Functional Analysis:

Functional Analysis is the study of sets with composite

structure , partially algebraic and partially topology.
Metric Space:
Metric space is defined as,
Let X be a non-empty set and d is metric on X defined as,
d :X ×X ⟶ R
M 1¿ d ( x , y ) ≥ 0

M 2 ¿ d ( x , y )=0 ⅈff x= y
M 3 ¿ d ( x , y ) =d ( y , x ) (symmetry)
M 4 ¿ d ( x , z ) ≤ d ( x , y )+ d ( y , z ) ( Triangular inequality )
d satisfies M 1 ¿ M 4 properties then it is called Metric on X , and ( X , d ) is called Metric
Space.
Examples:
1) Let X =R
d ( x , y )=| x− y|

This metric is called Usual metric on R .

2) Let X =R2 let x=( x 1 , x 2 ) , y=¿, y 2 ¿

d (x , y)= √ ¿ ¿

This metric is called Euclidean metric on R .

OR: Let X =R2
d ( x , y )=¿ x 1− y 1∨+ ¿ x 2− y 2|

OR: If X =R2
d ( x , y )=max ¿

3) If X =R3 let x=( x 1 , x 2 , x 3 ) , y=¿, y 2 , y 3 ¿

√ 2 2
d ( x , y)= ( x 1− y 1 ) + ( x 2− y 2 ) + ( x 3− y3 )
2

4) If X =Rn then x=( x 1 , x 2 , … , x n ) , y=¿, y 2 , … , y n ¿

 n
d ( x , y)=∑ |x i− y i|
i=1

OR:

√∑
n
2
d ( x , y)= ( x i− y i)
i=1

This is called Euclidean metric on Rn .

5) If X =C n ⟶ n‐dimensional unitary space

√
n
d ( x , y )= ∑ ¿ x i− y i ¿2
i=1

6) Sequence space l∞ ¿Space of all bounded sequences ¿

Let x , y ∈l ∞
x=( x 1 , x 2 , … … ) is bounded if there exist real number C x >0 such that
|x i|≤C x ∀ i≥ 1
The Metric defined on l is: ∞

d ( x , y )=¿ i≥ 1|x i− yi|

7) If S=space of sequences on C ( Bounded∨unbounded )

x=( x 1 , x 2 , … , x n , … )
∞
1 ¿
d ( x , y ) =∑ ¿ x j − y j∨ ¿
j=1 2j 1+ ¿ x j− y j ∨¿ ¿

Induced Metric:
~
A subspace ( Y , d ) of ( X , d ) is obtained if we take a⊂¿
Y ⊂ X∧restrict d ¿Y ×Y , Thus the metric on Y isthe restriction
~
d=d ¿Y × Y
~
d is called the metric induced on Y by d .

8) Space B ( A )
spaceof all fuctions which are defined∧bounded on A .
f : A⟶ R

d ( f , g)=¿x ∈ A|f ( x ) −g ( x )|

9) Function space C [ a ,b ]
space of all continous functions defined on closed interval [ a , b ]
 d ( f , g )=max|f ( t ) −g ( t )|
a≤ t ≤ b

∵ Continous functionson closed interval attainsits maximum∧minimum value .

10) Discrete metric space :
d ( x , x )=0 , d ( x , y )=1(x ≠ y )
( X , d ) isdiscrete metric space .

11) Space l p ¿

{ }
∞
l = x=( x 1 , x 2 , … ) : ∑ ¿ xi ¿ < ∞
p p
where p ≥1
i=1

[∑ ]
∞ 1
p p
d ( x , y)= ¿ x i− y i ¿
i=1

12) Space l2 ( Hilbert space )

If p=2 thenis called Hilbert space .
x=( ξ j ) , y=( η j )

√
∞
d ( x , y)= ∑ ¿ ξ j−η j ¿ 2
j=1

 Holder inequality:
p 1/p ( ∑ ❑|η |q )1/q
∞ ∞

∑ ❑|ξ j ηj |≤ ¿|ξ k| ) m
j=1 m=1

When p>1 and 1/p+1/q=1

Cauchy Schwarz Inequality:
If p=2 , and q=2 in holder inequality

√ |2
√∑ 2
∞ ∞ ∞
∑ ❑|ξ j ηj | ≤ ∑ ❑| ξ k ❑ |ηm |
j=1 k=1 m=1

Minkowski Inequality:
❑|ξ + η |p )1/p p 1/p + ( ∑ ❑|η |q )1/q
∞ ∞

(∑
j=1 j j ≤ ¿|ξ
k| ) m m=1
  Example: Let d be the metric on a non- empty X Show
that the function
d1 :X*X→R

Defined by d1(x ,y)=min{1, d(x ,y)} ,∀ x, y ϵX is metric on X.

Sol: d1 ( x ,y )=min{1, d(x ,y)}
M1): Since (X,d) is a metric space so
d(x ,y) ≥0
min (1, d(x ,y)) ≥0
d1(x ,y) ≥ 0

M2): d1(x,y)= 0 iff x=y ∀ x,y ϵ R

Consider d1(x,y)= 0
min{1, d(x ,y)}= 0
d(x,y)= 0 , ∀ x,y ϵ R
But d is metric on X .So x=y
Hence , d1(x,y)= 0 iff x=y
M3) : d1(x,y)= d1(y,x)
Since d1(x ,y)=min(1, d(x ,y))
d1(x ,y)=min(1, d(y ,x))
d1(x ,y)=d1(y ,x)
M4): d1(x ,z)≤d1(x ,y)+ d1(y,z)
d1(x ,z)=min{1, d(x ,z)}
d1 (x ,y)=min{1, d(x ,y)}
d1 (y,z)=min{1, d(y ,z)}
Let : d1(x ,z)=1 , d1 (x ,y)=1 , d1(y,z)=1
d1(x ,z)≤d1(x ,y)+ d1(y,z)
1 ≤ 1+ 1
1≤2
Hence, d1(x ,y) is metric on X .
Example: Show that the function
d2 :X*X →R defined by
d ( x , y)
d2(x,y) = 1+ d ( x , y )
, ∀ x,y ϵ X is metric on X

Sol: M1): d(x ,y) ≥ 0

Also 1+d ( x , y) ≥0
d (x , y)
≥0
1+ d (x , y )

d2(x,y) ≥0

M2 ): To verify d2(x,y)= 0 iff x = y , ∀ x,y ϵ X

Consider , d2(x,y)= 0
d (x , y)
⟺
1+ d ( x , y )
=0
⟺ d(x,y)= 0
Since d is metric on X
⟺ x=y
Hence, d2 (x,y) = 0 ⟺ x = y , ∀ x,y ϵ X
M3): d2(x,y)= d2(y,x)
d ( x , y)
d2(x,y)= 1+ d ( x , y )

d ( y , x)
= 1+ d ( y , x )

= d2(y,x)
M4 ): d2(x ,z)≤d2(x ,y)+ d2(y,z)
Since d is metric on X so,
d(x ,z)≤d(x ,y)+ d(y,z)
Adding 1 on B.S:
1+d(x ,z)≤ 1+ ¿d(x ,y)+ d(y,z)
Taking reciprocal:
1 1
≥
1+d(x ,z) 1+ d(x ,y)+ d(y,z) 

Multiply by (-1) and add (1) on B.S:

1 1
1- 1+d(x ,z) ≥1- 1+d(x ,y)+ d(y,z) 
1 +d(x ,z)−1 1-1+d(x ,y)+ d(y,z)
≤
1+d(x ,z) 1+d(x ,y)+ d(y,z) 

d(x ,z) d(x ,y)+ d(y,z)
≤
1+d(x ,z) 1+d(x ,y)+ d(y,z) 
d(x ,z) d(x ,y)  d(y,z)
1+d(x ,z)
≤
1+d(x ,y)+ d(y,z)  + 1+d(x ,y)+ d(y,z) 
d2(x ,y) ≤ d(x ,y)
1+ d(x ,y)+ d(y,z) 
+  d(y,z)
1+d(x ,y)+ d(y,z) 
……….(A)
As
d(x ,y) d(x ,y)
1+d(x ,y)+ d(y,z) 
≤
1+d(x ,y)  …….(1)

And :
 d(y,z)  d(y,z)
1+d(x ,y)+ d(y,z) 
≤
1+ d(y,z) 
………(2)

Adding 1 and 2:
d(x ,y)  d(y,z) d(x ,y)  d(y,z)
1+d(x ,y)+ d(y,z)  + 1+d(x ,y)+ d(y,z)  ≤ 1+d(x ,y)  + 1+ d ( y,z) 

d(x ,y)  d(y,z)
1+d(x ,y) 
+ 1+ d(y,z)  ≤d (x ,y)+ d (y,z)
2 2
From (A):
d2(x ,z)≤d2(x ,y)+ d2(y,z)
Hence, d2 is metric on X.

• Open Ball:
let (X,d) be a metric space and x 0 ∈ X and r >0 such that;
B( x 0 ,r ¿={ y ∈ X : d(y, x 0)<r}
is called open ball with center x 0and radius r.
• Closed Ball:
B( x 0 ,r ¿={ y ∈ X : d(y, x 0)≤r}
is called closed ball with center x 0and radius r.
• Sphere:
S ( x 0 ,r ¿={ y ∈ X: d(y, x 0)¿r}
 Examples:
Let X= R2

√ 2
d ( x , y)= ( x 1− y 1 ) + ( x 2− y 2 )
2

x 0=(0,0) , r =1

Find open ball?

B ( x 0 ,r ¿={ y ∈ X: d (y, x 0)<r}
={( y 1 , y 2): √ ( y 1−0 ) + ( y 2−0 ) <1 }
2 2

={( y 1 , y 2): y 12+ y 22<1 }

Set of points that lies

inside the circle.

 Example 2:
Let X= R2
d ( x , y)= max { |x 1− y 1|,|x 2− y 2|}

x 0=(0,0) , r =1

Find open ball?

