INNER PRODUCT SPACES

Hoang Hai Ha

HCMUT-OISP
Email: [email protected]

September 17, 2023

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 1 / 29

O UTLINE

1 R EAL WORLD PROBLEMS

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 2 / 29

O UTLINE

1 R EAL WORLD PROBLEMS

2 R EAL INNER PRODUCT SPACE

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 2 / 29

O UTLINE

1 R EAL WORLD PROBLEMS

2 R EAL INNER PRODUCT SPACE

3 O RTHOGONALITY

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 2 / 29

O UTLINE

1 R EAL WORLD PROBLEMS

2 R EAL INNER PRODUCT SPACE

3 O RTHOGONALITY

4 M AT L AB

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 2 / 29

 →
−
W ORK DONE BY A FORCE F

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 3 / 29

 →
−
W ORK DONE BY A FORCE F

→
− −
W = F .→
s = F.s. cos α

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 3 / 29

A real vector space V is called a real Euclidean inner product space if
〈·〉· : V × V → R
(x, y) 7−→ 〈x, y〉− which is called inner product of 2 vectors.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 4 / 29

A real vector space V is called a real Euclidean inner product space if
〈·〉· : V × V → R
(x, y) 7−→ 〈x, y〉− which is called inner product of 2 vectors.
The following 4 axioms are satisfied

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 4 / 29

A real vector space V is called a real Euclidean inner product space if
〈·〉· : V × V → R
(x, y) 7−→ 〈x, y〉− which is called inner product of 2 vectors.
The following 4 axioms are satisfied
1 〈x, y〉 = 〈y, x〉, ∀x, y ∈ V

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 4 / 29

A real vector space V is called a real Euclidean inner product space if
〈·〉· : V × V → R
(x, y) 7−→ 〈x, y〉− which is called inner product of 2 vectors.
The following 4 axioms are satisfied
1 〈x, y〉 = 〈y, x〉, ∀x, y ∈ V
2 〈x + y, z〉 = 〈x, z〉 + 〈y, z〉, ∀x, y, z ∈ V

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 4 / 29

A real vector space V is called a real Euclidean inner product space if
〈·〉· : V × V → R
(x, y) 7−→ 〈x, y〉− which is called inner product of 2 vectors.
The following 4 axioms are satisfied
1 〈x, y〉 = 〈y, x〉, ∀x, y ∈ V
2 〈x + y, z〉 = 〈x, z〉 + 〈y, z〉, ∀x, y, z ∈ V
3 〈αx, y〉 = α〈x, y〉, ∀x, y ∈ V, ∀α ∈ R

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 4 / 29

A real vector space V is called a real Euclidean inner product space if
〈·〉· : V × V → R
(x, y) 7−→ 〈x, y〉− which is called inner product of 2 vectors.
The following 4 axioms are satisfied
1 〈x, y〉 = 〈y, x〉, ∀x, y ∈ V
2 〈x + y, z〉 = 〈x, z〉 + 〈y, z〉, ∀x, y, z ∈ V
3 〈αx, y〉 = α〈x, y〉, ∀x, y ∈ V, ∀α ∈ R
4 〈x, x〉 > 0, x ̸= 0 and 〈x, x〉 = 0 ⇔ x = 0.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 4 / 29

A real vector space V is called a real Euclidean inner product space if
〈·〉· : V × V → R
(x, y) 7−→ 〈x, y〉− which is called inner product of 2 vectors.
The following 4 axioms are satisfied
1 〈x, y〉 = 〈y, x〉, ∀x, y ∈ V
2 〈x + y, z〉 = 〈x, z〉 + 〈y, z〉, ∀x, y, z ∈ V
3 〈αx, y〉 = α〈x, y〉, ∀x, y ∈ V, ∀α ∈ R
4 〈x, x〉 > 0, x ̸= 0 and 〈x, x〉 = 0 ⇔ x = 0.

