VIETNAM NATIONAL UNIVERSITY-HCMC

International University
Lecture Notes/Slides for

APPLIED LINEAR ALGEBRA

Chapter 3A. Vector Spaces

Chapter 3A. Vector Spaces Page 1/70

What is Rn ?

Notation and Terminology

R denotes the set of real numbers.
R2 denotes the set of all column vectors with two entries.
R3 denotes the set of all column vectors with three entries.
In general, Rn denotes the set of all column vectors with n entries.

Chapter 3A. Vector Spaces What is Rn? Page 2/70

 Scalar quantities versus vector quantities

A scalar quantity has only magnitude; e.g. time, temperature.

A (non-zero) vector quantity has both magnitude and direction; e.g.
displacement, force, wind velocity.

Whereas two scalar quantities are equal if they are represented by the same value,
two vector quantities are equal if and only if they have the same magnitude and
direction.

Chapter 3A. Vector Spaces What is Rn? Page 3/70

R2 and R3
Vectors in R2 and R3 have convenient geometric representations as position
vectors of points in the 2-dimensional (Cartesian) plane and in 3-dimensional
space, respectively.

Chapter 3A. Vector Spaces What is Rn? Page 4/70

 z

R3 c

R2
(a, b, c)
y

b (a, b)

0 b
0 a x y

x
 
a
The vector .  
b a
The vector  b .
c

Chapter 3A. Vector Spaces What is Rn? Page 5/70

Notation
If P is a point in Rn with coordinates (p1 , p2 , ..., pn ) we denote this by
P = (p1 , p2 , ..., pn ).
If P = (p1 , p2 , . . . , pn ) is a point in Rn , then
 
p1
−
→   p2 

0P =  . 
 .. 
pn

is often used to denote the position vector of the point.

Instead of using a capital letter to denote the vector (as we generally do
with matrices), we emphasize the importance of the geometry and the
direction with an arrow over the name of the vector.

Chapter 3A. Vector Spaces What is Rn? Page 6/70

 Notation and Terminology
−
→
The notation 0P emphasizes that this vector goes from the origin 0 to the
point P. We can also use lower case letters for names of vectors. In this
−
→
case, we write 0P = p~.
Any vector  
x1
 x2 
~x =  ..  in Rn
 
 . 
xn
is associated with the point (x1 , x2 , . . . , xn ).
Often, there is no distinction made between the vector ~x and the point
(x1 , x2 , . . . ,xn ), and we say that both (x1 , x2 , . . . , xn ) ∈ Rn and
x1
 x2 
~x =  .  ∈ Rn .
 
 .  .
xn

Chapter 3A. Vector Spaces What is Rn? Page 7/70

Geometric Vectors in R2 and R3

Let A and B be two points in R2 or R3 .

y
−→
B • AB is the geometric vector from A to B.
−→
• A is the tail of AB.
−→
• B is the tip of AB.
−→
• the magnitude of AB is its length, and is
−→
denoted ||AB||.
0 x

Chapter 3A. Vector Spaces Geometric Vectors Page 8/70

 Equality of geometric vectors
y
B
−→
• AB is the vector from A = (1, 0).
to B = (2, 2).
D −→
• CD is the vector from C = (−1, −1)
to D = (0, 1).

0 A x −→ −→
• AB = CD because the vectors have
the same length and direction.
C

The fact that the points A and B are different from the points C and D is not
important. For geometric vectors, the location of the vector in the plane (or in
3-dimensional space) is not important; the important properties are its length and
direction.

Chapter 3A. Vector Spaces Geometric Vectors Page 9/70

Coordinatizing Vectors – Part 1

y
P B

−
→
0P is the position vector for P = (1, 2),
 
−→ 1
and 0P = .
2

0 A x

 
−→ − → −→ 1
Since AB = 0P, it should be the case that AB = . This can be seen by
2
−→
moving AB so that its tail is at the origin.

A geometric vector is coordinatized by putting it in standard position, meaning

with its tail at the origin, and then identifying the vector with its tip.

Chapter 3A. Vector Spaces Geometric Vectors Page 10/70

 Algebra in Rn
Addition in Rn
Since vectors in Rn are n × 1 matrices, addition in Rn is precisely matrix addition
using column matrices, i.e.,
If u~ and ~v are in Rn , then u~ + ~v is obtained by adding together
corresponding entries of the vectors.
The zero vector in Rn is the n × 1 zero matrix, and is denoted ~0.

Example

  
1 4
Let u~ = 2 and ~v = 5 . Then,
  
3 6
     
1 4 5
u~ + ~v = 2 + 5 = 7 
    
3 6 9

Chapter 3A. Vector Spaces Algebra in Rn Page 11/70

Properties of Vector Addition

~ be vectors in Rn . Then the following properties hold.

