assume that vectors (1; et ; e2t ; e3t ) are linearly dependent i.e. there are constants
( ; ; ; ), not all equal to zero, such that
+ et + e2t + e3t = 0 holds for all t 2 [a; b]

(10)

Taking derivative on both the sides, the above equality implies

et ( + 2 et + 3 e2t ) = 0 holds for all t 2 [a; b]
Since et > 0 holds for all t 2 [a; b], the above equation implies that
+ 2 et + 3 e2t = 0 holds for all t 2 [a; b]

(11)

Taking derivative on both the sides, the above equality implies

et (2 + 6 et ) = 0 holds for all t 2 [a; b]
which implies that
2 + 6 et = 0 holds for all t 2 [a; b]
) et =

(12)

holds for all t 2 [a; b]

which is absurd. Thus, equality (12) holds only for = = 0 and vectors (1; et ) are linearly
independent on any interval [a; b]. With = = 0, equality (11) only when = 0 and
equality (10) holds only when = 0: Thus, vectors (1; et ; e2t ; e3t ) are linearly independent.
Example 23 Consider system of linear
2
1
6
Ax = 4 1
0

algebraic equations
32
3
0 1
x1
76
7
1 0 5 4 x2 5 = b
1 1
x3

Show that the set of all solutions of this equation for arbitrary vector b is same as R3 :
It is easy to see that matrix A has rank equal to three and columns (and rows) are linearly
independent. Since the columns are linearly independent, a unique solution x 2R3 can be
7

found for any arbitrary vector b 2 R3 : Now, let us nd a general solution x for an arbitrary
vector b by computing A 1 as follows
2
3
1=2
1=2
1=2
6
7
x = A 1 b = 4 1=2 1=2
1=2 5 b
1=2
1=2 1=2
2
3
2
3
2
3
1=2
1=2
1=2
6
7
6
7
6
7
= b1 4 1=2 5 + b2 4 1=2 5 + b3 4 1=2 5
1=2
1=2
1=2
= b1 v(1) + b2 v(2) + b3 v(3)

By denition
b1 v(1) + b2 v(2) + b3 v(3) 2 span v(1) ; v(2) ; v(3)
for an arbitrary b 2 R3 ; and, since vectors v(1) ; v(2) ; v(3) are linearly independent, we have
span v(1) ; v(2) ; v(3) = R3
i.e. set of all possible solutions x of the system of equations under considerations is identical
to the entire space R3 :
Example 24 Consider the ODE-BVP
d2 u(z)
+ 2 u(z) = 0
f or 0 < z < 1
2
dz
B:C:1 (at z = 0) : u(0) = 0
B:C:2 (at z = 1) : u(1) = 0
The general solution of this ODE-BVP, which satises the boundary conditions, is given by
u(z) =

sin( z) +

sin(2 z) +

sin(3 z) + ::: =

1
X

sin(i z)

i=1

where ( 1 ; 2 ; :::) 2 R are arbitrary scalars. The set of vectors fsin( z); sin(2 z); sin(3 z); :::g
is linearly independent and form a basis for C (2) [0; 1]; i.e. the set of twice dierentiable continuous functions in interval [0; 1] i.e.
C (2) [0; 1] = span fsin( z); sin(2 z); sin(3 z); :::g
Example 25 Consider system of linear algebraic equations
2
32
3 2 3
1
2
4
x1
0
6
76
7 6 7
Ax = 4 1
2 4 5 4 x2 5 = 4 0 5
2
4
8
x3
0
8

Show that solutions of this equation forms a two dimensional subspace of R3 :

It is easy to see that matrix A has rank equal to one and columns (and rows) are linearly
dependent. Thus, it is possible to obtain non-zero solutions to the above equation, which can
be re-written as follows
2
3
2
3
2
3
2 3
1
2
4
0
6
7
6
7
6
7
6 7
4 1 5 x1 + 4 2 5 x2 + 4 4 5 x3 = 4 0 5
2
4
8
0
Two possible solutions are

x(1)

3
2
6
7
=4 1 5
0

and

x(2)

3
4
6 7
=4 0 5
1

In fact, x(1) and x(2) and linearly independent and any linear combination of these two
vectors, i.e.
x = x(1) + x(2)
for any scalars ( ; ) 2 R satises
Ax = A

Ax(1) +

x(1) + x(2) =

Ax(2) = 0:

Thus, the solutions can be represented by a set S = span x(1) ; x(2) ; which forms a two
dimensional subspace of R3 :
Example 26 Consider a third order linear ordinary dierential equation
d2 u
du
d3 u
+
6
+ 11 + 6u = 0
2
2
dz
dz
dz
dened over C (3) [0; 1]; i.e. set of thrice dierentiable continuous functions over [0; 1]: Show
that the general solution of the ODE forms a 3 dimensional subspace of C (3) [0; 1]:
Roots of the characteristic polynomial i.e.
p3 + 6p2 + 11p + 6 = 0
are p =

1, p =

2 and p =

3. Thus, general solution of the ODE can be written as

u(z) = e

+ e

2z

+ e

3z

where ( ; ; ) 2 R are arbitrary scalars. Since vectors fe z ; e 2z ; e 3z g are linearly independent, the set of solutions can be represented as S = span fe z ; e 2z ; e 3z g ; which forms
a three dimensional sub-space of C (3) [0; 1]:
9

A vector space having nite basis (spanned by set of vectors with nite number of elements) is said to be nite dimensional. All other vector spaces are said to be innite
dimensional. We characterize a nite dimensional space by number of elements in a basis.
Any two basis for a nite dimensional vector space contain the same number of elements.
Let X and Y be two vector spaces. Then their product space, denoted by X Y , is an
ordered pair (x; y) such that x 2 X; y 2 Y: If z(1) = (x(1) ; y(1) ) and z(2) = (x(2) ; y(2) ) are
two elements of X Y; then it is easy to show that z(1) + z(2) 2 X Y for any scalars
( , ). Thus, product space is a linear vector space.
Example 27 Let X = C[a; b] and Y = R; then the product space X Y = C[a; b] R
forms a linear vector space. Such product spaces arise in the context of ordinary dierential
equations.

