MATH1550

Lecture 9: Vector space and subspace

Warning: the note is for reference only. It may contain typos. Read at your own risk.
The lecture is based on
Beezer, A first course in Linear algebra. Ver 3.5
Downloadable at http://linear.ups.edu/download.html.
The print version can be downloaded at
http://linear.ups.edu/download/fcla-3.50-print.pdf.

Textbook Reading:
Beezer, Ver 3.5 Section VO (print version p57 - p63)
Subsection VS, EVS (print version p197-203) Strang, Section 2.1
Exercise
Exercises with solutions can be downloaded at
http://linear.ups.edu/download/fcla-3.50-solution-manual.pdf

1 Vectors
Notation: R is the set of real numbers.
If X is a set, x ∈ X means x is an element of the set X.

Definition 1 (Vector Space of Column Vectors) The vector space Rm is the set of
all column vectors of size m with entries from the set of real numbers, R.
Rm is also called the Euclidean m-space.
Definition 2 (Column Vector Equality) Suppose that u, v ∈ Rm . Then u and v
are equal, written u = v if

[u]i = [v]i 1≤i≤m

That is,
u1 v1
   
 u2   v2 
 . = . 
 ..   .. 
um vm
if
ui = vi 1 ≤ i ≤ m.

Example 1 The system of linear equations

−7x1 − 6x2 − 12x3 = −33
5x1 + 5x2 + 7x3 = 24
x1 + 4x3 = 5

1
can be rewritten as    
−7x1 − 6x2 − 12x3 −33
 5x1 + 5x2 + 7x3  =  24  .
x1 + 4x3 5

Definition 3 (Column Vector Addition) Suppose that u, v ∈ Rm . The sum of u
and v is the vector u + v defined by

[u + v]i = [u]i + [v]i 1 ≤ i ≤ m.

That is
u1 v1 u1 + v1
     
 u2   v2   u2 + v2 
 . + . = .. .
 ..   ..   . 
um vm um + vm

Example 2 Addition of two vectors in R4

If
  
2 −1
−3 5
u= 4
 v=
2

2 −7

then        
2 −1 2 + (−1) 1
−3  5   −3 + 5   2 
 4 + 2 = 4+2 = 6 
u+v =       
2 −7 2 + (−7) −5


Definition 4 (Column Vector Scalar Multiplication) Suppose u ∈ Rm and α ∈

R, then the scalar multiple of u by α is the vector αu defined by

[αu]i = α [u]i 1 ≤ i ≤ m.
That is
u1 αu1
   
 u2   αu2 
α
 ...  =  ...  .
  

um αum

Example 3 If 

3
1
 
−2
u= 
4
−1

2
and α = 6, then      
3 6(3) 18
 1   6(1)   6 
     
αu = 6 −2  = 6(−2) = −12 .
     
 4   6(4)   24 
−1 6(−1) −6

Example 4 The system of linear equations
−7x1 − 6x2 − 12x3 = −33
5x1 + 5x2 + 7x3 = 24
x1 + 4x3 = 5

can be written as
       
−7 −6 −12 −33
x1  5  + x2  5  + x3  7  =  24  .
1 0 4 5


2 Vector Space Properties

With definitions of vector addition and scalar multiplication we can state, and prove,
several properties of each operation, and some properties that involve their interplay.
We now collect ten of them here for later reference.
Theorem 5 Vector Space Properties of Column Vectors
Suppose that Rm is the set of column vectors of size m with addition and scalar
multiplication as defined in Definition 3 and Definition 4. Then
1. ACC Additive Closure, Column Vectors
If u, v ∈ Rm , then u + v ∈ Rm .
2. SCC Scalar Closure, Column Vectors
If α ∈ R and u ∈ Rm , then αu ∈ Rm .
3. CC Commutativity, Column Vectors
If u, v ∈ Rm , then u + v = v + u.
4. AAC Additive Associativity, Column Vectors
If u, v, w ∈ Rm , then u + (v + w) = (u + v) + w.
5. ZC Zero Vector, Column Vectors
There is a vector, 0, called the zero vector, such that u + 0 = u for all u ∈ Rm .
6. AIC Additive Inverses, Column Vectors
If u ∈ Rm , then there exists a vector −u ∈ Rm so that u + (−u) = 0.
7. SMAC Scalar Multiplication Associativity, Column Vectors
If α, β ∈ R and u ∈ Rm , then α(βu) = (αβ)u.