B ( x 0 ,r ¿={ y ¿ ¿, y 2)∈ R 2: d (y, x 0)<r}
B ( x 0 ,1 ¿={¿, y 2); max { | y 1−0|,| y 2−0|<1}
B ( x 0 ,1 ¿={( y 1 , y 2 ) :max ⁡{| y 1|,| y 2|}<1 }
Let max {| y 1| ,| y 2|}=| y 1|=1
 max {| y 1| ,| y 2|}=| y 1|=1

 Example 3:
If X= R2
d ( x , y)=|x 1− y 1|+|x 2− y 2|

x 0=(0,0) , r =1

B ( x 0 ,1 ¿={( y 1 , y 2 ) ∈ R 2:| y 1|+| y 2|< 1}

1. If y1 ≥ 0 , y2 ≥ 0

Then y 1 + y 2 <1
Suppose if y 1 + y 2=1

2. If y 1 ≥ 0∧ y 2 ≤ 0

3. If y 1 ≤ 0 then y 2 ≥ 0

4. If y 1 ≤ 0∧ y 2 ≤ 0

 Open Set:
Let ( X , d) be metric space & M⊆X are say that M is open set if it contains open
ball at each of its points
 Closed Set:
N is closed set if N c is open.
∈- Neighbourhood of x 0 ∈ X :

Let x 0 ∈ X where (X,d) is metric space

B ( x 0,∈) = {y∈x:d(y, x 0)<∈}
 Neighborhood of x 0:
By a Neighborhood of x 0 we mean that a subset of X that contains an ε -
Neighbourhood.
 Interior Point:
Let M⊆X we say x 0 ∈ M is an interior point of M if M is neighbourhood of x 0.
Let if interior points of M are denoted by M 0
(It is largest open set contains in M).
 Continuous Mapping:
Suppose that we have two metric spaces
(X , d x) , (Y , d y )
T: X→Y
We say that this mapping T is continuous at x 0 ∈ X
If for each ε > 0there exist S>0 .
d y ( T x ,T x 1 ) < ε for x satisfying d x ( x , x 0 ) < S

T ( B ( x 0 , S ) ) ⊆ B (T x , ∈)
  Theorem:
T : X→Y is continuous iff the inverse image of every open set in Y under T is open
in X.
 Limit Point:
Let (x , d) be a metric space and M⊆ X and x 0 ∈ X then x 0 is Limit point of M if each
ε -Neighbourhood contains of x 0 contains a point of M distinct from x 0.

 Closure of set :
Let M be a subset of X. the set consisting of the points of M and the limit points of
M is called the closure of M and is denoted by M .
M =MUM d

It is the smallest closed super set of M.

 Dense subset :
A subset M of metric space X is said to be dense in X if M = X.
If we have a countable set M then this space X is called separable.

• Example:
The set R is separable becauseQ is dense in R .i.e.Q=R and Q is countable.
⇒ R with usual metric is separable.

1)The complex plane C is separable. A countable dense subset of C is the set of all
complex numbers whose real and imaginary parts are both rational.
2)A discrete metric space X is separable iff X is countable.
3)l p is separable but l ∞ is not separable.
 • Convergent sequence: A sequence { x n } in a metric space X =( X , d)
is said to be convergent if there is an x ∈ X such that
lim ⅆ ( x n , x )=0
n→∞

x is called the limit of ( x n ) and we write

lim x n=x
n→∞

Or x n → x
It means that for every ε > 0 ∃ N ∈ N
Such that ⅆ ( x n , x ) <¿ ε ∀ n ≥ N
• Cauchy Sequence: Let { x n } ⊆( X , d ) is Cauchy if for each ε > 0 there exist N ∈ N
such that
ⅆ ( x m , x n ) ¿ ε ∀ n , m≥ N

• Diameter: Let M ⊆( X , d ) We define diameter of M as

δ ( M ) ={ⅆ ( x , y ); x , y∈ M } ¿

If diameter of M is finite then M is bounded.

• Lemma:
Let X =( X , d) be a metric space then
a)A convergent sequence in X is bounded and its limit is unique.
b)If x n → x and y n → y in X , then ⅆ ( x m , x n ) → ⅆ ( x , y )
Proof:
Suppose that
xn→ x ∈ X

We have to show that ( x n ) is bounded and its limit is unique

Since x n → x so for each ε > 0 ∃ N ∈ N such that
ⅆ ( xn , x )< ε ∀ N ≥ N

Let
 a=max { ⅆ ( x 1 , x ) , ⅆ ( x 2 , x ) , … … , ⅆ ( x n , x ) }

ⅆ ( x n , x ) <  1+ a

This shows that ¿ ¿) is bounded.

Now for uniqueness
Let xn → x , xn → z

⇒ 0 ≤ ⅆ ( x , z ) ≤ ⅆ ( x , xn )+ ⅆ ( xn , z )

⇒0 ≤ ⅆ ( x , z )≤ 0

⇒ x=z

(b)Suppose that if x n → x and y n → y then to show that

ⅆ ( xn , yn ) → ⅆ ( x , y )

By using triangular inequality

⇒ ⅆ ( xn , yn )≤ ⅆ ( xn , x )+ ⅆ ( x , y ) + ⅆ ( yn , y )

⇒ ⅆ ( x n , y n )−ⅆ ( x , y ) ≤ ⅆ ( x n , x ) + ⅆ ( y n , y )

Similarly, Inequality by interchanging x n and x as well as y n and y and multiply by

-1
We get
⇒∨ⅆ ( x n , y n ) −ⅆ ( x , y )∨≤ ⅆ ( x n , x ) + ⅆ ( y n , y ) As x n → x and y n → y
⇒ ⅆ ( xn , yn ) → ⅆ ( x , y )

• Complete Metric Space: The space X is said to be complete if every Cauchy

sequence in X converges in X.
Examples: R , C , Rn ,C n ,l ∞ , l p spaces are complete metric spaces.
Theorem: Let M be a non empty subset of a metric space ( X , d ) and M be its
closure. Then
a) x ∈ M Iff there is a sequence ( x n) in M such that x n → x
Proof:
Let x ∈ M = M M d
⇒ x ∈ M or x ∈ M d

If x ∈ M then sequence of this type ( x , x , x , … … x ¿=(x n ) converges to x .

If x ∈ M d that is x is limit point of M ,then each ε -neighborhood of x contains a point
of M other than x .
Now for ε =1
The neighborhood { x ∈ X ; ⅆ ( x 1 , x ) <1 } of x contains a point x 1(say) of M
1
Similarly, for ε = 2 then ∃ x 2 ≠ x of M

{ 1
Neighborhood of x= x 2 ∈ X ; ⅆ ( x 2 , x ) < 2 }
In general

{ 1
}
For each n then Neighborhood of x= x n ∈ X ;ⅆ ( xn , x ) < n contains a point x n of M
different forms.
Thus
1
⇒0 ≤ ⅆ ( x n , x ) ≤ ∀n∈N
n

⇒ lim ( 0 ) ≤ lim ( x ¿¿ n , x )¿ ≤ lim ¿ 1

n→∞ n →∞ n→∞ n
⇒ 0 ≤ lim ( x ¿¿ n , x)¿ ≤ 0
n →∞

By Sandwich Theorem
⇒ lim ( x¿¿ n , x)=0 ¿
n→∞

⇒ xn→ x

Hence Proved!

Conversely:

Suppose that x n ⟶ x then by the definition,

For every ε>0 ∃ Nє N such that

 d( x n, x ) < ε ∀ nϵN

⟹d( x n, x ) =0

If d( x n, x ) =0 ⟺ x n= x
As

x nє M

⟹ xєM
⟹xє M ∴ M⊆M
b) M is closed if and only if x n ϵM , x n ⟶ x , xϵM

Proof:

Suppose that M is closed , x n ϵM , x n ⟶ x

We have to prove that xϵM

As M is closed then

M = M ……….1)
As ( x n) in M , x n ⟶ x

⟹ xϵ M ∴ xϵ M ⟺ ( x n) in M such that x n ⟶ x
⟹ xϵM From 1)

Conversely:

Suppose that x n ϵM , x n ⟶ x , xϵM

We have to prove that M is closed i.e

M =M
Let xϵ M ∴ xϵ M ⟺ ( x n) in M such that x n ⟶ x
⟹ xϵM
⟹ M ⊆ M …………1)
Also

M ⊆ M ………….2)
From 1) and 2)

M =M
⟹ M is closed
Theorem:

A subspace M of a complete metric space X is itself complete if and only if M

is closed in X.

Proof:

Suppose that (X,d) is complete and M is complete subspace of X. We have to prove that M

is closed.

Let xϵ M then ∃ a sequence ( x n) in M such that x n ⟶ x

Since every convergent sequence is Cauchy so ( x n) is Cauchy sequence in M .

Since M is complete. So ( x n) is convergent in M i.e x n ⟶ x

Therefore xϵM

⟹ M ⊆ M …………1)
Also

M ⊆ M ………….2)
From 1) and 2)

M =M
⟹ M is closed in X
Conversely:

Let M is closed and ( x n) be a Cauchy in M . But M is subspace of X.

Since X is complete so ∃ xϵX such that x n ⟶ x

⟹ xϵ M ∴ xϵ M ⟺ ( x n) in M such that x n ⟶ x
xϵM ∵ M = M
Therefore ( x n) is arbitrary Cauchy sequence in M and it converges

in M .

Hence M is complete subspace.

n n
R ∨C is complete metric space

Proof:

Let ( x m ¿ be a Cauchy sequence in Rn .

Then for every ε > 0 ∃ Nϵ N such that

 d( x m, x r ) < ε ∀ m ,r≥ N

The usual metric on Rn is

√
n

∑ ( x (m)
i −x i )
2
d( x m, x r )= (r )

i=1

Since d( x m, x r) < ε ∀ m ,r≥ N

√
n

∑ ( x (mi )−x (ri ) )

2
⟹ <ε
i=1

Taking square on both side,

n

∑ ( x(m)
2
i −xi ) < ε ………1)
(r) 2

i=1

⟹ ( x (m)
i −x i ) < ε
(r ) 2 2

Taking square on both side,

|x (m)
i −x i |<ε
(r )
for each i

⟹ ( x(m )
i ) is Cauchy sequence in R . Since R is complete then every Cauchy sequence in R

is convergent so x i in R such that

(m )
xi ⟶ xi
(m )
⟹ lim x i =x i
m⟶∞

Similarly,

lim x (r)
i = xi
r⟶∞

If we fix m for each i ,

lim ¿¿ ¿ - x(r ))= x (m )- x i for each i ∵ lim an =a

r⟶∞ i i n⟶ ∞

(If f is continuous function and x n ⟶ x then f ( x n ) ⟶ f (x ))

Since |.| is continuous function.

So,

lim | xi −x i |=|x (im ) -  x i|

(m) (r)

r⟶∞

Taking sum on both side,

 n n

∑|x ( m)
i −x i| = lim
r⟶∞
∑ | i −xi |
x(m) (r)

i=1 i=1

Taking square on both side,

n n
2
∑(x ∑
2 2
(m)
i −xi ) = lim ( x(m)
i −x i ) < r ⟶ ∞
lim ε From 1)
i=1 r⟶∞ i=1

∑ ( x(m)
2
i −xi ) < ε
2

i=1

Taking square root on both side,

√∑ (
n
2
i −x i )
x (m) <ε
i=1

⟹ d ¿, x ¿ <ε ∀ m≥ N
⟹ x m is converges to x .

Since ( x m) is arbitrary Cauchy sequence in Rn which is convergent.

So every Cauchy sequence in Rn is convergent.

Hence Rn is complete.