Let u~, ~v , and w
1 u~ + ~v = ~v + u~ (vector addition is commutative).
2 u + ~v ) + w
(~ ~ = u~ + (~v + w
~) (vector addition is associative).
3 u~ + ~0 = u~ (existence of an additive identity).
4 u ) = ~0
u~ + (−~ (existence of an additive inverse).

Chapter 3A. Vector Spaces Algebra in Rn Page 12/70

 Scalar Multiplication

Since vectors in Rn are n × 1 matrices, scalar multiplication in Rn is precisely

matrix scalar multiplication using column matrices, i.e., If u~ is a vector in Rn and
k ∈ R is a scalar, then k u~ is obtained by multiplying every entry of u~ by k.

Example

1
Let u~ =  2  and k = 4. Then,
3
   
1 4
k u~ = 4  2  =  8 
3 12

Chapter 3A. Vector Spaces Algebra in Rn Page 13/70

Properties of Scalar Multiplication

Let u~, ~v ∈ Rn be vectors and k, p ∈ R be scalars. Then the following properties

hold.
1 u + ~v ) = k u~ + k~v (scalar multiplication distributes over vector addition).
k(~
2 u = k u~ + p~
(k + p)~ u (addition distributes over scalar multiplication).
3 k(p~
u ) = (kp)~
u (scalar multiplication is associative).
4 1~
u = u~ (existence of a multiplicative identity).

Chapter 3A. Vector Spaces Algebra in Rn Page 14/70

 Some notation you may encounter

  
1 0
Often, in R2 the notation ~i = , ~j = is used. Whereas in R3 the
0 1
     
1 0 0
~   ~
notation is i = 0 , j = 1 , k = 0 
  ~ 
0 0 1
So we have  
a
= a~i + b~j
b
and  
a
 b  = a~i + b~j + c ~k
c

Chapter 3A. Vector Spaces Algebra in Rn Page 15/70

The Geometry of Vector Addition

1 Vector Equality. The vectors have the same length and direction.
2 The zero vector, ~0 has length zero and no direction.
3 Addition. Let u~, ~v be vectors. Then u~ + ~v is the diagonal of the
parallelogram defined by u~ and ~v , and having the same tail as u~ and ~v .

u~ + ~v

u~

~v

Chapter 3A. Vector Spaces The Geometry of Vector Addition Page 16/70
 Tip-to-Tail Method for Vector Addition
For points A, B and C ,
−→ −→ −→
AB + BC = AC .

C
−
→
AB

−→ −→
AC BC

−
→
BC B
−→
A AB

Chapter 3A. Vector Spaces The Geometry of Vector Addition Page 17/70

The Geometry of Vector Subtraction

Let u~ and ~v be vectors in R2 or R3 . The vector u~ − ~v = u~ + (−~v ) is obtained from
the parallelogram defined by u~ and ~v by taking the vector from the tip of ~v to the
tip of u~, i.e., the diagonal of the parallelogram, directed towards the tip of u~.
u~

~v u~ − ~v −~v

u~

Chapter 3A. Vector Spaces The Geometry of Vector Addition Page 18/70
 Coordinatizing Vectors – Part 2
Let A = (x1 , y1 , z1 ) and B = (x2 , y2 , z2 ) be two points in R3 .
B z

0 y
A
x
−
→ −→ − →
We see from the figure that 0A + AB = 0B, and hence
     
x2 x1 x2 − x1
−→ − → − →
AB = 0B − 0A =  y2  −  y1  =  y2 − y1  .
z2 z1 z2 − z1

Chapter 3A. Vector Spaces The Geometry of Vector Addition Page 19/70

Length of a Vector, R2

 
x1 y
If ~x = ∈ R2 ,
x2 X = (x1 , x2 )
~
x

x
0

then the length of the vector ~x is the distance from the origin 0 to the point
X = (x1 , x2 ) given by d(0, X ).

The length of ~x , denoted k~x k, is given by:

q
d(0, X ) = k~x k = x12 + x22 .

Chapter 3A. Vector Spaces Length of a Vector Page 20/70

 Length of a Vector, R3

 
x1
This extends clearly to ~x =  x2  ∈ R3 .
x3
The length of ~x is the distance from the origin 0 to the point X = (x1 , x2 , x3 )
given by d(0, X ).
q
d(0, X ) = k~x k = x12 + x22 + x32 .

Suppose we want to find the distance between points other than the origin?

Chapter 3A. Vector Spaces Length of a Vector Page 21/70

Length of a Vector, R3
Consider two arbitrary points in R3 , A = (x1 , y1 , z1 ) and B = (x2 , y2 , z2 ). Then
the distance between them is written d(A, B) and is given by the distance formula.

Distance Formula
p
d(A, B) = (x2 − x1 )2 + (y2 − y1 )2 + (z2 − z1 )2 .

