Linear Algebra

VECTOR SPACES
By
Vallampati Ramachandra Prasad
 4.1 The vector Space Rn
Definition 1.
Let (u1 , u2 , ..., un ) be a sequence of n real numbers. The set of all
such sequences is called n-space (or n-dimensional. space) and
is denoted Rn.
u1 is the first component of (u1 , u2 , ..., un ) .
u2 is the second component and so on.

Example 1
• R2 is the collection of all sets of two ordered real numbers.
For example, (0, 0) , (1, 2) and (-2, -3) are elements of R2.
• R3 is the collection of all sets of three ordered real numbers.
For example, (0,0, 0) and (-1,3, 4) are elements of R3.

Ch04_2
 n
Vectors in R
 An ordered n-tuple:
a sequence of n real number ( x1 , x2 ,, xn )

n
 n-space: R
the set of all ordered n-tuple
 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 )
 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:

x1 , x2  x1 , x2 

0,0
a point a vector
Definition 2.
Let u  (u1 , u2 , ..., un ) and v  (v1 , v2 , ..., vn ) be two elements of Rn.
We say that u and v are equal if u1 = v1, …, un = vn.
Thus two elements of Rn are equal if their corresponding
components are equal.

Definition 3.
Let u  (u1 , u2 , ..., un ) and v  (v1 , v2 , ..., vn ) be elements of Rn
and let c be a scalar. Addition and scalar multiplication are
performed as follows:
Addition: u  v  (u1  v1 , ..., un  vn )
Scalar multiplication : cu  (cu1 , ..., cun )
Ch04_6
► The set Rn with operations of componentwise addition and
scalar multiplication is an example of a vector space, and its
elements are called vectors.
We shall henceforth interpret Rn to be a vector space.
(We say that Rn is closed under addition and scalar multiplication).
► In general, if u and v are vectors in the same vector space, then u + v is
the diagonal of the parallelogram defined by u and v.

Figure 4.1

Ch04_7
 Example 2
Let u = ( –1, 4, 3) and v = ( –2, –3, 1) be elements of R3.
Find u + v and 3u.
Solution: u + v = (–1, 4, 3) + (– 2, –3, 1) = (-3 ,1 ,4)
3u = 3 (–1, 4, 3) = (-3 ,12 ,9)
Example 3
In R2 , consider the two elements
(4, 1) and (2, 3).
Find their sum and give a geometrical
interpretation of this sum.
we get (4, 1) + (2, 3) = (6, 4).
The vector (6, 4), the sum, is the
diagonal of the parallelogram.
Figure 4.2
Ch04_8
 Example 4
Consider the scalar multiple of the vector (3, 2) by 2, we get
2(3, 2) = (6, 4)
Observe in Figure 4.3 that (6, 4) is a vector in the same direction
as (3, 2), and 2 times it in length.

Figure 4.3
Ch04_9
 Zero Vector
The vector (0, 0, …, 0), having n zero components, is called the
zero vector of Rn and is denoted 0.

Negative Vector
The vector (–1)u is writing –u and is called the negative of u.
It is a vector having the same length (or magnitude) as u, but
lies in the opposite direction to u.
u

-u
Subtraction
Subtraction is performed on element of Rn by subtracting
corresponding components.

Ch04_10
 Theorem 4.1
Let u, v, and w be vectors in Rn and let c and d be scalars.
(a) u + v = v + u
(b) u + (v + w) = (u + v) + w
(c) u + 0 = 0 + u = u
(d) u + (–u) = 0
(e) c(u + v) = cu + cv
(f) (c + d)u = cu + du
(g) c(du) = (cd)u
(h) 1u = u
Figure 4.4
Commutativity of vector addition
u+v=v+u

Ch04_11
 Example 5
Let u = (2, 5, –3), v = ( –4, 1, 9), w = (4, 0, 2) in the vector space R3.
Determine the vector 2u – 3v + w.

Solution
2u - 3v  w  2(2, 5, - 3) - 3(-4, 1, 9)  (4, 0, 2)
 (4, 10, - 6) - (-12, 3, 27)  (4, 0, 2)
 (4  12  4, 10 - 3  0, - 6 - 27  2)
 (20, 7, - 31)

Ch04_12
 Column Vectors
Row vector: u  (u1 , u2 , ..., un )
 u1 
Column vector: 
 
u n 

We defined addition and scalar multiplication of column vectors

in Rn in a componentwise manner:
 u1   v1   u1  v1   u1   cu1 
        and c      
un  vn  un  vn  un  cun 
         

Ch04_13
 4.2 Dot Product, Norm, Angle, and Distance

Definition
Let u  (u1 , u2 , ..., un ) and v  (v1 , v2 , ..., vn ) be two vectors in Rn.
The dot product of u and v is denoted u.v and is defined
by u  v  u1v1 .   un vn
The dot product assigns a real number to each pair of vectors.