Completeness of l ( The space of all bounded sequences):

∞

Proof:
Let ( x m ) be a Cauchy sequence in l∞. Then for each ε > 0there exist n 0 ∈ N such that
d ( x m , x n ) < ε ∀ n , m≥ n o ⟶(1)
Now metric on l∞ is d ( x m , x n ) =|x |
(m ) ( n)
−x i
i
¿

As x m= { x(m)
1 , x2 , … , x n } ∈l .
(m) (m ) ∞

⇒|x i |≤ Μ
(m)
∴ ¿ is bounded)

Put metric of l∞ in 1,
|x(mi )−x(in)|
d ( xm , xn ) = ¿ ε ∀ n , m ≥n 0 , ∀ i ≥ 1

⇒|x i −x i |<ε ∀ n , m≥ n0
(m) (n )

As x(m
i
)
is a Cauchy sequence in R . Since R is complete. So, every Cauchy sequence
in R is convergent. Therefore x(mi
)
is convergent. So ∃ x i ∈ Rsuch that;
(m ) i
x i ⟶ x for each i.

In this way we get an element x i=x=( x 1 , x 2 , x 3 , … ) .

Now since x m ∈ l∞ . So x m is bounded.
⇒ There exist k m ∈ R such that

|x (m)
i |≤ k m for each i ⟶(2)

As |x (m)
i −x i |≤ ε ∀ n , m≥ n0
(n)

If we fix m and letting n → ∞. Then we get,

|x (m)
i −x i|≤ ε ∀ m≥ n 0 → ( 3 )

Now consider,
|x i|=|xi −x(im) + x (mi )|
By using Triangular Inequity (|a+b|≤|a|+|b|),
|x i|≤|x i−x (m)
i |+|x i |
(m)

|x i|≤ ε +k m ∴ ( Using 2 & 3)

⇒ xi ∈ l∞

⇒ x ∈l ∴¿ is bounded)
∞

Now we prove that x(m )

i →x .
From (3);
|x (m)
i −x i|≤ ε ∀ n , m≥ n0 ∧∀ i

|x (m)
i −x i|
⇒ <ε ∀ n , m≥ n0∧¿for each i.

Now x(m
i
)
converges to x i according to l∞ .
⇒ d ( x m , x ) < ε ∀ m ≥ n0

⇒ ( xm) ⟶ x

Since x m is arbitrary Cauchy sequence in l ∞. And it is convergent so every Cauchy

sequence in l∞ is convergent.
Hence l∞ is complete metric space.

Completeness of C (The space of all convergent sequences):

Proof:
As we know that C is subspace of l∞ . So, to prove C is complete we have to show
that C is closed in l∞ .
“A subspace M of a complete metric space X is itself complete iff the set M is
closed in X.”
To prove C is closed we will prove that C=C . As we know that C ⊆ C So we only
prove that C ⊆ C .
So, for this we consider;
x=( ξ j ) ∈ C

By using theorem (If x ∈ Μ then ∃a sequence x n in Μ such that x n → x ¿ .

So there exist x n=( ξ(n)
j ) ∈C such that ξ j → ξ j.
(n)

Hence for every ε > 0 ∃n ∈ N such that;

d ( xn , x )< ε ∀ n ≥ N

¿|ξ (nj )−ξ j|< ε ∀ n≥ N , ∀ j ≥1

ε
⇒|ξ j −ξ j|< ∀ n ≥ N
(n)
⟶(1)
3
 In particular for n ¿ Ν ∀ j;
Since x n ∈ C its terms ξ (N)
j form a convergent sequence. Since every convergent
sequence is Cauchy sequence. So ( x n )
is Cauchy, So ∃n 1 ∈ N such that ∀ j we have;
ε
|ξ (n)j −ξ (N)
k |<
3
⟶(2)

Consider;
|ξ j−ξ k|=|ξ j−ξ(Nj ) +ξ(Nj )−ξ(Nk ) +ξ(kN) −ξ k|
Now by using Triangular Inequality;
|ξ j−ξ k|≤|ξ j−ξ(Nj )|+|ξ (N)
j −ξ k |+|ξ k −ξk|
( N) (N )

ε ε ε
¿ + + ∴(from 1 &2)
3 3 3

|ξ j−ξ k|< ε
This shows that x=( ξ j )is convergent. So x ∈ C . Since x ∈ C was arbitrary.
So;
C ⊆C

⇒ C=C

⇒C is closed in l ∞ .

Hence C is complete.
Completeness of lP Space:
Let x , y ϵ l psuch that
x=¿ ¿ , x 2,…) , y=¿¿ , y 2,…) where
∞ ∞

∑| x j| <∞ and ∑| y j| <∞

p P

j=1 j=1
∞ 1

d( x , y ¿=∑ (| x j− y j| )
P P

j =1

Let ( x m ) be Cauchy sequence in l P so by definition for every ε > 0there exist

n° ϵ N such that

d ¿)¿ ε ∀ m ,n ≥ n°
∞ 1

[ P P
⟹ ∑ | x(jm)−x (jn)|
j=1
] < ε ⟶(eq 1)
∞
P
⟹ ∑ |x (jm )−x (jn )| <ε p ⟶( eq 2)
j=1
 p
⟹|x (jm )−x (jn)| < ε p ∀ m , n ≥n ° ⟶ (eq 3)

⟹|x j −x j |< ε ∀ m , n ≥n ° j = 1,2,…..

(m ) ( n)

⟹ ( x j (m )) is a Cauchy sequence in C∨R∧C is complete so it is convergent.

Let x (jm ) ⟶ x j
Using the infinite limits we have
x=(x1 , x2 , …) lim x (jm ) → x j
n→∞

We have to show that xϵ l for this will prove ∑ | x j| < ∞ p P

j=1

Taking n ⟶ ∞ and m is fix in (2) we get

∞
P
∑| x(jm)−x j| < ε P for m ≥n ° ⟶( ¿)
j=1

∞
p
x m = ( x j (m ))∈l p so ∑ | x(jm)| < ∞⟶ ¿
j=1

( ) ( )
∞ 1 ∞ 1
( )P P
Now ∑| x j|
p P
= ∑|x j−x jm + x mj |
( )

j=1 j=1

By using Minkowski inequality

(∑ | ) (∑ | | )
∞ 1 ∞ 1
(m ) p P (m ) p P
≤ x j−x j | + x j
j=1 j=1

By using (¿ ¿∧¿ (¿∗¿

¿ε+∞
∞

∑|x j| <∞
p
⟹ ⟹ xϵ l P
j=1

(∑ ( )
∞ 1
p P
x −x j )
( m)
Taking n ⟶ ∞ in (1) we get j < ε ∀ m ≥n °
j=1

⟹ d ( xm , x ) < ε ∀ m≥ n° ⟹ x m ⟶ x Since ( x m) is arbitrary Cauchy sequence in l p so every

Cauchy sequence in l p is convergent.

⟹ l Space is complete.
P
Completeness of C[a,b]:
(The Space of all continuous functions on [ a , b ] ¿
Let x , y ϵ C[a , b]
The metric on C [a , b] is d ( x , y )=max ( x (t )− y (t ))
a≤ t ≤ b

Let ( x m) be a Cauchy sequence in C [a , b] so by definition for every ε > 0there exist

n° ϵ N

d ¿, x n)¿ ε ∀ m ,n ≥ n°

⟹ max | x m ( t )−x n ( t )|< ε ∀ m , n≥ n °

a ≤t ≤ b

For t = t °
⟹|x m ( t 0 ) −x n ( t 0 )|< ε ∀ m , n ≥n °

Where t ° is any fixed element of [ a , b ]. As ( x m ( t0 ) ) is Cauchy sequence in R .

Since R is complete so it is convergent.
Let x m(t o)⟶ x ¿)
Using these infinite limits define a function
X :[a,b]⟶ R
x=x ( t ) on t ϵ [a , b]

Taking n ⟶ ∞ and m is fixed we get

max ¿
a ≤t ≤b

⟹ d ( xm , x ) <ε for m≥ n° ∵ xm (t)⟶ x=x (t ) uniformly so x is convergent.

⟹ xm⟶ x

( x m) is arbitrary Cauchy sequence. Hence C [ a , b] is complete.

Examples of incomplete metric space:

1. Space Q is incomplete.
On contrary we suppose that Q is complete and we know that Q ⊆R∧R is also
complete.
 As we know that (A subspace M of a complete metric space X is itself complete
iff M is closed in X).
⟹ Q is closed.

⟹ Q=Q

But Q is dense in R ∵ Q=R So Q ≠Q

Hence our assumption is wrong so Q is incomplete metric space.

2. Space of all polynomials is incomplete.
Let X be the Set of all polynomials consider as function of t on some finite closed
interval J=[a , b] and define a metric d on X by
d ( x , y )=max| x ( t )− y ( t )|
t ∈J

This metric on X is not complete.

∞ m
x
Let Pn ( x ) = ∑
m=0 m!
2 3
x x
e = 1+ x+
x
+ +…
2 ! 3!
S1 = 1

S2 = 1+ x
2
S3 = 1+ x
2!

lim S n = e x ∉ P x2 x3
1+ x+ + + …Converges to e x but e x is not
n→∞
2 ! 3!
polynomial.
∞ m
x
Pn (x ) = ∑ m! ( Pn( x ¿ ¿ is Cauchy sequence in X but converges not in X .
m=0

Let X be the set of all continuous real valued functions on J= [ 0,1 ] and Let
 1
d ( x , y )=∫|x ( t )− y ( t )| ⅆt
0

The metric space (X , d ) is not complete.

Proof:

The function x m form a Cauchy sequence , Because d (x m , xn ) is the area of triangle.

1 1
As n> m⇒ n < m

1
Adding 2 on both sides

1 1 1 1
⇒ + < +
2 n 2 m
⇒ an ¿ am
1
d ¿ x n ) ¿ ∫|x m ( t )− xn ( t )| ⅆt
0

The integral give Area of triangle.

1
Area of triangle ¿ 2 (a m- a n) (1)

¿ −(
1 1 1
2 m n )
1 1 1
¿ − <
2m 2n 2 m
 1 1
¿ ¿ ¿ε
2m m
1
From m> ε

1
d ¿, x n)¿ ε ∀ m ,n>
ε

Suppose x m → x
x ∈ C [ 0,1 ]
1
⇒ d ¿, x ¿=∫|x m ( t ) −x   (t )|ⅆt
0

∵ x m ( t )=0if t ∈ 0 , [ ] 1
2

∵ x m ( t )=1 if t ∈ [ am , 1 ]
1
2 am 1

d ( x m ,  x )=∫|x m ( t )−x   ( t )| ⅆt +∫|x m ( t )−x   ( t )| ⅆt +∫| xm ( t )−x   ( t )| ⅆt

0 1 am
2

∵ xm is Cauchy .
1
2 1

0=∫ |x   ( t )| ⅆt + ¿∫|1−x   ( t )| ⅆt ¿
0 am

It is possible only if both integrals are zero.

x   ( t ) =0 if t∈¿
x   ( t ) =1 if t∈(1/2,1]
1 1
As a m= 2 + m

1
a m>
2

which is a contradiction to our supposition i.e. x m → x & x ∈ C [a , b ] So t∈(1/2,1]

This is impossible for a continuous function.
So Hence,
xm ↛ x

So, this metric space is incomplete.