Now let P = (x2 − x1 , y2 − y1 , z2 − z1 ), and

 
x2 − x1
p~ =  y2 − y1 
z2 − z1

Then the length of p~ is equal to the distance between the origin and P, which are
both equal to the distance between points A and B
p
p k = d(0, P) = d(A, B) = (x2 − x1 )2 + (y2 − y1 )2 + (z2 − z1 )2 .
k~

Chapter 3A. Vector Spaces Length of a Vector Page 22/70

 Length of a Vector, Rn

More generally, if P = (p1 , p2 , . . . , pn ) and Q = (q1 , q2 , . . . , qn ) are points in Rn ,

−→ −→
then the distance between P and Q is the length of the vector PQ, written kPQk.
−→ p
d(P, Q) = kPQk = (q1 − p1 )2 + (q2 − p2 )2 + · · · + (qn − pn )2 .

The formula for calculating the length of a vector generalizes to Rn : if

 
x1
 x2 
~x =  .  ∈ Rn ,
 
 .. 
xn

then q
k~x k = x12 + x22 + · · · + xn2 ,
which represents the distance from the origin to the point (x1 , x2 , . . . , xn ).

Chapter 3A. Vector Spaces Length of a Vector Page 23/70

Properties of Distance

Let P and Q be two points in Rn , and d(P, Q) the distance between them. Then
the following properties hold.
1 The distance between P and Q is equal to the distance between Q and P,
i.e., d(P, Q) = d(Q, P).
2 d(P, Q) ≥ 0 with equality if and only if P = Q.

Example
−1, 3) and Q = (3, 1,√0), the distance between P and Q is
For P = (1,p
d(P, Q) = 22 + 22 + (−3)2 = 17.

Chapter 3A. Vector Spaces Length of a Vector Page 24/70

 Example

  
  3 −6
−3
Let p~ = ~ =  −1 . Then −2~
and q q =  2 .
q = (−2)~
4
−2 4
The lengths of these vectors are
p √
p k = (−3)2 + 42 = 9 + 16 = 5,
k~

p √ √
k~
qk = (3)2 + (−1)2 + (−2)2 = 9+1+4= 14,
and
p
k − 2~
qk = (−6)2 + 22 + 42
√
= 36 + 4 + 16
√ √
= 56 = 4 × 14
√
= 2 14 = 2k~q k.

Chapter 3A. Vector Spaces Length of a Vector Page 25/70

The Geometry of Scalar Multiplication

6 ~0 and a ∈ R, a 6= 0, then a~v has length

Scalar Multiplication. If ~v =
ka~v k = |a| · k~v k, and
I has the same direction as ~v if a > 0;
I has direction opposite to ~v if a < 0.
Parallel Vectors. Two nonzero vectors are called parallel if they have the
same direction or opposite directions. It follows that nonzero vectors ~v and
~ are parallel if and only if one is a scalar multiple of the other.
w

Chapter 3A. Vector Spaces The Geometry of Scalar Multiplication Page 26/70
 Problem
Let P = (1, −2, 1),Q = (−3, 0, 5), X = (2, −1, 5) and Y = (4, −2, 3) be points in
−→ −→ −→ −−→
R3 . Is PQ parallel to XY ? Is PX parallel to QY ?

Solution
   
−4 2
−→  −→ −→ −→
PQ = 2 , XY =  −1 , and these vectors are parallel if PQ = k XY for
4 −2
some scalar k, i.e.,
       
−4 2 −4 2k
 2  = k  −1  or  2  =  −k  .
4 −2 4 −2k
This gives a system of three equations in one variable, which is consistent, and
−→ −→
has unique solution k = −2. Therefore, PQ is parallel to XY .
   
1 7
−→   −−→  −→ −−→
PX = 1 , QY = −2 , and these vectors are parallel if PX = `QY for
4 −2
−→ −−→
some scalar `. You will find that no such ` exists, so PX is not parallel to QY .
Chapter 3A. Vector Spaces The Geometry of Scalar Multiplication Page 27/70

Unit Vectors

Definition
A unit vector is a vector of length one.

Example
     √2 
 
1 0 0 2 
 0 ,  1 ,  0 ,  0
 √ , are examples of unit vectors.
0 0 1 2
2

Example
If ~v 6= ~0, then
1
~v
k~v k
is a unit vector in the same direction as ~v .

Chapter 3A. Vector Spaces Unit Vectors Page 28/70

 Example
 
−1 √
~v =  3  is not a unit vector, since k~v k = 14. However,
 −1 
2 √
14
1  √3 
u~ = √ ~v =  14 
14 2 √
14
is a unit vector in the same direction as ~v , i.e.,
1 1 √
u k = √ k~v k = √
k~ 14 = 1.
14 14

Example
If ~v and w
~ are nonzero that have
k~
vk
the same direction, then ~v = wk w ;
k~
~
k~
vk
opposite directions, then ~v = − k~
wk w .
~

Chapter 3A. Vector Spaces Unit Vectors Page 29/70

The Dot Product

Definition
   
x1 x2
Let u~ =  y1  and ~v =  y2  be vectors in R3 . The dot product of u~ and ~v is
z1 z2

u~ • ~v = x1 x2 + y1 y2 + z1 z2 ,

i.e., u~ • ~v is a scalar.

