Linear Algebra

L. Benferhat & A. Cherchem

ENSIA 2024
 Introduction
• Linear Algebra can be viewed as the study of vector spaces and linear
maps between them.
• Linear Algebra has applications in almost all of the scientific fields.
• In particular, Linear Algebra plays a fundamental role in many areas of
computer science, including artificial intelligence and machine
learning.
• This course is devoted to the study of four principal objects : Vector
spaces, Linear maps, Matrices and Systems of linear equations. We
will see how these objects are all related to each other.
 Bibliography
• Benali Benzaghou, Introduction à l’Algèbre Linéaire, OPU, 2005
• Mark Peter Deisenroth, A. Aldo Faisal, Cheng Soon Ong, Mathematics
for Machine Learning, Cambridge University Press, 2020
• David C. Lay, Steven R. Lay, Judi J. McDonald, Linear Algebra and its
Applications, Pearson, 2015
• Gilbert Strang, Introduction to Linear Algebra, Wellesley-Cambridge
Press, 2016
Lecture 1 : Vector Spaces
ENSIA, April 2024
 Contents
• Introduction
• Vector space structure
• Subspaces
• Linear combination and Spanning set
• Linear independence
• Bases and Dimension
• Rank of a family of vectors
• Row Echelon Form
• Direct sum
 Introduction

The beauty and power of Linear Algebra will be seen more clearly
when you view ℝ𝑛 as only one of a variety of vector spaces that arise
naturally in applied problems. Actually, a study of vector spaces is not
much different from a study of ℝ𝑛 itself, because you can use your
geometric experience with ℝ2 and ℝ3 to visualize many general
concepts. [LLM]
 Vector Space Structure
Throughout this lecture, 𝐾 will denote an arbitrary field.
Definition 1
A vector space over 𝑲, or a 𝑲- vector space, is a commutative group (𝑉,⊕)
equipped with a function 𝐾 × 𝑉 → 𝑉; 𝛼, 𝑥 ↦ 𝛼 ⊗ 𝑥 such that
(𝑖) ∀ 𝛼, 𝛽 ∈ 𝐾, ∀𝑥 ∈ 𝑉, (𝛼 + 𝛽) ⊗ 𝑥 = (𝛼 ⊗ 𝑥) ⊕ (𝛽 ⊗ 𝑥);
𝑖𝑖 ∀ 𝛼, 𝛽 ∈ 𝐾, ∀𝑥 ∈ 𝑉, (𝛼 ∙𝛽) ⊗ 𝑥 = 𝛼 ⊗ 𝛽 ⊗ 𝑥 ;
(𝑖𝑖𝑖) ∀𝑥, 𝑦 ∈ 𝑉, ∀𝛼 ∈ 𝐾, 𝛼 ⊗ 𝑥 ⊕ 𝑦 = (𝛼 ⊗ 𝑥) ⊕ (𝛼 ⊗ y);
𝑖𝑣 1 ⊗ 𝑥 = 𝑥, ∀𝑥 ∈ 𝑉.
 Vector Space Structure
Remark 1 : The elements of 𝑉 are called vectors and the elements of 𝐾
are called scalars. The multiplication ⊗ is called multiplication by a
scalar.

Remark 2 : In the definition of a 𝐾- vector space 𝑉, if we replace the

field 𝐾 by a commutative ring 𝐴, we obtain the definition of a module
over the ring 𝐴.