Example 1
Find the dot product of
u = (1, –2, 4) and v = (3, 0, 2)
Solution
u  v  (1 3)  (-2  0)  (4  2)
 3 08
 11 Ch04_14
 Properties of the Dot Product
Let u, v, and w be vectors in Rn and let c be a scalar. Then
1. u.v = v.u
2. (u + v).w = u.w + v.w
3. cu.v = c(u.v) = u.cv
4. u.u  0, and u.u = 0 if and only if u = 0
Proof
1. Let u  (u1 , u2 , ..., un ) and v  (v1 , v2 , ..., vn ). We get
u  v  u1v1    un vn
 v1u1    vn un by the commutativ e property of real numbers
 v u
4. u  u  u1u1    unun  u1 2    un 2
u1 2    un 2  0, thus u  u  0.
u1     un   0, if and only if u1  0, , un  0.
2 2

Thus u  u  0 if and only if u  0. Ch04_15

 Norm of a Vector in Rn

Definition
The norm (length or magnitude)
of a vector u = (u1, …, un) in Rn
is denoted ||u|| and defined by
u  u1 2    u n 2

Note:
The norm of a vector can also
be written in terms of
the dot product u  u  u
Figure 4.5 length of u

Ch04_16
 Example 2
Find the norm of each of the vectors u = (1, 3, 5) of R3
and v = (3, 0, 1, 4) of R4.
Solution u  (1) 2  (3) 2  (5) 2  1  9  25  35

v  (3) 2  (0) 2  (1) 2  (4) 2  9  0  1  16  26

Definition
A unit vector is a vector whose norm is 1.
If v is a nonzero vector, then the vector
1
u v

is a unit vector in the direction of v.

v

This procedure of constructing a unit vector in the same direction

as a given vector is called normalizing the vector.

Ch04_17
 Example 3
(a) Show that the vector (1, 0) is a unit vector.
(b) Find the norm of the vector (2, –1, 3). Normalize this vector.
Solution
(a) (1, 0)  12  0 2  1. Thus (1, 0) is a unit vector. It can be
similarly shown that (0, 1) is a unit vector in R2.
(b) (2, - 1, 3)  2 2  (-1) 2  32  14. The norm of (2, –1, 3) is 14.
The normalized vector is
1
(2, - 1, 3)
14
 2 -1 3 
The vector may also be written  , , .
 14 14 14 
This vector is a unit vector in the direction of (2, –1, 3).

Ch04_18
Angle between Vectors ( in R2)

The law of cosines gives:

← Figure 4.6

Ch04_19
 Angle between Vectors (in R n)

Definition
Let u and v be two nonzero vectors in Rn.
The cosine of the angle  between these vectors is
uv
cos   0  
u v

Example 4
Determine the angle between the vectors u = (1, 0, 0) and
v = (1, 0, 1) in R3.
Solution u  v  (1, 0, 0)  (1, 0, 1)  1
u  1  0  0 1
2 2 2
v  1  0 1  2
2 2 2

uv 1
Thus cos    , the angle between u and v is 45.
u v 2
Ch04_20
 Orthogonal Vectors

Definition
Two nonzero vectors are orthogonal if the angle between them is a
right angle .

Theorem 4.2
Two nonzero vectors u and v are orthogonal if and only if u.v = 0.

Proof
u, v are orthogonal  cos  0  u  v  0

Ch04_21
Example 5
Show that the following pairs of vectors are orthogonal.
(a) (1, 0) and (0, 1).
(b) (2, –3, 1) and (1, 2, 4).
Solution
(a) (1, 0).(0, 1) = (1  0) + (0  1) = 0.
The vectors are orthogonal.
(b) (2, –3, 1).(1, 2, 4) = (2  1) + (–3  2) + (1  4) = 2 – 6 + 4 = 0.
The vectors are orthogonal.

Ch04_22
Note

(1, 0), (0,1) are orthogonal unit vectors in R2.

(1, 0, 0), (0, 1, 0), (0, 0, 1) are orthogonal unit
vectors in R3.
(1, 0, …, 0), (0, 1, 0, …, 0), …, (0, …, 0, 1) are
orthogonal unit vectors in Rn.

Ch04_23
 Example 6
Determine a vector in R2 that is orthogonal to (3, –1). Show that
there are many such vectors and that they all lie on a line.
Solution
Let the vector (a, b) be orthogonal to (3, –1).
We get (a, b)  (3, - 1)  0
(a  3)  (b  (-1))  0
3a - b  0
b  3a
Thus any vector of the form (a, 3a)
is orthogonal to the vector (3, –1).
Any vector of this form can be written
a(1, 3)
The set of all such vectors lie on the
line defined by the vector (1, 3).
Figure 4.7
Ch04_24
 Theorem 4.3
Let u and v be vectors in Rn.
(a) Triangle Inequality:
||u + v||  ||u|| + ||v||.
(a) Pythagorean theorem :
If u.v = 0 then ||u + v||2 = ||u||2 + ||v||2.

Figure 4.8(a) Figure 4.8(b)

Ch04_25
 Distance between Points
Let x  ( x1 , x2 , ..., xn ) and y  ( y1 , y2 , ..., yn ) be two points in Rn.
The distance between x and y is denoted d(x, y) and is defined by
d (x, y )  ( x1 - y1 ) 2    ( xn - yn ) 2
x
Note: We can also write this distance as follows. x-y
d (x, y )  x - y y

Note: It is clear that d (x, y)  d (y, x) (the symmetric property)

Example 7. Determine the distance between the points

x = (1,–2 , 3, 0) and y = (4, 0, –3, 5) in R4.
Solution d (x, y )  (1 - 4) 2  (-2 - 0) 2  (3  3) 2  (0 - 5) 2
 9  4  36  25
 74

Ch04_26
 4.3 General Vector Spaces
Our aim in this section will be to focus on the algebraic
properties of Rn.

Definition
A vector space is a set V of elements called vectors, having
operations of addition and scalar multiplication defined on it
that satisfy the following conditions.
Let u, v, and w be arbitrary elements of V, and c and d are scalars.
• Closure Axioms
1. The sum u + v exists and is an element of V. (V is closed
under addition.)
2. cu is an element of V. (V is closed under scalar multiplication.)

Ch04_27
 Example 1
(1) V={ …, -3, -1, 1, 3, 5, 7, …}
V is not closed under addition because 1+3=4  V.

