Download as pdf or txt
Download as pdf or txt
You are on page 1of 32

11.

Determinant, inverse & systems of linear equations

University of Bristol

Julia Wirtz

ECON10005 - Mathematics for Economics

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 1 / 32
The Determinant

Consider the pair of linear equations with coefficient matrix A:


(  
ax1 + bx2 = y a b
A=
cx1 + dx2 = z c d

Solving the system in the usual way yields:


dy − bz az − cy
x1 = x2 =
ad − bc ad − bc
Common denominator (ad − bc) is the determinant of matrix A.
The determinant is denoted by det(A) or |A|.
To have a valid solution, the denominator cannot be zero.
→ Thus: system of equations only has a solution if det(A) ̸= 0

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 2 / 32
Geometric interpretation

 
a b
Consider the matrix A =
c d

Determinant: Pink area

You can check that this is given by


det(A) = |A| = ad − bc

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 3 / 32
Determinant for 1 × 1 and 2 × 2 matrices

1 × 1 matrix Example:
▶ A = [a11 ] ▶ A = [−3]
▶ The determinant is: |A| = a11 ▶ The determinant is: |A| = −3

2 × 2 matrix Example:
   
a b 2 1
▶ A= ▶ A=
c d −1 4
▶ The determinant is: ▶ The determinant is:
a b 2 1
|A| = = ad − cb |A| = = 2 · 4 − (−1) · 1 = 9
c d −1 4

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 4 / 32
Determinant of larger matrices

1. Rule of Sarrus (Only for 3 × 3 matrices)


2. Laplace expansion (Expansion by cofactors)

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 5 / 32
Rule of Sarrus – Only for 3 × 3 matrices

 
a11 a12 a13
Consider A = a21 a22 a23 
a31 a32 a33

To find the determinant do the


following:
▶ Red: multiply and add.
▶ Blue: multiply and deduct.
▶ The sum equals the determinant of A:

|A| =a11 a22 a33 + a12 a23 a31 + a13 a21 a32
− a13 a22 a31 − a11 a23 a32 − a12 a21 a33

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 6 / 32
Example
 
−2 5 −6
A= 1 0 4 
3 2 1

 
−2 5 −6 −2 5
⇒ 1 0 4 1 0 
3 2 1 3 2

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 7 / 32
Laplace expansion
 
a11 a12 a13
Consider A = a21 a22 a23 
a31 a32 a33

1. Select a row or column for expansion


▶ We select the row 1

2. Find the minor for each element in the selected row (column)
▶ Submatrix for element aij : delete row i and column j of A.
▶ Minor: determinant of this submatrix
a22 a23
▶ Element a11 : M11 = = a22 a33 − a32 a23
a32 a33

a21 a23
▶ Element a12 : M12 = = a21 a33 − a31 a23
a31 a33
a21 a22
▶ Element a13 : M13 = = a21 a32 − a31 a22
a31 a32
University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 8 / 32
Laplace expansion II

3. Find the cofactors, Cij for each element in the selected row (column)

Cij = (−1)i+j Mij

 
+ − +
Note: for a 3 × 3 matrix the signs are: − + −
+ − +

The cofactor of a11 is: C11 = (−1)1+1 M11 = M11

The cofactor of a12 is: C12 = (−1)1+2 M12 = −M12

The cofactor of a13 is: C13 = (−1)1+3 M13 = M13

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 9 / 32
Laplace expansion III

Then we have:
Determinant of A

|A| = a11 C11 + a12 C12 + a13 C13


= a11 M11 − a12 M12 + a13 M13
= a11 (a22 a33 − a23 a32 ) − a12 (a21 a33 − a23 a31 ) + a13 (a21 a32 − a22 a31 )

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 10 / 32
Example

 
−2 5 −6
A= 1 0 4 
3 2 1

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 11 / 32
Basic rules for determinants I
Let A be an n × n matrix.

1. If all elements in a row (or column) are 0 then |A| = 0

0 0
= 0b − 0a = 0
a b

2. If two rows (or columns) interchange, the determinant changes sign but not absolute
value.
a b c d
= ad − cb = −(cb − ad) = −
c d a b
3. If a multiple of one row (or column) is added to a different row (or columns) the
value of the determinant remains unchanged.

a b a b
= a(b + d) − (a + c)b = ab + ad − ab − cb = ad − cb =
a+c b+d c d

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 12 / 32
Basic rules for determinants II
Let A be an n × n matrix and k a real number.

4. If any two rows (or columns) are proportional [i.e., one is the multiple of another],
then |A| = 0
a b
= a · kb − ka · b = 0
ka kb
5. |kA| = k n |A|
ka kb a b
= k 2 ad − k 2 cb = k 2 (ad − cb) = k 2
kc kd c d
6. If all elements in a single row (or column) are multiplied by k, then the determinant
is multiplied by k

ka kb a b
= kad − ckb = k(ad − cb) = k
c d c d

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 13 / 32
Basic rules for determinants III

Let A and B be n × n matrices.

7. |AT | = |A|
a c a b
= ad − cb =
b d c d

8. |AB| = |A||B|
(See problem set)

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 14 / 32
The Inverse

The inverse of matrix A is denoted by: A−1

Real numbers: x−1 = 1


x and x · x−1 = 1.