Problem
 T  T
Find u~ • ~v for u~ = 1 2 0 −1 , ~v = 0 1 2 3 .

Solution

u~ • ~v = (1)(0) + (2)(1) + (0)(2) + (−1)(3)

= 0 + 2 + 0 + −3 = −1

Chapter 3A. Vector Spaces The Dot Product Page 30/70

 Note
If    
u1 v1
 u2   v2 
u~ =  ..  , and ~v =  ..
   

 .   . 
un vn
are in Rn , then another way to think about the dot product u~ • ~v is as the 1 × 1
matrix
 
v1
 v2  
 
T
 
u~ ~v = u1 u2 · · · un  .  = u1 v1 + u2 v2 + · · · + un vn
 .. 
vn

which is treated as a scalar given by u1 v1 + u2 v2 + · · · + un vn

Chapter 3A. Vector Spaces The Dot Product Page 31/70

Properties of the Dot Product

Theorem
~ be vectors in Rn and let k ∈ R.
Let u~, ~v , w
1 u~ • ~v is a real number
2 u~ • ~v = ~v • u~
3 u~ • ~0 = 0
4 u k2
u~ • u~ = k~
5 (k u~) • ~v = k(~
u • ~v ) = u~ • (k~v )
6 u~ • (~v + w
~ ) = u~ • ~v + u~ • w
~
u~ • (~v − w
~ ) = u~ • ~v − u~ • w
~

u k2 , we have an alternate (but equivalent) expression

Since, for u~ ∈ Rn , u~ • u~ = k~
for the length of u~: √
k~
u k = u~ • u~.

Chapter 3A. Vector Spaces The Dot Product Page 32/70

 Length of a Vector
We can use the properties of the dot product to find the length of a vector.

Problem
 
1
 3 
Find the length of the vector u~ = 
 5 .


Solution
u k2 = u~ • u~.
By the properties of the dot product, k~

u~ • u~ = (1)(1) + (3)(3) + (5)(5) + (2)(2)

= 1 + 9 + 25 + 4
= 39
√ √
Therefore, k~
u k = u~ • u~ = 39

Chapter 3A. Vector Spaces The Dot Product Page 33/70

Two Important Inequalities

Theorem
The Cauchy-Schwarz Inequality is given as follows. For u~, ~v ∈ Rn ,

|~
u • ~v | ≤ k~
u kk~v k

Equality is obtained if one vector is a scalar multiple of the other.

Theorem
The Triangle Inequality is given as follows. For u~, ~v ∈ Rn ,

k~
u + ~v k ≤ k~
u k + k~v k

Equality is obtained if one vector is a non-negative scalar multiple of the other.

Chapter 3A. Vector Spaces The Dot Product Page 34/70

 The Included Angle

Let u~ and ~v be two vectors in Rn (n ≥ 2), positioned so they have the same tail.
Then there is a unique angle θ between u~ and ~v with 0 ≤ θ ≤ π. This angle θ is
called the included angle.

u~ θ
θ u~ ~v
~v
Theorem
Let u~ and ~v be nonzero vectors, and let θ denote the angle between u~ and ~v .
Then
u~ • ~v = k~
u k k~v k cos θ.

Chapter 3A. Vector Spaces The Geometric Significance of the Dot Product Page 35/70

Finding the included angle for nonzero vectors

As a consequence of the Theorem, if u~ and ~v are nonzero vectors with included
angle θ, then k~
uk =
6 0 and k~v k =
6 0,
and
u~ • ~v
cos θ = .
k~
u kk~v k

1 If 0 ≤ θ < π2 , then cos θ > 0, implying that u~ • ~v > 0. Conversely, if

u~ • ~v > 0, then 0 ≤ θ < π2 .
2 If θ = π2 , then cos θ = 0, implying that u~ • ~v = 0. Conversely, if u~ • ~v = 0,
then θ = π2 .
3 If π2 < θ ≤ π, then cos θ < 0, implying that u~ • ~v < 0. Conversely, if
u~ • ~v < 0, then π2 < θ ≤ π.

Chapter 3A. Vector Spaces The Geometric Significance of the Dot Product Page 36/70
 Included Angle

Problem
  
1 0
Find the angle between u~ =  0  and ~v =  1 .
−1 −1

Solution
√ √
u~ • ~v = 1, k~
uk = 2 and k~v k = 2.
Therefore,
u~ • ~v 1 1
cos θ = =√ √ = .
k~
u k k~v k 2 2 2
π
Since 0 ≤ θ ≤ π, θ = 3.