N ormed Space:

Let X be a vector space. Now we have a function ‖.‖:X→ R

It is normed on X if it satisfy the following conditions.
N 1 ¿ ‖x‖≥ 0 ∀ x ∈ X

N 2 ¿‖x‖=0iif x=0

N 3 ¿‖αx‖=|α|‖x‖

N 4 ¿ ‖x + y‖≤ ‖x‖+‖ y‖

Here x and y are arbitrary vectors in X and α is any scalor .

A norm on X defines a metric d on X which is given by
d ( x , y )=‖x− y‖

It is called metric induced by the norm.

It is denoted by ( X ,‖.‖) ∨X .
Examples:1
Let X =¿ R
‖x‖=|x|
‖.‖: R → R isnormed on R  2.
Examples:2
Let X =¿ R  2
x=¿ ¿ ,ξ 2)

1) ‖x‖1 =|ξ 1|+|ξ 2|

is normed on R  2.
1) ‖x‖2 =√|ξ 1|2+|ξ 2|2
is normed on R  2.
3) ‖x‖3 =max {|ξ1|,|ξ 2|}
Examples:3
Let X=l p
If xε l p

[ ]
∞ 1

‖x‖= ∑|ξi|
p p

i=1

Is normed on X.
Examples:4
Let X=l∞
If x∈ l∞ x=(ξ 1 , ξ 2 , …)

‖x‖=S up|ξ i| is normed on X.

i ≥1

Examples:5
Let X =C [a ,b ]
‖x‖=Max| x(t)|
a ≤t ≤b

Remark:
1) Every norm ‖.‖ on a vector space induces a metric given by d ( x , y )=‖x− y‖
but if we given metric on a vector space then it is not always that it is
coming from normed space.
2) A complete normed space is Banach space.
3) Norm give us length of a vector.
4) If a metric d on a vector space induced from a norm
‖.‖ .Thenwill satisfy the following :
ⅈ ¿ d ( x+ a , y+ a )=d ( x , y )

ⅈⅈ ¿ d ( αx , αy )=|α |d ( x , y )

Proof:
If a metric d induced from a normed space .Then
 d ( x , y )=‖x− y‖

ⅈ ¿ d ( x+ a , y+ a )=‖( x+ a )−( y +a)‖

¿‖x +a− y−a‖

¿‖x− y‖

¿d (x, y)

ⅈⅈ ¿ d ( αx , αy )=‖αx−αy‖

¿‖α ( x− y)‖

¿|α |‖x− y‖

¿|α |d ( x , y )

Hence proved .

Example: if x=¿) then ∥ x ∥=√ x 21+ x 22 + x 23 is normed space .

Sol:
N 1) ∥ x ∥≥ 0 ∴ √ x 21 + x 22+ x 23 ≥ 0

N 2) ∥ x ∥= 0

⟺ √ x21 + x 22 + x 23= 0

⟺ x 1+ x 2 + x 3 = 0 ( sum is only zero if they are zero )

2 2 2

⟺ x 1=0, x 2=0 , x 3= 0
⟺ x=( 0,0,0 ) =¿)
N 3)
∥ αx ∥=∥ α ¿ )∥=∥(α x1 , α x 2 , α x 3 )∥

=√ α 2 x 21+ α 2 x 22+ α 2 x23

=√|α |2 ¿¿ ¿ ) =|α|√ x21 + x 22+ x23

⇒∥ αx ∥=|α |∥ x ∥

N 4 )∥ x + y ∥ =∥ ( x 1 , x 2 , x 3 ) +( y 1 , y 2 , y 3 )∥
=∥( x ¿ ¿ 1 + y 1 , x 2 + y 2 , x3 + y 3 )∥ ¿

=√ ( x 1+ y1 ) + ( x 2 + y 2 ) + ( x 3 y 3 )
2 2 2
†

√
3
= ∑ ( x i+ y i )2
i=1

by using Minkowski Inequality

√ √ √
3 3 3
= ∑ ( x i+ y i ) ≤ ∑ x + ∑ y 2i
2 2
i
i=1 i=1 i=1

≤ √ x 21 + x 22+ x23 + √ y 21 + y 22+ y 23 ∥ x + y ∥≤ ∥ x ∥+ ∥ y ∥ hence,

∥ . ∥ is normed space.

Question: Prove that|∥ y ∥−∥ x ∥∨≤ ∥ y−x ∥ and the norm is continuous function.
Proof: Consider
∥ y ∥=∥ y−x + x ∥

≤ ∥ y−x ∥ +∥ x ∥

∥ y ∥−∥ x ∥ ≤∥ y −x ∥ ⟶ (1)

similarly,
Consider ,
∥ x ∥=∥ x− y + y ∥ ≤ ∥ x− y ∥ +∥ y ∥

∥ x ∥−∥ y ∥ ≤∥ x− y ∥

-{ ∥ y ∥−∥ x ∥} ≤∥−1( y−x )∥

-{ ∥ y ∥−∥ x ∥ } ≤|−1|∥ y −x ∥

−{ ∥ y ∥−∥ x ∥ } ≤∥ y −x ∥ ⟶ (2)
from (1)and (2)
|∥ y ∥−∥ x ∥|≤ ∥ y−x ∥ (proved).

Now we have to prove that norm is continuous function.

As we know that,

|∥ y ∥ -∥ x ∥|≤ ∥ y−x ∥ ⟶ (¿ ¿

let x n → x , then from (¿)

we can write

|∥ x n ∥−∥ x ∥ │≤ ∥ xn −x ∥ =d ( x n , x)
Since x n → x so d( x n , x ¿=0 ∴ ¿d is metric induced by norm)

⟹|∥ x n ∥−∥ x ∥ |→ 0

⟹ ∥ x n ∥ →∥ x ∥

⟹ ∥ .∥ is continuous function.

(For continuous function)

∴ if x n → x f( x n ¿ →f( x )

Theorem :
A subspace y of a Banach space x is complete if and only if the set y is
closed in x .

Convergence of sequence :
A sequence ( x n) in a normed space X is convergent
if X contains an x such that
lim ∥ x n−x ∥=0 or xn→ x
n→∞

Cauchy sequence in Normed space:

A sequence ( x n ¿
 in a normed space X is Cauchy if for every ε > 0their existn° ϵ N such that

∥ x m −xn ∥< ε ∀ n , m≥ n°

Question :
show that C ⊂ l∞is a vector subspace of l∞and so C ° the space of all
sequence of scalars converging to zero.
Proof :
first we prove that C (space of all converging sequence of complex no is
vector subspace of l∞)
let x , y ϵ C x=(x¿¿ n) ¿ , y=( y ¿¿ n)¿

⇒ ( x n) & ( y n) are converging sequence of complex numbers.

Let α , βϵ C
we have to show that α x n + β y n ϵ C

suppose x n → a∧ y n → b

lim (x n )= a∧lim (¿ n → ∞ y n )= b
n→∞

lim ¿)=α nlim ( x n ) + β lim ¿ )

→∞
¿ αa + βb ⟹ α x n+ β y n →α a+ βb

⇒ α xn + β y n ϵ C hence C is the vector subspace of l ∞. Now to show C °

(space of all sequence of scalars converging to zero ) is vector subspace of l∞ .
Let (ξ (in ) ¿∧(η(in )) are sequence in C °.then they are
converges to zero .
ξ (in ) → 0∧η(in ) → 0 lim ( ξ i )=0∧lim ( ηi ) =0
( n) (n )

n→∞ n→∞

Consider,
lim ( α ξ (in) + β η(in )) =α lim (ξ ¿ ¿ i (n) )+ β lim ¿ ¿ ¿)
n→∞ n→ ∞ n→∞

¿ α ( 0 )+ β ( 0 ) =0

( n) (n )
α ξ i + β ηi ϵ C °
Hence,
∞
C°⊆ l proved.

Question # 2:
Show that C 0 (space of all sequences of scalars converging to zero) is closed
subspace of l∞ so that C 0 is complete.
Proof:
We have to show that c o is closed for this we show
co ⊆ co

Let x ∈ c 0 then ∃ ( xin) ∈c0 s . t

x ni → x

By definition of convergence for every ε > 0 ∃n ∈ N s .t

‖x ni − x‖< ε ∀ n ≥ N
¿i ≥1|x i − x i|<ε ∀ n ≥ N
n

|x ni − x i|<ε ∀ n ≥ N ∀ i (1)

As ( x ni ) ∈ c o
n n n
x i =( x1 , x 2 , … … …)

& x ∈ c 0 ⊆ l∞
x=( x1 , x2 , … ..)

For each i we have sequence of scalors and this sequence is convergent and every
convergent sequence is Cauchys sequence
For n=N

‖x − x ‖< ε3 (2)
N
i
N
i

From (1) for n=N

 ε
¿ x i‖ ¿ ∀ i (3)
3

|x i− x i|=¿ ¿
| N
|
≤ x i− xi |+ xi −x i |+| xi − x i|
N N N

ε ε ε
¿ +¿ +¿ from 2 & 3
3 3 3
¿ ¿ε

(x ¿¿ i) ¿ is Cauchy sequence

Since it is complete so every Cauchy sequence is convergent so (x ¿¿ i)¿ converges

Now to show x i ∈ c o so it must converges to zero
¿

|x i−0|≤|x i − x Ni |+|x Ni |
ε
¿ ε /3+ 2
3

|x i−0|<ε
(x i)→ 0

x=x i → 0 ⇒ x ∈ c o

As c o ⊆c o
co = co

Hence c o is closed hence complete.

Question #3:
show that in a normed space X vector addition & scalor multiplication are
continuous operations w.r.t the norm i.e the mapping defined by
( x , y ) → x + y∧(α , x )→ αx are continuous
Proof:
Mapping is defined as
‖.‖: XxX → X by
( x , y ) → x + y∧(α , x )→ αx
Now to show that continuous
As x n → x∧ y n → y
 lim ¿n →∞‖x n −x‖=0∧lim ¿n →∞‖ y n− y‖=0 ¿ ¿
‖( x n + y n ) −(x + y )‖=‖x n−x + y n− y‖≤‖x n−x‖+‖ y n− y‖
lim ¿n →∞‖( x n+ y n) −( x+ y )‖¿ ≤ 0+0

( xn+ yn) → ( x + y )
 ‖α xn −αx‖=|α|‖x n−x‖=|α|( 0 )=0
α x n → αx

Both operations are continuous w.r.t normed space

Question # 4:
In l∞ ,let y be the subset of all sequences with only finitely many non zero terms.
Show that y is a subspace of l∞ but not a closed subspace
Proof:
let x=( x i ) , y=( y i ) ∈Y x i ≠ 0 ≠ y i
Then
x=( x 1 , x 2 , x 3 , … … , x n , 0,0 , … … )

y=( y 1 , y 2 , y 3 , … … . , y n ,0,0 … .)

