Lecture 4

Vector Spaces
4.1 Vectors in Rn
4.2 Vector Spaces
4.3 Subspaces of Vector Spaces
4.4 Spanning Sets and Linear Independence
4.5 Basis and Dimension
4.6 Rank of a Matrix and Systems of Linear Equations
4.7 Coordinates and Change of Basis

4.0 Vectors as we have learned in highschool

n An oriented interval of a straight line:

§ Used to represent speed of motions, forces,….

§ Addition:

→
! →
"

!#"
 n
4.1 Vectors in R
n An ordered n-tuple:
a sequence of n real number ( x1 , x2 , !, xn )

n
§ n-space: R
the set of all ordered n-tuple

We call a sequence of n real number ( x1 , x2 , !, xn ) a vector of Rn

n Ex:
1
n=1 R = 1-space
= set of all real number
2
n=2 R = 2-space
= set of all ordered pair of real numbers ( x1 , x2 )
3
n=3 R = 3-space
= set of all ordered triple of real numbers ( x1 , x2 , x3 )

4
n=4 R = 4-space
= set of all ordered quadruple of real numbers ( x1 , x2 , x3 , x4 )
 n Notes:
n
(1) An n-tuple ( x1 , x2 , !, xn ) can be viewed as a point in R
with the xi’s as its coordinates.
(2) An n-tuple ( x1 , x2 , !, xn ) can be viewed as a vector
x = ( x1 , x2 , !, xn ) in Rn with the xi’s as its components.
§ Ex:

u = (u1 , u2 , !, un ), v = (v1 , v2 , !, vn ) (two vectors in Rn)

§ Equal:
u = v if and only if u1 = v1 , u2 = v2 , !, un = vn

§ Vector addition (the sum of u and v):

u + v = (u1 + v1 , u2 + v2 , !, un + vn )

§ Scalar multiplication (the scalar multiple of u by c):

cu = (cu1 , cu2 , ! , cun )

§ Notes:
The sum of two vectors and the scalar multiple of a vector
n
in R are called the standard operations in Rn.
 § Negative:
- u = (-u1 ,-u2 ,-u3 ,...,-un )
§ Difference:
u - v = (u1 - v1 , u2 - v2 , u3 - v3 ,..., un - vn )

§ Zero vector:
0 = (0, 0, ..., 0)

§ Notes:
(1) The zero vector 0 in Rn is called the additive identity in Rn.
(2) The vector –v is called the additive inverse of v.

n Thm 4.1: (Properties of vector addition and scalar multiplication)

n
Let u, v, and w be vectors in R , and let c and d be scalars.
(1) u+v is a vector in Rn
(2) u+v = v+u
(3) (u+v)+w = u+(v+w)
(4) u+0 = u
(5) u+(–u) = 0
(6) cu is a vector in Rn
(7) c(u+v) = cu+cv
(8) (c+d)u = cu+du
(9) c(du) = (cd)u
(10) 1(u) = u
 n Ex 5: (Vector operations in R4)
Let u=(2, – 1, 5, 0), v=(4, 3, 1, – 1), and w=(– 6, 2, 0, 3) be
4
vectors in R .
Find x in each of the following equations.
(a) x = 2u – (v + 3w)
(b) 3(x+w) = 2u – v+x
Sol: (a) x = 2u - ( v + 3w )
= 2u - v - 3w
= (4, - 2, 10, 0) - (4, 3, 1, - 1) - (-18, 6, 0, 9)
= (4 - 4 + 18, - 2 - 3 - 6, 10 - 1 - 0, 0 + 1 - 9)
= (18, - 11, 9, - 8).

(b) 3(x + w ) = 2u - v + x
3x + 3w = 2u - v + x
3x - x = 2u - v - 3w
2x = 2u - v - 3w
x = u - 12 v - 32 w
= (2,1,5,0 ) + (- 2, -23 , -21 , 12 ) + (9,-3,0, -29 )
= (9, -211 , 92 ,-4 )
n Thm 4.2: (Properties of additive identity and additive inverse)
n
Let v be a vector in R and c be a scalar. Then the following is true.
(1) The additive identity is unique. That is, if u+v=v, then u = 0
(2) The additive inverse of v is unique. That is, if v+u=0, then u = –v
(3) 0v=0
(4) c0=0
(5) If cv=0, then c=0 or v=0
(6) –(– v) = v

11

n Linear combination:
The vector x is called a linear combination of v 1 , v 2 ,..., v n,
if it can be expressed in the form
x = c1 v 1 + c2 v 2 + ! + cn v n c1 , c2 , ! , cn : scalar
§ Ex 6:
Given x = (– 1, – 2, – 2), u = (0,1,4), v = (– 1,1,2), and
3
w = (3,1,2) in R . Show that x is a linear combination of u, v and w.
That is, we need to find a, b, and c such that x = au+bv+cw.
Sol:
- b + 3c = - 1
a + b + c = -2
4a + 2b + 2c = - 2
Þ a = 1, b = -2, c = -1 Thus x = u - 2 v - w
 § Notes:
A vector u = (u1 , u2 ,!, un ) in R n can be viewed as:

a 1×n row matrix (row vector): u = [u1 , u2 , !, un ]

or éu1 ù
êu ú
a n×1 column matrix (column vector): u = ê ú
2

ê! ú
ê ú
ëun û

§ Observation: The matrix operations of addition and scalar

multiplication give the same results as the corresponding
vector operations

13

Vector addition Scalar multiplication

u + v = (u1 , u2 , ! , un ) + (v1 , v2 , ! , vn ) cu = c(u1 , u2 , ! , un )
= (u1 + v1 , u2 + v2 , ! , un + vn ) = (cu1 , cu2 , ! , cun )
Row matrix addition Row matrix multiplication

u + v = [u1 , u2 , ! , un ] + [v1 , v2 , ! , vn ] cu = c[u1 , u2 , ! , un ]

= [u1 + v1 , u2 + v2 , ! , un + vn ] = [cu1 , cu2 , ! , cun ]
Column matrix addition Column matrix multiplication
éu1 ù év1 ù éu1 + v1 ù éu1 ù écu1 ù
êu ú êv ú êu + v ú êu ú êcu ú
u+ v = ê 2ú + ê 2ú = ê 2 2ú cu = c ê 2 ú = ê 2 ú
ê! ú ê! ú ê ! ú ê! ú ê! ú
ê ú ê ú ê ú ê ú ê ú
u v u
ë nû ë nû ë n nû + v ëun û ëcun û
 4.2 Vector Spaces
Definition
Let V be a set on which two operations (addition and scalar
multiplication) are defined. If the following axioms are satisfied
for every u, v, and w in V and every scalar (real number) c and d,
then V is called a vector space and elements of V is called vectors

Addition:
(1) u+v is in V
(2) u+v=v+u
(3) u+(v+w)=(u+v)+w
(4) V has a an element, denoted by 0, such that for every u in V,
we have u+0=u. This element 0 is called zero vector.
(5) For every u in V, there is a vector in V denoted by –u
such that u+(–u)=0
15

