ALGEBRA - First Year - Computer Science Prof. dr.

Septimiu Crivei

Course 3: 15.10.2020

Chapter 2 VECTOR SPACES

This chapter deals with vector spaces (also called linear spaces), that are algebraic structures having also
an “external operation” beside a usual operation, as it was the case in the previous chapter. They are the
bricks of Linear Algebra and have numerous applications in different branches of Mathematics, but also
in Physics, Computer Science and in other fields of Natural Sciences. Some of their algebraic applications
will be studied in the next chapter, dedicated to the study of matrices and linear systems of equations.
Throughout the present chapter K will always denote a field.

2.1 Basic definitions, properties and examples

The reader surely remembers the notion of a vector, as an object met in elementary Physics, characterized
by an origin, a direction, a sense and a length. We might wonder if the notion of a vector space yet to
be defined is somehow connected to or generalizes the classical notion of a vector. The answer is positive
and we are going to see that in a forthcoming example.
But let us begin with the definition of the key notion of a vector space.

Definition 2.1.1 A vector space over K (or a K-vector space) is an abelian group (V, +) together with
a so-called external operation
· : K × V → V , (k, v) 7→ k · v ,

satisfying the following axioms:

(L1 ) k · (v1 + v2 ) = k · v1 + k · v2 ;
(L2 ) (k1 + k2 ) · v = k1 · v + k2 · v;
(L3 ) (k1 · k2 ) · v = k1 · (k2 · v);
(L4 ) 1 · v = v,
for every k, k1 , k2 ∈ K and every v, v1 , v2 ∈ V .
In this context, the elements of K are called scalars and the elements of V are called vectors.
Sometimes a vector space is also called a linear space.

We denote the fact that V is a vector space over K either by K V or by (V, K, +, ·), since the addition
in the abelian group and the external multiplication are the significant operations.

Remark 2.1.2 (1) Notice that in the definition of a vector space there are present four operations, two
denoted by the same symbol “ + ” and two denoted by the same symbol “ · ”. Of course, they are not
the same, but as we have already done it several times before, we use the convention to denote them
identically for the sake of simplicity of writing. There are 3 operations in the classical sense, namely the
addition and the multiplication in the field K and the addition in the group V and, on the other hand,
there is also an external operation of multiplication.
(2) The axioms (L1 ) and (L2 ) look like some distributive laws and the axiom (L3 ) looks like an
associative law, but they are not, since the elements intervening are not taken from the same set.
(3) The definition we have just given is that of a left vector space. It is also possible to give the
definition of a right vector space by considering an external operation

·:V ×K →V , (v, k) 7→ v · k,

satisfying some similar axioms, but on the right hand side.

Since one can show that there is a bijection between the left and the right vector spaces of the field
K, we are going to study only the left vector spaces and omit the adjective “left”.

Let us now see several important examples of vector spaces.

Page 1
ALGEBRA - First Year - Computer Science Prof. dr. Septimiu Crivei

Example 2.1.3 (a) Let V2 be the set of all vectors (in the classical sense) in the plane with a fixed origin
O. Then V2 is a vector space over R (or a real vector space), where the addition is the usual addition
of two vectors by the parallelogram rule and the external operation is the usual scalar multiplication of
vectors by real scalars.
If we consider two coordinate axes Ox and Oy in the plane, each vector in V2 is perfectly determined
by the coordinates of its ending point. Therefore, the addition of vectors and the scalar multiplication of
vectors by real numbers become:
(x, y) + (x0 , y 0 ) = (x + x0 , y + y 0 ) ,
k · (x, y) = (k · x, k · y) ,
0 0
∀k ∈ R and ∀(x, y), (x , y ) ∈ R × R. Thus, (R2 , R, +, ·) is a vector space.
Similarly, one can consider the real vector space V3 of all vectors in the space with a fixed origin.
Moreover, a further generalization is possible, as we may see in the following example.
(b) Let n ∈ N∗ . Define
(x1 , . . . , xn ) + (y1 , . . . , yn ) = (x1 + y1 , . . . , xn + yn ) ,
k · (x1 , . . . , xn ) = (kx1 , . . . , kxn ) ,
∀(x1 , . . . , xn ), (y1 , . . . , yn ) ∈ K and ∀k ∈ K. Then (K n , K, +, ·) is a vector space, called the canonical
n

vector space.
Let us discuss some particular cases. For n = 1, we get that K K is a vector space. Hence, as far as
the classical numerical fields are concerned, Q Q, R R and C C are vector spaces.
(c) If V = {e} is a single element set, then we know that there is a unique structure of an abelian
group for V , namely that one defined by e + e = e. Then we can define a unique scalar multiplication,
namely k · e = e, ∀k ∈ K. Thus, V is a vector space, called the zero (null) vector space and denoted by
{0}.
(d) Let A be a subfield of the field K. Then K is a vector space over A, where the addition and the
scalar multiplication are just the addition and the multiplication of elements in the field K.
In particular, Q R, Q C and R C are vector spaces.
(e) (K[X], K, +, ·) is a vector space, where the addition is the usual addition of polynomials and the
scalar multiplication is defined as follows: ∀f = a0 + a1 X + · · · + an X n ∈ K[X], ∀k ∈ K,
kf = (ka0 ) + (ka1 )X + · · · + (kan )X n .