Letα , β ∈ F
αx+ βy=α ( x 1 , x 2 , x 3 , … … … , x n , 0,0 , … . ) + β ( y 1 , y 2 , y 3 , … . y n , 0,0 … . )

¿( α x 1 , α x 2 , x 3 , ….. α x n , 0,0 … . …) +( β y 1 , β y 2 , β y 3 , ….. β y n , 0,0 … )

¿( α x 1+ β y 1 , α x 2 + β y 2 , … … , α x n + β y n , 0,0 ,.. … .)

It has finitely many non zero terms so αx+ βy ∈Y

Now to show that Y is not closed
Consider

( 1 1
2 3
1
x=( x n ) = 1 , , ,… … ,0,0 … ∈Y
n )
To show that ( x n )is Cauchy sequence

( 1 1 1
( x m ) =¿ 1 , 2 , 3 ,… … m , 0,0 … ) 1
if n< m n > m
1

{ 1 1 1
= ¿ m+1 , m+ 2 , … … n , 0,0 … . , , }

1
= m+1

1
as m→ ∞ →∞
m+ 1

‖x n− x m‖=0
⇒ (x n) is a Cauchy sequence

Now to prove that x n → y ∈ y

since (x ¿¿ n)→ y ¿ ¿ l∞ ; l ∞is complete
Now consider,

( 1 1
y= 1 , , , , , , , , ,,
2 3 )
‖x n− y‖ = ¿(|1−1|, |12− 12|, … …|1n − 1n|,|0− n+11 |,|0− n+21 |,… .)
1 1
= ¿( n+1 , n+2 , … …)

1
= n+1

1
As n → ∞ n+1 → ∞

xn → y

so y ∉Y
Y is not complete hence Y is not closed
Question :06
show that the closure Ῡ of subspace y of a normed space x is again vector
subspace.

Proof :

we have to prove that Ῡ is subspace .As у is a subspace of ꓫ.

Let x , y ϵῩ we have to show that αx+βy ϵ Ῡ ( α, β are scalars )

if x ϵῩ then their exist a sequence( x n ¿ in y such that x n→x

if y ϵῩ then their exist a sequence ( y n ¿in y such that y n→y

as ( x n)and ( y n) in y and y is subspace so α x n + β y n ϵ y

as α x n + β y n→αx + βy ( x n→ x & y n → y )
∴if x ∈ M iff their exist ( x n ¿ in M such that x n → x
x n in M x n → x ( x ∈ M ¿

as α x n + β y n → αx+ βy

⟹ αx + βy ∈ y

⇒ y is vector subspace

Show that of x n→ x & y n→ y

⟹ x n+ y n→ x + y , show that if α n→α & x n→ x
then α n x n →αx

proof:
since x n→ x & y n→ y then by definition
lim ‖x n−x‖ =0 & lim ‖ y n − y‖ =0
n→∞ n→∞

to prove x n + y n → x+ y
lim ¿= lim ‖( x n−x)+( y n− y )‖
n→∞ n→∞

≤ lim ‖x n−x‖+ lim ‖ y n− y‖ ≤0

n →∞ n→∞

⇒ xn + yn → x + y

now if α n → α & x n → x then nlim

→∞
‖αn−α‖=0
&
lim ‖x n−x‖=0
n→∞

to prove α n x n → αx

lim ∥ α n x n−αx∥=lim ∥ α n x n−α n x+ α n x−αx ∥

n→∞ n→∞

≤ nlim
→∞
‖α n x n−α n x‖+ nlim
→∞
‖α n x−αx‖

≤ lim ¿
n →∞

‖ ‖
≤|α n| lim x n−x + lim |α n−α | x
n→ ∞ n →∞ ‖‖
=0 as α n x n → αx
Question :
show that convergence of ‖ y 1‖+‖ y 2‖+‖ y 3‖+⋯
may not apply convergence of y 1 + y 2+ y 3+ ⋯

sol:
let y be the space of all sequence having finite terms are non zero.

Let y 1=¿ 0, 0 …)

y 2=¿ …)

y 3=¿ ,0,0,…)… y n=¿..)

 1
∥ y 1 ∥ =| | ¿ 1 y 1=( ,O , 0 …)
y1
2
1
1 1 1 1
∥ y 2 ∥=|y | ¿
2
= & ∥ y 3 ∥ = s uP| y 3| = 2 =
⋮
2 4
2
3 9
⋮

1
‖ y n‖=|y | ¿
n

2
n

‖ y 1‖+‖ y 2‖+‖ y 3‖+…

1 1 1
1+ 2
+ 2 +… 2
2 3 n
∞ ∞
1
∑ ‖ y i‖=∑ n
2
i=1 i=1

so it is convergent by p- series test (∴p=2>1)

∞

Now, ∑ yi = y 1+ y 2+ y 3 + … y n + …
i=1

s1= y 1 , s 2= y 1 + y 2 …. sn= y 1+ y 2 +… y n

sn=¿…)

lim s n=lim ¿,0,0,…)

n→∞

lim ¿( 1 , 1 , 1 ...¿ ∉ y
2 2 2
n→∞
1 2 3

sn converges in l ∞ but it is not in y so that converges of ‖ y 1‖+‖ y 2‖+¿ … does not

imply the converges of y 1 + y 2+ y 3+ … hence y is space of all sequences having
finite terms are non zero.

QUESTION :09
show that in a Banach space an absolutely
convergent series is convergent .
Proof :

let x be the Banach space and let ∑ x n be an absolutely convergent series

n =1

in x. then we have to show that

∞

∑ x nis convergent series .

n =1

let ε > ¿0 & let t n=‖x 1‖+‖x 2‖+…+‖ x n‖be nth partial sum of the series
∞ ∞

∑ ‖x n‖=‖x 1‖+‖x 2‖+…+‖x n‖ ∴ ∑ x n is absolutely convergent series so

n =1 n=1

∑ ‖x n‖ converges
n =1

so t nas a sequence of partial sum is convergent .

as
w e know t h at every convergent sequence isCauc h y so for every ε >0 t h eir exist n° ϵ N suc h t h at ∥t m−t n ∥<ε ∀ m ,n

∥¿)-(∥ x 1 ∥+∥ x 2 ∥+ …+∥ x n ∥ ¿ ∥< ε

∥ x n+1 ∥+∥ x n+2 ∥+… ∥ x m ∥<ε ……(eq¿ ¿

let sn=x 1 + x 2 +…+ x n then sm =x1 + x 2+ …+ x m

∥ sm- sn ∥ = ∥( x 1 + x 2 + …+ x m)-( x 1 + x 2 + …+ x n)∥

= ∥ x n+1+ x n+2 +…+ x m ∥

≤ ∥ xn +1 ∥+∥ x n+2 ∥+…+∥ x m ∥< ε (from eq ¿ ¿

∥ sm- sn ∥< ε

(sm )is a Cauchy sequence in x ans x is Banach space so( sm ) is convergent and sn is
∞

nth partial sum of the series ∑ x n is converges .

i=1
hence this series is convergent therefore every absolutely convergent series is
convergent .chy sequence in x and x is Banach space
(in Banach space every Cauchy sequence is convergent )

QUESTION 08:
If a normed space x, absolute convergence of any series always
implies convergence
of the series then show that x is complete. (Banach
space ).
Proof :
let ( x n)be a Cauchy sequence in x . Then for

each ε > 0 their exist a natural number n° ϵ N such that

∥ x m−¿ x ∥< ε ∀ ¿ m , n ≥ n°
n

1 1
in particular take ε = 2 then their exist n1 ϵ N such that ∥ x m- x n ∥< 2 ∀ m , n≥ n1

1
if we take ε = n ϵ N such that
2 then their exist 2
2

1
∥ x m −xn ∥< 2
∀ m, n ≥ n2
2

similarly , we have
1
∥ x m−x n ∥< ∀ m, n ≥ n3
23

1
∥ x m −xn ∥< ∀ m, n ≥ nk
2k

without any loss of generality, we assume that:

 n1<n2<n3<……..<nk<nk+1
So we have :
ǁ x n - x n ǁ <1/2
2 1

ǁ x n - x n ǁ <1/22
3 2

ǁ x n - x n ǁ <1/2k
k+ 1 k

Now consider the series

∞

∑ ǁxn - xn ǁ
k+ 1 k
i=1

Now comparing this series by series ∑ 1 /2 k

k =1

We note that ǁ x n – x n ǁ <1/2k , ∀ k k+ 1 k

But ∑ (1/2 k )is convergent ,

k =1

Now by comparison test:

∞

∑ ǁ x n - x n ǁ convergent
k+ 1 k
k =1

∑ ¿- x n ) …………….. (1)
k
k =1

is convergent.
Thus we conclude that the kth partial sum of the series …(1) has a limit say S.
Sk =( x n – x n )+( x n – x n ) +………..+( x n - x n )
2 1 3 2 k+ 1 k

Sk = x n - x n k+ 1 1

Now the kth partial sum has a limit

i.e, S=klim
→∞
¿ ¿S )= lim ¿- x n ¿
k k→∞ 1

lim x n = S+ x n
k+ 1
k→∞ 1
 Since , x n is convergent subsequence if the Cauchy sequence. Thus x n is
k+ 1

convergent. x n is arbitrary Cauchy sequence so it is complete.

