Sunteți pe pagina 1din 20

Unit 1: Matrix Algebra

Matrix Algebra

Unit 1: Matrix Algebra

1 / 20

Unit 1: Matrix Algebra

Abstract

Elementary matrix operations

Determinant of a matrix

Rank of a matrix

Inverse of a matrix

Systems of equations. Rouche-Fr


obeniuss Theorem. Cramers rule.

Unit 1: Matrix Algebra

2 / 20

1. Elementary matrix operations


Examples of matrices


3 1 2
A=
is a matrix of dimension 2 3. (2 rows and 3
4
0 1
columns). The entries of this matrix are
a11 = 3, a12 = 1, a13 = 2,
a21 = 4, a22 = 0, a23 = 1

1 4 2
2 is a matrix of dimension 3 3. (3 rows and 3
B= 3 0
1 1
3
columns). The entries of this matrix are
b11 = 1, b12 = 4, b13 = 2,
b21 = 3, b22 = 0, b23 = 2,
b31 = 1, b32 = 1, b33 = 3
If the number of rows equals the number of columns we say that the
matrix is a square matrix. So, B is a square matrix. In short, we say that
B is a square matrix of order 3.
Unit 1: Matrix Algebra

3 / 20

Addition or subtraction of matrices They need to have the same


dimension.
It is done
If

 element by
 element. Example.

1 3 1
1 3
2
A=
and B =
then the sum can be
4 0
2
1
5 1
done since they have
 the same dimension. Furthermore, the sum is
2 0 1
given by A + B =
5 5 1
Product by a constant Always can be carried out. Multiply
each


1 3
then
entry of the matrix by the constant. Example. If A =
4 0


2 6
2A =
8 0

Unit 1: Matrix Algebra

4 / 20

Product of two matrices If A is of dimension m n and B is of


dimension n p then the product AB of these two matrices can be
carried out, and the product matrix P = AB is of dimension m p.
The element pij of P is obtained multiplying the row i of A by the j
th column of B.



4 2
5 3 4
Example. Consider A =
and B = 0 5 .
0
1 2
1 3
A is a matrix of dimension 2 3 and B is 3 2 . Then the product
AB is possible (the same underline numbers). AB is a matrix of
dimension 22 . (Or what is the same, of order 2.) One can obtain


24 7
that AB =
2 11

Unit 1: Matrix Algebra

5 / 20

Exercise 2. Consider the matrices





1 2
1 1 0
0 1

0 5
A=
, B=
and C =
.
0 2 3
2
6
3 8
Obtain if possible: a) ABC b) 2AB c) A(B C) d) B 3C





6 17
2 6
Solution. (a)
(b)
28
75
18 28

12 39
(c) It is not possible (d) 30 90
48 135

Unit 1: Matrix Algebra

6 / 20


1 2
Exercise 4. Obtain all matrices that commute with A =
,
3 4
that is, obtain all matrices B of order two such that AB = BA.

Solution. All matrices B of the form B =

1
1

2
3


+

1 0
0 1


,

for any , R.


Exercise 5. Consider the matrix A =

2
5
2 1

. Obtain the values




1 0
2
a R and b R such that A + aA + bI = N, where I =
0 1


0 0
is the indentity matrix of order two and N =
denotes the
0 0
null matrix of order 2.
Solution. a = 1, b = 12.

Unit 1: Matrix Algebra

7 / 20

2. Determinant of a matrix
This operation is only for square matrices. Its denoted by det(A) or |A|.


a11 a12
Matrix of order 2. If A =
then
a21 a22


3 1
det(A) = a11 a22 a12 a21 . Example. If A =
then
4
2
det(A) = 10.
Matrix of order 3. To obtain
the determinat
use the so-called Sarruss

2 0
1
rule. Example. If A = 1 3 2 then using the Sarruss rule
3 4 4
one can obtain that det(A) = 3.

Unit 1: Matrix Algebra

8 / 20

Determinant of a square matrix A of order n.


1

The complementary minor of the element aij is denoted by ij and


defined as the determinant obtained by removing the row i and
column j from A.