3
 8. DVAC Distributivity across Vector Addition, Column Vectors
If α ∈ R and u, v ∈ Rm , then α(u + v) = αu + αv.
9. DSAC Distributivity across Scalar Addition, Column Vectors
If α, β ∈ R and u ∈ Rm , then (α + β)u = αu + βu.
10. OC One, Column Vectors
If u ∈ Rm , then 1u = u.
Proof. You can skip the proof
While some of these properties seem very obvious, they all require proof. However, the
proofs are not very interesting, and border on tedious. We will prove one version of
distributivity very carefully, and you can test your proof-building skills on some of the
others. We need to establish an equality, so we will do so by beginning with one side of
the equality, apply various definitions and theorems (listed to the right of each step) to
massage the expression from the left into the expression on the right. Here we go with
a proof of Property 9 (DSAC)
For 1 ≤ i ≤ m,
[(α + β)u]i = (α + β) [u]i Definition 4
= α [u]i + β [u]i
= [αu]i + [βu]i Definition 4
= [αu + βu]i Definition 3
Since the individual components of the vectors (α + β)u and αu + βu are equal for
all i, 1 ≤ i ≤ m, Definition 2 tells us the vectors are equal. 

Many of the conclusions of our theorems can be characterized as identities, especially

when we are establishing basic properties of operations such as those in this section. Most
of the properties listed in Theorem 5 are examples. So some advice about the style we
use for proving identities is appropriate right now.
Be careful with the notion of the vector −u. This is a vector that we add to u so that
the result is the particular vector 0. This is basically a property of vector addition. It
happens that we can compute −u using the other operation, scalar multiplication. We
can prove this directly by writing that
[−u]i = − [u]i = (−1) [u]i = [(−1)u]i
We will see later how to derive this property as a consequence of several of the ten
properties listed in Theorem 5.
Similarly, we will often write something you would immediately recognize as vector
subtraction. This could be placed on a firm theoretical foundation – as you can do
yourself with exercise T30.
A final note. Theorem 5 Property 4 (AAC) implies that we do not have to be careful
about how we parenthesize the addition of vectors. In other words, there is nothing
to be gained by writing (u + v) + (w + (x + y)) rather than u + v + w + x + y, since
we get the same result no matter which order we choose to perform the four additions.
So we will not be careful about using parentheses this way.

3 Subspaces
Definition 6 Let V be vector space. A subset W of V is said to be a subspace of V if

4
 1. W is nonempty.
2. For v, w ∈ W , then v + w ∈ W .
3. For α ∈ R, v ∈ W , then αv ∈ W .
We will prove several theorem first before we give examples.
Proposition 7 Let V be a vector spaceand W a subspace of V . Then 0 is in W .
Proof. By Definition 6, Condition 1, W is nonempty. Let w ∈ W .
By Definition 6, Condition 3, with α = −1, (−1)w = −w ∈ W .
By Definition 6, Condition 2, with v = −w, then v+w ∈ W . But v+w = (−w)+w = 0.
So 0 ∈ W . 

Theorem 8 Let V be a vector space and W a subset of V , then W is a subspace if and

only if
1. W is nonempty.
2. For any α ∈ R, v, w ∈ W , αv + w ∈ W .
Proof.
(⇒) By Definition 6, Condition 1, W is nonempty.
Next, for α ∈ R, v, w ∈ W . By Definition 6, Condition 3, αv ∈ W . By Definition 6,
Condition 2, αv + w ∈ W .
(⇐) By Condition 1, Definition 6, Condition 1 is true.
Because W is nonempty, let x ∈ W . Let v = w = x and α = −1. Then by condition 2,
(−1)w + w = 0 ∈ W .
Now we want to check Definition 6 Condition 2, suppose v, w ∈ W . In condition 2, let
α = 1, then v + w = αv + w ∈ W .
Next we want to check Definition 6 Condition 3, suppose v ∈ W , α ∈ R. Let w = 0,
then αv = αv + w ∈ W . 

4 Examples
To verify if W is a subspace, we can either use Definition 6 or Theorem 8. Usually we
use the equivalent definition given in Theorem 8 because it involves only one condition.
Example 5 V = Rm , W = {0}.
W consists of one element. It is called the zero subspace of V .
Check that it is a subspace: W is nonempty. For any α ∈ R, v, w ∈ W , v = w = 0,
αv + w = 0 ∈ W . Thus by Theorem 8, W is a subspace. 
Example 6 V = Rm , W = V .
Then obviously W is a subspace. 
Example 7 V = Rm , W = {v ∈ V | [v]1 = 0}.
Check that W is a subspace: First all all, 0 ∈ W , so W is nonempty. For any α ∈ R,
v, w ∈ W . Then [v]1 = [w]1 = 0. So [αv + w]1 = α[v]1 + [w]1 = 0. Hence αv + w ∈ W .
Thus by Theorem 8, W is a subspace. 