From now on, we will denote the operation ⊕ by + and the operation
⊗ by ∙ .
 Examples
Example 1.
Take 𝐾 = ℝ and V = ℝ2 , and set
∀ 𝑥1 , 𝑥2 , 𝑦1 , 𝑦2 ∈ ℝ2 , 𝑥1 , 𝑥2 + 𝑦1 , 𝑦2 = 𝑥1 + 𝑦1 , 𝑥2 + 𝑦2
and
∀ 𝛼 ∈ ℝ, ∀ 𝑥1 , 𝑥2 ∈ ℝ2 , 𝛼 𝑥1 , 𝑥2 = 𝛼𝑥1 , 𝛼𝑥2 .
Then ℝ2 is an ℝ-vector space.
 Examples
Example 2.
More generally, let 𝐾 be any field and let 𝑛 be a positive integer.
The set 𝐾 𝑛 of 𝑛-tuples of elements of 𝐾 is a vector space over 𝐾 with
(𝑥1 , 𝑥2 , . . . , 𝑥𝑛 ) + (𝑦1 , 𝑦2 , . . . , 𝑦𝑛 ) = (𝑥1 + 𝑦1 , 𝑥2 + 𝑦2 , . . . , 𝑥𝑛 + 𝑦𝑛 ),
and
𝛼(𝑥1 , 𝑥2 , . . . , 𝑥𝑛 ) = (𝛼𝑥1 , 𝛼𝑥2 , . . . , 𝛼𝑥𝑛 ),
for all elements (𝑥1 , 𝑥2 , . . . , 𝑥𝑛 ) and (𝑦1 , 𝑦2 , . . . , 𝑦𝑛 ) of 𝐾 𝑛 and for all
elements 𝛼 of 𝐾.
In particular, 𝐾 is a vector space over itself.
 Examples
Example 3.
Let 𝐾 𝑋 be the set of polynomials over 𝐾 with an indeterminate 𝑋.
Let 𝑃 𝑋 = 𝑎0 + 𝑎1 𝑋 + ⋯ +𝑎𝑛 𝑋 𝑛 , 𝑄 𝑋 = 𝑏0 + 𝑏1 𝑋 + ⋯ +𝑏𝑚 𝑋 𝑚 ∈ 𝐾 𝑋 and
𝛼 ∈ 𝐾.
We define on 𝐾 𝑋 the following operations :
𝑃(𝑋) + 𝑄(𝑋) = 𝑟𝑖=0 𝑎𝑖 + 𝑏𝑖 𝑋 𝑖 , where 𝑟 = max 𝑚, 𝑛 .

𝛼𝑃 𝑋 = 𝑎0 + (𝛼𝑎1 )𝑋 + ⋯ +(𝑎𝑛 )𝑋 𝑛 .

Endowed with these two operations, 𝐾 𝑋 is a 𝐾- vector space.

 Examples
Example 4.
Let 𝑋 be a set and 𝑉 be a 𝐾- vector space. Denote by ℱ(𝑋, 𝑉) the set of
functions from 𝑋 to 𝑉. Let 𝑓, 𝑔 ∈ ℱ(𝑋, 𝑉) and α ∈ 𝐾. We define the
sum of 𝑓 and 𝑔 by
𝑓 + 𝑔 𝑥 = 𝑓 𝑥 + 𝑔 𝑥 , ∀𝑥 ∈ 𝑋,
and the multiplication by a scalar by
𝑓 𝑥 = 𝑓 𝑥 , ∀𝑥 ∈ 𝑋.
Endowed with these two operations, ℱ(𝑋, 𝑉) is a 𝐾- vector space.
In particular, with these operations, the set ℱ ℕ, ℝ of real sequences
has an ℝ-vector space structure.
 Properties

Proposition 1
Let 𝑉 be a 𝐾-vector space. Then we have : ∀𝑥 ∈ 𝑉, ∀𝛼 ∈ 𝐾,
1- 0𝐾 ∙ 𝑥 = 0𝑉 ;
2- 𝛼 ∙ 0𝑉 = 0𝑉 ;
3- 𝛼 ∙ 𝑥 = 0𝑉 ⇔ 𝛼 = 0𝐾 or 𝑥 = 0𝑉 ;
4- (−𝛼) ∙ 𝑥 = 𝛼 ∙ (−𝑥) = − 𝛼 ∙ 𝑥 .
 Subspaces
Definition 2
Let 𝑉 be a 𝐾- vector space and 𝑈 be a subset of 𝑉. We say that 𝑈 is a
subspace of 𝑉 if
1- ∀𝑢, 𝑣 ∈ 𝑈, 𝑢 + 𝑣 ∈ 𝑈;
2- ∀𝑢 ∈ 𝑈, ∀𝛼 ∈ 𝐾, 𝛼 ∙ 𝑢 ∈ 𝑈;
3- 0𝑉 ∈ 𝑈.
Remark 3
This definition means that 𝑈 is a 𝐾-vector space under the inherited
operations. Check it !
 Examples

Example 5
𝑉 and 0 are subspaces of the vector space 𝑉.