Normed Linear Spaces and Banach Spaces

In three dimensional space, we use lengths or magnitudes to compare any two vectors. Generalization of the concept of length / magnitude of a vector in three dimensional vector space
to an arbitrary vector space is achieved by dening a scalar valued function called norm of
a vector.
Denition 28 (Normed Linear Vector Space): A normed linear vector space is a vector
space X on which there is dened a real valued function which maps each element x 2 X
into a real number kxkcalled norm of x. The norm satises the fallowing axioms.
1. kxk

0 for all x 2 X ; kxk = 0 if and only if x =0 (zero vector)

2. kx + yk

kxk + kykfor each x; y 2 X: (triangle inequality).

3. k xk = j j : kxk for all scalars

and each x 2 X

Example 29 Vector norms:

1. (Rn ; k:k1 ) :Euclidean space Rn with 1-norm: kxk1 =
2. (Rn ; k:k2 ) :Euclidean space Rn with 2-norm:
kxk2 =

"

10

N
X
i=1

(xi )2

N
P

i=1

# 21

jxi j

3.

Rn ; k:kp :Euclidean space Rn with p-norm:

kxkp =

"

N
X
i=1

jxi jp

# p1

(13)

where p is a positive integer

4. (Rn ; k:k1 ) :Euclidean space Rn with 1 norm: kxk1 = max jxi j
5. n-dimensional complex space (C n ) with p-norm:
kxkp =

"

N
X
i=1

jxi jp

# p1

(14)

; where p is a positive integer

6. Space of innite sequences (l1 ) with p-norm: An element in this space, say x 2 l1 , is
an innite sequence of numbers
x = fx1; x2 ; ::::::::; xn ; ::::::::g

(15)

such that p-norm dened as

kxkp =

"

1
X
i=1

jxi jp

# p1

<1

(16)

is bounded for every x 2 l1 ; where p is an integer.

7. (C[a; b]; kx(t)k1 ) : The normed linear space C[a; b] together with innite norm
kx(t)k1 =

max
a t

jx(t)j

(17)

It is easy to see that kx(t)k1 dened above qualies to be a norm

max jx(t) + y(t)j

max[jx(t)j + jy(t)j]

max jx(t)j + max jy(t)j

max j x(t)j = max j j jx(t)j = j j max jx(t)j

(18)
(19)

8. Other types of norms, which can be dened on the set of continuous functions over
[a; b] are as follows
Zb
kx(t)k1 = jx(t)j dt
(20)
a

11

3 21
2b
Z
kx(t)k2 = 4 jx(t)j2 dt5

(21)

Example 30 Determine whether(a) max jdf (t)=dtj (b) max jx(t)j + max jx0 (t)j (c) jx(a)j +
max jx0 (t)j and (d) jx(a)j max jx(t)j can serve as a valid denitions for norm in C(2) [a; b]:
Solution: (a) max jdf (t)=dtj : For this to be a norm function, Axiom 1 in the denition
of the normed vector spaces requires
kf (t)k = 0 ) f (t) is the zero vector in C(2) [a; b] i.e. f (t) = 0 for all t 2 [a; b]
However, consider the constant function i.e. g(t) = c for all t 2 [a; b] where c is some
non-zero value. It is easy to see that
max jdg(t)=dtj = 0
even when g(t) does not correspond to the zero vector. Thus, the above function violates
Axiom 1 in the denition of a normed vector space and, consequently, cannot qualify as a
norm.
(b) max jx(t)j + max jx0 (t)j : For any non-zero function x(t) 2 C(2) [a; b], Axiom 1 is
satised. Axiom 2 follows from the following inequality
kx(t) + y(t)k = max jx(t) + y(t)j + max jx0 (t) + y0 (t)j

[max jx(t)j + max jy(t)j] + [max jx0 (t)j + max jy0 (t)j]

[max jx(t)j + max jx0 (t)j] + [max jy(t)j + max jy0 (t)j]

kx(t)k + ky(t)k
It is easy to show that Axiom 3 is also satised for all scalars : Thus, given function
denes a norm on C (2) [a; b]
(c) jx(a)j + max jx0 (t)j : For any non-zero function x(t) 2 C(2) [a; b], Axiom 1 is satised.
Axiom 2 follows from the following inequality
kx(t) + y(t)k = jx(a) + y(a)j + max jx0 (t) + y0 (t)j

[jx(a)j + jy(a)j] + [max jx0 (t)j + max jy0 (t)j]

[jx(a)j + max jx0 (t)j] + [jy(a)j + max jy0 (t)j]

kx(t)k + ky(t)k
Axiom A3 is also satised for any

as

k x(t)k = j x(a)j + max j x0 (t)j

= j j [jx(a)j + max jx0 (t)j]