5
   
 x 
3 3
Example 8 V = R , W =  y ∈ R x + 2y + 3z = 0 .

 z 
Check that W is a subspace:
    0 ∈ W , so W is nonempty.
Obviously
v1 w1
For any α ∈ R, v = v2 , w = w2  ∈ W . Then
  
v3 w3
v1 + 2v2 + 3v3 = 0,
w1 + 2w2 + 3w3 = 0.
So
0 = α(v1 + 2v2 + 3v3 ) + (w1 + 2w2 + 3w3 ) = (αv1 + w1 ) + 2(αv2 + w2 ) + 3(αv3 + w3 ).
Hence  
αv1 + w1
αv + w = αv2 + w2  ∈ W.
αv3 + w3
Thus by Theorem 8, W is a subspace. 
Theorem 9 Let A ∈ Mmn , then W = N (A) is a subspace of Rn .
Proof. Because 0 ∈ N (A), so W is nonempty.
For α ∈ R, v, w ∈ W . Then
Av = 0, Aw = 0.
Then
A(αv + w) = αAv + Aw = α0 + 0 = 0.
So
αv + w ∈ N (A) .
Thus by Theorem 8, W = N (A) is a subspace. 

Theorem 10 Skipped for now, until you learn the definition of span
Let S = {u1 , . . . , uk } ⊆ V = Rm . Then hSi is a subspace of V .
Proof. Obviously hSi is nonempty.
For α ∈ R, v, w ∈ W = hSi. Then there exists α1 , . . . , αk , β1 , . . . , βk such that
v = α1 u1 + · · · + αk uk ,
w = β1 u1 + · · · + βk uk .
Then
αv + w = (αα1 + β1 )u1 + · · · + (ααk + βk )uk
is in hSi. Thus by Theorem 8, W is a subspace. 

Corollary 11 Skipped for now, until you learn the definition of column space
Let A ∈ Mmn , then C(A) is a subspace of Rm .

6
Proof. C(A) = h{A1 , . . . , An }i. So by the previous theorem, C(A) is a subspace of Rm .
Alternate proof: Suppose For α ∈ R, v, w ∈ W = C(A). Recall
C(A) = {Ax | x ∈ Rm }.
Then there exist x, y such that Ax = v, Ay = w.

αv + w = αAx + Ay = A(αx + y) ∈ C(A) .

Thus by Theorem 8, W is a subspace. 

5 Non examples
To show that W is not a subspace of V , it suffices to show that it violates Definition
6 condition 1 or condition 2. This can be done by finding counter examples to either
condition. Usually before checking those conditions, we quickly check if 0V ∈ W (see
Proposition 7).
Example 9 V = Rm , W = {v ∈ V | [v]1 = 1}.
Method 1 Obviously 0 is not in W . So by proposition 7, W is not a subspace.
Method 2 For Suppose v, w ∈ W . Then [v + w]1 = [v]1 + [w]1 = 2. So v + w 6∈ W .
So W violates Definition 6 condition 1 and hence not a subspace. 
Example 10 V = Rm , W = {v ∈ V | ni=1 [v]i = 1}.
P
Method 1 (the easiest method) Obviously 0 is not in W . So by Proposition 7, W is
not a subspace.
Method 2 We will find an explicit counter example, let
1
 
0
v=w=  ...  .


0
Then both v and w are in W .
2
 
0
v+w =
 ...  .


0
Obvious v + w ∈
/ W . Therefore W violates Definition 6 condition 1 and hence not a
subspace.

Example 11 V = Rn , W = {v ∈ V | [v]1 ≥ 0}.
Let α = −1 and  
1
v= .
0
Then [αv]1 = α[v]1 = α = −1 < 0. So αv ∈
/ W . Thus W violates Theorem 6 condition
3 and hence not a subspace. 

7
  
v
Example 12 V = R , W = { 1 ∈ V | v1 v2 ≥ 0}.
2
v2
     
1 0 1
Let v = , w = ∈ W. v + w = . Because 1 × (−1) = −1 < 0. So
0 −1 −1
v+w ∈ / W . Thus W violates Definition 6 condition 2 and hence not a subspace.
In fact, we can show that W satisfies Definition 6 condition 3 but fails condition 2.