Example 6
The set
𝑈= 𝑥, 𝑦 ∈ ℝ2 , 2𝑥 + 𝑦 = 0
is a subspace of ℝ2 .
 Examples

Example 7
Let 𝑛 be a positive integer. Then the set 𝐾𝑛 𝑋 of polynomials over 𝐾 of
degree ≤ 𝑛 is a subspace of 𝐾 𝑋 .

Example 8
The set
𝑈= 𝑥, 𝑦 ∈ ℝ2 , 𝑥 + 𝑦 = 1
is not a subspace of ℝ2 .
 Intersection of subspaces

Proposition 2
The intersection of two subspaces of a 𝐾- vector space 𝑉 is a subspace
of 𝑉.

Remark 4 . The union of two subspaces of a 𝐾- vector space 𝑉 is not in

general a subspace of 𝑉. (See exercise 5, Worksheet 1)
 Sum of subspaces
Definition 3
Let 𝑈 and 𝑊 be two subspaces of a 𝐾-vector space 𝑉. The set
𝑈 + 𝑊 = 𝑢 + 𝑤: 𝑢 ∈ 𝑈, 𝑤 ∈ 𝑊
is called sum of 𝑈 and 𝑊.

Proposition 3
If 𝑈 and 𝑊 are two subspaces of a 𝐾-vector space 𝑉, then 𝑈 + 𝑊 is a
subspace of 𝑉.
 Linear combination of vectors

Definition 4.
Let 𝑉 be a 𝐾- vector space and 𝑣 ∈ 𝑉. We say that 𝑣 is a linear
combination of 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 ∈ 𝑉 if there exist 𝛼1 , 𝛼2 , ⋯ , 𝛼𝑛 ∈ 𝐾
such that
𝑣 = 𝑛𝑖=1 𝛼𝑖 𝑣𝑖 .
Example 9
Take 𝑉 = ℝ2 , 𝑣 = 2,3 , 𝑒1 = 1,0 and 𝑒2 = 0,1 . We have
𝑣 = 2 1,0 + 3 0,1 .
Then 𝑣 is a linear combination of 𝑒1 and 𝑒2 .
 Spanning set
Definition 5
Let 𝑉 be a 𝐾-vector space and 𝐴 ⊆ 𝑉 a subset of 𝑉 . Then the span of 𝑨
in 𝑽 is the set of all finite linear combinations of elements of 𝐴.
Notation :
The span of 𝐴 in 𝑉 is denoted by 𝐴 , or Span (𝐴).
When 𝐴 is a finite set of vectors, that is 𝐴 = 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 , we write
simply 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 , or Span 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 .
Convention :
By convention, we set ∅ = 0 .
 Spanning set
From the previous definition, we have

𝑛
𝐴 = 𝛼𝑖 𝑢𝑖 : 𝛼𝑖 ∈ 𝐾, 𝑢𝑖 ∈ 𝐴, 𝑛 ∈ ℕ∗ .
𝑖=1

Proposition 4
Let 𝑉 be a 𝐾-vector space and 𝐴 ⊆ 𝑉 a subset of 𝑉 . Then the span of 𝐴
in 𝑉 is the smallest subspace of 𝑉 containing 𝐴.
 Spanning set

Example 10
ℝ𝑛 is spanned by the vectors
𝑒1 = 1,0, ⋯ , 0 , 𝑒2 = 0,1, ⋯ , 0 , ⋯ , 𝑒𝑛 = (0,0, ⋯ , 1) ∈ ℝ𝑛 .

Example 11
𝐾𝑛 𝑋 is spanned by the vectors
1, 𝑋, ⋯ , 𝑋 𝑛 .
 Linear independence

Definition 6.
Let 𝑉 be a 𝐾- vector space and 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 ∈ 𝑉. We say that the
vectors 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 are linearly independent (𝐿. 𝐼. ) if we have : for
all 𝛼1 , 𝛼2 , ⋯ , 𝛼𝑛 ∈ 𝐾,
𝑛
𝑖=1 𝛼𝑖 𝑣𝑖 =0𝑉 ⟹ 𝛼1 =𝛼2 =⋯ = 𝛼𝑛 =0𝐾 .
Otherwise, we say that 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 are linearly dependent 𝐿. 𝐷. .
 Linear dependence

From the previous definition, by taking the negation, we can see that
the vectors 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 are linearly dependent if there exist 𝑛 scalars
𝛼1 , 𝛼2 , ⋯ , 𝛼𝑛 ∈ 𝐾, not all zero, for which

𝑛
𝛼𝑖 𝑣𝑖 = 0𝑉 .
𝑖=1
 Examples

Example 12
Set 𝑉 = ℝ2, 𝑣1 = 1,2 , 𝑣2 = −1,1 .
The vectors 𝑣1 and 𝑣2 are linearly independent.