Scalar multiplication:
(6) cu is in V.
(7) c(u + v) = cu + cv
(8) (c + d )u = cu + du

(9) c(du) = (cd )u

(10) 1(u) = u
 n Notes:
(1) A vector space consists of four entities:
a set of vectors, a set of scalars, and two operations
V：nonempty set
c：scalar – real number (it might be complex
numbers)
+ (u, v ) = u + v：addition
• (c, u) = cu： scalar multiplication
(V , +, •) is called a vector space

(2) V = {0}： zero vector space

17

n Examples of vector spaces:

(1) n-tuple space: Rn
(u1 , u2 ,!, un ) + (v1 , v2 ,!, vn ) = (u1 + v1 , u2 + v2 ,!, un + vn ) vector addition
k (u1 , u2 , !, un ) = (ku1 , ku2 , !, kun ) scalar multiplication

(2) Matrix space: V = M m´n (the set of all m×n matrices with real values)
Ex: ：(m = n = 2)
éu11 u12 ù é v11 v12 ù é u11 + v11 u12 + v12 ù
êu u ú + êv v ú = êu + v u + v ú vector addition
ë 21 22 û ë 21 22 û ë 21 21 22 22 û
éu u ù é ku ku12 ù
k ê 11 12 ú = ê 11 ú scalar multiplication
u u ku
ë 21 22 û ë 21 ku 22 û
 V = Pn (x)

Given 2 polynomials of degree n

p ( x) = a0 + a1 x + ! + an x n
q ( x) = b0 + b1 x + ! + bn x n
Define:
p ( x) + q ( x) = (a0 + b0 ) + (a1 + b1 ) x + ! + (an + bn ) x n
kp( x) = ka0 + ka1 x + ! + kan x n
(4) Function space: V = c(-¥, ¥) (the set of all real-valued
continuous functions defined on the entire real line.)
( f + g )( x) = f ( x) + g ( x)
(kf )( x) = kf ( x)

19

§ Thm 4.3: (Properties of scalar multiplication)

Let v be any element of a vector space V, and let c be any
scalar. Then the following properties are true.
(1) 0v = 0
(2) c0 = 0
(3) If cv = 0, then c = 0 or v = 0
(4) (-1)v = - v
 n Notes: To show that a set is not a vector space, you need
only find one axiom that is not satisfied.
§ Ex 6: The set V of all integer is not a vector space.
Pf: 1 Î V , 12 Î R
( 12 )(1) = 12 Ï V (V is not closed under scalar multiplication)
­ ­ ­
scalar noninteger
integer

§ Ex 7: The set of all second-degree polynomials is not a vector space.

Pf: Let p( x) = x 2 and q( x) = - x 2 + x + 1

Þ p( x) + q( x) = x + 1Ï V
(it is not closed under vector addition)

21

n Ex 8:
V=R2=the set of all ordered pairs of real numbers
We define new operations on V as follow.

vector addition: (u1 , u2 ) + (v1 , v2 ) = (u1 + v1 , u2 + v2 )

scalar multiplication: c(u1 , u2 ) = (cu1 ,0)
Verify V is not a vector space.
Sol:
!1(1, 1) = (1, 0) ¹ (1, 1)
\ the set V (together with the two given operations) is
not a vector space
 4.3 Subspaces of Vector Spaces
n Subspace:
(V ,+,•) : a vector space
W ¹f ü
ý : a nonempty subset
W ÍVþ
(W ,+,•) ：a vector space (under the same operations of addition
and scalar multiplication defined in V)
Þ W is a subspace of V
§ Trivial subspace:
Every vector space V has at least two subspaces.
(1) Zero vector space {0} is a subspace of V.
(2) V is a subspace of V.

23

n Thm 4.4: (Test for a subspace)

If W is a nonempty subset of a vector space V, then W is
a subspace of V if and only if the following conditions hold.

(1) If u and v are in W, then u+v is in W.

(2) If u is in W and c is any scalar, then cu is in W.
 n Ex: Subspace of R2
(1) {0} 0 = (0, 0 )
(2) Lines through the origin
(3) R 2

n Ex: Subspace of R3
(1) {0} 0 = (0, 0, 0 )
(2) Lines through the origin
(3) Planes through the origin
(4) R 3

25

§ Ex 2: (A subspace of M2×2)
Let W be the set of all 2×2 symmetric matrices. Show that
W is a subspace of the vector space M2×2, with the standard
operations of matrix addition and scalar multiplication.
Sol:
!W Í M 2´2 M 2´2 : vector sapces
Let A1, A2 Î W ( A1T = A1, A2T = A2 )
A1 Î W, A2 Î W Þ ( A1 + A2 )T = A1T + A2T = A1 + A2 ( A 1 + A 2 Î W )
k Î R , A Î W Þ (kA)T = kAT = kA (k A Î W )
\W is a subspace of M 2´2
 n Ex 3: (The set of singular matrices is not a subspace of M2×2)
Let W be the set of singular matrices of order 2. Show that
W is not a subspace of M2×2 with the standard operations.

Sol:
é1 0ù é0 0 ù
A=ê ú Î W , B = ê0 1 ú Î W
ë 0 0 û ë û

é1 0ù
\A+ B = ê ú ÏW
ë 0 1 û
\W2 is not a subspace of M 2´2

27

§ Ex 4: (The set of first-quadrant vectors is not a subspace of R2)

Show that W = {( x1 , x2 ) : x1 ³ 0 and x2 ³ 0} , with the standard
operations, is not a subspace of R2.

Sol:
Let u = (1, 1) Î W
! (- 1)u = (- 1)(1, 1) = (- 1, - 1) Ï W
(not closed under scalar
multiplication)

\W is not a subspace of R 2
 § Ex 6: (Determining subspaces of R2)
Which of the following two subsets is a subspace of R2?
(a) The set of points on the line given by x+2y=0.
(b) The set of points on the line given by x+2y=1.
Sol:
(a) W = {( x, y ) x + 2 y = 0} = {(-2t , t ) t Î R}
Let v1 = (- 2t1 , t1 ) Î W v2 = (- 2t 2 , t 2 ) Î W

! v1 + v2 = (- 2(t1 + t 2 ),t1 + t 2 ) Î W (closed under addition)

kv1 = (- 2(kt1 ), kt1 ) Î W (closed under scalar multiplication)

\W is a subspace of R 2

29

(b) W = {( x, y ) x + 2 y = 1} (Note: the zero vector is not on the line)

Let v = (1,0) Î W

! (- 1)v = (- 1,0 ) Ï W
\W is not a subspace of R 2
 § Ex 8: (Determining subspaces of R3)
Which of the following subsets is a subspace of R 3?
(a) W = {( x1 , x2 ,1) x1 , x2 Î R}
(b) W = {( x1 , x1 + x3 , x3 ) x1 , x3 Î R}
Sol:
(a) Let v = (0,0,1) Î W
Þ (-1) v = (0,0,-1) Ï W
\W is not a subspace of R 3
(b) Let v = ( v1 , v1 + v 3 , v 3 ) Î W , u = (u1 , u1 + u 3 , u 3 ) Î W
! v + u = (v1 + u1 , (v1 + u1 ) + (v 3 + u 3 ), v 3 + u 3 ) Î W
kv = (kv1 , (kv1 ) + (kv 3 ), kv 3 ) Î W
\W is a subspace of R 3

31

§ Thm 4.5: (The intersection of two subspaces is a subspace)

If V and W are both subspaces of a vector space U,
then the intersection of V and W (denoted by V ∩ W )
is also a subspace of U.