π
Therefore, the angle between u~ and ~v is 3.

Chapter 3A. Vector Spaces The Geometric Significance of the Dot Product Page 37/70

Problem
  
3 −2
Find the included angle for u~ =  −6  and ~v =  1 .
−3 −1

Solution
√ √ √
u~ • ~v = − 9, k~
uk = 54 = 3 6, and k~v k = 6.
Let θ denote the included angle for u~ and ~v . Then

u~ • ~v −9 −9 1
cos θ = = √ √ = =− .
k~
u kk~v k 3 6× 6 18 2
2π
Since 0 ≤ θ ≤ π, the included angle is θ = 3 .

Chapter 3A. Vector Spaces The Geometric Significance of the Dot Product Page 38/70
 Problem
   
7 1
Find the included angle for u~ =  −1  and ~v =  4 .
3 −1

Solution
Let θ denote included angle.
u~ • ~v = 0.
π
Regardless of k~
u k and k~v k, cos θ = 0, and therefore the included angle is θ = 2.

Chapter 3A. Vector Spaces The Geometric Significance of the Dot Product Page 39/70

Orthogonal Vectors

Definition
Vectors u~ and ~v are orthogonal, also called perpendicular, if and only if u~ = ~0 or
~v = ~0 or θ = π2 .

Theorem
Nonzero vectors u~ and ~v are orthogonal if and only if u~ • ~v = 0.

Proof
We have u~ ⊥ ~v if and only if k~
u − ~v k = k~
u + ~v k (see the picture).
This is equivalent to

u − ~v ) • (~
(~ u − ~v ) = (~
u + ~v ) • (~
u + ~v )

which gives −2~

u • ~v = 2~
u • ~v and therefore u~ • ~v = 0.

Chapter 3A. Vector Spaces The Geometric Significance of the Dot Product Page 40/70
 Problem
    
x −1 0
Find all vectors ~v =  y  orthogonal to both u~ =  −3  and w
~ = 1 
z 2 1

Solution
There are infinitely many such vectors.
Since ~v is orthogonal to both u~ and w
~,

~v • u~ = −x − 3y + 2z = 0
~v • w~ = y +z =0

Chapter 3A. Vector Spaces The Geometric Significance of the Dot Product Page 41/70

Solution (continued)
This is a homogeneous system of two linear equation in three variables.
   
−1 −3 2 0 1 0 −5 0
→ ··· →
0 1 1 0 0 1 1 0
 
  5t
1 0 −5 0
implies that ~v =  −t  for t ∈ R.
0 1 1 0
t
 
5
Therefore, ~v = t −1  for all t ∈ R.

1

Chapter 3A. Vector Spaces The Geometric Significance of the Dot Product Page 42/70
 Problem
Are A(4, −7, 9), B(6, 4, 4) and C (7, 10, −6) the vertices of a right angle triangle?

Solution
    
2 3 1
−→  −→ −→
AB = 11  , AC =  17  , BC =  6 
−5 −15 −10

−→ −→
AB • AC = 6 + 187 + 75 6= 0.
−→ −→ −→ −→
BA • BC = (−AB) • BC = −2 − 66 − 50 6= 0.
−→ −→ −→ −→ −→ −→
CA • CB = (−AC ) • (−BC ) = AC • BC = 3 + 102 + 150 6= 0.
π
None of the angles is 2, and therefore the triangle is not a right angle triangle.

Chapter 3A. Vector Spaces The Geometric Significance of the Dot Product Page 43/70

Projections

Theorem
Given nonzero vectors ~v and u~ in Rn (for n = 2, 3...), there exist unique vectors
~v|| , ~v⊥ such that ~v can be written as a sum

~v = ~v|| + ~v⊥

where ~v|| is parallel to u~ and ~v⊥ is orthogonal to u~.

~v ~v
~v⊥

u~ ~v||
~v|| is the projection of ~v onto u~, written ~v|| = proju~ ~v and ~v⊥ = ~v − ~v|| .

Chapter 3A. Vector Spaces Projections Page 44/70

 Projections
A formula for proju~ ~v
The defining properties of ~v|| and ~v⊥ are
1 ~v|| is parallel to u~;
2 ~v⊥ is orthogonal to u~;
3 ~v|| + ~v⊥ = ~v .
Since ~v|| is parallel to u~, ~v|| = t u~ for some t ∈ R. Furthermore, ~v⊥ = ~v − ~v|| and
~v⊥ is orthogonal to u~, so

0 = ~v⊥ • u~ = (~v − ~v|| ) • u~ = (~v − t u~) • u~ = ~v • u~ − t(~

u • u~).