Example 13
Set 𝑉 = 𝐾 𝑋 , 𝑃1 = 𝑋, 𝑃2 = 𝑋 + 1.
The polynomyals 𝑃1 and 𝑃2 are linearly independent.
 Examples

Example 14
Let ℱ(ℝ, ℝ) be the set of functions from ℝ to ℝ, and consider the two functions
defined by
𝑓 𝑥 = 𝑒 𝑥 , 𝑔 𝑥 = 𝑒 2𝑥 for all 𝑥 ∈ ℝ.
Then 𝑓 and 𝑔 are linearly independent.

Example 15
The 𝑛 vectors
𝑒1 = 1,0, ⋯ , 0 , 𝑒2 = 0,1, ⋯ , 0 , ⋯ , 𝑒𝑛 = (0,0, ⋯ , 1) ∈ ℝ𝑛
are linearly independent.
 Examples

Example 16
Set 𝑉 = ℝ2 , 𝑣1 = 1,2 , 𝑣2 = −1,1 and 𝑣3 = 3,3 .
The vectors 𝑣1 , 𝑣2 , 𝑣3 are linearly dependent. Indeed,
2𝑣1 − 𝑣2 − 𝑣3 = 0,0 .
 Finite-dimensional vector spaces
Definition 7
A vector space 𝑉 is said to be finite-dimensional if it is spanned by a finite
number of vectors.

Example 17
ℝ𝑛 is spanned by the 𝑛 vectors
𝑒1 = 1,0, ⋯ , 0 , 𝑒2 = 0,1, ⋯ , 0 , ⋯ , 𝑒𝑛 = (0,0, ⋯ , 1),
so it is a finite-dimensional vector space.
Example 18
𝐾 𝑋 is not a finite-dimensional vector space.
 Basis of a vector space

Definition 8
Let 𝑉 be a 𝐾- vector space and 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 ∈ 𝑉. We say that the set
*𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 + is a basis of 𝑉 if we have
1-The vectors 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 are linearly independent ;
2-The vectors 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 span 𝑉.
 Examples
Example 19
The vectors
𝑒1 = 1,0, ⋯ , 0 , 𝑒2 = 0,1, ⋯ , 0 , ⋯ , 𝑒𝑛 = (0,0, ⋯ , 1) ∈ ℝ𝑛
form a basis of ℝ𝑛 , called standard basis.

Example 20
The set 1, 𝑋, ⋯ , 𝑋 𝑛 is a basis of 𝐾𝑛 ,𝑋-.
 Basis of a vector space
Theorem 1
Let 𝑉 be a 𝐾- vector space. Then, the vectors 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 of 𝑉
constitute a basis of 𝑉 if, and only if , given any vector 𝑢 of 𝑉, there
exist uniquely determined scalars 𝛼1 , 𝛼2 , ⋯ , 𝛼𝑛 ∈ 𝐾 such that
𝑢 = 𝑛𝑖=1 𝛼𝑖 𝑣𝑖 .

Remark 4
If *𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 + is a basis of 𝑉 and 𝑢 = 𝑛𝑖=1 𝛼𝑖 𝑣𝑖 , then the scalars
𝛼1 , 𝛼2 , ⋯ , 𝛼𝑛 are called coordinates or components of 𝑢 in the
basis *𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 +.
 Existence of basis

Theorem 2
Let 𝑉 be a finite-dimensional 𝐾-vector space, and 𝐴 be a finite subset
of 𝑉 which spans 𝑉. Let 𝐿 be a subset of linearly independent vectors
of 𝐴. Then, there exists a basis 𝐵 of 𝑉 so that
𝐿 ⊆ 𝐵 ⊆ 𝐴.
Theorem 3
Any nonzero finite-dimensional 𝐾-vector space has a basis.
 Incomplete basis Theorem