§Homework : Verify the above with examples from geometry

of plan and space.
 4.4 Spanning Sets and Linear Independence
§ Linear combination: (recall)
A vector v in a vector space V is called a linear combination of
the vectors u1,u 2 ,! ,u k in V if v can be written in the form

v = c1u1 + c2u 2 + ! + ck u k c1,c2 ,! ,ck : scalars

33

§Ex 2-3: (Finding a linear combination)

v1 = (1,2,3) v 2 = (0,1,2) v 3 = (-1,0,1)
Prove (a) w = (1,1,1) is a linear combination of v1 , v 2 , v 3
(b) w = (1,-2,2) is not a linear combination of v1 , v 2 , v 3
Sol:
(a) w = c1 v1 + c2 v 2 + c3 v 3 , where w = (1,1,1)

(1,1,1) = c1 (1,2,3) + c2 (0,1,2) + c3 (- 1,0,1)

= (c1 - c3 , 2c1 + c2 , 3c1 + 2c2 + c3 )

c1 - c3 =1
Þ 2c1 + c2 =1
3c1 + 2c2 + c3 =1
 é1 0 - 1 1ù é1 0 - 1 1 ù
Þ êê2 1 0 1úú "Gauss−Jordan
"""""" Elimination
→ ê0 1 2 - 1ú
ê ú
êë3 2 1 1úû êë0 0 0 0 úû

Þ c1 = 1 + t , c2 = -1 - 2t , c3 = t

(this system has infinitely many solutions)

t =1
Þ w = 2 v1 - 3 v 2 + v 3

35

(b)
w = c1 v1 + c2 v 2 + c3 v 3 , where w = (1, - 2,2)

é1 0 - 1 1 ù é1 0 - 1 1 ù
Þ êê2 1 0 - 2úú " """"""→ ê0 1 2 - 4 ú
Gauss−Jordan Elimination
ê ú
êë3 2 1 2 úû êë0 0 0 7 úû

Þ this system has no solution (! 0 ¹ 7)

Þ w ¹ c1 v1 + c2 v 2 + c3 v 3
 § the span of a set: span (S)
If S={v1, v2,…, vk} is a set of vectors in a vector space V,
then the span of S is the set of all linear combinations of
the vectors in S,
span(S ) ={c1 v1 + c2 v 2 + ! + ck v k "ci Î R}
(the set of all linear combinations of vectors in S )

§ a spanning set of a vector space:

If every vector in a given vector space V can be written as a
linear combination of vectors in a given set S, then S is
called a spanning set of the vector space V.
In other words, S is a spanning set of V if and only if
Span(S) = V.

37

§ Notes: In case S is a spanning set of V we use following notation

span ( S ) = V
Þ S spans (generates) V
V is spanned (generated) by S
S is a spanning set of V

§ Notes:
(1) span(f ) = {0} , we agree by consent!
(2) S Í span( S )
(3) S1 , S 2 Í V
S1 Í S 2 Þ span( S1 ) Í span( S 2 )
 § Ex 5: (A spanning set for R3)
Show that the set S = {(1,2,3), (0,1,2), (-2,0,1)} spans R 3
Sol:
We must determine whether an arbitrary vector u = (u1 , u2 , u3 )
in R 3 can be written as a linear combination of v1 , v 2 , and v 3 .
u Î R 3 Þ u = c1 v1 + c2 v 2 + c3 v 3
Þ c1 - 2c3 = u1
2c1 + c2 = u2
3c1 + 2c2 + c3 = u3
The problem thus reduces to determining whether this system
is consistent for all values of u1 , u2 , and u3 .

39

1 0 -2
!A=2 1 0 ¹0
3 2 1

Þ Ax = b has exactly one solution for every u.

Þ span( S ) = R 3
 § Thm 4.6: (Properties of Span(S))

If S={v1, v2,…, vk} is a set of vectors in a vector space V, then

(a) span (S) is a subspace of V.

(b) Every subspace of V that contains S must contain span (S).

That is, span (S) is the smallest subspace of V that contains S.

41

§ Linear Independence (L.I.) and Linear Dependence (L.D.):

S = {v1 , v 2 , !, v k } : a set of vectors in a vector space V
We consider the equation c1 v1 + c2 v 2 + ! + ck v k = 0 , where
c1 , c2 , ! ck are unknowns.
(1) Obviously the equation has the solution (c1 = c2 = ! = ck = 0)

(2) If the equation has only the trivial solution (c1 = c2 = ! = ck = 0)

then S is called linearly independent.
(3) If the equation has a nontrivial solution (i.e., not all zeros),
then S is called linearly dependent.
 § Notes:
(1) f is linearly independent we agree by consent!
(2) 0 Î S Þ S is linearly dependent.
(3) v ¹ 0 Þ {v} is linearly independent
(4) S1 Í S 2
S1 is linearly dependent Þ S 2 is linearly dependent

S 2 is linearly independent Þ S1 is linearly independent

43

§ Ex 8: (Testing for linearly independent)

Determine whether the following set of vectors in R3 is L.I. or L.D.
S = {(1, 2, 3), (0, 1, 2 ), (- 2, 0, 1)}
v1 v2 v3
Sol: c1 - 2c3 = 0
c1 v1 + c2 v 2 + c3 v 3 = 0 Þ 2c1 + c2 + =0
3c1 + 2c2 + c3 = 0
é1 0 - 2 0 ù é1 0 0 0ù
Þ êê2 1 0 0úú ¾Gauss
¾¾ ¾¾¾¾ ¾® êê0 1 0 0úú
- Jordan Elimination

