Mathematical Methods
in Economics I
Vector Space
Larson 4.1-4.3
Graphical Representation of a Vector
𝑢𝑢1
𝑢𝑢 = 𝑢𝑢 𝑢𝑢2
2
e.g.
2 3
𝑢𝑢 =
3
𝑢𝑢1
2
Length of a Vector
Length of a vector is called norm
𝑢𝑢2
and is denoted by ⋅ . For real
vectors, it is defined as:
3
𝑢𝑢⃑ = 𝑢𝑢12 + ⋯+ 𝑢𝑢𝑛𝑛2
2 𝑢𝑢1
𝑢𝑢 = 2
3
𝑢𝑢⃑ = 22 + 32 = 13
Graphical Representation of
Vector Addition
𝑢𝑢2
2
𝑢𝑢 =
3 4
𝑣𝑣
3
𝑣𝑣⃑ = 𝑢𝑢
1 𝑢𝑢 𝑢𝑢 + 𝑣𝑣
5
𝑢𝑢 + 𝑣𝑣⃑ = 𝑣𝑣
4 5
𝑢𝑢1
Graphical Representation of
Vector Addition
2 𝑢𝑢2
𝑢𝑢 =
3
𝑢𝑢
−2
−𝑢𝑢 =
−3 𝑢𝑢1
0 0
𝑢𝑢 + −𝑢𝑢 = −𝑢𝑢 0
0
Graphical Representation of
Scalar Multiplication
𝑢𝑢2
3
𝑢𝑢 = , 𝑐𝑐 = 2
1
6
𝑐𝑐𝑢𝑢 =
2 2
𝑐𝑐𝑐𝑐
𝑢𝑢
𝑢𝑢1
6
Dot Product of Perpendicular Vectors
An interesting fact: if two vectors
are perpendicular to each other, 𝑢𝑢2
their dot product is zero:
2 0
𝑢𝑢 = , 𝑣𝑣⃑ =
0 3
𝑢𝑢 ⋅ 𝑣𝑣⃑ = 2 × 0 + 0 × 3 = 0 𝑦𝑦 𝑣𝑣 𝑥𝑥
1 −1
𝑥𝑥⃑ = , 𝑦𝑦⃑ = 𝑢𝑢1
1 1 𝑢𝑢
𝑥𝑥⃑ ⋅ 𝑦𝑦⃑ = 1 × −1 + 1 × 1 = 0
We say the vectors are
orthogonal to each other
Properties of Vector Addition
2
and Scalar Multiplication in ℝ
• 𝑢𝑢 + 𝑣𝑣⃑ is a vector in ℝ2 • 𝑐𝑐𝑢𝑢 is a vector in ℝ2
• 𝑢𝑢 + 𝑣𝑣⃑ = 𝑣𝑣⃑ + 𝑢𝑢 • 𝑐𝑐 𝑢𝑢 + 𝑣𝑣⃑ = 𝑐𝑐𝑢𝑢 + 𝑐𝑐 𝑣𝑣⃑
• 𝑢𝑢 + 𝑣𝑣⃑ + 𝑤𝑤 • 𝑐𝑐 + 𝑑𝑑 𝑢𝑢 = 𝑐𝑐𝑢𝑢 + 𝑑𝑑𝑢𝑢
= 𝑢𝑢 + 𝑣𝑣⃑ + 𝑤𝑤 • 𝑐𝑐 𝑑𝑑𝑢𝑢 = 𝑐𝑐𝑐𝑐 𝑢𝑢
• 𝑢𝑢 + 0 = 𝑢𝑢 • 1 𝑢𝑢 = 𝑢𝑢
0 is called a zero vector