Theorem 4
Let 𝑉 be a finite-dimensional 𝐾- vector space, and 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑝 ∈ 𝑉
linearly independent vectors. Then, one can find vectors of 𝑉, 𝑣𝑝+1 ,
𝑣𝑝+2 , ⋯ , 𝑣𝑛 , such that *𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛 + is a basis of 𝑉.
 Dimension
Theorem 5
Let 𝑉 be a finite-dimensional 𝐾-vector space. Then All basis of 𝑉 have
the same number of vectors.
Definition 9
The dimension of a finite-dimensional vector space is defined to be
number of elements in any basis of that vector space. The dimension is
defined to be zero in the case where the vector space consists of just
the zero element.
Notation
The dimension of 𝑉 is denoted by dim 𝑉.
 Examples

Example 21.
ℝ𝑛 is of dimension 𝑛.
Example 22.
𝐾𝑛 ,𝑋- is of dimension 𝑛 + 1.
Example 23.
The subspace 𝑈 = 𝑥, 𝑦 ∈ ℝ2 : 𝑥 + 𝑦 = 0 of ℝ2 is of dimension 1.
 Basis with known dimension

Theorem 6
Let 𝑉 be a 𝐾-vector space of dimension 𝑛 ≥ 1, then we have
1- Any 𝑛 linearly independent vectors of 𝑉 form a basis of 𝑉.
2- Any 𝑛 vectors of 𝑉 which span 𝑉 form a basis of 𝑉.
 Minimal spanning set and
Maximal linearly independent set
Definition 10
A spanning set of a vector space 𝑉 is said to be minimal if it has no proper
subset which forms a spanning set of 𝑉.
A set *𝑣1, 𝑣2, ⋯ , 𝑣𝑛 + of linearly independent vectors of 𝑉 is said to be
maximal if for all 𝑣 ∈ 𝑉, the vectors 𝑣, 𝑣1, 𝑣2, ⋯ , 𝑣𝑛 are linearly dependent.

Theorem 7
Let 𝑉 be a finite-dimensional vector space. Then we have :
1- Any minimal spanning set of 𝑉 forms a basis of 𝑉.
2- Any maximal set of linearly independent vectors of 𝑉 forms a basis of 𝑉.
 Dimension and subspaces

Theorem 8
Let 𝑉 be a finite-dimensional 𝐾-vector space and let 𝑈 be a subspace of
𝑉. Then we have
1- dim 𝑈 ≤ dim 𝑉 ;
2- dim 𝑈 = dim 𝑉 ⟹ 𝑈 = 𝑉
 Rank of a family of vectors

Definition 11
Let 𝑉 be a finite dimensional 𝐾-vector space and 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑚 ∈ 𝑉.
The rank of 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑚 is the dimension of the subspace of 𝑉
spanned by 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑚 .
 Computing the rank of a family of vectors
To compute the rank of 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑚 , we will use Gaussian
Elimination to obtain Row Echelon Form.
Set
𝑣1 = 𝑎11 𝑎12 ⋯ 𝑎1𝑗 ⋯ 𝑎1𝑛
𝑣2 = 𝑎21 𝑎22 ⋯ 𝑎2𝑗 ⋯ 𝑎2𝑛
⋮
𝑣𝑖 = 𝑎𝑖1 𝑎𝑖2 ⋯ 𝑎𝑖𝑗 ⋯ 𝑎𝑖𝑛
⋮
𝑣𝑚 = 𝑎𝑚1 𝑎𝑚2 ⋯ 𝑎𝑚𝑗 ⋯ 𝑎𝑚𝑛 ,
where the 𝑎𝑖𝑗 , 1 ≤ 𝑗 ≤ 𝑛, are the components of 𝑣𝑖 in a fixed basis.
 Row Echelon Form
Definition 12
A family of vectors is said to be in the row-echelon form (REF) if the
two following conditions are satisfied :
1- If a row is zero, then all the following rows are zeros.
2- If the row 𝑖 has its first nonzero coefficient in column 𝑗, then the first
nonzero coefficient of row 𝑖 + 1 is in column 𝑘 > 𝑗.
Definition 13
When the vectors are in the REF, the first nonzero element of each row
is called a pivot.
 Elementary row operations

To obtain the REF, we use a sequence of elementary row operations.