The adjoint of the element aij is denoted by Aij and is defined by


Aij = (1)i+j ij .
Example.
Determine the
adjoints A32 and A22 of
2 1
0
1 2 . Solution. A32 = 4, A22 = 2.
A= 0
2
0
1
The determinant of A is equal to the sum of the elements of any row
or column of A multiplied
by its corresponding
adjoints. Example. Let

1 3
1 1
2 0
0 1

us consider A =
4 2 3 0 . It results det(A) = 1.
0 1
1 1

Unit 1: Matrix Algebra

9 / 20

3. Rank of a matrix
Definition
Any determinant of order k formed from any arbitrary k rows and k
columns of a matrix A is called a minor of order k of the matrix A.

1 1
1
Example. Give a minor of order two of A = 0 0 2 .
2 2 1


1
0 1
Example. Obtain all minors of A =
0 2 3

Definition
The rank of a matrix is the order of the greatest minor different from zero.
We can conclude that if A is of dimension m n then rank(A) is less or
equal minimum(m, n). Furthermore, if A is a square matrix of order n
with det(A) 6= 0 then rank(A)= n.
Unit 1: Matrix Algebra

10 / 20

Exercise 8. Obtain the rank of the following


matrices.

2
1 1 3
1 1
3
0
1
0
4
1
1 1
1 , G =
A= 2
4 2
4
1
0 3
7 1
0
0
3 2

1 0
0 2 2
3 1 1 4
0

H=
2 1 1 6 2
6 2 2 1
0

Solution. rank(A)=2, rank(G)=4, rank(H)=3

Unit 1: Matrix Algebra

11 / 20

4. Inverse of a matrix
Suppose that A is a square matrix of order n.
The inverse matrix of A is another matrix denoted by A1 such that
AA1 = A1 A = I, where I is the identity matrix of order n.
Not all square matrices have an inverse. If A has inverse then is called
regular or invertible, in other case is called singular.
A is invertible if and only if det(A) 6= 0.
Properties: (A1 )1 = A and (AB)1 = B 1 A1 .
The adjoint
Adj(A) is another matrix given by
matrix of A denoted by
A11 A12 . . . A1n
A21 A22 . . . A2n

Adj(A)=
..
.. where Aij = (1)i+j ij is the

.
.
An1 An2 . . . Ann
adjoint of aij and ij is the complentary minor of aij .

Unit 1: Matrix Algebra

12 / 20

Let B be a matrix of dimension m n. The transpose of B, denoted


by B t , is a matrix of dimension n m where its (ij)-th element is
equal to the (ji)-th element of B. In practice to obtain it change rows
by columns (or columns by rows)



3 0
3 3 3
Example. If B =
then B t = 3 1
0
1 2
3 2
a matrix A is said to be symmetric if At = A.
1
If A is invertible (det(A) 6= 0) then A1 =
(Adj(A))t
det(A)

Exercise 9. Determine if the following matrices have inverse. If they


have
obtain it by
the adjoint procedure.
1 0 3
E= 2 1 4
0 1 3

Solution. E 1

1/5
3/5 3/5
3/5
2/5
= 6/5
2/5 1/5
1/5
Unit 1: Matrix Algebra

13 / 20

Exercise 11. Suppose det(A) 6= 0 and det(C) 6= 0, solve then the


matrix
 equationAXC =B. As application

 obtain the
 matrix X for
2
3
0 3
1 3
A=
, B=
, C=
.
4 1
4 2
1 2
Solution. X = A1 BC 1 . Particular case, X =

33/70 27/70
4/35
16/35

Exercise 13. Express the following systems in matrix form and then
solve them by means of the inverse of the matrix
coefficients.

3x + y z = 1
2x + 3y = 18
a)
c) x 3y + z = 9
3x + 5y = 29
x y + 4z = 3
Solution. (a) x = 3, y = 4 (c) x =

16
63
34
,y = ,z =
19
19
19

Unit 1: Matrix Algebra

14 / 20

5. Systems of equations. Rouche-Frobeniuss Theorem.


Cramers rule.
Consider a system of m linear equations with n unknowns
a11 x1 + a12 x2 + + a1n xn = b1
a21 x1 + a22 x2 + + a2n xn = b2
..
.
am1 x1 + am2 x2 + + amn xn = bm
Its matrix form is given by Ax = b where

a11 a12 . . . a1n


a21 a22 . . . a2n

A=
..
..