êë3 2 1 0úû êë0 0 1 0úû

Þ c1 = c2 = c3 = 0 (only the trivial solution )
Þ S is linearly independent
 n Ex 9: (Testing for linearly independent)
Determine whether the following set of vectors in P2 is L.I. or L.D.
S = {1+x – 2x2 , 2+5x – x2 , x+x2}
v1 v2 v3
Sol:
c1v1+c2v2+c3v3 = 0
i.e. c1(1+x – 2x2) + c2(2+5x – x2) + c3(x+x2) = 0+0x+0x2

c1+2c2 =0 é1 2 0 0ù é1 2 0 0ù
ê1 ú ¾G.¾J.® ê 1 ú
c1+5c2+c3 = 0 ê 5 1 0 ú ê1 1 3 0ú
–2c1 – c2+c3 = 0 êë- 2 - 1 1 0úû ê0 0 0 0 ú
ë û
This system has infinitely many solutions.
(i.e., This system has nontrivial solutions.)
S is linearly dependent. (Ex: c1=2 , c2= – 1 , c3=3)

45

n Ex 10: (Testing for linearly independent)

Determine whether the following set of vectors in 2×2
matrix space is L.I. or L.D.
ìé2 1ù é3 0ù é1 0ù ü
S = íê ú , ê 2 1 ú , ê 2 0ú ý
îë 0 1û ë û ë ûþ
v1 v2 v3
Sol:
c1v1+c2v2+c3v3 = 0

é2 1ù é 3 0ù é1 0 ù é0 0 ù
c1 ê
0 1ú 2 ê 2 1 ú 3 ê 2 0 ú = ê0 0 ú
+ c + c
ë û ë û ë û ë û
 2c1+3c2+ c3 = 0
c1 =0
2c2+2c3 = 0
c1+ c2 =0

é2 3 1 0ù é1 0 0 0ù
ê1 0 0 0úú ê0
ê ê 1 0 0úú
ê0 ¾¾¾¾¾¾¾¾
Gauss - Jordan Elimination
®
2 2 0ú ê0 0 1 0ú
ê ú ê ú
ë1 1 0 0û ë0 0 0 0û

c1 = c2 = c3= 0 (This system has only the trivial solution.)

S is linearly independent.

47

n Thm 4.7: (A property of linearly dependent sets)

A set S = {v1,v2,…,vk}, k > 1, is linearly dependent if and
only if at least one of the vectors vj in S can be written as
a linear combination of the other vectors in S.

Pf:
(=>) c1v1+c2v2+…+ckvk = 0

! S is linearly dependent

∃ ci ≠ 0 for some i
%! %"#! %"$! %%
v$ = − v& −%%% − v$'& − v$#& −%%% − v(
%" %" %" %"
 (Ü)
Let vi = d1v1+…+di-1vi-1+di+1vi+1+…+dkvk
Þ d1v1+…+di-1vi-1-vi+di+1vi+1+…+dkvk = 0

Þ c1=d1, …,ci-1=di-1, ci=-1,ci+1=di+1,…, ck=dk (nontrivial solution)

Þ S is linearly dependent

n Corollary to Theorem 4.7:

Two vectors u and v in a vector space V are linearly dependent
if and only if one is a scalar multiple of the other. That is, u = kv,
for some scalar k.

49

4.5 Basis and Dimension

n Basis:
Linearly
V：a vector space Generating
Bases Independent
Sets
S ={v1, v2, …, vn} ⊂ V Sets

ì(a ) S spans V (i.e., span(S) = V )

í
î(b) S is linearly independent
S is called a basis for V

§ Notes:
(1) Ø (empty set) is a basis for {0} we agree by consent!
(2) the standard basis for R3:
{i, j, k} i = (1, 0, 0), j = (0, 1, 0), k = (0, 0, 1)
 n
(3) the standard basis for R :
{e1, e2, …, en} e1=(1,0,…,0), e2=(0,1,…,0), en=(0,0,…,1)
Ex: R4 {(1,0,0,0), (0,1,0,0), (0,0,1,0), (0,0,0,1)}
(4) the standard basis for m×n matrix space:
{ Eij | i = 1,…, m , j = 1,…. , n }, where
ì a =1
{ Eij | 1£i£m , 1£j£n }, and in Eij í ij
îother entries are zero
ìé1 0ù é0 1ù é0 0ù é0 0ù ü
Ex: 2´ 2 matrix space: íê ú, ê0 0ú, ê1 0ú, ê0 1ú ý
îë 0 0 û ë û ë û ë ûþ
(5) the standard basis for Pn(x):
{1, x, x2, …, xn}
Ex: P3(x) {1, x, x2, x3}

51

n Thm 4.8: (Uniqueness of basis representation)

If S = {v1 , v 2 , !, v n } is a basis for a vector space V, then every
vector in V can be written in one and only one way as a linear
combination of vectors in S.
Pf:
ì 1. span(S) = V
! S is a basis Þ í
î 2. S is linearly independent
! span(S) = V Let v = c1v1+c2v2+…+cnvn
v = b1v1+b2v2+…+bnvn
==> 0 = (c1–b1)v1+(c2 – b2)v2+…+(cn – bn)vn

! S is linearly independent
==> c1= b1 , c2= b2 ,…, cn= (i.e., uniqueness)
bn
 nThm 4.9: (Bases and linear dependence)
If S = {v1 , v 2 , !, v n } is a basis for a vector space V, then every
set containing more than n vectors in V is linearly dependent.

Pf:
Let S1 = {u1, u2, …, um} , m > n
! span( S ) = V
u1 = c11 v1 + c21 v 2 + ! + cn1 v n
ui ∈V Þ u 2 = c12 v1 + c22 v 2 + ! + cn 2 v n
"
u m = c1m v1 + c2 m v 2 + ! + cnm v n

53

Let k1u1+k2u2+…+kmum= 0
=> d1v1+d2v2+…+dnvn= 0 (where di = ci1k1+ci2k2+…+cimkm)
! S is L.I.
=> di=0 ∀ i i.e. c11k1 + c12 k 2 + ! + c1m k m = 0
c21k1 + c22 k 2 + ! + c2 m k m = 0
"
cn1k1 + cn 2 k 2 + ! + cnm k m = 0
! Known: If the homogeneous system has fewer equations
than variables, then it must have infinitely many solution.
m > n => k1u1+k2u2+…+kmum = 0 has nontrivial solution
==> S1 is linearly dependent
 n Thm 4.10: (Number of vectors in a basis)
If a vector space V has one basis with n vectors, then every
basis for V has n vectors. (All bases for a finite-dimensional
vector space has the same number of vectors.)
Pf:
S ={v1, v2, …, vn}
two bases for a vector space
S'={u1, u2, …, um}
S is a basis ü Thm.4.9 ü
ý Þ n ³ m ï
S ' is L.I. þ ï
ýÞ n = m
S is L.I. ü Thm .4.9
ï
ý Þ n £ mï
S ' is a basis þ þ

55

n Finite dimensional:
A vector space V is called finite dimensional,
if it has a basis consisting of a finite number of elements.
n Infinite dimensional:
If a vector space V is not finite dimensional,
then it is called infinite dimensional.
n Dimension:
The dimension of a finite dimensional vector space V is
defined to be the number of vectors in a basis for V.
V: a vector space S: a basis for V
dim(V) = #(S) (the number of vectors in S)
 § Notes: dim(V) = n
Linearly
(1) dim({0}) = 0 = #(Ø) Generating Bases Independent
Sets Sets