Definition 14
There are three types of elementary row operations :
1- Exchanging two rows ;
2- Multiplying a row by a nonzero scalar ;
3- Adding a multiple of one row to another row.
 Computing the rank of a family of vectors

Proposition 5
The rank of 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑚 is the number of nonzero rows of the Row
Echelon Form.
 Gaussian Elimination Example
Example 24
Determine the rank of the vectors :
𝑣1 = 1, 2, 1, 2, 1 , 𝑣2 = 2, 4, 2, 1, 5 , 𝑣3 = 1, 0, 1, 1, 0 , 𝑣4 = 0, 1, 0, 0, 1 .

𝑣′1 = 𝑣1, 𝑣′2 = 𝑣2 − 2𝑣1, 𝑣 ′3 = 𝑣3 − 𝑣1, 𝑣′4 = 𝑣4

1 2 1 2 1 1 2 1 2 1
2 4 2 1 5 0 0 0 −3 3
⟶
1 0 1 1 0 0 −2 0 −1 −1
0 1 0 0 1 0 1 0 0 1
 Gaussian Elimination Example

𝑣′′1 = 𝑣′1 , 𝑣′′2 = 𝑣′4 , 𝑣′′3 = 𝑣′3 , 𝑣′′4 = 𝑣′2

1 2 1 2 1 1 2 1 2 1
0 0 0 −3 3 0 1 0 0 1
⟶
0 −2 0 −1 −1 0 −2 0 −1 −1
0 1 0 0 1 0 0 0 −3 3
 Gaussian Elimination Example

𝑣′′′1 = 𝑣′′1 , 𝑣′′′2 = 𝑣′′2 , 𝑣′′′3 = 𝑣′′3 + 2𝑣′′2 , 𝑣′′′4 = 𝑣′′4

1 2 1 2 1 1 2 1 2 1
0 1 0 0 1 0 1 0 0 1
⟶ .
0 −2 0 −1 −1 0 0 0 −1 1
0 0 0 −3 3 0 0 0 −3 3
 Gaussian Elimination Example

𝑣′′′′1 = 𝑣′′′1 , 𝑣′′′′2 = 𝑣′′′2 , 𝑣′′′′3 = 𝑣′′′3 , 𝑣′′′′4 = 𝑣′′′4 − 3𝑣′′′3

1 2 1 2 1 1 2 1 2 1
0 1 0 0 1 0 1 0 0 1
⟶ .
0 0 0 −1 1 0 0 0 −1 1
0 0 0 −3 3 0 0 0 0 0

Then, since we have 3 nonzero rows, rank 𝑣1 , 𝑣2 , 𝑣3 , 𝑣4 = 3.

 Direct sum of subspaces
Definition 15
Let 𝑊1 , 𝑊2 be two subspaces of a 𝐾-vector space 𝑉. The sum 𝑊1 +𝑊2 is
called direct if 𝑊1 ∩ 𝑊2 = 0𝑉 .
In particular, a vector space 𝑉 is said to be the direct sum of two subspaces
𝑊1 and 𝑊2 if 𝑉 = 𝑊1 + 𝑊2 and 𝑊1 ∩ 𝑊2 = 0𝑉 .
When 𝑉 is a direct sum of 𝑊1 and 𝑊2, we write 𝑉 = 𝑊1 ⊕ 𝑊2 .
Theorem 9
Suppose 𝑊1 and 𝑊2 are subspaces of a vector space 𝑉. Then 𝑉 = 𝑊1 ⊕ 𝑊2
if, and only if, every vector in 𝑉 can be written in a unique way as 𝑤1 + 𝑤2,
where 𝑤1 ∈ 𝑊1 and 𝑤2 ∈ 𝑊2 .
 Dimension of sum of subspaces

Theorem 10
If 𝑊1 , 𝑊2 are subspaces of a finite-dimensional vector space 𝑉, then
we have
dim 𝑊1 + 𝑊2 = dim 𝑊1 + dim 𝑊2 − dim(𝑊1 ∩ 𝑊2 ).