Algebra of matrices

 Matrices
 Types of matrices
 Operations of matrices
 Properties of matrices
 Determinants
 Inverse of a matrix
Why use it?
 Matrix algebra makes mathematical
expression and computation easier. Ex
solve systems of linear eqns.
 It allows you to get rid of cumbersome
notation, concentrate on the concepts
involved and understand where your results
come from.
Where we use matrices cont…
 In economics: input-output analysis,
 Insurance/Social protection: Pricing products
 Banking: Portfolio analysis
 In statistics: forecasting, time series
 In survival model: Application of Markov
processes
 Computer graphics: store inventory data,
transportation routing for storing connections,
etc
 Engineering, geology, etc
Definitions - Matrix
 A matrix is an array of numbers enclosed by a
pair of brackets.

A=  a11 a12 a13 

a 21 a 22 a 23
 
 Denoted with a bold Capital letter
 All matrices have an order (or dimension):
that is, the number of rows  the number of
columns. So, A is 2 by 3 or (2  3).
Definitions cont...
1 3 1 
2 3 7
A  B   2 1 4
 1 1 5   4 7 6 

Both A and B are examples of matrix.

A is of order 2 x 3 and B is order 3 x 3

Why matrix?

5
Matrices cont…..
 numbers aij are called elements. First
subscript indicates the row; second
subscript indicates the column. The matrix
consists of mn elements
 It is called “the m  n matrix A = [aij]” or
simply “the matrix A ” if number of rows
and columns are understood.
What is it?
 Matrix algebra is a means of making
calculations upon arrays of numbers (or
data).
 Most data sets are matrix-type
Types of Matrices
A square matrix is a matrix that has the
same number of rows and columns (n  n)
When m = n,  a11 a12  a1n 
a a2 n 
i.e., A   21
 
a22 


 
 an1 an 2 ann 

A is called a “square matrix of order n” or

“n-square matrix”
elements a11, a22, a33,…, ann called diagonal
elements.
n

 a
i 1
ii  a11  a22  ...  ann
is called the trace of 8
 Types cont...
Row matrix

 A matrix with single row, it is called a row

matrix, i.e.,

A= 1 2 3 4 5

9
 Types cont...
Column matrix

 A matrix with single column, it is called a

column matrix, i.e.,

1 
B=  2
 3
 4
5

10
Types cont...
Equal matrices

Two matrices A = [aij] and B = [bij] are said to

be equal (A = B) iff each element of A is equal
to the corresponding element of B, i.e., aij = bij
for 1  i  m, 1  j  n.
iff pronouns “if and only if”
if A = B, it implies aij = bij for 1  i  m, 1  j  n;
if aij = bij for 1  i  m, 1  j  n, it implies A = B.
11
 Equal matrices cont....

 1 0 a b 
Example: A  and B 
  4 2   c d 
Given that A = B, find a, b, c and d.

if A = B, then a = 1, b = 0, c = -4 and d = 2.

12
 Types cont….
Triangular matrix

 A square matrix whose elements aij = 0, for

i > j is called upper triangular, i.e., a11 a12  a1n 
0 a22 a2 n 

   
 
0 0 ann 

A square matrix whose elements aij = 0, for i

< j is called lower triangular, i.e.,  a11 0  0 
a a22 0 
 21
   
 
 an1 an 2 ann 
13
 Types cont...
Diagonal matrix

Both upper and lower triangular, i.e., aij = 0, for

i  j , i.e.,  a11 0  0 
0 a22 0 
D
   
 
0 0 ann 

is called a diagonal matrix, simply

D  diag[a11 , a22 ,..., ann ]

14
Types cont...
Scalar matrix

Is a diagonal matrix with equal diagonal

entries.

 3 0 0
 0 3 
A  0 
 0 0  3

15
 Types cont...
Identity matrix

Is diagonal matrix with ones on the diagonal

In particular, a11 = a22 = … = ann = 1, the
matrix is called identity matrix.
Properties: AI = IA = A

Examples of identity matrices:

1 0 0
1 0 0
0 1 0 
 1  
I2 = I3 =  0 0 1 
16
 Types cont...
Zero matrices

Every element of a matrix is zero, it is called a

zero matrix, i.e.,
0 0  0 
0 0 0 
A 
  
 
 0 0 0 

17
Types cont..
The transpose of a matrix -Not really type!!

The matrix obtained by interchanging the

rows and columns of a matrix A is called the
transpose of A (write AT).

1 2 3
A 
Example:  4 5 6  1 4
The transpose of A is AT   2 5 
 3 6 

For a matrix A = [aij], its transpose AT = [aji], for

all i and j
18
Types cont...
Symmetric matrix and Skew – symmetric matrix

A matrix A such that AT = A is called symmetric,

i.e., aji = aij for all i and j.
A + AT must be symmetric. Why?
1 2 3 
Example: A   2 4 5 is symmetric.
 3 5 6 

A matrix A such that AT = -A is called skew-

symmetric, i.e., aji = -aij for all i and j.
A - AT must be skew-symmetric. Why? 19
Powers of matrices

 Matrix multiplication can also be written in

exponent form.
 This requires that we have a square matrix.
Like real number multiplication and
exponents.