(2) Z={ …, -2, -1, 0, 1, 2, 3, 4, …}

Z is closed under addition because
for any a, b  Z, a + b Z.
Z is not closed under scalar multiplication because
½ is a scalar, for any odd a  Z, (½)a  Z.

Ch04_28
 Definition of Vector Space (continued)
• Addition Axioms
3. u + v = v + u (commutative property)
4. u + (v + w) = (u + v) + w (associative property)
5. There exists an element of V, called the zero vector, denoted 0,
such that u + 0 = u.
6. For every element u of V there exists an element called the
negative of u, denoted -u, such that u + (-u) = 0.
• Scalar Multiplication Axioms
7. c(u + v) = cu + cv
8. (c + d)u = cu + du
9. c(du) = (cd)u
10. 1u = u

Ch04_29
 A Vector Space in R3

Let W  { a (1, 0, 1) | a  R }. Prove that W is a vector space.

Proof
Let u  a(1,0,1) and v  b(1,0,1) W , for some a, bR.
Axiom 1: u  v  a(1,0,1)  b(1,0,1)  (a  b)(1,0,1)
u + v W. Thus W is closed under addition.
Axiom 2: cu  ca (1,0,1)W.
Thus W is closed under scalar multiplication.
Axiom 5: Let 0 = (0, 0, 0) = 0(1,0,1),
then 0 W and 0+u = u+0 = u for any u  W.
Axiom 6: For any u = a(1,0,1)  W. Let -u = -a(1,0,1),
then -u W and (-u)+u = 0.
Axiom 3,4 and 7~10: trivial Ch04_30
Vector Spaces
Vector spaces:
Let V be a set on which two operations (vector 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.
Addition:
(1) u+v is in V
(2) u+v=v+u
(3) u+(v+w)=(u+v)+w
(4) V has a zero vector 0 such that for every u in V, u+0=u
(5) For every u in V, there is a vector in V denoted by –u
such that u+(–u)=0
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
 Vector Spaces of Matrices (Mmn)
 p q
Let M 22 {  | p , q , r , s  R }. Prove that M22 is a vector space.
r s

Proof a b  e f
Let u    and v     M 22 .
c d  g h
Axiom 1:
f  a  e b f 
uv  
a b e
 
 c d   g h  c  g d  h 

u + v is a 2  2 matrix. Thus M22 is closed under addition.

► Question: Prove Axiom 2 and Axiom 7 .
Axiom 3 and 4:
From our previous discussions we know that 2  2 matrices are
commutative and associative under addition (Theorem 2.2).
Ch04_33
Axiom 5:
 0 0
The 2  2 zero matrix is 0  0 0 , since
 
u0   a b  0 0   a b 
  u
 c d  0 0  c d 
Axiom 6:
- a - b
If u  
a b
, then - u   , since
 c d   - c - d 
 a b   - a - b   a - a b - b  0 0 
u  ( -u )     0
 c d   - c - d   c - c d - d  0 0 

In general: The set of m  n matrices, Mmn, is a vector space.

p q
Is the set W  {   | p ,q ,r ,s  0} a vector space?
r s
Ch04_34
 Vector Spaces of Functions
Prove that F = { f | f: R R } is a vector space.

Let f, g  F, c  R. For example: f: R R, f(x)=2x,

Axiom 1: g: R R, g(x)=x2+1.
f + g is defined by (f + g)(x) = f(x) + g(x).
 f + g : R R
 f + g  F. Thus F is closed under addition.

Axiom 2:
cf is defined by (cf)(x) = c f(x).
 cf : R R
 cf  F. Thus F is closed under scalar multiplication.

Ch04_35
 Vector Spaces of Functions (continued)
Axiom 5:
Let 0 be the function such that 0(x) = 0 for every xR.
0 is called the zero function.
We get (f + 0)(x) = f(x) + 0(x) = f(x) + 0 = f(x) for every xR.
Thus f + 0 = f. (0 is the zero vector.)

Axiom 6:
Let the function –f defined by (-f )(x) = -f (x).
[ f  (- f )]( x)  f ( x)  (- f )( x)
 f ( x) - [ f ( x)]
0
 0( x)
Thus [f + (-f )] = 0, -f is the negative of f.
Is the set F ={ f | f(x)=ax2+bx+c for some a,b,c R} a vector space?
Ch04_36
 Theorem 4.4 (useful properties)
Let V be a vector space, v a vector in V, 0 the zero vector of V,
c a scalar, and 0 the zero scalar. Then
(a) 0v = 0
(b) c0 = 0
(c) (-1)v = -v
(d) If cv = 0, then either c = 0 or v = 0.

Ch04_37
4.4 Subspaces

Figure 4.9 Ch04_38

Note:
► In general, a subset of a vector space may or may not satisfy
the closure axioms.
► However, any subset that is closed under both of these
operations satisfies all the other vector space properties.

Definition
Let V be a vector space and U be a nonempty subset of V.
U is said to be a subspace of V if it is closed under addition
and under scalar multiplication.

Ch04_39
 Example 1
Let U be the subset of R3 consisting of all vectors of the form
(a, 0, 0) (with zeros as second and third components and aR ),
i.e., U = {(a, 0, 0)  R3 }.
Show that U is a subspace of R3.
Solution
Let (a, 0, 0), (b, 0, 0)  U, and let k R.
We get
(a, 0, 0) + (b, 0, 0) = (a + b, 0, 0)  U
k(a, 0, 0) = (k a, 0, 0)  U
The sum and scalar product are in U.
Thus U is a subspace of R3. #

Geometrically, U is the set of vectors that lie on the x-axis.

Ch04_40
 Example 2
Let V be the set of vectors of of R3 of the form (a, a2, b), namely
V = {(a, a2, b)  R3 }.
Show that V is not a subspace of R3.