= j j : kxk
12

(d) jx(a)j max jx(t)j : Consider a non-zero function x(t) in C (2) [a; b] such that x(a) = 0
and max jx(t)j 6= 0: Then, Axiom 1 is not satised for all vector x(t) 2 C (2) [a; b] and the
above function does not qualify to be a norm on C (2) [a; b]:
In a normed linear space X; the set of all vectors x 2X such that kx xk 1 is called
unit ball centered at x:A unit ball in (R2 ; k:k2 ) is the set of all vectors in the circle with the
origin at the center and radius equal to one while a unit ball in (R3 ; k:k2 ) is the set of all
points in the unit sphere with the origin at the center. Schematic representation of a unit
ball in C[0,1] when maximum norm is used is shown in Figure 1. The unit ball in C[0,1] is
set of all functions f(z) such that jf (z)j 1 where z 2 [0; 1]:

Figure 1: Schematic representation of a unit ball in C[0,1]

Once we have dened a norm in a vector space, we can proceed to generalize the concept
of convergence of a sequence of vector. Concept of convergence is central to all iterative
numerical methods.
Denition 31 (Cauchy sequence): A sequence x(k) in normed linear space is said to
be a Cauchy sequence if x(n) x(m) ! 0 as n; m ! 1:i.e. given an " > 0 there exists an
integer N such that x(n) x(m) < " for all n; m N:
Denition 32 (Convergence): In a normed linear space an innite sequence of vectors
x(k) : k = 1; 2; ::::::: is said to converge to a vector x if the sequence x
x(k) ; k = 1; 2; :::
of real numbers converges to zero. In this case we write x(k) ! x :
13

In particular, a sequence x(k) in Rn converges if and only if each component of the

vector sequence converges. If a sequence converges, then its limit is unique.
Example 33 Convergent sequences: Consider the sequence of vectors represented as

(k)

6
6
=6
4

1 + (0:2)k
1 + (0:9)k
3= 1 + ( 0:5)k
(0:8)k

7
6 1
7
6
7!6
5
4 3
0

3
7
7
7
5

(22)

for k = 0, 1, 2,.... is a convergent sequence with respect to any p-norm dened on R4 : It can
be shown that it is a Cauchy sequence. Note that each element of the vector converges to a
limit in this case.
Every convergent sequence is a Cauchy sequence. Moreover, when we are working in
Rn or C n ;all Cauchy sequences are convergent. However, all Cauchy sequences in a general
vector space need not be convergent. Cauchy sequences in some vector spaces exhibit such
strange behavior and this motivates the concept of completeness of a vector space.
Denition 34 (Banach Space): A normed linear space X is said to be complete if every
Cauchy sequence has a limit in X. A complete normed linear space is called Banach space.
Examples of Banach spaces are
(Rn ; k:k1 ) ; (Rn ; k:k2 ) ; (Rn ; k:k1 )
(Cn ; k:k1 ) ; (Cn ; k:k2 ) ; (l1 ; k:k1 ) ; (l1 ; k:k2 )
Concept of Banach spaces can be better understood if we consider an example of a vector
space where a Cauchy sequence is not convergent, i.e. the space under consideration is an
incomplete normed linear space. Note that, even if we nd one Cauchy sequence in this
space which does not converge, it is su cient to prove that the space is not complete.
Example 35 Let X = (Q; k:k1 ) i.e. set of rational numbers (Q) with scalar eld also as the
set of rational numbers (Q) and norm dened as
kxk1 = jxj

(23)

A vector in this space is a rational number. In this space, we can construct Cauchy sequences
which do not converge to a rational numbers (or rather they converge to irrational numbers).

14

For example, the well known Cauchy sequence

x(1) = 1=1
x(2) = 1=1 + 1=(2!)
:::::::::
x

(n)

= 1=1 + 1=(2!) + ::::: + 1=(n!)

converges to e, which is an irrational number. Similarly, consider sequence

x(n+1) = 4

(1=x(n) )

Starting from initial point x(0) = 1; we can generate the sequence of rational numbers
3=1; 11=3; 41=11; ::::
p

which converges to 2 + 3 as n ! 1:Thus, limits of the above sequences is outside the space
X and the space is incomplete.
Example 36 Consider sequence of functions in the space of twice dierentiable continuous
functions C (2) ( 1; 1)
1 1
f (k) (t) = + tan 1 (kt)
2
dened in interval 1 < t < 1; for all integers k. The range of the function is (0,1). As
k ! 1; the sequence of continuous function converges to a discontinuous function
u( ) (t) = 0

1<t<0

= 1

0<t<1

Example 37 Let X = (C[0; 1]; k:k1 ) i.e. space of continuous function on [0; 1] with one
norm dened on it i.e.
Z1
kx(t)k1 = jx(t)j dt
(24)
0

and let us dene a sequence [2]

8
>
< 0
(n)
x (t) =
n(t
>
:
1

(0
1
) + 1 ( 12
2
(t

( 21
1
) t
n
1
)
2
t

1
)
n
1
)
2

9
>
=
>
;

(25)

Each member is a continuous function and the sequence is Cauchy as

x(n)

x(m) =

1 1
2 n

1
!0
m

(26)

However, as can be observed from Figure 2, the sequence does not converge to a continuous
function.
15

Figure 1: Figure 2: Sequence of continuous functions

The concepts of convergence, Cauchy sequences and completeness of space assume importance in the analysis of iterative numerical techniques. Any iterative numerical method
generates a sequence of vectors and we have to assess whether the sequence is Cauchy to
terminate the iterations. To a beginner, it may appear that the concept of incomplete vector
space does not have much use in practice. It may be noted that, when we compute numerical
solutions using any computer, we are working in nite dimensional incomplete vector spaces.
In any computer with nite precision, any irrational number such as or e; is approximated
by an rational number due to nite precision. In fact, even if we want to nd a solution
in Rn ; while using a nite precision computer to compute a solution, we actually end up
working in Qn and not in Rn :