Q10: Show that if a normed space has a schauder basis it is separable.
Proof: Let (en) be schauder basis for a normed space X. Then for each xϵ X there
exist unique set of scalars (α n)n=1∞❑ such that
∞

X=∑ ¿n en)
n =1

Then x=klim ❑S where S = ∑ ∞¿

en)
→∞ k, k n=1 n
¿
¿

then for every ε >0 , ε /2 > 0 there exist a natural number N such that
ǁ Sk –x ǁ < ε /2 ∀ k≥ N
In particular, we choose k=N
∑
Then SN = ǁn=1
∞α
n en - xǁ < ε /2 …………..(1)
¿
¿

∑
We define M={ n=1
∞β
n en :Nϵ N, β 1 , β 2 ,……. β N ϵ Q+Qi }
¿
¿

Note that M is countable since it is union of countable sets .We only show that M
is dense in X for this we will show that M = X
We choose β n then,
¿ β n - α n ¿< ε /2N ǁ en ǁ…………..(2)
N
∑
Now ǁ∑ β n en - x ǁ = ǁn=1
N β
n
∑
en -n=1
Nα
n
∑
en +n=1
Nα
n en - x ǁ
¿ ¿ ¿
n =1 ¿ ¿ ¿

∑
≤ ǁn=1
Nβ
n
∑
en -n=1
Nα
n
∑
en ǁ+ ǁn=1
Nα
n en - x ǁ
¿ ¿ ¿
¿ ¿ ¿

∑
= n=1
N¿β
n -α n
∑
¿ǁen ǁ+ ǁ n=1
Nα
n en - x ǁ
¿ ¿
¿ ¿

From eq 1 and 2
∑
< n=1
N
ε /2N + ε /2= ε
¿
¿
 ∑
ǁn=1
Nβ
n en - x ǁ< ε
¿
¿

∑
For x ϵ X there exist n=1
N β
n en ϵ M, such that , ǁ y - x ǁ <ε
¿
¿

M = X, Also M is countable , X Is separable

Q11:Show that (en) where en=(Snj) is a schauder basis for l p where 1≤p<∞ .
Proof: (en) is the sequence in X . (en) is schauder basis if for each x there is a
∑
unique set of scalars α n such that x=i=1
n¿¿
)
¿
¿

e1=(1,0,0,……)
e2=(0,1,0,……)
en=(0,0,…0,1,0….)
Show en is schauder basis for l p
Let x ϵ l p
∞
¿x1¿p + ¿x2¿p +……=∑ ❑∨¿xi¿p <∞
i=1

Also x=(x1 , x2 ,…..xn-1 , xn , xn+1 ,….) where x1 , x2 … are scalars

∞

Now we show that ∑ ❑xi ei

i=1

For this we show that the nth partial sum of the series converges to x and
remainder after n-terms is zero.Now
Sn= { ¿x1¿p + ¿x2¿p +……+ ¿xn¿p }
n
lim ❑S = lim ¿[∑ ❑∨¿x ¿p ]=S (say)
n→∞ n n→∞ i 1
i=1

lim ❑Sn =S1

n→∞

Remainder after n terms is zero so,

∞

Rn= ∑ ❑ ¿xi¿p
i> n+1
 ∞
lim ¿ R =[ ∑ ❑ ¿xi¿p ]
n→∞ n
i> n+1

lim ¿ R = 0
n
∑
, As Sn =i=1
n¿¿
)
n→∞ ¿
¿

Sn = α 1e1+ α 2 e2 +…….+ α n en
Sn= x1(1,0,0,……)+x2(0,1,0,……)+……+xn(0,0,…0,1,0….)
Sn=( x1 , x2 ,…..,xn ,0,0….)
∞

ǁ Sn - x ǁ=[ ∑ ❑ ¿Si –xi ¿p ]1/p

i=1

=[0+0….+ ¿xn+1¿p + ¿xn+2¿p …..]1/p

∞

=[ ∑ ❑ ¿xi¿p ]1/p =[ Rn ]1/p

i=n+1

lim ¿ ǁ S - x ǁ=0
n→∞ n

Sn → x , nth partial sum is convergent to x

∞
So x = ∑ ❑xi ei
i=1

Hence (en) is schauder basis for l p.

Q 12:A Semi norm on a vector space X is a mapping P : X → R , satisfies (N1),(N3)

and (N4). Show that P ( 0 )=0. Such that;
|P ( y )−P ( x )|≤ P ¿)
Solution :
As P : X → R satisfies three conditions of normed space stated as;

i. P ( x) ≥ 0

ii. P ( αx )=|α| P ( x )

iii. P ( x+ y ) ≤ P ( x ) + P ( y )

Firstly we have to prove that P ( 0 )=0

From condition (ii):

 P ( αx )=|α| P ( x )
Put α =0 (∴As α is any scalar)

⇒ P ( 0( x ) )=0 ( P ( x ) )

⇒ P ( 0 )=0

Now we have to prove |P ( y )−P ( x )|≤ P ¿)

Consider,

P ( y )= P ( y−x + x )

P ( y )= P (( y −x)+ x ) ≤ P ¿+ P ¿ (∴From Condition iii)

⇒ P ( y ) ≤ P¿ + P ¿
⇒ P ( y )−P ( x ) ≤ P¿ …(A)

Now consider,

P ( x ) = P ( x− y + y )

P ( x ) = P (( x− y)+ y ) ≤ P ¿+ P ¿ (∴From Condition iii)

⇒ P ( x )−P ( y ) ≤ P¿
Now,

P ( x ) −P( y )≤ P[ (−1 )( y −x ) ]
P ( x ) −P( y )≤|−1| P ( y−x )
P ( x ) −P( y )≤ P ( y −x ) ( ∴From Condition ii)

Multiply by ‘-1’ on both sides, we get

⇒ P ( y )−P ( x ) ≥−P ¿ …(B)

From (A) and (B):

⇒|P ( y )−P ( x )|≤ P ( y−x ) ∴|n|≤ a

−a ≤ n ≤ a
(Proved)

Q 13: Show that the element x ∈ X such that P ( x ) =0 from a subspace N of X and a norm on X /N
defined by ∥ ^x ∥ 0=P( x ) where x ∈ ^
X  and
^x ∈ X /N
Proof:

First we will prove that N={x ∈ X :P ( x )=0} is subspace of X .

Let x , y ∈ N

Since x ∈ N ⇒ P ( x ) =0

and y∈ N ⇒ P ( y )=0

Consider,

P ( x+ y ) ≤ P ( x ) + P ( y ) = 0+0 = 0 (∴From Condition iii)

So,

0 ≤ P ( x + y ) ≤0 ( ∴From Condition i)

⇒ P ( x+ y )=0
⇒ x+ y ∈N
Now consider α ∈ F and x ∈ N

P ( αx )=|α| P( x )

¿|α | (0) = 0
⇒ P ( αx ) =0
⇒ αx ∈ N
Hence, N is subspace of X

Now as;

X /N ¿ { x+ N : x ∈ X }
x + N= { x +n :n ∈ N }
As ^x =x + N

We have to prove that ∥ ^x ∥ 0 or ∥ x + N ∥ 0= P( x ) is Norm on X /N

N 1 ¿: ∥ ^x ∥ 0=P ( x ) ≥ 0 (∴ P ( x ) ≥ 0 )

⇒∥ ^x ∥0 ≥0

  N 2 ¿: ∥ ^x ∥ 0=0 ⇔ ^x =0

Consider,

∥ ^x ∥ 0=0

⇒ P ( x )=0
 ⇒x∈N
⇒ x + N=N
⇒ x^ =N =0
⇒ x^ =0
Thus,

∥ ^x ∥ 0=0 ⇔ x^ =0

N 3 ¿ and N 4 ¿ are satisfied obviously because ∥ ^x ∥ 0=P( x )

(Proved)

Q 1 4 : Let Y be a closed subspace of a Normed space ( X , ∥ ∎∥) . Show that a norm ∥ ∎ ∥0 on X /Y is

defined by ∥ ^x ∥ 0=inf x ∈ ^X ∥ x ∥, Where

^
X =X /Y
Proof:

Given;

by ∥ ^x ∥ 0=inf x ∈ ^X ∥ x ∥

∥ x + y ∥=inf {∥ x + y ∥ : y ∈ Y }

Where, ^
X =X /Y and X /Y = {x +Y : x ∈ X } and x +Y = { x + y 1 : y 1 ∈Y }

N 1 ¿: ∥ ^x ∥ 0 ≥ 0

Since ∥ x ∥ ≥ 0

⇒ inf ∥ x ∥ ≥0
Thus,

⇒ ∥ ^x ∥0 ≥0

  N 2 ¿: To prove is ∥ ^x ∥ 0=0 ⇔ x^ =0

Consider,

^x =0
⇒ x +Y =Y
⇒ x ∈Y
So,
 ∥ x +Y ∥ 0=inf y ∈Y ∥ x+ y ∥

= ∥ x + (−x ) ∥=0

⇒ ∥ x + y ∥=0
⇒ ∥ ^x ∥0=0 ⇔ ^x =0

Conversely,

Consider,

∥ ^x ∥ 0=0

⇒ inf ⁡{∥ x+ y ∥ : y ∈Y }

By definition of infimum, ∃ a sequence ( y n ) in Y such that;

lim ∥ x + y n ∥=0 ⇒ lim ( x+ y n)=0

n⟶∞ n⟶∞

⇒ x + nlim
⟶∞
y n¿0

(∴ ∥x ∥=0 ⇔ x=0

So, ∥ x + y n ∥=0 ⇒ x + y n=0 )

⇒ nlim
⟶∞
y n ¿−x

⇒−x ∈ Y (Because y n ∈Y and Y is closed)

Since, Y is Subspace,

(−1)(−x)∈Y … ¿ is any scalar)

⇒ x ∈Y
⇒ x +Y =Y
⇒ x^ =0
N 3 ¿ : To prove is ∥ α x^ ∥0=|α|∥ x^ ∥0

L.H.S = ∥α x^ ∥0

= ∥ αx+Y ∥ 0 … (' Y ' is identity)

= inf ⁡{∥ αx+ y ∥ : y ∈Y }

y
= inf ⁡{∥(α )( x+ )∥ : y ∈Y }
α
y
= inf |α|{∥ x + ∥ : y ∈Y }
α
 y y
= |α|inf {∥ x + ∥ : y ∈Y } … (Y is subspace, ⇒ ∈Y
α α
= |α|∥ ^x ∥ 0

=R.H. S

∥ α x ∥=inf { ∥αx + y ∥: y ∈ y }
1
¿ inf {|α ∨∥ x + y ∥ : yϵy }
α
1
¿∨α∨inf {∥ x + y ∥ : yϵy }
α
1 1
¿∨α∨inf {∥ x + y ∥ y ∈ y} ∴ yϵy
α : α
1
y is subspacemeanthat y∈ y
α
Hence
‖a x^‖0=¿ |α|inf {‖x + y‖: y ∈ y }
‖α ^x‖=|α|‖x^‖0
Hence N 3 is satisfied .
‖( x + y ) +( z + y )‖0=‖( x+ z )+ y‖0
= inf {‖( x + z ) + y‖: y ∈ y }
y∈y

= inf
y∈y ‖( ) ( )‖
y
x+ + z+
2
y
2
y
so, 2 = y
'

=inf ¿)
By using 4 property of norm: ‖x + y‖≤‖x‖+‖ y‖

≤ yinf ‖x + y ‖+ inf ‖ z+ y ‖
' '
' ∈Y y'∈y

≤‖x + y‖0+‖z+ y‖0

So,
‖( x + y ) + ( z + y )‖0 ≤‖x+ y‖0 +‖z + y‖0
Hence proved .
Question#15 (product of normed space)
If( x 1,∥. ∥1 ¿∧¿,‖.‖2) are normed space
Show that the product vector
x= x 1* x 2becomes a normed space. If we define : ‖x‖=max (‖ x1‖1 ,‖x 2‖2 ).
Proof:
We have to show that
‖x‖=max (‖ x‖1 ,‖ x2‖2)
N 1:

‖x‖≥ 0
max (‖x 1‖1 ,‖ x2‖2) ≥ 0 ∥ x 1 ∥1 ≥ 0∧∥ x 2 ∥ 2 ≥ 0

N 2:

‖x‖=0
∥ x ∥=¿ max (‖x 1‖1 ,‖ x2‖2) =0

(‖x 1‖1 ,‖x 2‖2)=( 0,0 )

‖x 1‖1=0 ,‖ x2‖2=0
x 1=0 , x 2=0

( x 1, x 2)=0
So,
∥x∥=0
N3 :

‖αx‖=max (‖α x 1‖1 ‖α x2‖2)

'

=max (|α|‖ x1‖1 ,|α |‖x 2‖2 )

=|α|max ¿ ¿)
‖αx‖=|α|‖x‖
 N4
Let x=( x 1, x 2) , y=( y 1, y 2)
‖x + y‖=‖( x 1 , x 2 ) + ( y 1 , y 2 )‖

=‖( x ,+ y 1 ) , ( x2 + y 2 )‖
=max (‖x 1+ y1‖1 ,‖x 2 + y 2‖2 )
≤ max (‖ x1‖1+‖ y 1‖,‖x 2‖2 +‖ y 2‖2 )

≤ max (‖ x1‖1 ,‖x 2‖2 ) + max (‖ y 1‖1 ‖ y 2‖2 )

'

‖x + y‖≤‖x‖+‖ y‖
Hence ∥ . ∥isnormed space on X= X 1∗X 2 .
LEMMA:
LET { x , x 2 , ϰ3 … xn } be the linearly independent set of vectors of a normed space
X . The there exist a real number C >0 such that for every choice of scalars
α 1 , α 2 ,α 3 ⋯ α n . We have

‖α x1 +α 2 x 2 +⋯ αn x n‖≥ C (|α1|+|α 2|+…|α n|)

THEOREM:
Every finite dimensional subspace Y of a normed space X is complete . In
particular , every finite dimensional normed space is complete.
Proof:
m m m
Let y m be the Cauchy sequence∈Y . thentheir exist scalor α 1 , α 2 , ⋯ α n . such that
n
y m=α m1 e 1+ α 2m e 2 +⋯ α mn e n=∑ α i ei .
( m)

i=1

By definition of Cauchy sequence:

For every ε >0 their exist natural number N ∈ N such that

‖ y m − y r‖< ε ∀ m, n> N “ OR”

‖∑ ‖ ‖∑ ( ‖
n n n
α imi −∑ α ir e i = α i −α i ) e i
( ) ( ) m r
ⅇ

i=1 i=1 i=1

by usinglemma their exist C >0 such that :

‖α x1 +α 2 x 2 +⋯ αn x n‖≥ C (|α1|+|α 2|+…|α n|)

 ‖∑ ( ‖
n
ε>
i=1
[
α mi −α ri ) e i ≥C |α m1 −α r1|+|α m2 −α r2|+ …+|α mn −α rn| ]
n
¿ C ∑ |α i −α i|<ε
m r

i=1
n

∑|αim−α ri|< ε /C
i=1

|α mi −αri|<ε /C ∀ m, r > N

This show that for each ¿ i. {α i } s aCauchy sequence∈¿ R∨C .

(m )

Since R∨C are complete space , so {α i }is convergent .

(m )

(m) m 1 2 m
a i → α i ( say ) α i =α i α i ⋯ α i
' '

Now we define:
n

Let y=∑ α i e i i=1

⇒ y∈y
Now we have to show that :
ym → y

n n
∥ y m− y ∥=∥ ∑ α e −∑ α i−ei ∥ (m )
i i
i=1 i=1
n
¿ ∥ ∑ ( α (im)−α i ) e i ∥
i=1

¿ ∥ ( α −α 1 ) e 1+ ( α 2 −α 2 ) e 2 +…+ ( α n −α n ) en ∥
(m ) (m ) ( m)
1
∵ by using (N 4)
≤ ∥ ( α −α 1) e1 ∥+∥ ( α −α 2 ) e 2 ∥+ …+∥ ( α −α n ) e n ∥
(m ) ( m) (m )
1 2 n
∵ by using(N ¿¿ 3)¿
¿|α −α 1|∥ e1 ∥+|α −α 2|∥ e2 ∥+⋅⋅ ⋅+|α −α n|∥e n ∥
(m )
1
(m )
2
(m )
n
n
¿ ∑|α (im )−α i|‖e i‖
i=1
n
⟹ ∥ y m− y ∥ ≤ ∑ |α (im )−α i|‖e i‖ ¿¿
i=1
n
Now let k =max i=1∥ e1 ∥
n
then ( 1 ) ⟹ ∥ y m− y ∥≤ k ∑ |α (im )−α i|
i=1
 ∵ ¿ ‖e2‖≤ k ,… ,‖en‖≤ k ¿
Letting m ⟶ ∞ we get
lim ∥ y m − y ∥=0 ∵ α (im ) ⟶ α i
m⟶∞
⟹ y m ⟶ y=α 1 e 1+ ¿ α 2 e2 +∙∙ ∙+¿ α n en ∈Y
Since ( y m ) is arbitrary Cauchy sequence. So Y is complete. Since X
Is a finite dimensional normed space then X itself being the finite dimensional
subspace of X is complete.
Hence every finite dimensional normed space is complete
_________________________
Theorem:
Every finite dimensional subspace Y of a normed space X
is closed in X .
Proof:
By using Theorem “A subspace of complete normed space is complete if and
only if it is closed.”
By using Theorem “Every finite dimensional normed space is complete.”
⇒ Finite dimensional subspace Y is complete and hence it is closed in X .
______________________
Note:
Infinite dimensional subspace need not to be closed.
Example:
Let X =C [ 0,1 ] and Y =span ( x 1 , x 2 ,∙ ∙∙ ) where x i ( t ) ¿ t i
So that Y is the set of all polynomials. ∴ Y is not closed in X because
x i ( t ) → e x ∉Y .
Equivalent Norms:
Let X be a normed space . Let ‖∙‖1and ‖∙‖2be two norms.
The norm ‖∙‖1 is said to be equivalent to ‖∙‖2. If there exist positive real number a
and b such that
a‖ x‖2 ≤‖x‖1 ≤ b‖x‖2 ∀ x ∈ X
If the norm ‖∙‖1 is equivalent to another ‖∙‖2.It is written as
‖∙‖1 ∼‖∙‖2
Theorem:
On a finite dimensional vector space X ,any norm ‖∙‖ is equivalent to any
other norm ‖∙‖0 .
Proof:
Let dim X =n and { e 1 , e2 ,∙ ∙ ∙e n } be any basis for X .Then for each x ∈ X there
exist a set of scalars α i such that
n
x=α 1 e1 +α 2 e 2 +∙∙ ∙+α n en =∑ α i e i
i=1
 ‖ ‖
n

Thus ∥ x ∥= ∑ αi e i
i=1

‖x‖=‖α 1 e1 +α 2 e2 +∙ ∙∙+α n en‖_________(1)

Since {e 1 , e2 ,∙ ∙ ∙e n } are basis for X .By using Lemma ( 2.4 .1 )There exist C>0such that
‖α 1 e1 +α 2 e2 +∙ ∙∙+α n en‖≥ C (|α 1|+|α 1|+∙ ∙ ∙+|α 1|)
OR
C (|α 1|+|α 1|+∙ ∙ ∙+|α 1|) ≤ ‖α 1 e1 +α 2 e2 +∙ ∙∙+α n en‖
⇒ C (|α 1|+|α 2|+…+|α n|) ≤‖x‖∵ using(1)
n
⇒ C ∑ |α i|≤‖x‖ ___________( 2 )
i=1

Now

‖ ‖
n n
‖x‖0= ∑ αi e i 0 ≤ ∑ ‖α i e i‖ ∵ by ( N 4 )
i=1 i=1
n
⇒ ‖x‖0 ≤ ∑ |α i|‖e i‖0 ∵ by ( N ¿¿ 3)¿
i=1
n
⇒ ‖x‖0 ≤ k ∑ |α i|
i=1
n
1
⇒ ‖ x‖ ≤ ∑ |α i|
k 0 i=1
n
c
⇒   ‖x‖ ≤ c ∑ |α i| ____________( 3 )
k 0 i=1

Now combining ( 2 )and (3)

n
c
⇒   ‖x‖ ≤ c ∑ |α i| ≤‖x‖
k 0 i=1
c
⇒‖ x‖0 ≤‖x‖ where a= > 0
k
Similarly changing the rule of ‖∙‖ and ‖⋅‖0
there exist C ' >0 such that
C (|α 1|+¿ α 2∨+∙ ∙∙+|α n|)≤ ∥α 1 e 1+ α 2 e 2+ …+α n e n ‖ ₒ
'

⇒C ' (|α 1|+|α 2|+ …+|α n|) ≤ ‖x‖0 ∵ using( 1)

n
⇒C
'
∑|αi|≤‖x‖0
i=1

Also

‖ ‖
n n
‖x‖= ∑ αi ei˙ ≤ ∑ ‖ αi ei̇ ‖ ∵ using(N 4 )
i=1 i=1
n
‖x‖≤ ∑ |α i̇|‖e i‖ ∵ using( N 3)
i=1
n
‖x‖≤ k ∑|α i| where k =max‖ei‖
i=1
 n
1
‖ x‖ ≤ ∑|α i|
k i= I
' n
C
‖x‖≤ C ' ∑|α i|→¿)
k i=1

combining (5) & (6)

n
C'
‖x‖≤ C ' ∑|α i|≤‖ x‖0
k i=1
Ċ '
‖x‖≤‖ x‖0
k
k
‖x‖≤ ' ‖x‖0
C
k
‖x‖≤ b‖x‖0 → ( 7 ) where b= '
>0
C
combining (4) & (7) we have
a‖ x‖0 ≤‖x‖≤ b‖x‖0 ∀(n∈ x)
Hence,
‖․‖ ‖․‖ ₒ

Proved
Inner product space (Hilbert space)
Let X be a vector space . A mapping < . , . > : X∗X → C
C is called an inner product on X if satisfy
following conditions :

IP 1) < x + y , z > = < x , z > + < Y , z >

IP 2) ¿ αx , y >¿ α < x , y >¿

IP 3) ¿ x , y >¿< y , x >¿

IP 4 ) ¿ x , x >≥ 0

And ¿ x , x >¿ 0if and only if x=0

Then this mapping is inner product space on X .
Examples:
1) Rn isinner product space
n
If x=( x 1 , x 2 , … , x n ) ∈ R
 y=( y ¿ ¿ 1 , y 2 , y 3 , … , y n )¿ ∈ Rn

Then ,
¿ x , y >¿ ( x 1 y 1+ x 2 y 2+ x 3 y 3+ …+ xn yn )
n n
¿ ∑ x i y i Ip 1 ¿< x + y , z >¿ = ∑ ( xi + yi ) z i
i=1 i=1
n
= ∑ ( xi z i+ yi z i )
i=1
n n