(2) dim(V) = n , S a subset of V #(S) > n #(S) = n #(S) < n

S：a generating set => #(S) ≥ n

S：a L.I. set => #(S) ≤ n
S：a basis => #(S) = n

(3) dim(V) = n , W is a subspace of V => dim(W) ≤ n

57

n Ex:
(1) Vector space Rn - basis {e1 , e2 , … , en}
- dim(Rn) = n
(2) Vector space Mm×n - basis {Eij | i = 1,…, m , j = 1,…, n}
- dim(Mm×n)=mn
(3) Vector space Pn(x) - basis {1, x, x2, … , xn}
- dim(Pn(x)) = n+1
(4) Vector space P(x) - basis {1, x, x2, …}
- dim(P(x)) = ∞
 n Ex 9: (Finding the dimension of a subspace)
(a) W={(d, c–d, c): c and d are real numbers}
(b) W={(2b, b, 0): b is a real number}
Sol: (Note: Find a set of L.I. vectors that spans the subspace)
(a) (d, c– d, c) = c(0, 1, 1) + d(1, – 1, 0)
=> S = {(0, 1, 1) , (1, – 1, 0)}(S is L.I. and S spans W)
=> S is a basis for W
=> dim(W) = #(S) = 2
(b) ! (2b, b,0 ) = b(2,1,0 )
=> S = {(2, 1, 0)} spans W and S is L.I.
=> S is a basis for W
=> dim(W) = #(S) = 1

59

§ Ex 11: (Finding the dimension of a subspace)

Let W be the subspace of all symmetric matrices in M2x2.
What is the dimension of W?
Sol:
ìïéa b ù üï
W = íê ú a, b, c Î R ý
ïîëb c û ïþ
éa b ù é1 0ù é0 1ù é0 0ù
!ê ú = a ê0 0ú + b ê1 0ú + c ê0 1ú
ë b c û ë û ë û ë û
ìé1 0ù é0 1ù é0 0ù ü
Þ S = íê ú, ê1 0ú, ê0 1ú ý spans W and S is L.I.
îë 0 0 û ë û ë ûþ
=> S is a basis for W => dim(W) = #(S) = 3
 n Thm 4.11: (Basis tests in an n-dimensional space)
Let V be a vector space of dimension n.
(1) If S = {v1 , v 2 , !, v n } is a linearly independent set of
vectors in V, then S is a basis for V.
(2) If S = {v1 , v 2 ,!, v n } spans V, then S is a basis for V.

dim(V) = n

Generating Linearly
Sets Bases Independent
Sets

#(S) > n #(S) < n

#(S) = n

61

Recall:
§ row vectors: Row vectors of A
é a11 a12 " a1n ù é A(1) ù [a11 , a12 , ! , a1n ] = A(1)
êa ê ú
a22 " a2 n úú ê A(2 ) ú [a21 , a22 , ! , a2n ] = A(2)
A = ê 21 =
ê ! ! ! ú ê ! ú "
ê ú ê ú
ëam1 am 2 " amn û êë A(m ) úû [am1 , am2 , ! , amn ] = A( m )

§ column vectors: Column vectors of A

é a11 a12 " a1n ù é a11 ù é a12 ù é a1n ù
êa êa ú ê a ú êa ú
" a2 n úú
A = ê 21
ê !
a22
! ! ú
[
= A(1) ! A(2 ) !"! A(n ) ]ê 21 ú ê 22 ú " ê 2 n ú
ê ! úê ! ú ê ! ú
ê ú ê úê ú ê ú
ëam1 am 2 " amn û ëam1 û ëam 2 û ëamn û
|| || ||
(1) (2) (n)
A A A
 Four Fundamental subspaces
Let A be an m×n matrix.
§ Row space:
The row space of A is the subspace of Rn spanned by
the row vectors of A.
RS ( A) = {a 1 A(1) + a 2 A( 2 ) + ... + a m A( m ) | a 1 , a 2 ,..., a m Î R}
§ Column space:
The column space of A is the subspace of Rm spanned by
the column vectors of A.

CS ( A) = {b1 A(1) + b 2 A(2) + ! + b n A( n ) b1 , b 2 , ! b n Î R}

n Null space:
The null space of A is the set of all solutions of Ax=0 and
it is a subspace of Rn.
NS ( A) = {x Î R n | Ax = 0}

63

Four Fundamental subspaces

§ Left Null space:
The left null space of A is the null space of AT. It is the set of
all solutions of ATy=0 and is a subspace of Rm.
NS( AT ) = {y ∈ R m | AT y = 0}

§ Compare notations
Here In some textbook
RS(A) C(AT)
CS(A) C(A)
NS(A) N(A)
NS(AT) N(AT)
 4.6 Rank of a Matrix and Systems of Linear Equations
n Recall: A m×n matrix A is row equivalent to a m×n matrix B if
B can be obtained from A by applying a finite number
of elementary row operations
n Thm 4.12: (Row-equivalent matrices have the same row space)
If an m×n matrix A is row equivalent to an m×n matrix B,

then the row space of A is equal to the row space of B.

§ Note
(1) The row space of a matrix is not changed by elementary
row operations.
RS(r(A)) = RS(A) r: elementary row operations
(2) Elementary row operations can change the column space.

65

n Thm 4.13: (Basis for the row space of a matrix)

If a matrix A is row equivalent to a matrix B in row-echelon

form, then the nonzero row vectors of B form a basis for the

row space of A.
 § Ex 2: ( Finding a basis for a row space)
é 1 3 1 3ù
ê 0 1 1 0úú
ê
A = ê- 3 0 6 - 1ú Find a basis of row space of A
ê ú
ê 3 4 -2 1ú
êë 2 0 -4 2úû
Sol: é 1 3 1 3ù é1 3 1 3ù w 1
ê 0 ê0 1 1 0úú w 2
ê 1 1 0úú ê
A= ê- 3 - 1ú ê 0 0 1ú w 3
ê
0 6
ú ¾Gauss
¾¾ ¾ E.
® B = ê0 ú
ê 3 4 -2 1ú ê0 0 0 0ú
êë 2 0 -4 2úû êë0 0 0 0úû
a1 a 2 a3 a4 b1 b 2 b3 b 4

67

a basis for RS(A) = {the nonzero row vectors of B} (Thm 4.13)

= {w1, w2, w3} = {(1, 3, 1, 3), (0, 1, 1, 0), (0, 0, 0, 1)}

n Notes:
(1) b 3 = -2b1 + b 2 Þ a 3 = -2a1 + a 2
(2) {b1 , b 2 , b 4 } is L.I. Þ {a1 , a 2 , a 4 } is L.I.
 n Ex 3: (Finding a basis for a subspace)
Find a basis for the subspace of R3 spanned by
v1 v2 v3
S = {(-1, 2, 5), (3, 0, 3), (5, 1, 8)}
Sol: é - 1 2 5ù v 1 é 1 - 2 - 5ù w1
A = êê 3 0 3úú v 2 G.E.
B = êê0 1 3úú w2
êë 5 1 8úû v 3 êë0 0 0úû

a basis for span({v1, v2, v3})