Solution
Let (a, a2, b), (c, c2, d)  V.
(a, a2, b) + (c, c2, d) = (a+ c, a2 + c2, b + d)
 (a + c, (a + c)2, b + d) ,
since a2 + c2  (a + c)2 .
Thus (a, a2, b) + (c, c2, d)  V.
V is not closed under addition.
V is not a subspace.

Ch04_41
 Example 3
Prove that the set W of 2  2 diagonal matrices is a subspace of
the vector space M22 of 2  2 matrices.
Solution
(+) Let u  a 0 and v   p 0  W.
0 b  0 q 
 a 0  p 0   a  p 0 
We get u  v   

 

 0 b   0 q   0 b  q 
 u + v  W.
 W is closed under addition.

() Let c R. We get cu  c 

a 0 ca 0 

0 b  0 cb
 cu  W.
 W is closed under scalar multiplication.
 W is a subspace of M22.
Ch04_42
 The vector space of polynomials (Pn)
5. Let Pn denoted the set of real polynomial functions of
Example 4.
degree  n. Prove that Pn is a vector space if addition and scalar
multiplication are defined on polynomials in a pointwise manner.
Solution
Let f and g  Pn, where
f ( x)  an x n  an -1 x n-1  ...  a1 x  a0 and
g ( x)  bn x n  bn-1 x n -1  ...  b1 x  b0
►(+) ( f  g )( x)
 f ( x)  g ( x)
 [an x n  an -1 x n -1  ...  a1 x  a0 ]  [bn x n  bn -1 x n -1  ...  b1 x  b0 ]
 (an  bn ) x n  (an -1  bn -1 ) x n -1  ...  (a1  b1 ) x  (a0  b0 )
(f + g)(x) is a polynomial of degree  n.
Thus f + g  Pn.
Then Pn is closed under addition. Ch04_43
►() Let c R,(cf )( x)  c[ f ( x)]
 c[an x n  an -1 x n -1  ...  a1 x  a0 ]
 can x n  can -1 x n -1  ...  ca1 x  ca0

(cf )(x) is a polynomial of degree  n.

So cf  Pn.
Then Pn is closed under scalar multiplication.

In conclusion : By (+) and (), Pn is a subspace of the vector

space F of functions.
Therefore Pn is itself a vector space.

Ch04_44
 Theorem 4.5 (Very important condition)
Let U be a subspace of a vector space V.
U contains the zero vector of V.

Note. Let 0 be the zero vector of V.

If 0 U  U is not a subspace of V.
If 0 U  (+)() hold  U is a subspace of V.
(+)() failed U is not a subspace of V.
Caution. This condition is necessary but not sufficient.
(See, for instance, Example 2 above and Example 5 below)

Ch04_45
 Example 5
Let W be the set of vectors of the form (a, a, a+2).
Show that W is not a subspace of R3.
Solution
If (a, a, a+2) = (0, 0, 0), then a = 0 and a + 2 = 0 .
This system is inconsistent it has no solution.
Thus (0, 0, 0)  W. (The necessary condition does not hold)
 W is not a subspace of R3.

Ch04_46
 4.5 Linear Combinations of Vectors
W={(a, a, b) | a,b R}  R3

(a, a, b) = a (1,1,0) + b (0,0,1)

W is generated by (1,1,0) and (0,0,1).
e.g., (2, 2, 3) = 2 (1,1,0) + 3 (0,0,1)
(-1, -1, 7) = -1 (1,1,0) + 7 (0,0,1).

Definition
Let v1, v2, …, vm be vectors in a vector space V.
We say that v, a vector of V, is a linear combination of
v1, v2, …, vm , if there exist scalars c1, c2, …, cm such that
v can be written v = c1v1 + c2v2 + … + cmvm .

Ch04_47
 Example 1
The vector (5, 4, 2) is a linear combination of the vectors
(1, 2, 0), (3, 1, 4), and (1, 0, 3), since it can be written
(5, 4, 2) = (1, 2, 0) + 2(3, 1, 4) – 2(1, 0, 3)

Ch04_48
 Example 2
Determine whether or not the vector (-1, 1, 5) is a linear
combination of the vectors (1, 2, 3), (0, 1, 4), and (2, 3, 6).
Solution
Suppose c1 (1, 2, 3)  c2 (0, 1, 4)  c3 (2, 3, 6)  (-1, 1, 5)
(c1 , 2c1 , 3c1 )  (0, c2 , 4c2 )  (2c3 , 3c3 , 6c3 )  (-1, 1, 5)
(c1  2c3 , 2c1  c2  3c3 , 3c1  4c2  6c3 )  (-1, 1, 5)
 c1  2c3  -1

  2c1  c2  3c3  1  c1  1, c2  2, c3  -1
 3c  4c  6c  5
 1 2 3
Thus (-1, 1, 5) is a linear combination of (1, 2, 3), (0, 1, 4), and
(2, 3, 6), where (-1, 1, 5)  (1, 2, 3)  2(0, 1, 4) - 1(2, 3, 6).