E XAMPLE 2.1
On Rn , dot product is standard inner product
n
x i y i = x.y T
P
(x, y) 7−→ 〈x, y〉 =
i =1
where x = (x 1 , x 2 , . . . , x n ), y = (y 1 , y 2 , . . . , y n ).

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 4 / 29

E XAMPLE 2.2
On R2 we define the mapping (x, y) → 〈x, y〉 ∈ R as below:

(x, y) 7−→ 〈x, y〉 = x 1 .y 1 + 2x 2 .y 2 + x 1 y 2 + x 2 y 1

where x = (x 1 , x 2 ), y = (y 1 , y 2 ).
Explain why this mapping is inner product?
Find 〈x, y〉 where x = (1, 1), y = (2, −1).

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 5 / 29

E XAMPLE 2.2
On R2 we define the mapping (x, y) → 〈x, y〉 ∈ R as below:

(x, y) 7−→ 〈x, y〉 = x 1 .y 1 + 2x 2 .y 2 + x 1 y 2 + x 2 y 1

where x = (x 1 , x 2 ), y = (y 1 , y 2 ).
Explain why this mapping is inner product?
Find 〈x, y〉 where x = (1, 1), y = (2, −1).

〈y, x〉 = y 1 .x 1 + 2y 2 x 2 + y 1 x 2 + y 2 x 1 = 〈x, y〉.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 5 / 29

E XAMPLE 2.2
On R2 we define the mapping (x, y) → 〈x, y〉 ∈ R as below:

(x, y) 7−→ 〈x, y〉 = x 1 .y 1 + 2x 2 .y 2 + x 1 y 2 + x 2 y 1

where x = (x 1 , x 2 ), y = (y 1 , y 2 ).
Explain why this mapping is inner product?
Find 〈x, y〉 where x = (1, 1), y = (2, −1).

〈y, x〉 = y 1 .x 1 + 2y 2 x 2 + y 1 x 2 + y 2 x 1 = 〈x, y〉.

For z = (z 1 , z 2 ), 〈x + y, z〉 = 〈x, z〉 + 〈y, z〉.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 5 / 29

E XAMPLE 2.2
On R2 we define the mapping (x, y) → 〈x, y〉 ∈ R as below:

(x, y) 7−→ 〈x, y〉 = x 1 .y 1 + 2x 2 .y 2 + x 1 y 2 + x 2 y 1

where x = (x 1 , x 2 ), y = (y 1 , y 2 ).
Explain why this mapping is inner product?
Find 〈x, y〉 where x = (1, 1), y = (2, −1).

〈y, x〉 = y 1 .x 1 + 2y 2 x 2 + y 1 x 2 + y 2 x 1 = 〈x, y〉.

For z = (z 1 , z 2 ), 〈x + y, z〉 = 〈x, z〉 + 〈y, z〉.
〈αx, y〉 = α〈x, y〉.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 5 / 29

E XAMPLE 2.2
On R2 we define the mapping (x, y) → 〈x, y〉 ∈ R as below:

(x, y) 7−→ 〈x, y〉 = x 1 .y 1 + 2x 2 .y 2 + x 1 y 2 + x 2 y 1

where x = (x 1 , x 2 ), y = (y 1 , y 2 ).
Explain why this mapping is inner product?
Find 〈x, y〉 where x = (1, 1), y = (2, −1).

〈y, x〉 = y 1 .x 1 + 2y 2 x 2 + y 1 x 2 + y 2 x 1 = 〈x, y〉.

For z = (z 1 , z 2 ), 〈x + y, z〉 = 〈x, z〉 + 〈y, z〉.
〈αx, y〉 = α〈x, y〉.
〈x, x〉 = x 12 + 2x 22 + 2x 1 x 2 = (x 1 + x 2 )2 + x 22 ≥ 0 and
〈x, x〉 = 0 ⇔ x = (x 1 , x 2 ) = (0, 0).

