Introduction

After reading this chapter, you should be able to

1. define what a matrix is.
2. identify special types of matrices, and
3. identify when two matrices are equal.
What does a matrix look like?
Matrices are everywhere. If you have used a spreadsheet such as Excel or otus or written a
table, you have used a !atrix. Matrices !a"e presentation of nu!bers clearer and !a"e
calculations easier to progra!. oo" at the !atrix below about the sale of tires in a
#lowoutr$us store % given by &uarter and !a"e of tires.
'1 '2 '3 '(
Copper
Michigan
Tirestone

)
*
2*

1)
1+
2+

,
1*
3

2,
2*
2
If one wants to "now how !any Copper tires were sold in Quarter 4, we go along the row
Copper and colu!n Q4 and find that it is 2,.
So what is a matrix?
A matrix is a rectangular array of ele!ents. -he ele!ents can be sy!bolic expressions or
nu!bers. Matrix
. / A
is denoted by

=
mn m m
n
n
a a a
a a a
a a a
A
.......
.......
.......
. /
2 1
2 22 21
1 12 11

0ow i of
. / A
has ! ele!ents and is
[ ]
in i i
a a a ....
2 1

and colu!n
j
of
. / A
has n ele!ents and is

mj
j
j
a
a
a

2
1
1
Each !atrix has rows and colu!ns and this defines the si1e of the !atrix. If a !atrix
. / A
has
m
rows and
n
colu!ns, the si1e of the !atrix is denoted by
n m
. -he !atrix
. / A
!ay also be denoted by n m
A

. /
to show that
. / A
is a !atrix with
m
rows and
n
colu!ns.
Each entry in the !atrix is called the entry or ele!ent of the !atrix and is denoted by ij
a
where i is the row nu!ber and
j
is the colu!n nu!ber of the ele!ent.
-he !atrix for the tire sales exa!ple could be denoted by the !atrix /A. as

=
2, , 1) )
2* 1* 1+ *
2 3 2+ 2*
. / A
.
-here are 3 rows and ( colu!ns, so the si1e of the !atrix is ( 3 . In the above
. / A
!atrix,
2,
3(
= a
.
What are the special types of matrices?
2ector3 A vector is a !atrix that has only one row or one colu!n. -here are two types of
vectors % row vectors and colu!n vectors.
Row Vector:
If a !atrix
. /B
has one row, it is called a row vector
. / . /
2 1 n
b b b B =
and
n
is the
di!ension of the row vector.
Example
4ive an exa!ple of a row vector.
Sol!tion
. + 2 3 2+ 2* / . / = B
is an exa!ple of a row vector of di!ension *.
Col!mn "ector:
If a !atrix
. /C
has one colu!n, it is called a colu!n vector

=
m
c
c
C

1
. /

and
m
is the di!ension of the vector.
Example #
4ive an exa!ple of a colu!n vector.
2
Sol!tion

=
)
*
2*
. /C

is an exa!ple of a colu!n vector of di!ension 3.
S!$matrix:
If so!e row5s6 or7and colu!n5s6 of a !atrix
. / A
are deleted 5no rows or colu!ns !ay be
deleted6, the re!aining !atrix is called a sub!atrix of
. / A
.
Example %
8ind so!e of the sub!atrices of the !atrix

=
2 1 3
2 ) (
. / A
Sol!tion
[ ] [ ]

2
2
, ( , 2 ) ( ,
1 3
) (
,
2 1 3
2 ) (
are so!e of the sub!atrices of
. / A
. 9an you find other sub!atrices of
. / A
:
S&!are matrix:
If the nu!ber of rows
m
of a !atrix is e&ual to the nu!ber of colu!ns
n
of a !atrix
. / A
,
5
n m =
6, then
. / A
is called a s&uare !atrix. -he entries nn
a a a ,..., ,
22 11 are called the
diagonal elements of a s&uare !atrix. ;o!eti!es the diagonal of the !atrix is also called the
principal or main of the matrix.
Example '
4ive an exa!ple of a s&uare !atrix.
Sol!tion

=
, 1* )
1* 1+ *
3 2+ 2*
. / A

is a s&uare !atrix as it has the sa!e nu!ber of rows and colu!ns, that is, 3. -he diagonal
ele!ents of
. / A
are
, , 1+ , 2*
33 22 11
= = = a a a
.
(pper triang!lar matrix:
A
n m
!atrix for which
j i a
ij
> = , +
is called an upper triangular !atrix. -hat is, all the
ele!ents below the diagonal entries are 1ero.
3
Example )
4ive an exa!ple of an upper triangular !atrix.
Sol!tion

=
1*++* + +
) ++1 . + +
+ , 1+
. / A

is an upper triangular !atrix.
*ower triang!lar matrix:
A
n m
!atrix for which
i j a
ij
> = , +
is called a lower triangular !atrix. -hat is, all the
ele!ents above the diagonal entries are 1ero.