R: real number space
• 𝑢𝑢 + −𝑢𝑢 = 0 2: 2-d
Vector Space
Let 𝑉𝑉 be a set of vectors. 𝑉𝑉 is a vector space if the
following properties are satisfied for every vector 𝑢𝑢, 𝑣𝑣⃑
and 𝑤𝑤 and every real number 𝑐𝑐 and 𝑑𝑑:
• 𝑢𝑢 + 𝑣𝑣⃑ is a vector in 𝑉𝑉 • 𝑐𝑐𝑢𝑢 is a vector in 𝑉𝑉
• 𝑢𝑢 + 𝑣𝑣⃑ = 𝑣𝑣⃑ + 𝑢𝑢 • 𝑐𝑐 𝑢𝑢 + 𝑣𝑣⃑ = 𝑐𝑐𝑢𝑢 + 𝑐𝑐 𝑣𝑣⃑
• 𝑢𝑢 + 𝑣𝑣⃑ + 𝑤𝑤 • 𝑐𝑐 + 𝑑𝑑 𝑢𝑢 = 𝑐𝑐𝑢𝑢 + 𝑑𝑑𝑢𝑢
= 𝑢𝑢 + 𝑣𝑣⃑ + 𝑤𝑤
• 𝑐𝑐 𝑑𝑑𝑢𝑢 = 𝑐𝑐𝑐𝑐 𝑢𝑢
• 𝑢𝑢 + 0 = 𝑢𝑢
0 is called a zero vector • 1 𝑢𝑢 = 𝑢𝑢
• 𝑢𝑢 + −𝑢𝑢 = 0
Examples of Vector Spaces
• Real number spaces: ℝ, ℝ2 , ℝ3 …
same length
𝑎𝑎1 𝑏𝑏1 𝑎𝑎1 𝑏𝑏1 𝑎𝑎1 + 𝑏𝑏1
⋮ , ⋮ ∈ 𝑅𝑅𝑛𝑛 ⋮ + ⋮ = ⋮ ∈ 𝑅𝑅𝑛𝑛
𝑎𝑎𝑛𝑛 𝑏𝑏𝑛𝑛 𝑎𝑎𝑛𝑛 𝑏𝑏𝑛𝑛 𝑎𝑎𝑛𝑛 + 𝑏𝑏𝑛𝑛
• Polynomial spaces: any set containing all polynomials of the
same order or smaller
3𝑥𝑥 2 + 4𝑥𝑥, 5𝑥𝑥 2 + 𝑥𝑥 ∈ 𝑃𝑃2
3𝑥𝑥 2 + 4𝑥𝑥 + 5𝑥𝑥 2 + 𝑥𝑥 = 8𝑥𝑥 2 + 5𝑥𝑥 ∈ 𝑃𝑃2
• Matrix spaces: any set containing all matrices of the same size
1 2 3 7 0 9
, ∈ 𝑀𝑀2,3
4 5 6 2 8 1
1 2 3 7 0 9 8 2 12
+ = ∈ 𝑀𝑀2,3
4 5 6 2 8 1 6 13 7
Example: Is ℝ a Vector Space?
To verify a set is a vector space, we need to check all
10 properties, which are called axioms:
• 𝑢𝑢 + 𝑣𝑣⃑ is a vector in 𝑉𝑉
• 𝑢𝑢 + 𝑣𝑣⃑ = 𝑣𝑣⃑ + 𝑢𝑢
• 𝑢𝑢 + 𝑣𝑣⃑ + 𝑤𝑤
= 𝑢𝑢 + 𝑣𝑣⃑ + 𝑤𝑤
• 𝑢𝑢 + 0 = 𝑢𝑢
0 is called a zero vector
• 𝑢𝑢 + −𝑢𝑢 = 0
Example: Is ℝ a Vector Space?