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 5 / 29

L ENGTH OF A VECTOR IN AN INNER PRODUCT SPACE

D EFINITION 2.1
If V is a real inner product space, then the length of a vector x ∈ V is
denoted by ||x|| and is defined by

p
||x|| = 〈x, x〉 . (1)

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 6 / 29

L ENGTH OF A VECTOR IN AN INNER PRODUCT SPACE

D EFINITION 2.1
If V is a real inner product space, then the length of a vector x ∈ V is
denoted by ||x|| and is defined by

p
||x|| = 〈x, x〉 . (1)

D EFINITION 2.2
If V is a real inner product space, then the distance between two
vectors u, v ∈ V is denoted by d (u, v) and is defined by

p
d (u, v) = ||u − v|| = 〈u − v, u − v〉 (2)

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 6 / 29

D EFINITION 2.3
The angle α between two vectors x, y ∈ V is defined by

〈x, y〉
cos α = , (0 É α É π)
||x||.||y||

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 7 / 29

D EFINITION 2.3
The angle α between two vectors x, y ∈ V is defined by

〈x, y〉
cos α = , (0 É α É π)
||x||.||y||

E XAMPLE 2.3
On R2 the inner product is given

〈x, y〉 = 3x 1 y 1 + x 1 y 2 + x 2 y 1 + x 2 y 2

where x = (x 1 , x 2 ), y = (y 1 , y 2 ). Find the length of vector x = (1, 2). Find

the distance between u = (1, −1), v = (0, 2) and the angle between
them.
Answer.
p
∥x∥ = 11.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 7 / 29

D EFINITION 2.3
The angle α between two vectors x, y ∈ V is defined by

〈x, y〉
cos α = , (0 É α É π)
||x||.||y||

E XAMPLE 2.3
On R2 the inner product is given

〈x, y〉 = 3x 1 y 1 + x 1 y 2 + x 2 y 1 + x 2 y 2

where x = (x 1 , x 2 ), y = (y 1 , y 2 ). Find the length of vector x = (1, 2). Find

the distance between u = (1, −1), v = (0, 2) and the angle between
them.
Answer.
p
∥x∥ = 11. p
u − v = (1, −3), d (u,
p v) = 6.
〈u, v〉 = 0, ∥u∥ = 2, ∥v∥ = 2.
Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 7 / 29
M ATRIX FORM OF INNER PRODUCT

E VALUATING INNER PRODUCT USING MATRIX FORM

Every inner product in Rn can be written in matrix form, i.e, for every
pair of vectors x = (x 1 , x 2 , .., x n ), y = (y 1 , y 2 , ..., y n ) their inner product
is:

〈x, y〉 = x Ay T

where A is a symmetric square matrix whose coefficients a i j are

numbers standing before x i y j .

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 8 / 29

E XAMPLE 2.4
In R3 the inner product is given

〈x, y〉 = x 1 y 1 + 2x 1 y 2 + 2x 2 y 1 + 5x 2 y 2 + 3x 3 y 3

where x = (x 1 , x 2 , x 3 ), y = (y 1 , y 2 , y 3 ). And given u = (1, 1, 1), v = (1, 0, 1).

Determine 〈u, v〉 and angle between u, v.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 9 / 29

E XAMPLE 2.4
In R3 the inner product is given

〈x, y〉 = x 1 y 1 + 2x 1 y 2 + 2x 2 y 1 + 5x 2 y 2 + 3x 3 y 3

where x = (x 1 , x 2 , x 3 ), y = (y 1 , y 2 , y 3 ). And given u = (1, 1, 1), v = (1, 0, 1).

Determine 〈u, v〉 and angle between u, v.

Solution.
Rewrite theformulaby using the matrix form: 〈x, y〉 = x Ay T ,
1 2 0
where A = 2 5 0.