Since u~ 6= ~0, it follows that t = v •~

~
k~
u
u k2 . Therefore
   
~v • u~ ~v • u~
~v|| = u~, and ~v⊥ = ~v − u~.
u k2
k~ u k2
k~

Chapter 3A. Vector Spaces Projections Page 45/70

Projections

Theorem
Let ~v and u~ be vectors with u~ 6= ~0.
 
1 proj ~ ~
v •~
u
~ v = k~
u u k2 u
~
 
v •~
~ u
2 ~ v − k~uk2 u~ is orthogonal to u~.

Chapter 3A. Vector Spaces Projections Page 46/70

 Problem
   
2 3
Let ~v =  −1  and u~ =  1 . Find vectors ~v|| and ~v⊥ so that ~v = ~v|| + ~v⊥ ,
0 −1
with ~v|| parallel to u~ and ~v⊥ orthogonal to u~.

Solution
   
  3 15/11
~v • u~ 5 
~v|| = proju~ ~v = u~ = 1  =  5/11  .
u k2
k~ 11
−1 −5/11

       
2 3 7 7/11
5  1 
~v⊥ = ~v − ~v|| =  −1  − 1 = −16  =  −16/11  .
11 11
0 −1 5 5/11

Chapter 3A. Vector Spaces Projections Page 47/70

Distance from a Point to a Line

Problem
Let P = (3, 2, −1) be a point in R3 and L a line with equation
     
x 2 3
 y  =  1  + t  −1  .
z 3 −2

Find the shortest distance from P to L, and find the point Q on L that is closest
to P.

Solution
Let P0 = (2, 1, 3) be a point on L,
P T
and let d~ = 3 −1 −2

u~ .
−−→ −−→ −→ −−→ −−→
Then P0 Q = projd~ P0 P, 0Q = 0P0 + P0 Q,
L
P0 and the shortest distance from P to L is
−→ −→ −−→ −−→
the length of QP, where QP = P0 P − P0 Q.
0 Q

Chapter 3A. Vector Spaces Projections Page 48/70

 Solution (continued)

  
1 3
−−→ 
P0 P = 1 , d~ =  −1 .
−4 −2

−−→ ~ !
   
3 15
−−→ −−→ P0 P • d ~ 10  1
P0 Q = projd~ P0 P = d= −1  =  −5  .
~ 2
kdk 14 7
−2 −10

Therefore,      
2 15 29
−→   1  1
OQ = 1 + −5  =  2  ,
7 7
3 −10 11
29 2 11


so Q = 7 , 7, 7 .

Chapter 3A. Vector Spaces Projections Page 49/70

Solution (continued)
−→
Finally, the shortest distance from P(3, 2, −1) to L is the length of QP, where
     
1 15 −4
−→ −−→ −−→  1 2
QP = P0 P − P0 Q = 1  −  −5  =  6  .
7 7
−4 −10 −9

Therefore the shortest distance from P to L is

−→ 2p 2√
kQPk = (−4)2 + 62 + (−9)2 = 133.
7 7

Chapter 3A. Vector Spaces Projections Page 50/70

 Equations of Planes

Given a point P0 and a nonzero vector n~, there is a unique plane containing P0
and orthogonal to n~.
~
n

P0

Definition
A nonzero vector n~ is a normal vector to a plane if and only if n~ • ~v = 0 for every
vector ~v in the plane, i.e., n~ is orthogonal to every vector in the plane.

Chapter 3A. Vector Spaces Equations of Planes Page 51/70

Consider a plane containing a point P0 and orthogonal to vector n~, and let P be
−−→
an arbitrary point on this plane. Then n~ • P0 P = 0,
~
n

P0

or, equivalently,
−
→ −−→
n~ • (0P − 0P0 ) = 0,
and is called a vector equation of the plane. The vector equation can also be
written as
−
→ −−→
n~ • 0P = n~ • 0P0 .

Chapter 3A. Vector Spaces Equations of Planes Page 52/70

 Suppose a plane contains a fixed point P0 = (x0 , y0 , z0 ) and has normal vector
 
a
n~ =  b  .
c
−
→ −−→
Let P = (x, y , z) denote an arbitrary point on the plane. Since n~ • 0P = n~ • 0P0 ,
       
a x a x0
 b  •  y  =  b  •  y0  .
c z c z0

Thus
ax + by + cz = ax0 + by0 + cz0 ,
where d = ax0 + by0 + cz0 is simply a scalar.
A scalar equation of the plane has the form

ax + by + cz = d, where a, b, c, d ∈ R.

Chapter 3A. Vector Spaces Equations of Planes Page 53/70

Problem
Find an equation of the plane containing P0 (1, −1, 0) and orthogonal to
 T
n~ = −3 5 2 .

Solution
A vector equation of this plane is
   
−3 x −1
 5  •  y + 1  = 0.
2 z

Thus, a scalar equation of this plane is

−3x + 5y + 2z = −3(1) + 5(−1) + 2(0) = −8,

i.e., the plane has scalar equation

−3x + 5y + 2z = −8.