Inner Product Spaces and Hilbert Spaces

Similar to magnitude / length of a vector, another important concept in three dimensional

space that needs to be generalized is angle between any two vectors. Given any two unit
b and y
b;the angle between these two vectors is dened using inner (or dot)
vectors in R3 , say x

16

product of two vectors as

T

x
y
kxk2
kyk2
= x
b1 yb1 + x
b2 yb2 + x
b3 yb3

b=
cos( ) = (b
x)T y

(27)
(28)

The fact that cosine of angle between any two unit vectors is always less than one can be
stated as
bij 1
jcos( )j = jhb
x; y
(29)

Moreover, vectors x and y are called orthogonal if (x)T y = 0: Orthogonality is probably

the most useful concept while working in three dimensional Euclidean space. Inner product
spaces and Hilbert spaces generalize these simple geometrical concepts in three dimensional
Euclidean space to higher or innite dimensional vector spaces.
Denition 38 (Inner Product Space): An inner product space is a linear vector space
X together with an inner product dened on X X. Corresponding to each pair of vectors
x; y 2 X the inner product hx; yi of x and y is a scalar. The inner product satises following
axioms.
1. hx; yi = hy; xi (complex conjugate)
2. hx + y; zi = hx; zi + hy; zi
3. h x; yi =

hx; yi

hx; yi =

hx; yi

4. hx; xi

0 and hx; xi = 0 if and only if x = 0:

Denition 39 (Hilbert Space): A complete inner product space is called as an Hilbert

space.
Here are some examples of commonly used inner product and Hilbert spaces.
Example 40 Inner Product Spaces
1. X

Rn with inner product dened as

n
X
hx; yi = x y = xi yi
T

(30)

i=1

hx; xi =
is a Hilbert space.

n
X
i=1

17

(xi )2 = kxk22

(31)

2. X

Rn with inner product dened as

hx; yiW = xT W y

(32)

where W is a positive denite matrix is a Hilbert space. The corresponding 2-norm is

p
p
dened as kxkW;2 = hx; xiW = xT W x

3. X

C n with inner product dened as

hx; yi =
hx; xi =
is a Hilbert space.

n
X

n
X

xi xi =

i=1

xi yi

(33)

i=1

n
X
i=1

jxi j2 = kxk22

(34)

4. The set of real valued square integrable functions on interval [a; b] with inner product
dened as
Zb
hx; yi = x(t)y(t)dt
(35)
a

is an Hilbert space and denoted as L2 [a; b]: Well known examples of spaces of this type
are the set of continuous functions on L2 [ ; ] or L2 [0; 2 ], which are considered while
developing Fourier series expansions of continuous functions on [ ; ] or [0; 2 ] using
sin(n ) and cos(n ) as basis functions.
5. Space of polynomial functions on [a; b]with inner product

hx; yi =

Zb

x(t)y(t)dt

(36)

is a inner product space. This is a subspace of L2 [a; b]:

6. Space of complex valued square integrable functions on [a; b] with inner product

hx; yi =

Zb

is an inner product space:

18

x(t)y(t)dt

(37)

Axioms 2 and 3 imply that the inner product is linear in the rst entry. The quantity
1
hx; xi 2 is a candidate function for dening norm on the inner product space: Axioms 1 and
3 imply that k xk = j j kxk and axiom 4 implies that kxk > 0 for x 6= 0: If we show that
p
p
hx; xisatises triangle inequality, then hx; xi denes a norm on space X . We rst prove
Cauchy-Schwarz inequality, which is generalization of equation (29), and proceed to show
p
p
that hx; xi denes the well known 2-norm on X; i.e. kxk2 = hx; xi.
Lemma 41 (Cauchey- Schwarz Inequality): Let X denote an inner product space. For
all x; y 2 X ,the following inequality holds
[hx; xi]1=2 [hy; yi]1=2

jhx; yij

(38)

The equality holds if and only if x = y or y = 0

Proof: If y = 0, the equality holds trivially so we assume y 6= 0: Then, for all scalars
;we have
0

hx

y; x

In particular, if we choose
we have

yi = hx; xi

hy; xi + j j2 hy; yi

(39)

hy; xi
; then, using axiom 1 in the denition of inner product,
hy; yi
=

hx; yi

hy; xi
hx; yi
=
hy; yi
hy; yi

hx; yi

hy; xi =

2 hx; yi hx; yi
=
hy; yi
)0

hx; xi

or j hx; yij

2 hx; yi hy; xi
hy; yi

2 jhx; yij2
hy; yi
jhx; yij2
hy; yi

(40)

(41)
(42)

(43)

hx; xi hy; yi

The triangle inequality can be can be established easily using the Cauchy-Schwarz inequality as follows

hx + y; x + yi = hx; xi + hx; yi + hy; xi + hy; yi :

(44)

hx; xi + 2 jhx; yij + hy; yi

p
hx; xi + 2 hx; xi hy; yi + hy; yi

(45)

19

(46)

p
p
p
hx + y; x + yi
hx; xi + hy; yi
(47)
p
Thus, the candidate function hx; xi satises all the properties necessary to dene a norm,
i.e.
p
hx; xi

p
0 8 x 2 X and hx; xi = 0 if f x = 0
p
p
h x; xi = j j hx; xi
p
p
p
hx + y; x + yi
hx; xi + hy; yi
(Triangle inequality)