Ch04_49
 Example 3
Express the vector (4, 5, 5) as a linear combination of the vectors
(1, 2, 3), (-1, 1, 4), and (3, 3, 2).
Solution
Suppose c1 (1, 2, 3)  c2 (-1, 1, 4)  c3 (3, 3, 2)  (4, 5, 5)

(c1 , 2c1 , 3c1 )  (-c2 , c2 , 4c2 )  (3c3 , 3c3 , 2c3 )  (4, 5, 5)

(c1 - c2  3c3 , 2c1  c2  3c3 , 3c1  4c2  2c3 )  (4, 5, 5)
 c1 - c2  3c3  4

 2c1  c2  3c3  5  c1  -2r  3, c2  r - 1, c3  r
3c  4c  2c  5
 1 2 3
Thus (4, 5, 5) can be expressed in many ways as a linear
combination of (1, 2, 3), (-1, 1, 4), and (3, 3, 2):
(4, 5, 5)  (-2r  3)(1, 2, 3)  (r - 1)(-1, 1, 4)  r (2, 3, 6)
Ch04_50
 Example 4
Show that the vector (3, -4, -6) cannot be expressed as a linear
combination of the vectors (1, 2, 3), (-1, -1, -2), and (1, 4, 5).
Solution
Suppose
c1 (1, 2, 3)  c2 (-1, - 1, - 2)  c3 (1, 4, 5)  (3, - 4, - 6)

 c1 - c2  c3  3

 2c1 - c2  4c3  -4
3c - 2c  5c  - 6
 1 2 3

This system has no solution.

Thus (3, -4, -6) is not a linear combination of the vectors
(1, 2, 3), (-1, -1, -2), and (1, 4, 5).

Ch04_51
 Example 5
- 1 7
Determine whether the matrix  is a linear combination
 8 - 1
of the matrices 1 0 ,  2 - 3 , and 0 1  in the vector space
 2 1  0 2   2 0
M22 of 2  2 matrices.
Solution
Suppose  1 0  2 - 3  0 1 - 1 7 
c1 c c 
2 1 2 0 2  3 2 0  8 - 1

Then
 c1  2c2 - 3c2  c3  - 1 7 
2c  2c  
 1 3
c1  2c2   8 - 1

Ch04_52
  c1  2c2  -1

- 3c2  c3  7

 2c1  2c3  8
 c1  2c2  -1
This system has the unique solution c1 = 3, c2 = -2, c3 = 1.
Therefore
- 1 7   31 0 - 22 - 3  0 1
 8 - 1 2 1 0 2  2 0

Ch04_53
 Example 6
Determine whether the function f ( x )  x 2  10 x - 7 is a linear
combination of the functions g ( x )  x 2  3 x - 1 and
h ( x )  2 x 2 - x  4.
Solution
Suppose c1 g  c2 h  f .
Then
c1 ( x 2  3x - 1)  c2 (2 x 2 - x  4)  x 2  10x - 7

(c1  2c2 ) x 2  (3c1 - c2 ) x - c1  4c2  x 2  10x - 7

 c1  2c2  1

  3c1 - c2  10  c1  3, c2  -1  f  3g - h.
- c  4c  -7
 1 2
Ch04_54
 Spanning Sets
Definition
The vectors v1, v2, …, vm are said to span a vector space if every
vector in the space can be expressed as a linear combination of
these vectors.
In this case {v1, v2, …, vm} is called a spanning set.

Ch04_55
 Example 7
Show that the vectors (1, 2, 0), (0, 1, -1), and (1, 1, 2) span R3.

Solution
Let (x, y, z) be an arbitrary element of R3.
Suppose ( x, y, z )  c1 (1, 2, 0)  c2 (0, 1, - 1)  c3 (1, 1, 2)

 ( x, y, z )  (c1  c3 , 2c1  c2  c3 , - c2  2c3 )

 c1  c3  x c1  3x - y - z
 
 2c1  c2  c3  y  c2  -4 x  2 y  z
 - c  2c  z c  -2 x  y  z
 2 3  3

 ( x, y, z)  (3x - y - z)(1,2,0)  (-4 x  2 y  z)(0, 1, - 1)  (-2 x  y  z )(1, 1, 2)

The vectors (1, 2, 0), (0, 1, -1), and (1, 1, 2) span R3.
Ch04_56
 Example 8
Show that the following matrices span the vector space M22 of
2  2 matrices. 1 0
  0 1  0 0  0 0 
0 0 0 0 1 0 0 1 
Solution

Let  a b
 M22 (an arbitrary element).
 c d 
We can express this matrix as follows:

a b   a 1 0  b 0 1  c 0 0  d 0 0
 c d  0 0 0 0 1 0 0 1
proving the result.

Ch04_57
 Theorem 4.6
Let v1, …, vm be vectors in a vector space V. Let U be the set
consisting of all linear combinations of v1, …, vm .
Then U is a subspace of V spanned by the vectors v1, …, vm .
U is said to be the vector space generated by v1, …, vm .
Proof

(+) Let u1 = a1v1 + … + amvm and u2 = b1v1 + … + bmvm  U.

Then u1 + u2 = (a1v1 + … + amvm) + (b1v1 + … + bmvm)
= (a1 + b1) v1 + … + (am + bm) vm
 u1 + u2 is a linear combination of v1, …, vm .
 u1 + u2  U.
 U is closed under vector addition.

Ch04_58
()Let c R. Then
cu1 = c(a1v1 + … + amvm)
= ca1v1 + … + camvm
 cu1 is a linear combination of v1, …, vm .
 cu1  U.
 U is closed under scalar multiplication.
Thus U is a subspace of V.
By the definition of U, every vector in U can be written as a
linear combination of v1, …, vm .
Thus v1, …, vm span U.

Ch04_59
 Example 9
Consider the vector space R3.
The vectors (-1, 5, 3) and (2, -3, 4) are in R3.
Let U be the subset of R3 consisting of all vectors of the form
c1(-1, 5, 3) + c2(2, -3, 4)
Then U is a subspace of R3 spanned by (-1, 5, 3) and (2, -3, 4).
The following are examples of some of the vectors in U, obtained
by given c1 and c2 various values.
c1  1, c2  0; vector (-1, 5, 3)
c1  0, c2  1; vector (2, - 3, 4)
c1  0, c2  0; vector (0, 0, 0)
c1  2, c2  3; vector (4, 1, 18)
We can visualize U. U is made up of all vectors in the plane
defined by the vectors (-1, 5, 3) and (2, -3, 4).
Ch04_60
Figure 4.10
Ch04_61
 We can generalize this result.
Let v1 and v2 be vectors in the
space R3.