0 0 3

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 9 / 29

E XAMPLE 2.4
In R3 the inner product is given

〈x, y〉 = x 1 y 1 + 2x 1 y 2 + 2x 2 y 1 + 5x 2 y 2 + 3x 3 y 3

where x = (x 1 , x 2 , x 3 ), y = (y 1 , y 2 , y 3 ). And given u = (1, 1, 1), v = (1, 0, 1).

Determine 〈u, v〉 and angle between u, v.

Solution.
T
Rewrite theformulaby using the matrix form: 〈x,  y〉
 = x Ay ,
1 2 0 ¡ ¢ 1
where A = 2 5 0 . Then 〈u, v〉 = 1 1 1 A 0
  
0 0 3 1
Find 〈u, v〉, 〈u, u〉, 〈v, v〉 by multiplying
µ ¶ 1 1 µ ¶ µ ¶
1 1 1  〈u, u〉 〈u, v〉 13 6
A 1 0 =  = .
1 0 1 〈v, u〉 〈v, v〉 6 4
1 1

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 9 / 29

O RTHOGONALITY

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 10 / 29

O RTHOGONALITY

D EFINITION 3.1
1 Two vectors x, y ∈ V in an inner product space V is called
orthogonal ⇔ 〈x, y〉 = 0. We denote it by x ⊥ y.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 11 / 29

O RTHOGONALITY

D EFINITION 3.1
1 Two vectors x, y ∈ V in an inner product space V is called
orthogonal ⇔ 〈x, y〉 = 0. We denote it by x ⊥ y.
2 Vector x is orthogonal to the set M ⊂ V if x is orthogonal to every
vector in M . We denote it by x ⊥ M .

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 11 / 29

E XAMPLE 3.1
On R2 the inner product is given

〈x, y〉 = 2x 1 y 1 − x 1 y 2 − x 2 y 1 + x 2 y 2

where x = (x 1 , x 2 ), y = (y 1 , y 2 ), and let u = (1, −1), v = (2, m). Find m

such that u ⊥ v.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 12 / 29

E XAMPLE 3.1
On R2 the inner product is given

〈x, y〉 = 2x 1 y 1 − x 1 y 2 − x 2 y 1 + x 2 y 2

where x = (x 1 , x 2 ), y = (y 1 , y 2 ), and let u = (1, −1), v = (2, m). Find m

such that u ⊥ v.
Ans: m = 3

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 12 / 29

T HEOREM 3.1
Vector x is orthogonal to a subspace F if and only if x is orthogonal to
a basis of F .

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 13 / 29

T HEOREM 3.1
Vector x is orthogonal to a subspace F if and only if x is orthogonal to
a basis of F .

E XAMPLE 3.2
On R3 given the standard inner product and the subspace
½ ¯ ¾
3 ¯ x 1 + 2x 2 − x 3 = 0
F = x = (x 1 , x 2 , x 3 ) ∈ R ¯
¯
2x 1 + 3x 2 + x 3 = 0

Find m such that vector v = (2, 3, m) is orthogonal to F .

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 13 / 29

T HEOREM 3.1
Vector x is orthogonal to a subspace F if and only if x is orthogonal to
a basis of F .

E XAMPLE 3.2
On R3 given the standard inner product and the subspace
½ ¯ ¾
3 ¯ x 1 + 2x 2 − x 3 = 0
F = x = (x 1 , x 2 , x 3 ) ∈ R ¯
¯
2x 1 + 3x 2 + x 3 = 0

Find m such that vector v = (2, 3, m) is orthogonal to F .

A basis of F is {e = (−5, 3, 1)}.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 13 / 29

T HEOREM 3.1
Vector x is orthogonal to a subspace F if and only if x is orthogonal to
a basis of F .

E XAMPLE 3.2
On R3 given the standard inner product and the subspace
½ ¯ ¾
3 ¯ x 1 + 2x 2 − x 3 = 0
F = x = (x 1 , x 2 , x 3 ) ∈ R ¯
¯
2x 1 + 3x 2 + x 3 = 0

Find m such that vector v = (2, 3, m) is orthogonal to F .