Chapter 3A. Vector Spaces Equations of Planes Page 54/70

 Shortest distance from a point to a plane
Problem
Find the shortest distance from the point P = (2, 3, 0) to the plane with equation
5x + y + z = −1, and find the point Q on the plane that is closest to P.

(wb example)

Solution
P = (2, 3, 0)
n~ Pick an arbitrary point P0 on the plane.
−→ −−→
Then QP = proj~n P0 P,
−→
kQPk is the shortest distance,
P0 Q −→ − → −→
and 0Q = 0P − QP.

   
5 2
−−→  
n~ = 1 . Choose P0 = (0, 0, −1). Then P0 P = 3 .
 
1 1
Chapter 3A. Vector Spaces Shortest distance from a point to a plane Page 55/70

Solution (continued)
P = (2, 3, 0)
n~
   
2 5
−−→  
P0 P = 3 and n~ = 1 .

1 1
P0 Q
−−→
 
5
!
−→ −−→ P0 P • n~ 14  
QP = proj~n P0 P = n~ = 1 .
nk2
k~ 27
1
−→ 14
√ √
14 3
√
14 3
Since kQPk = 27 27 = 9 , the shortest distance from P to the plane is 9 .

Chapter 3A. Vector Spaces Shortest distance from a point to a plane Page 56/70
 Solution (continued)
To find Q, we have
 
   
2 5 −16
−→ − → −→ 14   1 
0Q = 0P − QP =  3  − 1 = 67  .
27 27
0 1 −14

Therefore Q = − 16 67 14


,
27 27 , − 27 .

Chapter 3A. Vector Spaces Shortest distance from a point to a plane Page 57/70

The Cross Product

Definition
 T  T
Let u~ = u1 u2 u3 and ~v = v1 v2 v3 . Then
 
u2 v3 − u3 v2
u~ × ~v =  −(u1 v3 − u3 v1 )  .
u1 v2 − u2 v1

Note. u~ × ~v is a vector that is orthogonal to both u~ and ~v .

A helpful way to remember (once we cover determinants):

~i ~j ~k
     
1 0 0
~   ~
u~ × ~v = u1 u2 u3 , where i = 0 , j = 1 , k = 0  .
  ~ 
v1 v2 v3 0 0 1

Chapter 3A. Vector Spaces The Cross Product Page 58/70

 Computing the Cross Product
Problem

  
1 3
Find u~ × ~v for u~ =  −1  , ~v =  −2 .
2 1

Solution
We will use the equation:
 
u2 v3 − u3 v2
u~ × ~v =  −(u1 v3 − u3 v1 ) 
u1 v2 − u2 v1

Therefore,    
(−1)(1) − (2)(−2) 3
u~ × ~v =  −((1)(1) − (2)(3))  =  5 
(1)(−2) − (−1)(3) 1

Chapter 3A. Vector Spaces The Cross Product Page 59/70

Properties of the Cross Product

Theorem
~ be in R3 .
Let u~, ~v and w
1 u~ × ~v is a vector.
2 u~ × ~v is orthogonal to both u~ and ~v .
3 u~ × ~0 = ~0 and ~0 × u~ = ~0.
4 u~ × u~ = ~0.
5 u~ × ~v = −(~v × u~).
6 (k u~) × ~v = k(~
u × ~v ) = u~ × (k~v ) for any scalar k.
7 u~ × (~v + w
~ ) = u~ × ~v + u~ × w
~.
8 ~ ) × u~ = ~v × u~ + w
(~v + w ~ × u~.

Chapter 3A. Vector Spaces The Cross Product Page 60/70

 Problem
 T
Find all vectors orthogonal to both u~ = −1 −3 2 and
 T
~v = 0 1 1 .

Solution
~i −1 0 −5



u~ × ~v = ~j −3 1 = −5~i + ~j − ~k =  1  .
~k 2 1 −1

Any scalar multiple of u~ × ~v is also orthogonal to both u~ and ~v , so

 
−5
t  1  , t ∈ R,
−1

gives all vectors orthogonal to both u~ and ~v .

Chapter 3A. Vector Spaces The Cross Product Page 61/70

Problem
Let A = (1, −1, 2), B = (2, 0, −1) and C = (0, −2, 3) be points in R3 . These
points do no all lie on the same line (how can you tell?). Find an equation for the
plane containing A, B, and C .

(wb example)

Solution
~
n
−→ −→
C AB and AC lie in the plane, so
−→ −→
A ~ = AB × AC is a normal to the plane.
n
     
1 −1 −2
B −→  −→
AB = 1  , AC =  −1 , and n ~ =  2 .
−3 1 0
   
−2 0
Therefore − 2x + 2y =  2  •  −2  = −4
0 3
i.e. −2x + 2y = −4 is an equation of the plane.
Chapter 3A. Vector Spaces The Cross Product Page 62/70
 Area and Volume

The Lagrange Identity

If u~, ~v ∈ R3 , then
u × ~v k2 = k~
k~ u k2 k~v k2 − (~
u • ~v )2 .