Similarly for matrices:


Matrix Inverse
AA−1 = A−1 A = I
When you multiply a matrix with its inverse, the result is the identity matrix.

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 15 / 32
Existence of the inverse

Matrix must be square to have an inverse.

Invertible matrix
̸ 0 has a unique inverse A−1
Any square matrix A with determinant |A| =
(also called non-singular).

Singular matrix
A matrix A with determinant |A| = 0 does not have an inverse.

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 16 / 32
Inverse matrix properties

Let A and B be invertible square matrices and and c ̸= 0 a real number.

The inverse is unique

A−1 has the same dimensions as A

(A−1 )−1 = A

(AB)−1 = B −1 A−1

(AT )−1 = (A−1 )T

(cA)−1 = c−1 A−1

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 17 / 32
Methods for constructing the inverse

1. Simple formula for 2 × 2 matrices

2. Method of adjoint

3. Gauss-Jordan method

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 18 / 32
Constructing the inverse of a 2 × 2 matrix

 
a b
Consider a matrix: A =
c d
   
−1 1 d −b 1 d −b
Then the inverse is: A = |A| = ad−bc
−c a −c a
1
If |A| = 0, then |A| is not defined and the A does not have an inverse.

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 19 / 32
Example

     
a b 2 1 −1 1 d −b
A= = ⇒A =
c d −1 4 |A| −c a

2 1
Determinant: |A| = =
−1 4

Inverse: A−1 =

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 20 / 32
Method of adjoint (to derive inverse)

Adjoint matrix Adj(A): Transpose of the matrix of cofactors of square matrix A.


 
C11 C12 · · · C1m
 C21 C22 · · · C2m 
Matrix of cofactors: C=
 
.. 
 . 
Cm1 Cm2 · · · Cmm

Adjoint matrix: Adj(A) = C T

Inverse Adjoint Formula


1
A−1 = Adj(A)
|A|

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 21 / 32
Example

 
2 3 4
A = 4 3 1
1 2 4

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 22 / 32
Gauss-Jordan method

   
a b c 1 0 0 1 0 0 j k l
 d e f 0 1 0  −→  0 1 0 m n o 
g h i 0 0 1 0 0 1 p q r

Consider square matrix A of order n × n.

Step 1: Form the n × 2n matrix (A : I)


Step 2: Apply elementary row operations to transform it to a n × 2n matrix (I : B)

It follows that B = A−1

If the row operations are not possible there is no inverse.

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 23 / 32
Example

1 0 0 −2 45 9
   
2 3 4 1 0 0 5
 4 3 1 0 1 0  −→  0 1 0 3 − 4 − 14 
5 5
1 2 4 0 0 1 0 0 1 −1 15 6
5

(See problem set)

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 24 / 32
Solving systems of equations

1. Cramer’s rule

2. Matrix inversion

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 25 / 32
Cramer’s rule

Can be used to solve n × n systems of linear equations

Particularly helpful when solving for a specific variable

Uses matrix determinants

In a system of the form Ax = b, the ith variable, xi , can be found from xi = det(A
det(A)
i)

where Ai is the n × n matrix found by replacing column i by the RHS vector b


▶ Consider the system Ax = b
▶ Then the ith variable, xi is given by

det(Ai )
xi =
det(A)

where Ai is the n × n matrix found by replacing column i by the RHS vector b

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 26 / 32
Solving linear equations – Cramer’s rule
(  
a11 x1 + a12 x2 = b1 a a
Consider with coordinate matrix A = 11 12
a21 x1 + a22 x2 = b2 a21 a22

Then the solution is given by:

det(Ai )
xi = ,
det(A)
|A1 | |A2 |
i.e. x1 = x2 =
|A| |A|

Where:
b1 a12
▶ det(A1 ) =
b2 a22
a b
▶ det(A2 ) = 11 1
a11 b2

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 27 / 32
Example
(
−x1 + 9x2 = 20
Consider
3x1 − x2 = 70

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 28 / 32
Using the inverse to solve a system of equations

Solving an equation
Consider a simple equation with variable x and real numbers a, b:

ax = b.

We can solve: x = b
a = a−1 b (for a ̸= 0).

Solving a system of equations


Let A: invertible n × n matrix, x: vector of n of unknowns, b: vector of n constants.

Consider the system of equations: Ax = b

Solution: x = A−1 b

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 29 / 32
Rules
Solving a matrix equation
If |A| =
̸ 0 then:

AX = B ⇔ X = A−1 B
Y A = B ⇔ Y = BA−1

Example: (
ax + by =e
cx + dy =f

This can be written as Ax = B:


     
a b x e
A= ,x = ,B =
c d y f

We can solve: x = A−1 B


University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 30 / 32
Example
(
6x + 2y = 20
4x + 3y = 18

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 31 / 32
Reading

Main textbook
Sydaester, K. and Hammond, P., Essential Mathematics for Economic Analysis, chapters
15 and 16.

Additional textbooks
Ian Jacques, Mathematics for Economics and Business, chapter 7.
Shapoor Vali, Principles of Mathematical Economics, chapter 13.
Alpha C. Chiang and Kevin Wainwright, Fundamental Methods of Mathematical
Economics, chapters 15 and 16.

University of Bristol 11. Determinant, inverse & systems of linear equations ECON10005 32 / 32

You might also like