= ∑ x i zi + ∑ yi z i
i=1 i=1

¿ x+ y , z >¿< x , z >+ ¿ y , z> ¿

n
Ip 2 ¿ ¿ αx , y >¿ ∑ α i x i y i
i=1
n

= α ∑ xi yi
i=1

= α<x,y>
n n
Ip 3 ¿< x , y> ¿= ∑ x i y i=∑ y i xi
i=1 i=1

= <y,x>
n n
Ip 4 ¿ ¿ x , x >¿= ∑ x i xi =∑ x 2i
i=1 i=1
n

As , x 2i ≥ 0 ⇒ ∑ x 2i ≥ 0
i=1

⇒< x , x> ¿ 0
Now,
¿ x , x >¿ 0
n
⇔ ∑ x2i =¿ 0
i=1
2 2 2 2
⇔ x1 + x 2 + x3 + …+ x n=0
⇔ x 21=0 , x 22=0 , … , x2n =0
⟺ x=( x 1 , x 2 , x 3 , … , x n ) = 0
⟺ x=0
Hence all conditions are satisfied so < . , . > defines on inner
product space on Rn .
2) Example: if v=C n
< x , y > = x 1 y 1 + x 2 y 2 +…+ x n y n
n

= ∑ X i yi
i=1
3) Hilbert sequence space l2 :
Inner product space is linear in 1st variable and conjugate linear in 2nd
variable (factor) .
< αx , βy , z >¿< ax , z >¿ + ¿ βy , z >¿
= a< x , z>+ β< y , z >¿
= ¿ x ,αy >+¿ ¿ x , βz >¿
= ¿ x , αy + βz> ¿
= α ¿ x , y >+¿ β ¿ x , z >¿
Every inner product space is normed space * < x , x > = ‖x‖2.
⟹ complete inner product spaceis called Hilbert space.
Parallelogram equality :
The norm induced by an inner product satisfies following equality
‖x + y‖ +‖ x− y‖ =2 (‖x‖ +‖ y‖ )
2 2 2 2

Proof:
As norm coming form inner product space so,
2
‖x + y‖ =¿ x+ y , x+ y> ¿
= ¿ x+ y , x + y >¿
=¿ x , x >+ x , y >+¿ y , x>+ ¿ y , y >¿ → ( 1 )
& ‖x− y‖2=¿ x− y , x− y >
¿< x , x>−¿ x , y>−¿ y , x >+¿ y , y> ¿ →(2)
Adding (1) & (2)

2 2
‖x + y‖ +‖ x− y‖ =¿ x , x>+ ¿ x , x>+ ¿ y , x>+ ¿ y , y >+¿ x , x >−¿ x , y >−¿ y , x >+¿ y , y >¿
¿ 2< x , x >+2< y , y> ¿
= 2 ¿ ¿)
Hence proved
Note :
Every inner product space is normed space but converse is not true . Converse is
only true if normed space satisfies the parallelogram equality. Then normed space
is coming from inner product space .
Question :
1 2
Prove that ¿ x , y ≥ 4 (‖x + y‖¿ ¿2−‖x− y‖ ) ¿
Proof :
2
‖x + y‖ =¿ x+ y , x+ y> ¿
¿< x , x>+¿ x , y>+ ¿ y , x >+¿ y , y >¿ → (1)
 2
‖x− y‖ =¿ x− y , x− y >¿
¿< x , x>−¿ x , y>−¿ y , x >+¿ y , y> ¿ → (2)
Subtracting (2) from (1)
‖x + y‖ -‖x− y‖2 ¿< x , x>+¿ x , y>+ ¿ y , x >+¿ y , y >¿ -¿ x , x >+¿ x , y >+¿ y , x>−¿ y , y >¿
2

¿ 2< x , y >+2< y , x> ¿ ∴ ¿ y , x ≥¿ x , y >¿

¿ 4 < x , y >¿
1
¿ x , y >¿ ¿-‖x− y‖2 ¿
4

Polarization identity
Let v be a complex inner product space then
1
<x,y>= 4 [(||x+y|¿ -||x-y|¿ ¿+¿i(||x+iy|¿ −¿|x−iy|¿ ¿ → equation(1)
2 2 2 2

Proof
||x+y|¿2=<x+y,x+y>
=<x,x>+<x,y>+<y,x>+<y,y>
||x-y|¿2=<x-y,x-y>
=<x,x>-<x,y>-<y,x>+<y,y>
subtracting
|| x+y|¿2-||x-y|¿2 =¿2<x,y>+2<y,x>
=2 ¿ >) ∴ z+ z=2 Rez
=2[2Re<x,y>]
=4Re<x,y>
||x+iy|¿2=<x+iy,x+iy>
=< x,x >+< x+iy >+<iy+x>+<iy,iy>

=<x,x>-i<x,y>+i<y,x>+< y, y>
||x-iy|¿2 =<x-iy,x-iy>
= <x,x>+i<x,y>-i< y,x >+< y, y >
Subtracting
||x+iy|¿2- ||x-iy|¿2=<x,x>-i<x,y>+i<y,x> +<y,y>-<x,x> - i<x,y>+i<y,x>-<y,y>
||x+iy|¿2-||x-iy|¿2 =-2i<x,y>+2i<y,x>
 =-2i[<x,y>-⟨ x , y ⟩ ] ∴ z−z=2 iImg

=-2i[2i Img <x,y>]

=4Img <x,y>
Putting in equation (1)
1
<x,y>= 4 [ 4Re<x,y.+i4Im<x,y>]
=Re <x,y>+ iImg <x,y>
= <x,y>
Prove that l p P ≠ 2is not an inner products space
Proof
let
X=(1,1,0,0…)∈l p
Y=(1,-1,0,0…)∈l p
The norm on l p is
∞ 1 ∞

||x||=[∑ (|x i| ) ¿ ¿ ∴ ∑ ¿ xi ¿ p=2< ∞

P p

i=1 i=1

x +y = (2,0,0…)
x-y =(0,2,0…)
p p p
1 1
||x||=[|1¿ +|1¿ +|0¿ +….] p =2 p
1
||y|| = 2 p
p p p 1 1
||x+y||=[|2¿ +¿ |0¿ +¿ …] p =(2 ) p =2

p p p p 1 1
||x-y||=(|0¿ +|2¿ +¿ |0¿ +…) p =(2 ) p =2
Parrallelogram equality is

||x+y|¿2+||x-y|¿2 = 2[||x|¿2+||y|¿2]
 1 1
(2¿2+(2¿2=2[(2 P ¿2 +(2 P ¿2 ]
2 2
4+4 =2[2 p +2 p ]
2
8 =2[2.2 p ]
2
8=22.2 p
2
2+
8 ≠2 P
If p = 2 then
2
2+
8¿ 2 2

8=23
8=8
If p=2 then this is inner Hilbert space
Orthogonal vectors:
Let x be an inner product space we say x,y,ϵ x are orthogonal if their inner product
is zero i.e <x,y>=0 and we write x⊥ y . Let A,B,⊂X we write x⊥ A if x⊥ a ∀a∈ A and
A⊥ B if a⊥ b ∀a∈A , b∈ B
Example :
In R2 <x,y> = x 1 y 1+ x 2 y 2
x = ( x 1 , x2)
y =( y 1 , y 2)
(1,0) and (0,1) are orthogonal vectors
Question:
If X is real and ||x+y|¿2= ||x|¿2+||y|¿2 then x⊥y
Sol:
consider
||x+y|¿2-||x|¿2-||y|¿2=0
<x+y,x+y> -||x|¿2-||y|¿2=0
 <x,x>+ <x,y> +<y,x>+<y,y> -||x|¿2-||y|¿2 = 0
||x|¿2+2<x,y> +||y|¿2 -||x|¿2-||y|¿2 =0
2<x,y>= 0
<x,y> = 0

Prove:
1 1
¿|z−x|∨¿ ¿+¿|z− y|∨¿ = ||x− y|¿ +2|∨z − ( x + y ) ∨¿ ¿
2 2 2 2
2 2

Proof:
¿|x− y|∨¿2=¿ x− y , x− y >¿ ¿
=<x ,x > -<x,y>-¿ y , x >+¿ y , y >¿
=||x|¿2 −2 ℜ< x , y >+¿∨ y∨¿2
1 1 1
||z- 2 ( x+y ) |¿2=<z- 2 ( x+ y ) , z− 2 (x + y )>¿
1 1 1 1
=< z,z >-<z, 2 ( x+ y )> ¿ -< 2 ( x , y ) , z >+¿ 2 ( x + y ) , 2 ( x + y ) >¿
1 1 1
= <z,z>- 2 < z, x+y >- 2 <x+y,z>+ 4 <x+y,x+y>
1 −1 1 1 1 1
=<z,z>- 2 < z , x > 2 ⟨ z , y ⟩ - 2 <x,z>- 4 <x,x>+ 4 <y,x>+ 4 <y,y>
1 1 1
=||z|¿2-Re<z,x>- Re<z,y> + 4 ||x|¿2+ 2 Re<x,y>+ 4 ||y|¿2

RHS
1 2 1
¿|x− y|∨¿ ¿+2||z- (x+y)|¿2
2 2
1 1 1 +1
= 2 (||x|¿2-2Re<x,y>+||y|¿2)+2(||z|¿2 −ℜ< z , x >−¿ Re<z,y> + 4 ||x|¿2+ 2 ℜ< x , y> 4 ||
y|¿2)
1 1 1 1
= 2 ||x|¿2-Re<x,y> + 2 ||y|¿2+2||z¿2-2Re<z,x>-2Re<z,y>+ 2 ||x¿∨¿2 ¿+Re<x,y>+ 2 ||y|
¿2
=2||z¿∨¿2 ¿+||x|¿2+||y|¿2 -2Re<z,x>-2Re<z,y>
=(||Z|¿2 -2Re<z,x>+||x|¿2)+(||z|¿2 -2Re<z,y>+|y|¿2)
=||z−x∨¿ 2+||z-y|¿2
=L.H.S
This identify can also be obtained from the parallelogram equality
|| x ' + y' ∨¿2+ ¿∨x ' − y ' ∨¿ 2=2 ¿
Put x ' =z−x , y ' =z-y
||z-x+z-y|¿2+||z-x-z+y|¿2=2(||z-x|¿2 +¿∨z− y∨¿2)
||2z-x-y|¿2 +¿|−x + y|¿2=2(¿|z−x|¿2 +¿|z− y|¿2 )
1 2 2
||2[z- 2 ( x+ y )∨¿ +¿|x− y|¿ =2¿ +||z-y|¿ ¿
2

2
1 2 2 2 2
|2¿ ||z 2 (x+y)|¿ +||x-y|¿ =2(||z-x|¿ +||z-y|¿
1 2 2 2 2
4||z- 2 (x+y)|¿ +||x-y|¿ =2(||z-x|¿ +||z-y|¿ )
1 2
1 2 2 2
2||z- 2 (x+y)|¿ + 2 ||x-y|¿ =||z-x|¿ +||z-y|¿
Hence prove