A basis of F is {e = (−5, 3, 1)}.

v ⊥ F ⇔ v ⊥ e ⇔ −10 + 9 + m = 0.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 13 / 29

O RTHOGONAL AND O RTHONORMAL S ETS

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 14 / 29

D EFINITION 3.2
1 A
n set of two oromore vectors in a real inner product space
x 1 , x 2 , . . . , x n is called orthogonal ⇔ all pairs of distinct vectors
in the set are orthogonal.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 15 / 29

D EFINITION 3.2
1 A
n set of two oromore vectors in a real inner product space
x 1 , x 2 , . . . , x n is called orthogonal ⇔ all pairs of distinct vectors
in the set are orthogonal.
2 An orthogonal set in which each vector has norm 1 is said to be
orthonormal
||x k || = 1, (k = 1, 2, . . . , n)

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 15 / 29

D EFINITION 3.2
1 A
n set of two oromore vectors in a real inner product space
x 1 , x 2 , . . . , x n is called orthogonal ⇔ all pairs of distinct vectors
in the set are orthogonal.
2 An orthogonal set in which each vector has norm 1 is said to be
orthonormal
||x k || = 1, (k = 1, 2, . . . , n)

E XAMPLE 3.3
On Rn2 the standard
o inner product is given. Then the set
M = (1, −2), (2, 1) is orthogonal set.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 15 / 29

D EFINITION 3.2
1 A
n set of two oromore vectors in a real inner product space
x 1 , x 2 , . . . , x n is called orthogonal ⇔ all pairs of distinct vectors
in the set are orthogonal.
2 An orthogonal set in which each vector has norm 1 is said to be
orthonormal
||x k || = 1, (k = 1, 2, . . . , n)

E XAMPLE 3.3
On Rn2 the standard
o inner product is given. Then the set
M = (1, −2), (2, 1) is orthogonal set.
½µ ¶ µ ¶¾
1 2 2 1
N = p ,−p , p , p is the orthonormal set.
5 5 5 5

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 15 / 29

O RTHOGONAL C OMPLEMENTS

D EFINITION 3.3
If F is a subspace of a real inner product space V, then the set F ⊥ of all
vectors in V that are orthogonal to F is called the orthogonal
complement of F.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 16 / 29

O RTHOGONAL C OMPLEMENTS

D EFINITION 3.3
If F is a subspace of a real inner product space V, then the set F ⊥ of all
vectors in V that are orthogonal to F is called the orthogonal
complement of F.

T HEOREM 3.2
Let F be a subspace of a real inner product space V , then F ⊥ is a
subspace of V , and:

dimF + dimF ⊥ = dimV

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 16 / 29

F INDING F ⊥ OF A SUBSPACE F

1 Find a basis of F . Assume that basis of F contains vectors

{e 1 , e 2 , .., e m }
 ¯ 
 ¯ 〈e 1 , y〉 = 0 
 ¯ 

 ¯ 〈e 2 , y〉 = 0 

F ⊥ = (y 1 , y 2 , ...y n ) ¯ ⇒ find the d i m and a basis
¯
2
..



¯
¯ . 


 ¯ 〈e m , y〉 = 0 
of this null space.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 17 / 29

E XAMPLE 3.4
n o
Let F = (x 1 , x 2 , x 3 ) ∈ R3 : x 1 + x 2 + x 3 = 0 be the subspace of R3 and
given the inner product

〈x, y〉 = x 1 y 1 + 2x 1 y 2 + 2x 2 y 1 + 5x 2 y 2 + 3x 3 y 3

Find the dim and a basis of F ⊥ .

