MATH 304

Linear Algebra
Lecture 6:
Diagonal matrices.
Inverse matrix.
Matrices
Denition. An m-by-n matrix is a rectangular
array of numbers that has m rows and n columns:
_
_
_
_
_
a
11
a
12
. . . a
1n
a
21
a
22
. . . a
2n
.
.
.
.
.
.
.
.
.
.
.
.
a
m1
a
m2
. . . a
mn
_
_
_
_
_
Notation: A = (a
ij
)
1i n, 1j m
or simply A = (a
ij
)
if the dimensions are known.
Matrix algebra: linear operations
Addition: two matrices of the same dimensions
can be added by adding their corresponding entries.
Scalar multiplication: to multiply a matrix A by
a scalar r , one multiplies each entry of A by r .
Zero matrix O: all entries are zeros.
Negative: A is dened as (1)A.
Subtraction: A B is dened as A + (B).
As far as the linear operations are concerned, the
mn matrices can be regarded as mn-dimensional
vectors.
Properties of linear operations
(A + B) + C = A + (B + C)
A + B = B + A
A + O = O + A = A
A + (A) = (A) + A = O
r (sA) = (rs)A
r (A + B) = rA + rB
(r + s)A = rA + sA
1A = A
0A = O
Matrix algebra: matrix multiplication
The product of matrices A and B is dened if the
number of columns in A matches the number of
rows in B.
Denition. Let A = (a
ik
) be an mn matrix and
B = (b
kj
) be an np matrix. The product AB is
dened to be the mp matrix C = (c
ij
) such that
c
ij
=

n
k=1
a
ik
b
kj
for all indices i , j .
That is, matrices are multiplied row by column.
A =
_
_
_
_
_
a
11
a
12
. . . a
1n
a
21
a
22
. . . a
2n
.
.
.
.
.
.
.
.
.
.
.
.
a
m1
a
m2
. . . a
mn
_
_
_
_
_
=
_
_
_
_
_
v
1
v
2
.
.
.
v
m
_
_
_
_
_
B =
_
_
_
_
_
b
11
b
12
. . . b
1p
b
21
b
22
. . . b
2p
.
.
.
.
.
.
.
.
.
.
.
.
b
n1
b
n2
. . . b
np
_
_
_
_
_
= (w
1
, w
2
, . . . , w
p
)
= AB =
_
_
_
_
_
v
1
w
1
v
1
w
2
. . . v
1
w
p
v
2
w
1
v
2
w
2
. . . v
2
w
p
.
.
.
.
.
.
.
.
.
.
.
.
v
m
w
1
v
m
w
2
. . . v
m
w
p
_
_
_
_
_
Any system of linear equations can be represented
as a matrix equation:
_