= a basis for RS(A)
= {the nonzero row vectors of B} (Thm 4.13)
= {w1, w2}
= {(1, –2, – 5) , (0, 1, 3)}

69

n Ex 4: (Finding a basis for the column space of a matrix)

Find a basis for the column space of the matrix A given in Ex 2.
é 1 3 1 3ù
ê 0 1 1 0úú
ê
A = ê- 3 0 6 - 1ú
ê ú
ê 3 4 -2 1ú
êë 2 0 -4 - 2úû
Sol. 1st method:
é1 0 -3 3 2ù é1 0 -3 3 2ù w1
ê3 1 0 4 0úú ê0 1 9 - 5 - 6úú w 2
A =ê
T
¾¾® B = ê
G.E.

ê1 1 6 - 2 - 4ú ê0 0 1 - 1 - 1ú w3
ê ú ê ú
ë3 0 -1 1 - 2û ë0 0 0 0 0û
 ! CS(A)=RS(AT)
\ a basis for CS(A)
= a basis for RS(AT)
= {the nonzero vectors of B}
= {w1, w2, w3}
ìé 1 ù é 0 ù é 0 ùü
ïê ú ê 1 ú ê 0 úï
ïê 0 ú ê ú ê ú ïï
ï
= íê- 3ú, ê 9 ú, ê 1 ú ý (a basis for the column space of A)
ïê 3 ú ê ú ê úï
ïê ú ê - 5ú ê- 1ú ï
ïîêë 2 úû êë- 6úû êë- 1úû ï
þ
§ Note: This basis is not a subset of {c1, c2, c3, c4}.

71

n Sol. 2nd method:

é1 3 1 3ù é1 3 1 3ù
ê0 1 1 0 úú ê0 1 1 0úú
ê ê
A = ê- 3 0 6 - 1 ú ¾¾¾® B = ê0
G.E .
0 0 1ú
ê ú ê ú
ê3 4 -2 1 ú ê0 0 0 0ú
êë 2 0 - 4 - 2úû êë0 0 0 0úû
c1 c2 c3 c4 v1 v 2 v3 v 4

Leading 1 => {v1, v2, v4} is a basis for CS(B)

{c1, c2, c4} is a basis for CS(A)
§ Notes:
(1) This basis is a subset of {c1, c2, c3, c4}.
(2) v3 = –2v1+ v2, thus c3 = – 2c1+ c2 .
 n Thm 4.14: (Solutions of a homogeneous system)
If A is an m×n matrix, then the set of all solutions of the
homogeneous system of linear equations Ax = 0 is a subspace
of Rn called the nullspace of A. NS ( A) = {x Î R n | Ax = 0}
Pf:
NS(A) ⊆ R n
NS(A) ≠ φ (∵ A0 = 0)
Let x1, x 2 ∈NS(A) (i.e. Ax1 = 0, Ax 2 = 0)
Then (1)A(x1 + x 2 ) = Ax1 + Ax 2 = 0 + 0 = 0 Addition
(2)A(cx1) = c(Ax1) = c(0) = 0 Scalar multiplication
Thus NS(A) is a subspace of R n
§ Notes: The nullspace of A is also called the solution space of
the homogeneous system Ax = 0.
73

§ Ex 6: (Finding the solution space of a homogeneous system)

Find the nullspace of the matrix A. é1 2 - 2 1 ù
A = êê3 6 - 5 4úú
êë1 2 0 3úû
Sol: The nullspace of A is the solution space of Ax = 0.
é1 2 - 2 1 ù é1 2 0 3ù
A = êê3 6 - 5 4úú ¾G¾ ¾®êê0 0 1 1úú
. J .E x2 and x4 are
êë1 2 0 3úû êë0 0 0 0úû free variables
Þ x1 = –2s – 3t, x2 = s, x3 = –t, x4 = t
é x1 ù é- 2 s - 3t ù é- 2ù é- 3ù
ê x ú ê s ú ê 1ú ê 0ú
Þ x = ê 2ú = ê ú = s ê ú + t ê ú = sv 1 + t v 2
ê x3 ú ê - t ú ê 0ú ê - 1ú
ê ú ê ú ê ú ê ú
ë x4 û ë t û ë 0û ë 1û
Þ NS ( A) = {sv1 + tv 2 | s, t Î R}
 n Thm 4.15: (Row and column space have equal dimensions)
If A is an m×n matrix, then the row space and the column
space of A have the same dimension.
dim(RS(A)) = dim(CS(A))

§ Rank:
The dimension of the row (or column) space of a matrix A
is called the rank of A and is denoted by rank(A).

rank(A) = dim(RS(A)) = dim(CS(A))

75

§ Nullity:

The dimension of the nullspace of A is called the nullity of A.

nullity(A) = dim(NS(A))

§ Note: rank(AT) = rank(A)

Pf: rank(AT) = dim(RS(AT)) = dim(CS(A)) = rank(A)

 n Thm 4.16: (Dimension of the solution space)
If A is an m×n matrix of rank r, then the dimension of
the solution space of Ax = 0 is n – r. That is
n = rank(A) + nullity(A)

n Notes:
(1) rank(A): The number of leading variables in the solution of Ax=0.
(The number of nonzero rows in the row-echelon form of A)
(2) nullity (A): The number of free variables in the solution of Ax = 0.

77

n Notes:
If A is an m×n matrix and rank(A) = r, then

Fundamental Dimension
Subspaces
RS(A)=CS(AT) r
CS(A)=RS(AT) r
NS(A) n–r
NS(AT) m–r
 n Ex 7: (Rank and nullity of a matrix)
Let the column vectors of the matrix A be denoted by a1, a2, a3,
a4, and a5.
é 1 0 -2 1 0ù
ê 0 -1 - 3 1 3ú
A=ê ú
ê- 2 - 1 1 -1 3ú
ê ú
ë 0 3 9 0 - 12û
a1 a2 a3 a4 a5
(a) Find the rank and nullity of A.
(b) Find a subset of the column vectors of A that forms a basis for
the column space of A .
(c) If possible, write the third column of A as a linear combination
of the first two columns.

79

Sol: Let B be the reduced row-echelon form of A.