(48)
(49)
(50)

p
Thus, the function hx; xi indeed denes a norm on the inner product space X: In fact the
inner product denes the well known 2-norm on X; i.e.
kxk2 =
and the triangle inequality can be stated as
kx + yk22

hx; xi

kxk22 + 2 kxk2 : kyk2 + kyk22 : = [kxk2 + kyk2 ]2

or kx + yk2

Denition 42 (Angle) The angle

dened by

kxk2 + kyk2

(51)

(52)
(53)

between any two vectors in an inner product space is

= cos

hx; yi
kxk2 kyk2

(54)

Denition 43 (Orthogonal Vectors): In a inner product space X two vector x; y 2 X are

said to be orthogonal if hx; yi = 0:We symbolize this by x?y: A vector x is said to be
orthogonal to a set S (written as x?S) if x?z for each z 2 S:
Just as orthogonality has many consequences in three dimensional geometry, it has many
implications in any inner-product / Hilbert space [2]. The Pythagoras theorem, which is
probably the most important result the plane geometry, is true in any inner product space.
Lemma 44 If x?y in an inner product space then kx + yk22 = kxk22 + kyk22 .
Proof: kx + yk22 = hx + y; x + yi = kxk22 + kyk22 + hx; yi + hy; xi :
Denition 45 (Orthogonal Set): A set of vectors S in an inner product space X is said
to be an orthogonal set if x?y for each x; y 2 S and x 6= y: The set is said to be orthonormal
if, in addition each vector in the set has norm equal to unity.

20

Note that an orthogonal set of nonzero vectors is linearly independent set. We often prefer
to work with an orthonormal basis as any vector can be uniquely represented in terms of
components along the orthonormal directions. Common examples of such orthonormal basis
are (a) unit vectors along coordinate directions in Rn (b) function fsin(nt) : n = 1; 2; :::g
and fcos(nt) : n = 1; 2; :::g in L2 [0; 2 ]:
Example 46 Show that function hx; yiW : Rn

Rn ! R dened as

hx; yiW = xT W y
denes an inner product on when W is a symmetric positive denite matrix.
Solution: For hx; yiW = xT W y to qualify as inner product, it must satisfy the following
all four axioms in the denition of the inner product. We have,
hx; yiW = xT W y and hy; xiW = yT W x
Since W is symmetric, i.e.
W T = W , xT W y

= yT W T x = yT W x

Thus, axiom A1 holds for any x; y 2 Rn :

hx + y; ziW = (x + y)T W z = xT W z + xT W z = hx; ziW + hy; ziW
Thus, axiom A2 holds for any x; y; z 2 Rn :
h x; yi = ( x)T W y = (xT W y) = hx; yi

hx; yi = xT W ( y) = (xT W y) = hx; yi

Thus, axiom A3 holds for any x; y 2 Rn : Since W is positive denite, it follows that
hx; xiW = xT W x > 0 if x 6= 0 and hx; xiW = xT W x = 0 if x = 0: Thus, axiom A4 holds
for any x 2 Rn : Since all four axioms are satised, hy; xiW = yT W x is a valid denition of
an inner product.
Example 47 The triangle inequality asserts that, for any two vectors x and y belonging to
an inner product space
kx + yk2 jjyjj2 +jjxjj2
Does the Cauchy-Schwartz inequality follow from the triangle inequality? Under what condition Schwartz inequality becomes an equality?
Solution: Squaring both the sides, we have
kx + yk22 = hx + y; x + yi
21

[jjyjj2 +jjxjj2 ]2

jjyjj22 +jjxjj22 + 2jjyjj2 jjxjj2

hx; xi + hy; yi + 2 hx; yi

jjyjj22 +jjxjj22 + 2 hx; yi
Since, jjyjj22 + jjxjj22

jjyjj22 +jjxjj22 + 2jjyjj2 jjxjj2

0 for any x; y 2 X; the above inequality reduces to

hx; yi

jjyjj2 jjxjj2

(55)

The triangle inequality also implies that

kx

yk22 = hx

hx; xi + hy; yi
jjyjj22 +jjxjj22
Since, jjyjj22 + jjxjj22

y; x

yi

[jjyjj2 +jjxjj2 ]2

jjyjj22 +jjxjj22 + 2jjyjj2 jjxjj2

2 hx; yi

jjyjj22 +jjxjj22 + 2jjyjj2 jjxjj2

2 hx; yi

0 for any x; y 2 X; the above inequality reduces to

hx; yi

jjyjj2 jjxjj2

i.e.
jjyjj2 jjxjj2

hx; yi

(56)

Combining inequalities (55) and (56), we arrive at the Cauchy-Schwartz inequality

jjyjj2 jjxjj2

hx; yi

jjyjj2 jjxjj2

(57)

i.e.
jhx; yij

jjyjj2 jjxjj2

(58)

The Cauchy-Schwartz inequality reduces to equality when y = x:

Gram-Schmidt Process and Orthogonal Polynomials

Given any linearly independent set in an inner product space, it is possible to construct
an orthonormal set. This procedure is called Gram-Schmidt procedure. Consider a linearly
independent set of vectors x(i) ; i = 1; 2; 3:::::n in a inner product space we dene e(1) as
e(1) =
We form unit vector e(2) in two steps.
z(2) = x(2)

x(1)
kx(1) k2

(59)

x(2) ; e(1) e(1)

(60)

22

where x(2) ; e(1) is component of x(2) along e(1) :

e(2) =

z(2)
kz(2) k2

(61)

:By direct calculation it can be veried that e(1) ?e(2) : The remaining orthonormal vectors
e(i) are dened by induction. The vector z(k) is formed according to the equation
(k)

(k)

=x

k 1
X

x(k) ; e(i) :e(i)