_
a
11
x
1
+ a
12
x
2
+ + a
1n
x
n
= b
1
a
21
x
1
+ a
22
x
2
+ + a
2n
x
n
= b
2

a
m1
x
1
+ a
m2
x
2
+ + a
mn
x
n
= b
m
Ax = b,
where
A =
_
_
_
_
_
a
11
a
12
. . . a
1n
a
21
a
22
. . . a
2n
.
.
.
.
.
.
.
.
.
.
.
.
a
m1
a
m2
. . . a
mn
_
_
_
_
_
, x =
_
_
_
_
_
x
1
x
2
.
.
.
x
n
_
_
_
_
_
, b =
_
_
_
_
_
b
1
b
2
.
.
.
b
m
_
_
_
_
_
.
Properties of matrix multiplication:
(AB)C = A(BC) (associative law)
(A + B)C = AC + BC (distributive law #1)
C(A + B) = CA + CB (distributive law #2)
(rA)B = A(rB) = r (AB)
Any of the above identities holds provided that
matrix sums and products are well dened.
If A and B are nn matrices, then both AB and BA
are well dened nn matrices.
However, in general, AB = BA.
Example. Let A =
_
2 0
0 1
_
, B =
_
1 1
0 1
_
.
Then AB =
_
2 2
0 1
_
, BA =
_
2 1
0 1
_
.
If AB does equal BA, we say that the matrices A
and B commute.
Problem. Let A and B be arbitrary nn matrices.
Is it true that (A B)(A + B) = A
2
B
2
?
(A B)(A + B) = (A B)A + (A B)B
= (AA BA) + (AB BB)
= A
2
+ AB BA B
2
Hence (A B)(A + B) = A
2
B
2
if and only if
A commutes with B.
Diagonal matrices
If A = (a
ij
) is a square matrix, then the entries a
ii
are called diagonal entries. A square matrix is
called diagonal if all non-diagonal entries are zeros.
Example.
_
_
7 0 0
0 1 0
0 0 2
_
_
, denoted diag(7, 1, 2).
Let A = diag(s
1
, s
2
, . . . , s
n
), B = diag(t
1
, t
2
, . . . , t
n
).
Then A + B = diag(s
1
+ t
1
, s
2
+ t
2
, . . . , s
n
+ t
n
),
rA = diag(rs
1
, rs
2
, . . . , rs
n
).
Example.
_
_
7 0 0
0 1 0
0 0 2
_
_
_
_
1 0 0
0 5 0
0 0 3
_
_
=
_
_
7 0 0
0 5 0
0 0 6
_
_
Theorem Let A = diag(s
1
, s
2
, . . . , s
n
),
B = diag(t
1
, t
2
, . . . , t
n
).
Then A + B = diag(s
1
+ t
1
, s
2
+ t
2
, . . . , s
n
+ t
n
),
rA = diag(rs
1
, rs
2
, . . . , rs
n
).
AB = diag(s
1
t
1
, s
2
t
2
, . . . , s
n
t
n
).
In particular, diagonal matrices always commute.
Example.
_
_
7 0 0
0 1 0
0 0 2
_
_
_
_
a
11
a
12
a
13
a
21
a
22
a
23
a
31
a
32
a
33
_
_
=
_
_
7a
11
7a
12
7a
13
a
21
a
22
a
23
2a
31
2a
32
2a
33
_
_
Theorem Let D = diag(d
1
, d
2
, . . . , d
m
) and A be
an mn matrix. Then the matrix DA is obtained
from A by multiplying the i th row by d
i
for
i = 1, 2, . . . , m:
A =
_
_
_
_
_
v
1
v
2
.
.
.
v
m
_
_
_
_
_
= DA =
_
_
_
_
_
d
1
v
1
d
2
v
2
.
.
.
d
m
v
m
_
_
_
_
_
Example.
_
_
a
11
a
12
a
13
a
21
a
22
a
23
a
31
a
32
a
33
_
_
_
_
7 0 0
0 1 0
0 0 2
_
_
=
_
_
7a
11
a
12
2a
13
7a
21
a
22
2a
23
7a
31
a
32
2a
33
_
_
Theorem Let D = diag(d
1
, d
2
, . . . , d
n
) and A be
an mn matrix. Then the matrix AD is obtained
from A by multiplying the i th column by d
i
for
i = 1, 2, . . . , n:
A = (w
1
, w
2
, . . . , w
n
)
= AD = (d
1
w
1
, d
2
w
2
, . . . , d
n
w
n
)
Identity matrix
Denition. The identity matrix (or unit matrix) is
a diagonal matrix with all diagonal entries equal to 1.
The nn identity matrix is denoted I
n
or simply I .
I
1
= (1), I
2
=
_
1 0
0 1
_
, I
3
=
_
_
1 0 0
0 1 0
0 0 1
_
_
.
In general, I =
_
_
_
1 0 . . . 0
0 1 . . . 0
.
.
.
.
.
.
.
.
.
.
.
.
0 0 . . . 1
_
_
_
.
Theorem. Let A be an arbitrary mn matrix.
Then I
m
A = AI
n
= A.
Inverse matrix
Let M
n
(R) denote the set of all nn matrices with
real entries. We can add, subtract, and multiply
elements of M
n
(R). What about division?
Denition. Let A M
n
(R). Suppose there exists
an nn matrix B such that
AB = BA = I
n
.
Then the matrix A is called invertible and B is
called the inverse of A (denoted A
1
).
A non-invertible square matrix is called singular.
AA
1
= A
1
A = I
Examples
A =
_
1 1
0 1
_
, B =
_
1 1
0 1
_
, C =
_
1 0
0 1
_
.
AB =
_
1 1
0 1
__
1 1
0 1
_
=
_
1 0
0 1
_
,
BA =
_
1 1
0 1
__
1 1
0 1
_
=
_
1 0
0 1
_
,
C
2
=
_
1 0
0 1
__
1 0
0 1
_
=
_
1 0
0 1
_
.
Thus A
1
= B, B
1
= A, and C
1
= C.
Basic properties of inverse matrices:
If B = A
1
then A = B
1
. In other words, if A
is invertible, so is A
1
, and A = (A
1
)
1
.
The inverse matrix (if it exists) is unique.
Moreover, if AB = CA = I for some matrices
B, C M
n
(R) then B = C = A
1
.
Indeed, B = IB = (CA)B = C(AB) = CI = C.
If matrices A, B M
n
(R) are invertible, so is
AB, and (AB)
1
= B
1
A
1
.
(B
1
A
1
)(AB) = B
1
(A
1
A)B = B
1
IB = B
1
B = I ,
(AB)(B
1
A
1
) = A(BB
1
)A
1
= AIA
1
= AA
1
= I .
Similarly, (A
1
A
2
. . . A
k
)
1
= A
1
k
. . . A
1
2
A
1
1
.
Other examples
D =
_
0 1
0 0
_
, E =
_
1 1
1 1
_
.
D
2
=
_
0 1
0 0
__
0 1
0 0
_
=
_
0 0
0 0
_
.
It follows that D is not invertible as otherwise
D
2
= O = D
1
D
2
= D
1
O = D = O.
E
2
=
_
1 1
1 1
__
1 1
1 1
_
=
_
2 2
2 2
_
= 2E.
It follows that E is not invertible as otherwise
E
2
= 2E = E
2
E
1
= 2EE
1
= E = 2I .