2 3 2 n 1 n
A  A. A, A  A A.......... A  A A
Submatrix
 A matrix obtained by deleting some of the
rows and/or columns of a matrix is said to
be a submatrix of the given matrix.

1 4 5 
 For example , if A   
0 1 2
 Then
1 
1 , 2,  , 1 5, 0 2
0
 Are few submatrices of A.
Submatrices cont…..
1 4  1 4 
A    and A   
1 0  0 2 

 Are not submatrices of A. (Students

advised to give reasons)
 Example : Application

Three friends Asha, Benjamin & Chacha write

to each other from time to time. Last year Asha
received a letter from Benjamin and 5 letters
from Chacha. Benjamin received 2 letter from
Asha and 3 from Chacha, and Chacha received a
letter from Asha and 6 from Benjamin.
Task: (a) Summarize the information in matrix
form.
(b) Who wrote the most letters last year??.
Matrix Operations
 Addition and Subtraction
 Multiplication
 Inversion
Addition and Subtraction
 Two matrices may be added (or subtracted)
iff they are of the same order.

 Simply add (or subtract) the corresponding

elements. So, A + B = C yields
Addition and Subtraction (cont.)
a11 a12   b11 b12  c11 c12 
a a   b b   c c 
 21 22   21 22   21 22 

a31 a32  b31 b32  c31 c32 

Where a11  b11  c11

a12  b12  c12
a21  b21  c 21
a22  b22  c 22
a31  b31  c31
a32  b32  c32
Matrix Multiplication
 To multiply a scalar times a matrix, simply
multiply each element of the matrix by the
scalar quantity

 a11 a12   ka11 ka12 

k 
a21 a22  ka21 ka22 

Matrix Multiplication (cont.)
 To multiply a matrix times a matrix, we
write
• AB (A times B)
 This is pre-multiplying B by A, or post-
multiplying A by B.
Matrix Multiplication (cont.)
 In order to multiply matrices, they must be
CONFORMABLE
 that is, the number of columns in A must
equal the number of rows in B
 So,
A  B = C
(m  n)  (n  p) = (m  p)
Matrix Multiplication (cont.)
 (m  n)  (p  n) = cannot be done
 (1  n)  (n  1) = a scalar (1x1)
Matrix Multiplication (cont.)
 Thus
a11 a12 a13   b11 b12  c11 c12 
a a a  x b b   c c 
 21 22 23   21 22   21 22 
a31 a32 a33  b31 b32  c31 c32 
 where
c11  a11b11  a12 b21  a13b31
c12  a11b12  a12 b22  a13b32
c 21  a21b11  a22 b21  a23b31
c 22  a21b12  a22 b22  a23b32
c31  a31b11  a32 b21  a33b31
c32  a31b12  a32 b22  a33b32
Matrix Multiplication- an
example
 Thus
1 4 7 1 4 c11 c12  30 66
2 5 8 x 2 5  c c   36 81
     21 22   
3 6 9 3 6 c31 c32  42 96
 where
c11  1 * 1  4 * 2  7 * 3  30
c12  1 * 4  4 * 5  7 * 6  66
c 21  2 * 1  5 * 2  8 * 3  36
c 22  2 * 4  5 * 5  8 * 6  81
c31  3 * 1  6 * 2  9 * 3  42
c32  3 * 4  6 * 5  9 * 6  96
The Determinant of a Matrix
 Each n–square matrix A = [aij] is assigned a
special scalar called the determinant of A,
denoted by | A | (or det (A)).
 Determinants exist only for square matrices.
 They have a matrix characteristics.
 A matrix A is said to be a singular matrix if
det(A ) = 0. It is called non-singular if
det(A ) ≠ 0.
Why determinants???

 We use determinants in solving equations, to

find an inverse of a matrix etc.
The Determinant for a 1x1, 2x2, 3x3
and n x n matrices
Properties of Determinates
 Determinants have several mathematical
properties which are useful in matrix
manipulations.
1 |A|=|A'|.
2. If a row or column of A = 0, then |A|= 0.
3. If every value in a row or column is multiplied by k,
then |A| = k|A|.
4. If two rows (or columns) are interchanged the sign,
but not value, of |A| changes.
5. If two rows or columns are identical, |A| = 0.
6. If two rows or columns are linear combination of
each other, |A| = 0
Properties of Determinants
7. |A| remains unchanged if each element of a row
or each element multiplied by a constant, is
added to any other row.
8. |AB| = |A| |B|
9. Det of a diagonal matrix = product of the
diagonal elements
Minors
th
 The (n 1) order determinant formed by
deleting i th row and j th column.
 This determinant of order (n 1) is known as
the minor of and is denoted by M ij