(62)

k = 1; 2; :::::::::n

(63)

i=1

and
e(k) =

z(k)
kz(k) k2

It can be veried by direct computation that z(k) ?e(j) for all j < k as follows
(k)

z ;e

(j)

(k)

x ;e

k 1
X

(j)

i=1
(k)

x(k) ; e(j)

x(k) ; e(i) : e(i) ; e(j)

x ; e(j) = 0

(64)
(65)

Example 48 Gram-Schmidt Procedure in R3 : Consider X = R3 with hx; yi = xT y:

Given a set of three linearly independent vectors in R3
2 3
2 3
2 3
2
1
1
6 7
6 7
6 7
(2)
(1)
(3)
(66)
x = 4 0 5; x = 4 0 5; x = 4 1 5
0
0
1
we want to construct and orthonormal set. Applying Gram Schmidt procedure,
2 1 3
p

(1)

2
x(1)
6
7
= (1) : = 4 0 5
kx k2
1
p

z(2) = x(2)
x(2) ; e(1) :e(1)
2 3
2 1 3 2
1
p
1
2
2
1
6 7
7 6
6
= 4 0 5 p 4 0 5=4 0
2
p1
0
2
2

(2)

e(2) =

p1
2

z
6
:=4 0
kz(2) k2
23

(67)

p1
2

3
7
5

1
2

(68)

7
5
(69)

x(3) ; e(2)
z(3) = x(3)
x(3) ; e(1) :e(1)
2 3
2 1 3
2 1
p
p
2
2
2
6 7 p 6
7 p 6
= 4 1 5
24 0 5
24 0
p1
p1
0
2
2
e(3) =

h
iT
z(3)
:
=
0
1
0
kz(3) k2

:e(2)
3 2

3
0
7 6 7
5=4 1 5
0

(70)

Note that the vectors in the orthonormal set will depend on the denition of inner product.
Suppose we dene the inner product as follows
hx; yiW = xT W y
2
3
2
1 1
6
7
W =4 1 2
1 5
1
1 2

(71)

where W is a positive denite matrix. Then, length of x(1)

b
e(1) becomes
2 1 3

W;2

6 and the unit vector

b
e(1)

6
x(1)
7
6
= (1)
:=4 0 5
kx kW;2
1
p

(72)

The remaining two orthonormal vectors have to be computed using the inner product dened
by equation 71.
Example 49 Gram-Schmidt Procedure in C[a,b]: Let X represent set of continuous
functions on interval 1 t 1 with inner product dened as
hx(t); y(t)i =

Z1

(73)

x(t)y(t)dt

Given a set of four linearly independent vectors

x(1) (t) = 1; x(2) (t) = t;

x(3) (t) = t2 ;

x(4) (t) = t3

(74)

we intend to generate an orthonormal set. Applying Gram-Schmidt procedure

e(1) (t) =

(1)

x(1) (t)
1
=p
(1)
kx (t)k
2

(75)

Z1

(76)

(2)

e (t); x (t) =

24

t
dt = 0
2

x(2) ; e(1) :e(1) = t = x(2) (t)

z(2) (t) = t

e(2) =
2

(2)

z (t)

Z1

(77)

z(2)
kz(2) k

t3
t dt =
3

(78)
1

=
1

2
3

(79)

2
z (t) =
3
r
3
:t
e(2) (t) =
2
(2)

(80)
(81)

x(3) (t); e(2) (t) :e(2) (t)

x(3) (t); e(1) (t) :e(1) (t)
1
0r 1
01
1
Z
Z
1 @ 2 A (1)
3 3 A (2)
t dt e (t) @
t dt e (t)
2
2

z(3) (t) = x(3) (t)

= t2

= t2

1
3

1
3

0 = t2

e(3) (t) =

where

(3)

z (t)

(3)

z(3) (t)
kz(3) (t)k
(3)

z (t); z (t) =

(82)
(83)

Z1

1
3

(84)

dt

Z1

2 2 1
t +
3
9

t5
dt =
5

2t3
t
+
9
9

1
1

4 2
18 10
8
+ =
=
9 9
45
45
r
r
8
2 2
z(3) (t) =
=
(85)
45
3 5
The orthonormal polynomials constructed above are well known Legandre polynomials. It
turns out that
r
2n + 1
pn (t) ; (n = 0; 1; 2:::::::)
(86)
en (t) =
2
where
( 1)n dn
n
Pn (t) = n
1 t2
(87)
n
2 n! dt
=

2
3

25

are Legandre polynomials. It can be shown that this set of polynomials forms a orthonormal
basis for the set of continuous functions on [-1,1]. First few elements in this orthogonal set
are as follows
P0 (t) = 1;
P4 (t) =

1
1
P2 (t) = (3t2 1); P3 (t) = (5t3
2
2
1
30t2 + 3); P5 (t) = (63t5 70t3 + 15t)
8

P1 (t) = t;

1
(35t4
8

3t)

Example 50 Gram-Schmidt Procedure in other Spaces

1. Shifted Legandre polynomials: X = C[0; 1] and inner product dened as
hx(t); y(t)i =

Z1

(88)

x(t)y(t)dt

These polynomials are generated starting from linearly independent vectors

x(1) (t) = 1; x(2) (t) = t; x(3) (t) = t2 ;

x(4) (t) = t3

(89)

and applying Gram-Schmidt process.