To verify a set is a vector space, we need to check all
10 properties, which are called axioms:
• 𝑐𝑐𝑢𝑢 is a vector in 𝑉𝑉
• 𝑐𝑐 𝑢𝑢 + 𝑣𝑣⃑ = 𝑐𝑐𝑢𝑢 + 𝑐𝑐𝑣𝑣⃑
• 𝑐𝑐 + 𝑑𝑑 𝑢𝑢 = 𝑐𝑐𝑢𝑢 + 𝑑𝑑𝑢𝑢
• 𝑐𝑐 𝑑𝑑𝑢𝑢 = 𝑐𝑐𝑐𝑐 𝑢𝑢
• 1 𝑢𝑢 = 𝑢𝑢
Spanning Set
𝑢𝑢2
• You can represent any
𝑣𝑣
point on ℝ2 as the sum
of the scalar multiples
of two “distinct” vectors 𝑢𝑢
7
• E.g. 𝑢𝑢 = 2 , 𝑣𝑣⃑ = 3 , −𝑣𝑣
7
3 1
7 𝑢𝑢
= 2𝑢𝑢 + 𝑣𝑣⃑ −1
7 2 𝑣𝑣
𝑢𝑢1
−1
= 𝑢𝑢 − 𝑣𝑣⃑
2
Spanning Set
• All vectors in ℝ2 can be represented as a linear
combination of 𝑢𝑢 and 𝑣𝑣: ⃑
For any 𝑤𝑤 ∈ ℝ2 , there exist scalars 𝑐𝑐 and 𝑑𝑑 such that
𝑤𝑤 = 𝑐𝑐𝑢𝑢 + 𝑑𝑑 𝑣𝑣⃑
• We say 𝑢𝑢 = 2 and 𝑣𝑣⃑ = 3 span ℝ2
3 1
Or equivalently, the set of vectors (𝑢𝑢, 𝑣𝑣) ⃑ is a spanning
2
set of ℝ
• Notice that you can span ℝ2 with any two “distinct”
vectors in ℝ2
As long as 𝑢𝑢 ≠ 𝑐𝑐𝑣𝑣⃑ for any scalar 𝑐𝑐
= the vectors cannot be multiples of each other
= the vectors are pointing along different lines
Linear Combination
• If 𝑤𝑤 = 𝑐𝑐𝑢𝑢 + 𝑑𝑑 𝑣𝑣,
⃑ how do we find 𝑐𝑐 and 𝑑𝑑 given 𝑤𝑤, 𝑢𝑢 and 𝑣𝑣?
⃑
• This is a linear algebra problem, so you can use any of the tools
you have learnt so far to solve it: back substitution, Gaussian
Elimination,…
• E.g. 𝑢𝑢 = 2 , 𝑣𝑣⃑ = 3 , 𝑤𝑤 = 7
3 1 6
7 2 3
=c + 𝑑𝑑
6 3 1
7 = 2𝑐𝑐 + 3𝑑𝑑
�
6 = 3𝑐𝑐 + 𝑑𝑑
11 9
𝑐𝑐 = , 𝑑𝑑 =
7 7
Example: 挾公仔機
• Moving along 3 axes
is enough to reach
every point in a
3-dimensional space
• In other words, 3
vectors span 𝑅𝑅3
Subspace
• A nonempty set 𝑊𝑊 of a vector space 𝑉𝑉 is a subspace
of 𝑉𝑉 when 𝑊𝑊 is a vector space under the operations of
addition and scalar multiplication defined in 𝑉𝑉
• Example: 𝑊𝑊 = { 𝑥𝑥1 , 0, 𝑥𝑥3 |𝑥𝑥1 , 𝑥𝑥3 ∈ ℝ} is a subspace
of ℝ3
Under the addition and scalar multiplication rules of ℝ3 , all
the requirement of being a vector space is satisfied. E.g.:
𝑥𝑥1 , 0, 𝑥𝑥3 + 𝑥𝑥1′ , 0, 𝑥𝑥3′ = 𝑥𝑥1 + 𝑥𝑥1′ , 0, 𝑥𝑥3 + 𝑥𝑥3′ ∈ 𝑊𝑊
𝑐𝑐 𝑥𝑥1 , 0, 𝑥𝑥3 = 𝑐𝑐𝑥𝑥1 , 0, 𝑐𝑐𝑥𝑥3 ∈ 𝑊𝑊
Does 𝑊𝑊 equal ℝ3 ? No, because 𝑥𝑥1 , 𝑥𝑥2 , 𝑥𝑥3 ∉ 𝑊𝑊 if 𝑥𝑥2 ≠ 0
• The above 2 axioms are sufficient for proving that a set
is a subspace. No need to go through all 10 axioms.
Example: 挾公仔機
• The plane spanned
by any 2 axes is a
subspace of 𝑅𝑅3
Example: 挾公仔機
• The plane spanned
by any 2 axes is a
subspace of 𝑅𝑅3
Example: 挾公仔機
• The plane spanned
by any 2 axes is a
subspace of 𝑅𝑅3