 Example a a a 
11 12 13
 
A  a21 a22 a23 
a31 a32 a33 
 

a 22 a 23
M 11 
a32 a33
 Minors cont…
 4 7 8
If A = -9 0 0  , then
 2 3 4

M11 = Minor of a11 0 0

= =0
3 4

4 7
Similarly, M23 = Minor of a23 = = 12 -14 = -2
2 3

4 8
M32 = Minor of a32 = = 0+72 = 72 etc.
-9 0
cij   M ij Cofactors
Cofactor of an element cij is a signed minor,
mathematically obtained as:
i+ j
Cij = Cofactor of aij in A = -1 Mij ,

where Mij is minor of aij in A

Note: If i+j is even then, cij  M ij

If i+j is odd then, cij   M ij

Cofactors (Con.)
 4 7 8
A = -9 0 0 
 2 3 4 
0 0
C11 = Cofactor of a11 = (–1)1 + 1 M11 = (–1)1 +1 =0
3 4

4 7
C23 = Cofactor of a23 = (–1)2 + 3 M23 =  2
2 3

4 8
C32 = Cofactor of a32 = (–1)3 + 2M32 =- = - 72 etc.
-9 0
Cofactors (Con.)
 4 7 8
A = -9 0 0 
 2 3 4 
0 0
C11 = Cofactor of a11 = (–1)1 + 1 M11 = (–1)1 +1 =0
3 4

4 7
C23 = Cofactor of a23 = (–1)2 + 3 M23 =  2
2 3

C32 = Cofactor of a32 = (–1)3 + 2 M32 = - 4 8 = - 72

-9 0
Adjoint
 Then the adjoint of A , denoted by Adj(A) is
defined as the transposed cofactor matrix.
 Adj(A)= C t
 Example find Adj(A)
1 2 3
A 
 2 3 2 


3 3 4 
Laplace expansion
 For example, arbitrarily selecting row i and
column j respectively, the determinant can
be computed by:
n
A   (
j 1
1) i j
aij M ij
n
A   ( 1)
i 1
i j
aij M ij
The Inverse of a Matrix (A-1)
 Is a square matrix that satisfies the equations
BA  AB  I n
 IOWs: For an n  n matrix A, there may be a B
such that BA = I = AB. where B is A-1
 The inverse is analogous to a reciprocal
 A matrix which has an inverse is nonsingular.
 A matrix which does not have an inverse is
singular.
 An inverse exists only if A  0
Inverse of matrix cont…
 Generally inverse of a square matrix is
given by 1 Adj( A)
A  , for all A  0
A
 That is for inverse to exist determinant of a
matrix must be non zero.
Inverse cont….
For any 2x2 or order 2 matrix  a11 a12 
A
 a21 a22 

1  a22 a12 
Its inverse can be written as 1
A   a
A  21 a11 

 1 0 
Example: Find the inverse of A 
 1 2 

The determinant of A is -2
 1 0 
Hence, the inverse of A is 1
A 
1/ 2 1/ 2 
 
47
Inverse of order 3
1 2 3
Consider an example: A   4 5 6 
 7 8 9 
Its determinant can be obtained by:
1 2 3
4 5 1 2 1 2
A  4 5 6 3 6 9
7 8 7 8 4 5
7 8 9
 3  3  6  6   9  3  0

48
 Inverse cont…
1 2 3 
Cofactor matrix of A  0 4 5 
1 0 6 
The cofactor for each element of matrix A:
4 5 0 5 0 4
A11   24 A12   5 A13   4
0 6 1 6 1 0

2 3 1 3 1 2
A21   12 A22  3 A23   2
0 6 1 6 1 0

2 3 1 3 1 2
A31   2 A32    5 A33  4
4 5 0 5 0 4
49
 Inverse cont….
1 2 3 
Cofactor matrix of A  0 4 5  is then given by:
1 0 6 

 24 5 4 
 12 3 2 
 
 2 5 4 
50
Inverse cont….
1 2 3 
Inverse matrix of A  0 4 5  is given by:
1 0 6 

T
 24 5 4   24 12 2 
1 1   1  
A   12 3 2  5 3  5
A  22  
 2 5 4   4 2 4 

 12 11  6 11 1 11 
  5 22 3 22 5 22 
  2 11 1 11 2 11 
51
Properties of inverse matrices
 AB  1
 B A-1 -1

'

A' 
1
 A -1

 A 
-1 1
 A
For example,
 a = [1 21 12] and b = [1/3 7 4] are
linearly dependent

 A matrix A of dimension n  p (p < n) is of

rank p. Then A has maximum possible rank
and is said to be of full rank.
 In general, the maximum possible rank of an n
 p matrix A is min(n,p).
Reference
 Olver, P.J. and Shakiban, C. (2005)  Applied Linear
Algebra, Prentice Hall.
 Strang, G. (2005) Linear Algebra and its Applications, 4th
Edition, Thomson Publishing.
 Lay, D. C. Lay, S. R. and McDonald, J.J (2015) Linear
Algebra and its Applications, 5th Edition. Addison Wesley.
 Kolman, B. and Hill, D.R. (1993) Introductory Linear
Algebra with Applications, 6th Edition. MacMillan Publishing
Company.