2. Hermite Polynomials: X
L2 ( 1; 1); i.e. space of continuous functions over
( 1; 1) with 2 norm dened on it and
hx(t); y(t)i =

Z1

(90)

x(t)y(t)dt

Apply Gram-Schmidt to the following set of vectors in L2 ( 1; 1)

t2
) ; x(2) (t) = tx(1) (t) ;
2
x(3) (t) = t2 x(1) (t) ; ::::::x(k) (t) = tk 1 x(1) (t) ; ::::
x(1) (t) = exp(

(91)
(92)

First few elements in this orthogonal set are as follows

H0 (t) = 1;
H4 (t) = 16t4

H1 (t) = 2t;
48t2 + 12;

H2 (t) = 4t2

2; H3 (t) = 5t3

H5 (t) = 32t5

12t

160t3 + 120t

3. Laguerre Polynomials: X
L2 (0; 1); i.e. space of continuous functions over
(0; 1) with 2 norm dened on it and
hx(t); y(t)i =

Z1
0

26

x(t)y(t)dt

(93)

Apply Gram-Schmidt to the following set of vectors in L2 (0; 1)

t
) ; x(2) (t) = tx(1) (t) ;
2
x(3) (t) = t2 x(1) (t) ; ::::::x(k) (t) = tk 1 x(1) (t) ; ::::
x(1) (t) = exp(

(94)
(95)

The rst few Laguerre polynomials are as follows

L0 (t) = 1 ; L1 (t) = 1 t ; L2 (t) = 1 2t + (1=2)t2
3
2 3
1 3
1
L3 (t) = 1 3t + t2
t ; L4 (t) = 1 4t + 3t2
t + t4
2
6
3
24

Induced Matrix Norms

We have already mentioned that set of all m n matrices with real entries (or complex
entries) can be viewed a linear vector space. In this section, we introduce the concept of
induced norm of a matrix, which plays a vital role in the numerical analysis. A norm of
a matrix can be interpreted as amplication power of the matrix. To develop a numerical
measure for ill conditioning of a matrix, we rst have to quantify this amplication power of
the matrix.
Denition 51 (Induced Matrix Norm): The induced norm of a m
dened as mapping from Rm Rn ! R+ such that
kAk =

M ax kAxk
x 6= 0 kxk

n matrix A is

(96)

In other words, kAk bounds the amplication power of the matrix i.e.
kAxk
kxk

kAk for all x 2 Rn ; x 6= 0

(97)

The equality holds for at least one non zero vector x 2 Rn . An alternate way of dening
matrix norm is as follows
M ax
kAk =
kAb
xk
(98)
kb
xk = 1
b as
Dening x

b=
x

x
kxk

it is easy to see that these two denitions are equivalent. The following conditions are
satised for any matrices A; B 2 Rm Rn
27

1. kAk > 0 if A 6= [0] and k[0]k = 0

2. k Ak = j j.kAk
3. kA + Bk
4. kABk

kAk + kBk
kAk kBk

The induced norms, i.e. norm of matrix induced by vector norms on Rm and Rn , can be
interpreted as maximum gain or amplication factor of the matrix.

6.1

Computation of 2-norm

Now, consider 2-norm of a matrix, which can be dened as follows

jjAjj2 =

max jjAxjj2
x 6= 0 jjxjj2

(99)

Squaring both sides

jjAjj22 =

max xT Bx
max (Ax)T (Ax)
=
(xT x)
x 6= 0 (xT x)
x 6= 0

where B = AT A is a symmetric and positive denite matrix. Positive deniteness of matrix

B requires that
xT Bx > 0 if x 6= 0 and xT Bx = 0 if and only if x = 0

(100)

If columns of A are linearly independent, then it implies that

xT Bx = (Ax)T (Ax) > 0 if x 6= 0
= 0

if x = 0

(101)
(102)

Now, a positive denite symmetric matrix can be diagonalized as

T

B=

(103)

Where is matrix with eigen vectors as columns and is the diagonal matrix with eigenvalues of B (= AT A) on the main diagonal. Note that in this case
is unitary matrix
,i.e.,
T
1
= I i.e T =
(104)

28

and eigenvectors are orthogonal. Using the fact that

T

xT x = xT

is unitary, we can write

x = yT y

(105)

This implies that

yT y
xT Bx
=
(xT x)
(yT y)
where y =

x: Suppose eigenvalues
0<

i of

(106)

AT A are numbered such that

::::::

(107)

Then, we have
yT y
( 1 y12 + 2 y22 + :::::: + n yn2 )
=
(yT y)
(y12 + y22 + :::::: + yn2 )

(108)

which implies that

xT Bx
xT (AT A)x
yT y
= T
=
(yT y)
(x x)
(xT x)

(109)

The equality holds only at the corresponding eigenvector of AT A, i.e.,

v(n)

(AT A)v(n)

[v(n) ] v(n)

v(n)

nv

(n)

[v(n) ] v(n)

(110)

Thus, 2 norm of matrix A can be computed as follows

jjAjj22 =

max
jjAxjj2 =jjxjj2 =
x 6= 0

T
max (A A)

(111)

i.e.
jjAjj2 = [
where

6.2

T
max (A A)

T
1=2
max (A A)]

(112)

denotes maximum magnitude eigenvalue or the spectral radius of AT A.

Other Matrix Norms

Other commonly used matrix norms are

1-norm: Maximum over column sums
max
jjAjj1 =
1 j

"

"

1 norm: Maximum over row sums

jjAjj1

max
=
1 i
29

n
X
i=1

n
X
j=1

(113)

(114)

jaij j

jaij j

Remark 52 There are other matrix norms, such as Frobenious norm, which are not induced
matrix norms. Frobenious norm is dened as
jjAjjF =

"

n X
n
X
i=1 j=1

jaij j2

#1=2

Summary

In this chapter, some fundamental concepts from functional analysis have been reviewed.
We begin with the concept of a general vector space and dene various algebraic and geometric structures like norm and inner product. We then move to dene inner product, which
generalizes the concept of dot product, and angle between vectors. We also interpret the
notion of orthogonality in a general inner product space and develop Gram-Schmidt process,
which can generate an orthonormal set from a linearly independent set. Denition of inner
product and orthogonality paves the way to generalize the concept of projecting a vector
onto any sub-space of an inner product space. In the end, we discuss induced matrix norms,
which play an important role in the analysis of numerical schemes.