The subspace U generated by

v1 and v2 is the set of all
vectors of the form c1v1 + c2v2.

If v1 and v2 are not colinear,

then U is the plane defined by
v1 and v2 .
Figure 4.11

Ch04_62
 If v1 and v2 are vectors in R3
that are not colinear, then we
can visualize U as a plane in
three dimensions.

k1v1 and k2v2 will be vectors

on the same lines as v1 and v2.

Figure 4.12

Ch04_63
 Example 10
Let U be the subspace of R3 generated by the vectors (1, 2, 0) and
(-3, 1, 2). Let V be the subspace of R3 generated by the vectors
(-1, 5, 2) and (4, 1, -2). Show that U = V.
Solution
(U V) Let u be a vector in U. Let us show that u is in V.
Since u is in U, there exist scalars a and b such that
u  a(1, 2, 0)  b(-3, 1, 2)  (a - 3b, 2a  b, 2b)
Let us see if we can write u as a linear combination of (-1, 5, 2)
and (4, 1, -2).
u  p(-1, 5, 2)  q(4, 1, - 2)  (- p  4q, 5 p  q, 2 p - 2q)
Such p and q would have to satisfy
- p  4q  a - 3b
5 p  q  2a  b
2 p - 2q  2b
Ch04_64
This system of equations has unique solution p  a  b , q  a - 2b .
Thus u can be written 3 3
ab a - 2b
u (-1, 5, 2)  (4, 1, - 2)
3 3
Therefore, u is a vector in V.

(V U) Let v be a vector in V. Let us show that v is in U.

Since v is in V, there exist scalars c and d such that
v  c(-1, 5, 2)  d (4 , 1, - 2)
It can be shown that
v  (2c  d )(1, 2, 0)  (c - d )(-3, 1, 2)
Therefore, v is a vector in U and hence U=V.

Ch04_65
Figure 4.13

Ch04_66
 Example 11
Let U be the vector space generated by the functions f(x) = x + 1
and g(x) = 2x2 – 2x + 3. Show that the function h(x) = 6x2–10x+5
lies in U.
Solution
h will be in the space generated by f and g if there exist scalars a
and b such that
a(x + 1) + b(2x2 – 2x + 3) = 6x2 – 10x + 5
This given
2bx2 + (a – 2b)x + a + 3b = 6x2 – 10x + 5
2b = 6
 a – 2b = – 10
a + 3b = 5
This system has the unique solution a = – 4, b = 3.
Thus – 4(x + 1) + 3(2x2 – 2x + 3) = 6x2 – 10x + 5
Ch04_67
 4.6 Linear Dependence and Independence

The concepts of dependence and independence of vectors are

useful tools in constructing “efficient” spanning sets for vector
spaces – sets in which there are no redundant vectors.

Definition
(a) The set of vectors { v1, …, vm } in a vector space V is said to
be linearly dependent if there exist scalars c1, …, cm, not all
zero, such that c1v1 + … + cmvm = 0
(b) The set of vectors { v1, …, vm } is linearly independent if
c1v1 + … + cmvm = 0 can only be satisfied when
c1 = 0, …, cm = 0.
Ch04_68
 Example 1
Show that the set {(1, 2, 3), (-2, 1, 1), (8, 6, 10)} is linearly
dependent in R3.
Solution
Suppose c1 (1, 2, 3)  c2 (-2, 1, 1)  c3 (8, 6, 10)  0

 (c1 , 2c1 , 3c1 )  (-2c2 , c2 , c2 )  (8c3 , 6c3 , 10c3 )  0

(c1 - 2c2  8c3 , 2c1  c2  6c3 , 3c1  c2  10c3 )  0

 c1 - 2c2  8c3  0
 c1 = 4
  2c1  c2  6c3  0  c = -2
3c  c  10c  0
2
 1 2 3 c3 = -1

Thus 4(1, 2, 3) - 2(-2, 1, 1) - (8, 6, 10)  0

The set of vectors is linearly dependent.
Ch04_69
 Example 2
Show that the set {(3, -2, 2), (3, -1, 4), (1, 0, 5)} is linearly
independent in R3.
Solution
Suppose c1 (3, - 2, 2)  c2 (3, - 1, 4)  c3 (1, 0, 5)  0

 (3c1 , - 2c1 , 2c1 )  (3c2 , - c2 , 4c2 )  (c3 , 0, 5c3 )  0

(3c1  3c2  c3 , - 2c1 - c2 , 2c1  4c2  5c3 )  0

  3c1  3c2  c3  0

 - 2c1 - c2  0
2c  4c  5c  0
 1 2 3

This system has the unique solution c1 = 0, c2 = 0, and c3 = 0.

Thus the set is linearly independent.
Ch04_70
 Example 3
Consider the functions f(x) = x2 + 1, g(x) = 3x – 1, h(x) = – 4x + 1
of the vector space P2 of polynomials of degree  2.
Show that the set of functions { f, g, h } is linearly independent.
Solution
Suppose
c1 f + c 2 g + c3 h = 0
Since for any real number x,
c1 ( x 2  1)  c2 (3x - 1)  c3 (-4 x  1)  0
Consider three convenient values of x. We get
x  0 : c1 - c2  c3  0
x  1 : 2c1  2c2 - 3c3  0
x  -1 : 2c1 - 4c2  5c3  0

Ch04_71
It can be shown that this system of three equations has the unique
solution
c1 = 0, c2 = 0, c3 = 0
Thus c1f + c2g + c3h = 0 implies that c1 = 0, c2 = 0, c3 = 0.
The set { f, g, h } is linearly independent.