é 1 0 -2 1 0ù é1 0 -2 0 1ù
ê 0 -1 - 3 1 3ú ê0 1 3 0 - 4úú
A=ê ú B=ê
ê- 2 - 1 1 - 1 3ú ê0 0 0 1 - 1ú
ê ú ê ú
ë 0 3 9 0 - 12û ë0 0 0 0 0û
a1 a2 a3 a4 a5 b1 b2 b3 b4 b5

(a) rank(A) = 3 (the number of nonzero rows in B)

nullity(A) = n – rank(A) = 5-3 = 2
 (b) Leading 1
Þ {b1 , b 2 , b 4 } is a basis for CS ( B )
{a1 , a 2 , a 4 } is a basis for CS ( A)
é 1ù é 0ù é 1ù
ê 0ú ê- 1ú ê 1ú
a1 = ê ú, a 2 = ê ú, and a 4 = ê ú,
ê - 2ú ê- 1ú ê- 1ú
ê 0ú ê 3ú ê 0ú
ë û ë û ë û

(c) b 3 = -2b1 + 3b 2 Þ a 3 = -2a1 + 3a 2

81

§ Thm 4.17: (Solutions of a nonhomogeneous linear system)

If xp is a particular solution of the nonhomogeneous system
Ax = b, then every solution of this system can be written in
the form x = xp + xh , where xh is a solution of the corresponding
homogeneous system Ax = 0.
Pf: Let x be any solution of Ax = b.
Þ A(x - x p ) = Ax - Ax p = b - b = 0.
Þ (x - x p ) is a solution of Ax = 0
Let x h = x - x p
Þ x = x p + xh
 n Ex 8: (Finding the solution set of a nonhomogeneous system)
Find the set of all solution vectors of the system of linear equations.
x1 - 2 x3 + x4 = 5
3 x1 + x2 - 5 x3 = 8
x1 + 2 x2 - 5 x4 = -9

Sol:

é1 0 - 2 1 5ù é1 0 - 2 1 5ù
ê3 1 - 5 0 8 ú ¾G¾ ¾
. J .E
® ê0 1 1 - 3 - 7 ú
ê ú ê ú
êë 1 2 0 - 5 - 9úû êë0 0 0 0 0úû
s t

83

é x1 ù é 2 s - t + 5ù é 2ù é- 1ù é 5ù
ê x ú ê- s + 3t - 7 úú êê- 1úú êê 3úú êê- 7 úú
Þ x = ê 2ú = ê =s +t +
ê x3 ú ê s + 0t + 0ú ê 1ú ê 0ú ê 0ú
ê ú ê ú ê ú ê ú ê ú
ë x4 û ë 0 s + t + 0û ë 0û ë 1û ë 0û

= su1 + tu 2 + x p
é5ù
ê- 7 ú
i.e. x p = ê ú is a particular solution vector of Ax=b.
ê0ú
ê ú
ë0û

xh = su1 + tu2 is a solution of Ax = 0

 n Thm 4.18: (Solution of a system of linear equations)
The system of linear equations Ax = b is consistent if and only
if b is in the column space of A.
Pf.
Let
é a11 a12 " a1n ù é x1 ù é b1 ù
êa a22 " a2 n úú êx ú êb ú
A = ê 21 , x = ê 2ú, and b= ê 2ú
ê ! ! ! ú ê!ú ê!ú
ê ú ê ú ê ú
ë a m1 am 2 " amn û ë xn û ëbm û
be the coefficient matrix, the column matrix of unknowns,
and the right-hand side, respectively, of the system Ax = b.

85

Then é a11 a12 " a1n ù é x1 ù é a11 x1 + a12 x2 + " + a1n xn ù

êa a22 " a2 n úú êê x2 úú êê a21 x1 + a22 x2 + " + a2 n xn úú
Ax = ê 21 =
ê ! ! ! úê! ú ê ! ! ú
ê úê ú ê ú
ëam1 am 2 " amn û ë xn û ëam1 x1 + a m 2 x2 + " + amn xn û
é a11 ù é a12 ù é a1n ù
êa ú êa ú êa ú
= x1 ê ú + x2 ê ú + " + xn ê 2 n ú.
21 22

ê ! ú ê ! ú ê ! ú
ê ú ê ú ê ú
ëam1 û ë am 2 û ëamn û
So, b = Ax = x1A(1) + x2A(2)+… + xnA(n), which is a linear
combination of column vector of A.
Hence, Ax = b is consistent if and only if b is a linear combination
of the columns of A. That is, the system is consistent if and only if
b is in the subspace of Rm spanned by the columns of A.
 § Note:
If rank([A|b])=rank(A)
Then the system Ax=b is consistent.

n Ex 9: (Consistency of a system of linear equations)

x1 + x2 - x3 = -1
x1 + x3 = 3
3 x1 + 2 x2 - x3 = 1

Sol:
é 1 1 - 1ù é1 0 1ù
A = êê 1 0 1úú ¾G¾ ¾®êê0 1 - 2úú
. J .E .

êë3 2 - 1úû êë0 0 0úû

87

é 1 1 - 1 - 1ù é1 0 1 3ù
[ A ! b] = êê 1 0 1 3úú ¾G¾ ¾®êê0 1 - 2 - 4úú
. J .E .

êë3 2 - 1 1úû êë0 0 0 0úû

c1 c2 c3 b w1 w2 w3 v
! v = 3w1 - 4w 2
Þ b = 3c1 - 4c 2 + 0c3 (b is in the column space of A)

Þ The system of linear equations is consistent.

§ Check:
rank ( A) = rank ([ A b]) = 2
 n Summary of equivalent conditions for square matrices:
If A is an n×n matrix, then the following conditions are equivalent.
(1) A is invertible
(2) Ax = b has a unique solution for any n×1 matrix b.
(3) Ax = 0 has only the trivial solution
(4) rank(A) = n
(5) The n row vectors of A are linearly independent.
(6) The n column vectors of A are linearly independent.

89

4.7 Coordinates and Change of Basis

n Coordinate representation relative to a basis
Let B = {v1, v2, …, vn} be an ordered basis for a vector space V
and let x be a vector in V such that
x = c1 v1 + c2 v 2 + ! + cn v n .
The scalars c1, c2, …, cn are called the coordinates of x relative
to the basis B. The coordinate matrix (or coordinate vector)
of x relative to B is the column matrix in Rn whose components
are the coordinates of x.
é c1 ù
êc ú
[x]B = ê 2ú
ê!ú
ê ú
ëcn û
 n
§ Ex 1: (Coordinates and components in R )
Find the coordinate matrix of x = (–2, 1, 3) in R3
relative to the standard basis
S = {(1, 0, 0), ( 0, 1, 0), (0, 0, 1)}
Sol:
! x = (-2, 1, 3) = -2(1, 0, 0) + 1(0, 1, 0) + 3(0, 0, 1),

é - 2ù
\[x]S = êê 1úú.
êë 3úû

91

§ Ex 3: (Finding a coordinate matrix relative to a nonstandard basis)

Find the coordinate matrix of x=(1, 2, –1) in R3
relative to the (nonstandard) basis
B ' = {u1, u2, u3}={(1, 0, 1), (0, – 1, 2), (2, 3, – 5)}
Sol: x = c u + c u + c u Þ (1, 2, - 1) = c (1, 0, 1) + c (0, - 1, 2) + c (2, 3, - 5)
1 1 2 2 3 3 1 2 3

c 1
+ 2c = 1 3 é1 0 2ù é c ù é 1ù1

Þ - c + 3c = 2 2
i.e. êê0 - 1
3
3úú êêc úú = êê 2úú
2

c + 2c - 5c = - 1
1 2 3
êë 1 2 - 5úû êëc úû êë- 1úû 3

é1 0 2 1ù é1 0 0 5ù é 5ù
Þ êê0 - 1 3 2úú ® êê0 1 0 - 8úú Þ [ x ]B¢ = êê - 8úú
G.J. E.