A basis of F is {e 1 = (−1, 1, 0), e 2 = (−1, 0, 1)}.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 18 / 29

E XAMPLE 3.4
n o
Let F = (x 1 , x 2 , x 3 ) ∈ R3 : x 1 + x 2 + x 3 = 0 be the subspace of R3 and
given the inner product

〈x, y〉 = x 1 y 1 + 2x 1 y 2 + 2x 2 y 1 + 5x 2 y 2 + 3x 3 y 3

Find the dim and a basis of F ⊥ .

A basis of F is {e 1 = (−1, 1, 0), e 2 = (−1, 0, 1)}. 

1 2 0
The matrix in inner product formula is A = 2 5 0.
0 0 3

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 18 / 29

E XAMPLE 3.4
n o
Let F = (x 1 , x 2 , x 3 ) ∈ R3 : x 1 + x 2 + x 3 = 0 be the subspace of R3 and
given the inner product

〈x, y〉 = x 1 y 1 + 2x 1 y 2 + 2x 2 y 1 + 5x 2 y 2 + 3x 3 y 3

Find the dim and a basis of F ⊥ .

A basis of F is {e 1 = (−1, 1, 0), e 2 = (−1, 0, 1)}. 

1 2 0
The matrix in inner product formula is A = 2 5 0.
0 0 3
⊥
Let y = (y 1 , y 2 , y 3 ) ∈ F , we have
〈e 1 , y〉 = e 1 Ay T = 0, 〈e 2 , y〉 = e 2 Ay T = 0.
 
µ ¶ µ ¶ y1
−1 1 0 1 3 0  
⇒ Ay T = 0 ⇒ y 2 = 0. Thus,
−1 0 1 −1 −2 3
y3
⊥ ⊥
dim F = 1, a basis of F is {(9, −3, 1)}.
Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 18 / 29
E XAMPLE 3.5
In R3 given the inner product x = (x 1 , x 2 , x 3 ), y = (y 1 , y 2 , y 3 ) as:

〈x, y〉 = 4x 1 y 1 + 5x 2 y 2 − x 2 y 3 − x 3 y 2 + 4x 3 y 3 ,

and the subspace F = span {(1, 2, 1), (2, −1, 4), (1, −3, 3), (5, 0, 9)}. Find
the dim of F ⊥ and one its basis.

Solution.
A basis of F is E = {e 1 = (1, 2, 1), e 2 = (2, −1, 4)}.
Similarly to the previous example,
 we find
µ ¶ 4 0 0
1 2 1 
B= 0 5 −1, then F ⊥ ≡ nul l (B ).
2 −1 4
0 −1 4

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 19 / 29

O RTHOGONAL P ROJECTIONS

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 20 / 29

T HEOREM 3.3
If F is a finite-dimensional subspace of an inner product space V, then
every vector x in V can be expressed in exactly one way as

x = y + z,

where y ∈ F, z ∈ F ⊥ .

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 21 / 29

T HEOREM 3.3
If F is a finite-dimensional subspace of an inner product space V, then
every vector x in V can be expressed in exactly one way as

x = y + z,

where y ∈ F, z ∈ F ⊥ .

D EFINITION 3.4
Vector y is called the orthogonal projection of x on F . We denote it by
y = pr F x.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 21 / 29

F IND y = PRF x

1 Find a basis of F , let it be S = {e 1 , e 2 , ..., e m }.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 22 / 29

F IND y = PRF x

1 Find a basis of F , let it be S = {e 1 , e 2 , ..., e m }.

2 Since y ∈ F ⇒ y = λ1 e 1 + λ2 e 2 + ... + λm e m

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 22 / 29

F IND y = PRF x

1 Find a basis of F , let it be S = {e 1 , e 2 , ..., e m }.

2 Since y ∈ F ⇒ y = λ1 e 1 + λ2 e 2 + ... + λm e m
3 x = y + z, then

x = λ1 e 1 + λ2 e 2 + ... + λm e m + z

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 22 / 29

F IND y = PRF x

1 Find a basis of F , let it be S = {e 1 , e 2 , ..., e m }.

2 Since y ∈ F ⇒ y = λ1 e 1 + λ2 e 2 + ... + λm e m
3 x = y + z, then

x = λ1 e 1 + λ2 e 2 + ... + λm e m + z



 〈x, e 1 〉 = 〈e 1 , e 1 〉λ1 + 〈e 2 , e 1 〉λ2 + ... + 〈e m , e 1 〉λm

〈x, e 2 〉 = 〈e 1 , e 2 〉λ1 + 〈e 2 , e 2 〉λ2 + ... + 〈e m , e 2 〉λm

⇔ .
..