Ch04_72
 Theorem 4.7
A set consisting of two or more vectors in a vector space is
linearly dependent if and only if it is possible to express one
of the vectors as a linearly combination of the other vectors.

Example 4
The set of vectors {v1=(1, 2, 1) , v2 =(-1, -1, 0) , v3 = (0, 1,1)}
is linearly dependent, since v3 = v1 + v2 .
Thus, v3 is a linear combination of v1 and v2.

Ch04_73
 Linear Dependence of {v1, v2}

{v1, v2} linearly dependent; {v1, v2} linearly independent;

vectors lie on a line vectors do not lie on a line

Figure 4.14 Linear dependence and independence of {v1, v2} in R3.

Ch04_74
 Linear Dependence of {v1, v2, v3}

{v1, v2, v3} linearly dependent; {v1, v2, v3} linearly independent;
vectors lie in a plane vectors do not lie in a plane

Figure 4.15 Linear dependence and independence of {v1, v2, v3} in R3.

Ch04_75
 Theorem 4.8
Let V be a vector space. Any set of vectors in V that contains the
zero is linearly dependent.
Proof
Consider the set { 0, v2, …, vm }, which contains the zero vectors.
Let us examine the identity
c1 0  c2 v 2    cm v m  0
We see that the identity is true for c1 = 1, c2 = 0, …, cm = 0 (not
all zero).
Thus the set of vectors is linearly dependent, proving the theorem.

Ch04_76
 Theorem 4.9
Let the set {v1, …, vm} be linearly dependent in a vector space V.
Any set of vectors in V that contains these vectors will also be
linearly dependent.

Example 5
The set of vectors
{v1=(1, 2, 1) , v2 =(-1, -1, 0) , v3 = (0, 1,1) , v4 =( 1, 1, 1) }
is linearly dependent, since it contains the vectors v1 , v2 , v3
which are linearly dependent.

Ch04_77
 4.7 Bases and Dimension
Definition
A finite set of vectors {v1, …, vm} is called a basis for a vector
space V if the set spans V and is linearly independent.

Standard Basis
The set of n vectors
{(1, 0, …, 0), (0, 1, …, 0), …, (0, …, 1)}
is a basis for Rn. This basis is called the standard basis for Rn.
How to prove it?

Ch04_78
 Example 1
Show that the set {(1, 0, -1), (1, 1, 1), (1, 2, 4)} is a basis for R3.
Solution
(span)
Let (x1, x2, x3) be an arbitrary element of R3.
Suppose
( x1 , x2 , x3 )  a1 (1, 0, - 1)  a2 (1, 1, 1)  a3 (1, 2, 4)

 a1  a2  a3  x1  a1  2 x1 - 3x2  x3
a 2  2 a3  x 2 a2  -2 x1  5 x2 - 2 x3
- a1  a2  4a3  x3 a3  x1 - 2 x2  x3
Thus the set spans the space.

Ch04_79
(linearly independent)
Consider the identity
b1 (1, 0, - 1)  b2 (1, 1, 1)  b3 (1, 2, 4)  (0, 0, 0)
The identity leads to the system of equations
b1  b2  b3  0
b2  2b3  0
- b1  b2  4b3  0
 b1 = 0, b2 = 0, and b3 = 0 is the unique solution.
Thus the set is linearly independent.
 {(1, 0, -1), (1, 1, 1), (1, 2, 4)} spans R3 and is linearly
independent.
 It forms a basis for R3.

Ch04_80
 Example 2
Show that { f, g, h }, where f(x) = x2 + 1, g(x) = 3x – 1, and h(x)
= –4x + 1 is a basis for P2.
Solution
(linearly independent) see Example 3 of the previous section.
(span). Let p be an arbitrary function in P2.
p is thus a polynomial of the form
p( x)  bx 2  cx  d
Suppose p( x)  a1 f ( x)  a2 g ( x)  a3h( x)
for some scalars a1, a2, a3.
This gives
bx2  cx  d  a1 ( x 2  1)  a2 (3x - 1)  a3 (-4 x  1)
 a1 x 2  (3a2 - 4a3 ) x  (a1 - a2  a3 )
Ch04_81
 a1  b a1  b
 3a2 - 4a3  c  a2  4b - 4d - c
a1 - a2  a3  d a3  3b - 3d - c

Thus the polynomial p can be expressed

p( x)  a1 f ( x)  a2 g ( x)  a3h( x)
The functions f, g, and h span P2.
They form a basis for P2.

Ch04_82
 Theorem 4.10
Let {v1, …, vn } be a basis for a vector space V.
If {w1, …, wm} is a set of more than n vectors in V, then this set
is linearly dependent.
Proof
Suppose
c1w1    cm w m  0 (1)
We will show that values of c1, …, cm are not all zero.
The set {v1, …, vn} is a basis for V. Thus each of the vectors
w1, …, wm can be expressed as a linear combination of v1, …, vn.
Let
w1  a11 v1  a12 v 2    a1n v n

w m  am1 v1  am 2 v 2    amn v n

Ch04_83
Substituting for w1, …, wm into Equation (1) we get
c1 (a11v1  a12 v 2    a1n v n )    cm (am1v1  am 2 v 2    amn v n )  0
Rearranging, we get
(c1a11  c2 a21    cm am1 ) v1    (c1a1n  c2 a2 n    cm amn ) v n  0
Since v1, …, vn are linear independent,
a11c1  a21c2    am1cm  0

a1n c1  a2 n c2    amn cm  0
Since m > n, there are many solutions in this system.
Thus the set {w1, …, wm} is linearly dependent.