êë 1 2 - 5 - 1úû êë0 0 1 - 2úû êë- 2úû

 n Change of basis problem:
You were given the coordinates of a vector relative to one
basis B;
and you are asked to find the coordinates of the same vector
relative to another basis B'.
§ Ex: (Change of basis)
Consider two bases for a vector space V
B = {u1 , u 2 }, B¢ = {u1¢ , u¢2 }
éa ù éc ù
If [u1¢ ]B = ê ú, [u¢2 ]B = ê ú
ëb û ëd û
i.e., u1¢ = au1 + bu 2 , u¢2 = cu1 + du 2

93

é k1 ù
Let v Î V , [ v ]B¢ = ê ú
ëk 2 û
Þ v = k1u1¢ + k 2u¢2
= k1 (au1 + bu 2 ) + k 2 (cu1 + du 2 )
= (k1a + k 2 c)u1 + (k1b + k 2 d )u 2
So, coordinates of v relative to the basis B={u1,u2}
are k1a+k2c, k1b+k2d. That is,

é k a + k 2 c ù éa c ù é k1 ù
Þ [ v ]B = ê 1 ú =ê úê ú
ëk1b + k 2 d û ëb d û ëk 2 û
= [[u1¢ ]B [u¢2 ]B ] [v ]B¢
 n Transition matrix from B' to B:
Let B = {u1 , u 2 ,..., u n } and B¢ = {u1¢ , u¢2 ..., u¢n } be two bases
for a vector space V
If [v]B is the coordinate matrix of v relative to B
[v]B’ is the coordinate matrix of v relative to B'
then [ v ]B = P[ v ]B¢
= [[u1¢ ]B , [u¢2 ]B ,..., [u¢n ]B ] [v ]B¢
where
P = [[u1¢ ]B , [u¢2 ]B , ..., [u¢n ]B ]

is called the transition matrix from B' to B

95

n Remark: The bases B and B' have similar role. Therefore, we can
exchange B and B' and we have
n Thm 4.19: (The inverse of a transition matrix)

If P is the transition matrix from a basis B' to a basis B in Rn,

then
(1) P is invertible
–1
(2) The transition matrix from B to B' is P

n Notes:
B = {u1 , u 2 , ..., u n }, B ' = {u¢1 , u¢2 , ..., u¢n }
[v ]B = [[u¢1 ]B , [u¢2 ]B , ..., [u¢n ]B ] [v ]B¢ = P [v ]B¢
[v ]B¢ = [[u1 ]B¢ , [u 2 ]B¢ , ..., [u n ]B¢ ] [v ]B = P -1 [v ]B
 n Thm 4.20: (Transition matrix from B to B')
Let B={v1, v2, … , vn} and B' ={u1, u2, … , un} be two bases
n
for R . Then the transition matrix P from B' to B can be found
by using Gauss-Jordan elimination on the n×2n matrix 𝐵 ⋮ B′
as follows.

𝐵 ⋮ B′ I) ⋮ P

97

§ Ex 5: (Finding a transition matrix)

B={(–3, 2), (4,–2)} and B' ={(–1, 2), (2,–2)} are two bases for R2

(a) Find the transition matrix from B' to B.

é1 ù
(b) Let [ v ]B ' = ê ú, find [ v ]B
ë 2û
(c) Find the transition matrix from B to B' .
 Sol:
(a) é- 3 4 ! -1 2ù G.J.E. é 1 0 ! 3 - 2ù
ê 2 ê0 1 ! 2 - 1ú
ë -2 ! 2 - 2úû ë û
B B' I P
é 3 - 2ù
\P = ê ú (the transition matrix from B' to B)
ë2 - 1 û
(b)
é 1ù é3 - 2ù é1 ù é- 1ù
[v ]B¢ = ê ú Þ [v ]B = P [v ]B¢ = ê ú ê 2ú = ê 0 ú
Check : 2
ë û ë 2 - 1 ûë û ë û
é 1ù
[v ]B¢ = ê ú Þ v = (1)(-1,2) + (2)(2,-2) = (3,-2)
ë 2û
é- 1ù
[v ]B = ê ú Þ v = (-1)(3,-2) + (0)(4,-2) = (3,-2)
ë 0û

99

(c)
é- 1 2 ! -3 4ù é 1 0 ! - 1 2ù
ê 2 - 2 ! 2 - 2ú G.J.E.
ê0 1 ! - 2 3ú
ë û ë û
-1
B' B I P

é - 1 2ù
\ P -1 = ê ú (the transition matrix from B to B')
ë - 2 3û

n Check:
é 3 - 2ù é - 1 2ù é 1 0ù
PP -1 = ê ú ê ú =ê ú = I2
ë2 - 1û ë- 2 3û ë0 1û
 § Ex 6: (Coordinate representation in P3(x))
(a) Find the coordinate matrix of p = 3x3-2x2+4 relative to the
standard basis S = {1, x, x2, x3} in P3(x).
(b) Find the coordinate matrix of p = 3x3-2x2+4 relative to the
basis S = {1, 1+x, 1+ x2, 1+ x3} in P3(x).
é 4ù
Sol: ê 0ú
(a) p = (4)(1) + (0)( x) + (-2)( x 2 ) + (3)( x 3 ) Þ [ p ]B = ê ú
ê - 2ú
ê ú
ë 3û é 3ù
ê 0ú
(b) p = (3)(1) + (0)(1 + x) + (-2)(1 + x ) + (3)(1 + x ) Þ [ p ]B = ê ú
2 3

ê - 2ú
ê ú
ë 3û

101

§ Ex: (Coordinate representation in M2x2)

Find the coordinate matrix of x = éê5 6ùú relative to
ë7 8 û
the standard basis in M2x2.
B = ìé1 0ù, é0 1ù, é0 0ù, é0 0ù ü
íê ý
î ë 0 0úû êë0 0úû êë1 0úû êë0 1úû þ
Sol:
é5 6ù é1 0ù é0 1ù é0 0ù é0 0ù
x=ê ú = 5 ê0 0ú + 6 ê0 0ú + 7 ê1 0ú + 8ê0 1ú
ë 7 8 û ë û ë û ë û ë û
é5 ù
ê6 ú
Þ [x ]B = ê ú
ê7 ú
ê ú
ë8 û