.
.
am1 am2 . . . amn
{z
|
coefficients matrix

, x =

x1
x2
..
.

, b =

xm
| {z }
vector of
unknowns

Unit 1: Matrix Algebra

b1
b2
..
.

bm
{z }
vector of

independent terms
15 / 20

Some definitions. We say that a linear system is


incompatible if there is no solution.
a determinate system if there is a unique solution.
an indeterminate system if there are infinitely many solutions.

Rouche-Frobenius theorem
Consider a system of linear equations Ax = b. Let A+ be the augmented
matrix obtained by adding the column of independent terms to A. It holds,
1

if rank(A) 6= rank(A+ ) the system is incompatible: there is no


solution.
if rank(A) = rank(A+ )=number of unknowns the system is
determinate: there is a unique solution.
if rank(A) = rank(A+ ) < number of unknowns the system is
indeterminate: there are infinitely many solutions.

Unit 1: Matrix Algebra

16 / 20

Exercise 15. Discuss the following systems applying


Rouche-Frobeniuss theorem.
xyz =
1
2x + 3y z + 5t =

2x + y z =
2
7x 3t =
d)
e)
x + 2y 4z = 3
4z + t =

2x + 2y 4z =
0
2x y + 3z + 3t =

2x 5y + 4z + u = 3
x 2y + z u =
5
f)

x 4y + 6z + 2u = 10

0
0
0
0

Solution.
d) rank(A)=3 6= rank(A+ )=4 incompatible system: there is no solution.
e) rank(A)=rank(A+ )=4 = number of unknowns determinate system:
there is a unique solution.
f) rank(A)=rank(A+ )=3 < number of unknowns indeterminate
system: there are infinitely many solutions.

Unit 1: Matrix Algebra

17 / 20

Cramers rule
Consider a system Ax = b, of n linear equations with the same number n
of unknowns. Suppose that det(A) 6= 0. Then, we know that there exists a
unique solution. Cramers rule states that the solution is given by
xi =

det(Ai )
det(A)

i = 1, . . . , n,

where Ai is the matrix obtained by replacing the i-th column of A by the


column vector b.
Cramers rule can be extended to obtain the solution of any
indeterminate system. Next we give some examples.

Unit 1: Matrix Algebra

18 / 20

Exercise 16. Discuss the following systems of linear equations by


means of Rouche-Frobeniuss theorem. If there exists solution obtain
it by Cramers rule.

x+yz = 0
3x + y z = 2
d) 2x y + z = 1
e) 2x + y z = 1
x y + 2z = 2
x + 2y 2z = 3

3x + 2y 3z = 0
x y + 4z = 1
f)

2x + 3y 7z = 1
Solution. d) rank(A)=rank(A+ )=3 = number of unknowns determinate
system: there is a unique solution. The solution is x = 1, y = 1, z = 0.
e) rank(A)=rank(A+ )=2 < number of unknowns indeterminate
system: there are infinitely many solutions.
The solutions are x = 1/5, y = (5 + 7)/5, z = , for any R.
f) rank(A)=rank(A+ )=2 < number of unknowns indeterminate
system: there are infinitely many solutions.
1
1
1
1
The solutions are x = + , y = , z = + , for any R.
3
5
3
5
Unit 1: Matrix Algebra

19 / 20

Exercise 17. Discuss the following systems of linear equations by


Rouche-Frobeniuss theorem for the given values of the parameters. If
there exists solution obtain itby Cramers rule.
x+yz = m
(a) mx + 2y z = 3m
For m = 0, 1, 2

2x + my z = 6
Solution.
a) If m = 0 rank(A)=26= rank(A+ )=3 incompatible system: there is
no solution.
If m = 1 rank(A)=rank(A+ )=3=number of unknowns determinate
system: there is a unique solution. The solution is given by
x = 5, y = 2, z = 6.
If m = 2 rank(A)=rank(A+ )=2 < number of unknowns
indeterminate system: there are infinitely many solutions. The solutions are
given by x = 4 , y = , z = 2, for any R.

Unit 1: Matrix Algebra

20 / 20

S-ar putea să vă placă și