Ch04_84
 Theorem 4.11
Any two bases for a vector space V consist of the same number
of vectors.
Proof
Let {v1, …, vn} and {w1, …, wm} be two bases for V.
By Theorem 4.10,
m  n and n  m
Thus n = m.

Ch04_85
Definition
If a vector space V has a basis consisting of n vectors, then the
dimension of V is said to be n. We write dim(V) for the
dimension of V.

• V is finite dimensional if such a finite basis exists.

• V is infinite dimensional otherwise.

Ch04_86
 Example 3
Consider the set {{1, 2, 3), (-2, 4, 1)} of vectors in R3.
These vectors generate a subspace V of R3 consisting of all
vectors of the form
v  c1 (1, 2, 3)  c2 ( -2, 4, 1)
The vectors (1, 2, 3) and (-2, 4, 1) span this subspace.
Furthermore, since the second vector is not a scalar multiple of
the first vector, the vectors are linearly independent.
Therefore {{1, 2, 3), (-2, 4, 1)} is a basis for V.
Thus dim(V) = 2.
We know that V is, in fact, a plane through the origin.

Ch04_87
 Theorem 4.12
(a) The origin is a subspace of R3. The dimension of this
subspace is defined to be zero.
(b) The one-dimensional subspaces of R3 are lines through the
origin.
(c) The two-dimensional subspaces of R3 are planes through the
origin.

Figure 4.16 One and two-dimensional subspaces of R3 Ch04_88

Proof
(a) Let V be the set {(0, 0, 0)}, consisting of a single elements,
the zero vector of R3. Let c be the arbitrary scalar. Since
(0, 0, 0) + (0, 0, 0) = (0, 0, 0) and c(0, 0, 0) = (0, 0, 0)
V is closed under addition and scalar multiplication. It is thus
a subspace of R3. The dimension of this subspaces is defined
to be zero.
(b) Let v be a basis for a one-dimensional subspace V of R3.
Every vector in V is thus of the form cv, for some scalar c. We
know that these vectors form a line through the origin.
(c) Let {v1, v2}be a basis for a two-dimensional subspace V of R3.
Every vector in V is of the form c1v1 + c2v2. V is thus a plane
through the origin.

Ch04_89
 Theorem 4.13
Let {v1, …, vn} be a basis for a vector space V. Then each vector
in V can be expressed uniquely as a linear combination of these
vectors.
Proof
Let v be a vector in V. Since {v1, …, vn} is a basis, we can
express v as a linear combination of these vectors. Suppose we
can write
v  a1v1    an v n and v  b1v1    bn v n
Then
a1v1    an v n  b1v1    bn v n
giving (a1 - b1 ) v1    (an - bn ) v n  0
Since {v1, …, vn} is a basis, the vectors v1, …, vn are linearly
independent. Thus (a1 – b1) = 0, …, (an – bn) = 0, implying that
a1 = b1, …, an = bn. There is thus only one way of expressing v
as a linear combination of the basis. Ch04_90
 Theorem 4.14
Let V be a vector space of dimension n.
(a) If S = {v1, …, vn} is a set of n linearly independent vectors
in V, then S is a basis for V.
(b) If S = {v1, …, vn} is a set of n vectors V that spans V, then S
is a basis for V.

Let V be a vector space, S = {v1, …, vn} is a set of vectors in V.

(a) dim(V) = |S|.

(b) S is a linearly independent set.
S is a basis of V.
(c) S spans V.

Ch04_91
 Example 4
Prove that the set B={(1, 3, -1), (2, 1, 0), (4, 2, 1)} is a basis for R3.
Solution
Since dim(R3)=|B|=3. It suffices to show that this set is linearly
independent or it spans R3.
Let us check for linear independence. Suppose
c1 (1, 3, - 1)  c2 (2, 1, 0)  c3 (4, 2, 1)  (0, 0, 0)

This identity leads to the system of equations

c1  2c2  4c3  0
3c1  c2  2c3  0
- c1  c3  0
This system has the unique solution c1 = 0, c2 = 0, c3 = 0.
Thus the vectors are linearly independent.
The set {(1, 3, -1), (2, 1, 0), (4, 2, 1)} is therefore a basis for R3.
Ch04_92
 Theorem 4.15
Let V be a vector space of dimension n. Let {v1, …, vm} be a set
of m linearly independent vectors in V, where m < n.
Then there exist vectors vm+1, …, vn such that
{v1, …, vm, vm+1, …, vn } is a basis of V.

Ch04_93
 Example 5
State (with a brief explanation) whether the following statements
are true or false.
(a) The vectors (1, 2), (-1, 3), (5, 2) are linearly dependent in R2.
(b) The vectors (1, 0, 0), (0, 2, 0), (1, 2, 0) span R3.
(c) {(1, 0, 2), (0, 1, -3)} is a basis for the subspace of R3 consisting
of vectors of the form (a, b, 2a -3b).
(d) Any set of two vectors can be used to generate a two-
dimensional subspace of R3.
Solution
(a) True: The dimension of R2 is two. Thus any three vectors are
linearly dependent.
(b) False: The three vectors are linearly dependent. Thus they
cannot span a three-dimensional space.
Ch04_94
(c) True: The vectors span the subspace since
(a, b, 2a -3b) = a(1, 0, 2) + b(0, 1, -3)
The vectors are also linearly independent since they are not
colinear.
(d) False: The two vectors must be linearly independent.

Ch04_95