Exercise
1. While solving problems using a digital computer, arithmetic operations can be performed only with a limited precision due to nite word length. Consider the vector
space X R and discuss which of the laws of algebra (associative, distributive, commutative) are not satised for the oating point arithmetic in a digital computer.
2. Show that the solution of the dierential equation
d2 x
+x=0
dt2
is a linear space. What is the dimension of this space?
3. Show that functions 1, exp(t), exp(2t), exp(3t) are linearly independent over any interval [a,b].
4. Does the set of functions of the form
f (t) = 1=(a + bt)
constitute a linear vector space?
30

5. Give an example of a function which is in L1 [0; 1] but not in L2 [0; 1]:

6. Decide linear dependence or independence of
(a) (1,1,2), (1,2,1), (3,1,1)
(b) x(1)

x(2) ; x(2)

x(3) ; x(3)

x(4) ; x(4)

x(1) for any x(1) ; x(2) ; x(3) ; x(4)

(c) (1,1,0), (1,0,0), (0,1,1), (x,y,z) for any scalars x,y,z

7. Describe geometrically the subspaces of R3 spanned by following sets
(a) (0,0,0), (0,1,0), (0,2,0)
(b) (0,0,1), (0,1,1), (0,2,0)
(c) all six of these vectors
(d) set of all vectors with positive components
8. Consider the space X of all n n matrices. Find a basis for this vector space and show
that set of all lower triangular n n matrices forms a subspace of X:
9. Determine which of the following denitions are valid as denitions for norms in
C(2) [a; b]
(a) max jx(t)j + max jx0 (t)j

(b) max jx0 (t)j

(c) jx(a)j + max jx0 (t)j

(d) jx(a)j max jx(t)j

10. In a normed linear space X the set of all vectors x 2X such that kx xk
unit ball centered at x:

1 is called

(a) Sketch unit balls in R2 when 1, 2 and 1 norms are used.

(b) Sketch unit ball in C[0,1] when maximum norm is used.
(c) Can you draw picture of unit ball in L2 [0; 1]?
11. Two norms k:ka and k:kb are said to be equivalent if there exists two positive constants
c1 and c2 ;independent of x; such that
c1 kxka

kxkb

c2 kxka

Show that in Rn the 2 norm (Euclidean norm) and 1 norm (maximum norm) are
equivalent.
31

12. Show that

jkxk

kykj

kx

yk

13. A norm k:ka is said to be stronger than another norm k:kb if

lim
k!1

x(k)

=0)

lim
k!1

x(k)

=0

but not vice versa. For C[0,1], show that the maximum norm is stronger than 2 norm.
14. Show that function kxk2;W : Rn ! R dened as
kxk2;W =

xT W x

denes a norm on when W is a positive denite matrix.

15. Consider X = R3 with hx; yi = xT W y: Given a set of three linearly independent
vectors in R3
2 3
2 3
2 3
1
3
1
6 7
6 7
6 7
(3)
(2)
(1)
x = 4 2 5; x = 4 2 5; x = 4 2 5
3
1
1
we want to construct and orthonormal set. Applying Gram Schmidt procedure,
hx; yiW = xT W y
3
2
2
1 1
7
6
W =4 1 2
1 5
1
1 2

16. Gram-Schmidt Procedure in C[a,b]: Let X represent set of continuous functions

on interval 0 t 1 with inner product dened as
hx(t); y(t)i =

Z1

w(t)x(t)y(t)dt

Given a set of four linearly independent vectors

x(1) (t) = 1; x(2) (t) = t; x(3) (t) = t2 ;
nd orthonormal set of vectors if (a) w(t) = 1 (Shifted Legandre Polynomials) (b)
w(t) = t(1 t) (Jacobi polynomials).
32

17. Show that in C[a,b] with maximum norm, we cannot dene an inner product hx; yi
such that hx; xi1=2 = kxk1 : In other words, show that in C[a; b] the following function
hf (t);g(t)i =

max
jx(t)y(t)j
t

cannot dene an inner product.

18. In C (1) [a; b] is
hx; yi =

Zb

x0 (t)y0 (t)dt + x(a)y(a)

an inner product?
19. Show that in C (1) [a; b] is
hx; yi =

Zb

w(t)x(t)y(t)dt

with w(t) > 0 denes an inner product.

20. Show that parallelogram law holds in any inner product space.
kx + yk2 + kx

yk2 = 2 kxk2 + 2 kyk2

Does it hold in C[a,b] with maximum norm?

References
[1] Kreyzig, E.; Introduction to Functional Analysis with Applications,John Wiley, New
York, 1978.
[2] Luenberger, D. G.; Optimization by Vector Space Approach , John Wiley, New York,
1969.
[3] Pushpavanam, S.; Mathematical Methods in Chemical Engineering, Prentice Hall of
India, New Delhi, 1998.
[4] Strang, G.; Linear Algebra and Its Applications. Harcourt Brace Jevanovich College
Publisher, New York, 1988.
[5] Strang, G.; Introduction to Applied Mathematics. Wellesley Cambridge, Massachusetts,
1986.
33