〈x, e m 〉 = 〈e 1 , e m 〉λ1 + 〈e 2 , e m 〉λ2 + ... + 〈e m , e m 〉λm .


Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 22 / 29

D EFINITION 3.5
The distance between a vector x and the subspace F is defined by

d (x, F ) = ||x − projF x||. (3)

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 23 / 29

E XAMPLE 3.6
n o
On R3 given the dot product, the subspace F = span (1, 1, 1), (0, 1, 1)
and the vector x = (1, 1, 2) are given. Find the orthogonal projection
pr F x of x on F .

Solution.
Let y ∈ F be the projection vector, then y = α1 (1, 1, 1) + α2 (0, 1, 1).
Denote e 1 = (1, 1, 1), e 2 = (0, 1, 1).

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 24 / 29

E XAMPLE 3.6
n o
On R3 given the dot product, the subspace F = span (1, 1, 1), (0, 1, 1)
and the vector x = (1, 1, 2) are given. Find the orthogonal projection
pr F x of x on F .

Solution.
Let y ∈ F be the projection vector, then y = α1 (1, 1, 1) + α2 (0, 1, 1).
Denote e 1 = (1, 1, 1), e 2 = (0, 1, 1).
We solve the system
(
〈x, e 1 〉 = 〈e 1 , e 1 〉α1 + 〈e 1 , e 2 〉α2
〈x, e 2 〉 = 〈e 1 , e 2 〉α1 + 〈e 2 , e 2 〉α2
(
4 = 3α1 + 2α2
→ → Solve α1 , α2 → y.
3 = 2α1 + 2α2

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 24 / 29

E XAMPLE 3.7
In R3 given the inner product x = (x 1 , x 2 , x 3 ), y = (y 1 , y 2 , y 3 ) as:

〈x, y〉 = 4x 1 y 1 + 5x 2 y 2 − x 2 y 3 − x 3 y 2 + 4x 3 y 3 ,

and the subspace F = {(x 1 , x 2 , x 3 ) : −x 1 + x 2 + 2x 3 = 0}. Find the

orthogonal projection of the vector x = (2, 1, 5) onto F .

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 25 / 29

E XAMPLE 3.7
In R3 given the inner product x = (x 1 , x 2 , x 3 ), y = (y 1 , y 2 , y 3 ) as:

〈x, y〉 = 4x 1 y 1 + 5x 2 y 2 − x 2 y 3 − x 3 y 2 + 4x 3 y 3 ,

and the subspace F = {(x 1 , x 2 , x 3 ) : −x 1 + x 2 + 2x 3 = 0}. Find the

orthogonal projection of the vector x = (2, 1, 5) onto F .

A basis of F is {e 1 = (1, 1, 0), e 2 = (2, 0, 1)}.

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 25 / 29

E XAMPLE 3.7
In R3 given the inner product x = (x 1 , x 2 , x 3 ), y = (y 1 , y 2 , y 3 ) as:

〈x, y〉 = 4x 1 y 1 + 5x 2 y 2 − x 2 y 3 − x 3 y 2 + 4x 3 y 3 ,

and the subspace F = {(x 1 , x 2 , x 3 ) : −x 1 + x 2 + 2x 3 = 0}. Find the

orthogonal projection of the vector x = (2, 1, 5) onto F .

A basis of F is {e 1 = (1, 1, 0), e 2 = (2, 0, 1)}.