(f ) Let m, n ∈ N, m, n ≥ 2. Then (Mmn (K), K, +, ·) is a vector space, where the operations are the
usual addition and scalar multiplication of matrices.
(g) Let A be a non-empty set. Denote
K A = {f | f : A → K} .
Then (K A , K, +, ·) is a vector space, where the addition and the scalar multiplication are defined as
follows: ∀f, g ∈ K A , ∀k ∈ K, we have f + g ∈ K A , kf ∈ K A , where
(f + g)(x) = f (x) + g(x) ,
(kf )(x) = kf (x) , ∀x ∈ A .
As a particular case, we obtain the vector space (RR , R, +, ·) of real functions of a real variable.
(h) Let V and V 0 be K-vector spaces. Then the cartesian product V × V 0 is a K-vector space, where
the addition and the scalar multiplication are defined as follows:
(v1 , v10 ) + (v2 , v20 ) = (v1 + v10 , v2 + v20 ),
k(v1 , v10 ) = (kv1 , kv10 )
∀(v1 , v10 ), (v2 , v20 ) ∈ V × V 0 and ∀k ∈ K. This vector space is called the direct product of V and V 0 .

Page 2
ALGEBRA - First Year - Computer Science Prof. dr. Septimiu Crivei

Let us now state some computation rules in a vector space. Notice that we denote by 0 both the zero
scalar and the zero vector.

Theorem 2.1.4 Let V be a vector space over K. Then ∀k, k 0 ∈ K and ∀v, v 0 ∈ V we have:

(i) k · 0 = 0 · v = 0;

(ii) k(−v) = (−k)v = −kv;

(iii) k(v − v 0 ) = kv − kv 0 ;

(iv) (k − k 0 )v = kv − k 0 v.

Proof. (i) We have:

k · 0 + k · v = k(0 + v) = kv =⇒ k · 0 = 0,
0 · v + k · v = (0 + k)v = kv =⇒ 0 · v = 0.

(ii) We have:
kv + k(−v) = k(v − v) = k · 0 = 0 =⇒ k(−v) = −kv,
kv + (−k)v = (k − k)v = 0 · v = 0 =⇒ (−k)v = −kv.

(iii) We have:

k(v − v 0 ) + kv 0 = k(v − v 0 + v 0 ) = kv =⇒ k(v − v 0 ) = kv − kv 0 .

(iv) We have:

(k − k 0 )v + k 0 v = (k − k 0 + k 0 )v = kv =⇒ (k − k 0 )v = kv − k 0 v.

Theorem 2.1.5 Let V be a vector space over K and let k ∈ K and v ∈ V . Then

kv = 0 ⇐⇒ k = 0 or v = 0 .

Proof. =⇒. Assume kv = 0. Suppose that k 6= 0. Then k is invertible in the field K and we have

kv = 0 =⇒ k · v = k · 0 =⇒ v = 0 .

⇐=. This is Theorem 2.1.4 (i). 

Remark 2.1.6 Formally, the property given in Theorem 2.1.5 looks like a non-zero divisor property, but
it is not, since the elements are not in the same set.

2.2 Subspaces
Let us now discuss some special subsets of vector spaces, namely subspaces. We are going to define a
subspace in the same general way as we did for subgroups or subrings.

Definition 2.2.1 Let V be a vector space over K and let S ⊆ V . Then S is a subspace of V if:

(i) S 6= ∅;

(ii) ∀x, y ∈ S, x + y ∈ S;

(iii) ∀k ∈ K, ∀x ∈ S, kx ∈ S.

We usually denote by S ≤K V , or simply by S ≤ V , the fact that S is a subspace of the vector space
V over K.