Example +
4ive an exa!ple of a lower triangular !atrix.
Sol!tion

=
1 * . 2 ) . +
+ 1 3 . +
+ + 1
. / A

is a lower triangular !atrix.
,iagonal matrix:
A s&uare !atrix with all non<diagonal ele!ents e&ual to 1ero is called a diagonal !atrix, that
is, only the diagonal entries of the s&uare !atrix can be non<1ero, 5
j i a
ij
= , +
6.
Example -
4ive exa!ples of a diagonal !atrix.
Sol!tion

=
+ + +
+ 1 . 2 +
+ + 3
. / A
is a diagonal !atrix.
Any or all the diagonal entries of a diagonal !atrix can be 1ero. 8or exa!ple

=
+ + +
+ 1 . 2 +
+ + 3
. / A
is also a diagonal !atrix.
(
.dentity matrix:
A diagonal !atrix with all diagonal ele!ents e&ual to one is called an identity !atrix, 5
j i a
ij
= , +
and
1 =
ii
a
for all i 6.
Example /
4ive an exa!ple of an identity !atrix.
Sol!tion

=
1 + + +
+ 1 + +
+ + 1 +
+ + + 1
. / A

is an identity !atrix.
0ero matrix:
A !atrix whose all entries are 1ero is called a 1ero !atrix, 5
+ =
ij
a
for all i and
j
6.
Example 1
4ive exa!ples of a 1ero !atrix.
Sol!tion

=
+ + +
+ + +
+ + +
. / A

=
+ + +
+ + +
/#.

=
+
+
+

+ + +
+ + +
+ + +
. /C
[ ] + + + . / =
are all exa!ples of a 1ero !atrix.
Tridiagonal matrices:
A tridiagonal !atrix is a s&uare !atrix in which all ele!ents not on the following are 1ero <
the !a=or diagonal, the diagonal above the !a=or diagonal, and the diagonal below the !a=or
diagonal.
*
Example 2
4ive an exa!ple of a tridiagonal !atrix.
Sol!tion

=
) 3 + +
2 * + +
+ > 3 2
+ + ( 2
. / A

is a tridiagonal !atrix.
,o non3s&!are matrices ha"e diagonal entries?
?es, for a
n m
!atrix
. / A
, the diagonal entries are !! ! !
a a a a , ..., ,
1 , 1 22 11 where
@ , !inA n m ! =
.
Example
Bhat are the diagonal entries of

=
C . , ) . *
2 . 3 > . 2
, )
* 2 . 3
. / A
Sol!tion
-he diagonal ele!ents of
. / A
are . , and 2 . 3
22 11
= = a a
,iagonally ,ominant Matrix:
A
n n
s&uare !atrix
. / A
is a diagonally do!inant !atrix if

n
j i
j
ij ii
a a
1
D D
for all
n i ,....., 2 , 1 =
and

=
>
n
j i
j
ij ii
a a
1
D D
for at least one i ,
that is, for each row, the absolute value of the diagonal ele!ent is greater than or e&ual to the
su! of the absolute values of the rest of the ele!ents of that row, and that the ine&uality is
strictly greater than for at least one row. Eiagonally do!inant !atrices are i!portant in
ensuring convergence in iterative sche!es of solving si!ultaneous linear e&uations.
Example #
4ive exa!ples of diagonally do!inant !atrices and not diagonally do!inant !atrices.
)
Sol!tion

=
) 2 3
2 ( 2
, ) 1*
. / A

is a diagonally do!inant !atrix as
13 , ) 1* 1*
13 12 11
= + = + = = a a a
( 2 2 ( (
23 21 22
= + = + = = a a a
* 2 3 ) )
32 31 33
= + = + = = a a a
and for at least one row, that is 0ows 1 and 3 in this case, the ine&uality is a strictly greater
than ine&uality.

=
++1 . * 2 3
2 ( 2
> ) 1*
. /B
is a diagonally do!inant !atrix as
1* > ) 1* 1*
13 12 11
= + = + = = b b b
( 2 2 ( (
23 21 22
= + = + = = b b b
* 2 3 ++1 . * ++1 . *
32 31 33
= + = + = = b b b
-he ine&ualities are satisfied for all rows and it is satisfied strictly greater than for at least
one row 5in this case it is 0ow 36.
[ ]

=
1 12 1((
1 C )(
1 * 2*
C

is not diagonally do!inant as
)* 1 )( C C
23 21 22
= + = + = = c c c
Bhen are two !atrices considered to be e&ual:
-wo !atrices /A. and /B. are e&ual if the si1e of /A. and /B. is the sa!e 5nu!ber of rows and
colu!ns are sa!e for /A. and /B.6 and aij F bij for all i and j.
Example %
Bhat would !a"e

=
, )
3 2
. / A
to be e&ual to

=
22
11
)
3
. /
b
b
B
Sol!tion
-he two !atrices
. / A
and
. /B
should be e&ual if 2
11
= b and ,
22
= b .
,
4ey Terms:
"atrix
#ector
$ubmatrix
$quare matrix
%qual matrices
&ero matrix
'dentity matrix
iagonal matrix
(pper triangular matrix
)ower triangular matrix
*ri+diagonal matrix
iagonally dominant matrix
C