 
4 0 0
The matrix in inner product formula is A = 0 5 −1.
0 −1 4

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 25 / 29

E XAMPLE 3.7
In R3 given the inner product x = (x 1 , x 2 , x 3 ), y = (y 1 , y 2 , y 3 ) as:

〈x, y〉 = 4x 1 y 1 + 5x 2 y 2 − x 2 y 3 − x 3 y 2 + 4x 3 y 3 ,

and the subspace F = {(x 1 , x 2 , x 3 ) : −x 1 + x 2 + 2x 3 = 0}. Find the

orthogonal projection of the vector x = (2, 1, 5) onto F .

A basis of F is {e 1 = (1, 1, 0), e 2 = (2, 0, 1)}.

 
4 0 0
The matrix in inner product formula is A = 0 5 −1.
0 −1 4
     
2 1 5 1 2 8 35
Evaluate: 1 1 0 A 1 0 = 9 7 .
    
2 0 1 0 1 7 20

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 25 / 29

E XAMPLE 3.7
In R3 given the inner product x = (x 1 , x 2 , x 3 ), y = (y 1 , y 2 , y 3 ) as:

〈x, y〉 = 4x 1 y 1 + 5x 2 y 2 − x 2 y 3 − x 3 y 2 + 4x 3 y 3 ,

and the subspace F = {(x 1 , x 2 , x 3 ) : −x 1 + x 2 + 2x 3 = 0}. Find the

orthogonal projection of the vector x = (2, 1, 5) onto F .

A basis of F is {e 1 = (1, 1, 0), e 2 = (2, 0, 1)}.

 
4 0 0
The matrix in inner product formula is A = 0 5 −1.
0 −1 4
     
2 1 5 1 2 8 35
Evaluate: 1 1 0 A 1 0 = 9 7 .
    
2 0 1 0 1 7 20
Then the orthogonal
( projection is y = α1 e 1 + α2 e 2 , we find
8 = 9α1 + 7α2
α1 , α2 from: → find y.
35 = 7α1 + 20α2
Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 25 / 29
T HE STANDARD INNER PRODUCT ON Rn

1 < x, y >= d ot (x, y)

2 ||x|| = nor m(x)
3 d (x, y) = nor m(x − y)
4 cos α = d ot (x, y)/(nor m(x) ∗ nor m(y))

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 26 / 29

O RTHOGONAL C OMPLEMENT

1 f 1 , f 2 , . . . , f m : basis of F. A = [ f 1 ; f 2 ; . . . ; f m ]
 
f1
 f2  ⊥ ′ ′
A=  . . .  ⇒ Basis of F : nul l (A, r )


fm
2 If F is the solution subspace of homogeneous system AX = 0
then the basis of F ⊥ consists of all row vectors of matrix B

B = r r e f (A)

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 27 / 29

T HE STANDARD INNER PRODUCT ON Rn

Suppose
 the set of f 1 , f 2 , . . . , f m is a basis of F. 
d ot ( f 1 , f 1 ) d ot ( f 1 , f 2 ) . . . d ot ( f 1 , f m )
 d ot ( f , f ) d ot ( f , f ) . . . d ot ( f , f ) 
2 1 2 2 2 m 
A= ,

 ... ... ... ... 
d ot ( f m , f 1 ) d ot ( f m , f 2 ) . . . d ot ( f m , f m )
 
d ot (x, f 1 )
 d ot (x, f ) 
2 
B =  , λ = (λ1 , λ2 , . . . , λm )T = i nv(A) ∗ B

 ... 
d ot (x, f m )

1 Projection f = λ(1) ∗ f 1 + λ(2) ∗ f 2 + . . . + λ(m) ∗ f m

2 Distance ||g || = ||x − f || = nor m(x − f )
Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 28 / 29
 THANK YOU FOR YOUR ATTENTION

Hoang Hai Ha (HCMUT) INNER PRODUCT SPACES September 17, 2023 29 / 29