Proof.
  
x1 x2
Write u~ =  y1  and ~v =  y2 , and work out all the terms.
z1 z2

Chapter 3A. Vector Spaces The Cross Product Page 63/70

The length of the cross product

As a consequence of the Lagrange Identity and the fact that

u~ • ~v = k~
u k k~v k cos θ,

we have
u × ~v k2
k~ = u k2 k~v k2 − (~
k~ u • ~v )2
= u k2 k~v k2 − k~
k~ u k2 k~v k2 cos2 θ
= u k2 k~v k2 (1 − cos2 θ)
k~
= u k2 k~v k2 sin2 θ.
k~

Taking square roots on both sides yields,

k~
u × ~v k = k~
u kk~v k sin θ.

Note that since 0 ≤ θ ≤ π, sin θ ≥ 0.

If θ = 0 or θ = π, then sin θ = 0, and k~ u × ~v k = 0. This is consistent with our

earlier observation that if u~ and ~v are parallel, then u~ × ~v = ~0.

Chapter 3A. Vector Spaces The Cross Product Page 64/70

 Area of a Parallelogram
Theorem
Let u~ and ~v be nonzero vectors in R3 with included angle θ.
1 k~
u × ~v k = k~
u kk~v k sin θ, and is the area of the parallelogram defined by u~
and ~v .
2 u~ and ~v are parallel if and only if u~ × ~v = ~0.

Proof of area of parallelogram.

The area of the parallelogram defined by u~ and ~v is k~
u kh, where h is the height
of the parallelogram.
~
v
h
θ
~
u
h
sin θ = vk ,
k~ implying that h = k~v k sin θ. Therefore, the area is
k~
u kk~v k sin θ.

Chapter 3A. Vector Spaces The Cross Product Page 65/70

Area of a Triangle
Problem
Find the area of the triangle having vertices A(3, −1, 2), B(1, 1, 0) and
C (1, 2, −1).

Solution
−→
The area of the triangle is half the area of the parallelogram defined by AB and
−→
AC .    
−2 −2
−→  −→
AB = 2  and AC =  3 . Therefore
−2 −3
 
0
−→ −→ 
AB × AC = −2  ,
−2
−→ −→ √
so the area of the triangle is 12 kAB × AC k = 2.

Chapter 3A. Vector Spaces The Cross Product Page 66/70

 The Box Product
    
u1 v1 w1
Let u~ =  u2 , ~v =  v2 , and w
~ =  w2 . Then
u3 v3 w3
   
u1 v2 w 3 − v3 w 2
~ ) =  u2  •  −(v1 w3 − v3 w1 ) 
u~ • (~v × w
u3 v1 w 2 − v2 w 1
= u1 (v2 w3 − v3 w2 ) − u2 (v1 w3 − v3 w1 ) + u3 (v1 w2 − v2 w1 )
v2 w 2 v1 w1 v1 w 1
= u1 − u2 + u3
v3 w 3 v3 w 3 v2 w 2
u1 v1 w1
= u2 v2 w2 .
u3 v3 w3

Chapter 3A. Vector Spaces The Box Product Page 67/70

The Box Product

Theorem
     
u1 v1 w1
If u~ =  u2 , ~v =  v2 , and w
~ =  w2 . Then the box product is
u3 v3 w3
 
u1 v1 w1
u~ • (~v × w
~ ) = det u2
 v2 w2  .
u3 v3 w3
 
Shorthand: u~ • (~v × w
~ ) = det u~ ~v ~
w .

Theorem
The order of the box product is defined as follows:

u × ~v ) • w
(~ ~ = u~ • (~v × w
~ ).

Chapter 3A. Vector Spaces The Box Product Page 68/70

 The Volume of a Parallelepiped

Theorem
The volume of the parallelepiped determined by the three vectors u~, ~v , and w
~ in
3
R is
|~
u • (~v × w
~ )|.

Problem
 
2
Find the volume of the parallelepiped determined by the vectors u~ =  1 ,
−1
   
1 2
~v = 0 , and w
  ~ = 1 .

2 1

Chapter 3A. Vector Spaces The Box Product Page 69/70

Solution
The volume of the parallelepiped
 is |~
u • (~v × w
~ )|.
Since u~ • (~v × w
~ ) = det u~ ~v w ~ , and
 
2 1 2
det  1 0 1  = −2,
−1 2 1

the volume of the parallelepiped is | − 2| = 2 cubic units.

Chapter 3A. Vector Spaces The Box Product Page 70/70