Page 3
ALGEBRA - First Year - Computer Science Prof. dr. Septimiu Crivei

Remark 2.2.2 Notice that every subspace S of a vector space V over K is a subgroup of the additive
group (V, +), hence S must contain 0.
We have the following characterization theorem for subspaces.
Theorem 2.2.3 Let V be a vector space over K and let S ⊆ V . Then
(
S 6= ∅ (0 ∈ S)
S ≤ V ⇐⇒
∀k1 , k2 ∈ K , ∀x, y ∈ S , k1 x + k2 y ∈ S .
Proof. =⇒. Let k1 , k2 ∈ K and x, y ∈ S. By Definition 2.2.1 we have k1 x, k2 y ∈ S and then
k1 x + k2 y ∈ S.
⇐=. Choose k1 = k2 = 1 and then k2 = 0 and use Definition 2.2.1.
Example 2.2.4 (a) Every non-zero vector space V over K has two subspaces, namely {0} and V . They
are called the trivial subspaces.
(b) Let
S = {(x, y, z) ∈ R3 | x + y + z = 0} ,
T = {(x, y, z) ∈ R3 | x = y = z} .
We have S 6= ∅, because (0, 0, 0) ∈ S. Now let k1 , k2 ∈ R and v1 , v2 ∈ S. Then v1 = (x1 , y1 , z1 ) and
v2 = (x2 , y2 , z2 ) for some x1 , y1 , z1 , x2 , y2 , z2 ∈ R such that x1 + y1 + z1 = 0 and x2 + y2 + z2 = 0. It
follows that:
k1 v1 + k2 v2 = (k1 x1 + k2 x2 , k1 y1 + k2 y2 , k1 z1 + k2 z2 )
and we have
(k1 x1 + k2 x2 ) + (k1 y1 + k2 y2 ) + (k1 z1 + k2 z2 ) = k1 (x1 + y1 + z1 ) + k2 (x2 + y2 + z2 ) = 0.
Hence k1 v1 + k2 v2 ∈ S, and so S is a subspace of the real vector space R3 . Note that S is a plane passing
through the origin.
We have T 6= ∅, because (0, 0, 0) ∈ T . Now let k1 , k2 ∈ R and v1 , v2 ∈ T . Then v1 = (x1 , x1 , x1 ) and
v2 = (x2 , x2 , x2 ) for some x1 , x2 ∈ R. It follows that:
k1 v1 + k2 v2 = (k1 x1 + k2 x2 , k1 x1 + k2 x2 , k1 x1 + k2 x2 ).
Hence k1 v1 + k2 v2 ∈ T , and so T is a subspace of the real vector space R3 . Note that T is a line passing
through the origin.
(c) More generally, the only subspaces of R3 are {(0, 0, 0)}, any line containing the origin, any plane
containing the origin and R3 .
(d) Let n ∈ N and let
Kn [X] = {f ∈ K[X] | degree (f ) ≤ n} .
Then Kn [X] is a subspace of the polynomial vector space K[X] over K.
(e) Let I ⊆ R be an interval. By Example 2.1.3,
RI = {f | f : I → R}
is a real vector space, where the addition and the scalar multiplication are defined as follows: ∀f, g : I → R,
∀k ∈ K, we have f + g : I → R, kf : I → R, where
(f + g)(x) = f (x) + g(x) ,
(kf )(x) = kf (x) , ∀x ∈ I .
The subsets
C(I, R) = {f ∈ RI | f continuous on I},
D(I, R) = {f ∈ RI | f derivable on I}
are subspaces of RI , because they are nonempty and we have:
∀k1 , k2 ∈ R, ∀f, g ∈ C(I, R), k1 f + k2 g ∈ C(I, R),
∀k1 , k2 ∈ R, ∀f, g ∈ D(I, R), k1 f + k2 g ∈ D(I, R).

Page 4
ALGEBRA - First Year - Computer Science Prof. dr. Septimiu Crivei

Extra: Vernam’s cipher

Let n ∈ N∗ and consider the canonical vector space V = Zn 2 over Z2 . The vectors of V may be identified with
n-bit binary strings. Suppose that Alice needs to send an n-bit plaintext p ∈ Zn
2 to Bob.

Vernam’s cipher:
1. (Key establishment) Alice and Bob randomly choose a vector k ∈ Zn
2 as a key.

2. (Encryption) Alice computes the cyphertext c according to the formula

c = p + k,

where the sum is a vector in Zn

2.

3. (Decryption) Bob computes the plaintext p according to the formula

p = c − k = c + k,

where the sum is a vector in Zn

2.

Remark 2.2.5 The system satisfies perfect secrecy, but the key k has to be distributed in advance.

Example 2.2.6 Alice and Bob agree on the vector k = (0, 1, 1, 0, 1, 0, 0, 0, 0, 1) ∈ Z10
2 as a key.
Alice wants to send the message p = (0, 0, 0, 1, 1, 1, 0, 1, 0, 1) ∈ Z10
2 to Bob.
Alice encrypts the message by computing the ciphertext c as:

c = p + k = (0, 0, 0, 1, 1, 1, 0, 1, 0, 1) + (0, 1, 1, 0, 1, 0, 0, 0, 0, 1) = (0, 1, 1, 1, 0, 1, 0, 1, 0, 0) ∈ Z10

2 .

Bob decrypts the message by computing the plaintext p as:

p = c + k = (0, 1, 1, 1, 0, 1, 0, 1, 0, 0) + (0, 1, 1, 0, 1, 0, 0, 0, 0, 1) = (0, 0, 0, 1, 1, 1, 0, 1, 0, 1) ∈ Z10

2 .

Page 5