Matrices

Matrix: A matrix is a rectangular arrangement of numbers and alphabets written in

rows and columns, denoted by capital letters A, B, C…etc. The size of a matrix is denoted by
listing the number of rows followed by the number of columns. Matrix B is of order 3 × 4; it
has three rows and four columns. The number of rows is always stated first.

3 −1 2 0 3
1 3
Example: 𝐴 = 𝐵 = 10 15 21 −4 𝐶= 1
5 8 ×
4 0 −56 −9 × 5 ×

This convention lets us easily refer to a matrix using the letter assigned, here A,B.
The numerical values within a matrix are called elements or entries of the matrix. Each
matrix entry has an address/location given by the number of the row and the number of
the column in which it appears. Entries of a matrix are often denoted by a lower case letter
of its assigned name with a pair of subscripts denoting the row, column position of the
entry. We simply represent a matrix by writing 𝐴 = 𝑎 where 1 ≤ 𝑖 ≤ 𝑚, 1 ≤ 𝑗 ≤ 𝑛.
×

Note: Matrix is just a mathematical representation of numbers or alphabets and it has no

value.

Types of Matrices:

Let A be any matrix of matrix of order 𝑚 × 𝑛, then we say that A is a

a) Row matrix if it contains only one row. Example, 𝐴 = [1 3 5 2]

1
b) Column matrix if it contains only one column. Example, 𝐴 = 3
8
0 0
c) Zero matrix if all of its entries are zero. Example, 𝐵 =
0 0
d) Square matrix if it has same number of rows and columns, that is m=n.
e) The entries 𝑎 are called diagonal entries. A is an upper triangular matrix if all the
entries below the diagonal are zero and it is called a lower triangular matrix if entries
above the diagonal are zero.
f) Diagonal matrix if all non-diagonal entries are zero. If all the diagonal entries are equal
to one, it is called an identity matrix denoted by 𝑰𝒏×𝒏 . If all diagonal entries equal to 𝑘, it
is called a scalar matrix.

Inverse of a matrix: Let 𝑨 be any square matrix of order 𝑛 × 𝑛. A square matrix 𝑩 of order
𝑛 × 𝑛 is called an inverse of 𝑨 if 𝑨𝑩 = 𝑩𝑨 = 𝑰𝒏×𝒏 . Since B represents inverse of A, we shall
use notation 𝐴 to denote it.

Transpose of a matrix: Let A be any matrix, the matrix 𝐴 obtained by interchanging

rows and columns is called a transpose matrix of 𝐴.
 1 −2 5 1 6 73
Example: Let 𝐴 = 6 5 85 then 𝐴 = −2 5 −15
73 −15 0 5 85 0
Note: Let A be any rectangular matrix of order 𝑚 × 𝑛, then the transpose matrix will be of
order 𝑛 × 𝑚, and geometrically transpose of a matrix represents rotation of original matrix
by an angle 90 . Rotating again, we get original matrix, that is (𝐴 ) = 𝐴. Next we shall
consider some special class of graphs depending on the transpose of a matrix.

Symmetric and skew-symmetric matrices

Definition: Let A be any square matrix of order 𝑛 × 𝑛. We say that A is a symmetric matrix
if it is equal to its transpose. Formally, a matrix A is symmetric if 𝐴 = 𝐴 . Since equal
matrices have equal dimensions, only square matrices can be symmetric. The entries of a
symmetric matrix are symmetric with respect to the main diagonal since 𝒂𝒊𝒋 = 𝒂𝒋𝒊 for
all i and j. The following matrices are symmetric:
1 7 3 1 3 14
7 3
Examples: 𝐴 = 7 4 −5 B= 3 4 5 𝐶=
3 −5
3 −5 6 14 5 6
Note: If A and B are symmetric matrices of same order and k is any scalar, then

i. 𝐴 is symmetric
ii. 𝑘𝐴 is symmetric
iii. 𝐴 + 𝐵 and 𝐴 − 𝐵 are symmetric
iv. 𝐴𝐵 is symmetric if 𝐴𝐵 = 𝐵𝐴
v. If A is invertible, then 𝐴 is also symmetric.

Definition: Let A be any square matrix, then we say that A is a skew-symmetric matrix if it
is equal to negative of its transpose. Formally, matrix A is skew-symmetric if 𝐴 = −𝐴 .
Because equal matrices have equal dimensions, only square matrices can be skew-
symmetric. The entries of a skew-symmetric matrix are skew-symmetric with respect
to the main diagonal since 𝒂𝒊𝒋 = −𝒂𝒋𝒊 for all 𝒊 and 𝒋. In particular, 𝑎 = −𝑎 implies
that 𝑎 = 0, for all 𝑖. The following matrices are skew-symmetric:

0 −7 3 0 −3 −14
0 3
Examples: 𝐴 = 7 0 −5 B= 3 0 5 𝐶=
3 0
−3 5 0 14 −5 0
1. Every square diagonal matrix is symmetric, since all off-diagonal entries are zero.
Similarly, each diagonal element of a skew-symmetric matrix must be zero, since
each is its own negative.
2. Let A and B are any two skew-symmetric matrices of same order, then
a. 𝐴 , 𝑘𝐴 are all skew-symmetric.
b. 𝐴 + 𝐵 is skew-symmetric
c. AB is skew-symmetric 𝐴𝐵 = −𝐵𝐴.
 d. For any matrix 𝑃 of order 𝑚 × 𝑛, the matrix 𝑃 𝐴𝑃 is skew-symmetric.

Remark:

1. If A is any square matrix of order 𝑛, then 𝐴 + 𝐴 is symmetric and 𝐴 − 𝐴 is skew-

symmetric.
2. Every square matrix can be written as a sum of symmetric and skew-symmetric
matrix.

Let A be any square matrix of order n. We know that A + A and scalar multiple of a
symmetric are all symmetric. On the other hand A − A and scalar multiple of skew -
symmetric matrix are skew-symmetric. Hence, put C = which is real symmetric and

D= is skew-symmetric. Thus A = C + D where C is symmetric and D is skew-

symmetric gives representation of A as a sum of symmetric and skew-symmetric matrix.

Orthogonal Matrices: A matrix A is orthogonal if its transpose is equal to its inverse:

𝐴 = 𝐴 which requires 𝐴 𝐴 = 𝐴𝐴 = 𝐼, where I is the identity matrix.

⎡ ⎤
𝑐𝑜𝑠𝜃 −𝑠𝑖𝑛𝜃 ⎢ ⎥
Examples: 𝑨 = 𝑩= √ √
𝑪 = ⎢− ⎥
𝑠𝑖𝑛𝜃 𝑐𝑜𝑠𝜃 −
√ √ ⎢ ⎥
⎣ − ⎦

Normal Matrix: A square matrix A is said to be normal if it commutes with its transpose.
That is 𝐴𝐴 = 𝐴 𝐴. Clearly, every skew-symmetric matrix and any orthogonal,
symmetric matrix will be normal.

Note:
a. An orthogonal matrix A is necessarily invertible (with inverse 𝐴 = 𝐴 ), therefore
it is a normal matrix always, that is 𝐴𝐴 = 𝐴 𝐴.
b. Inverse of an orthogonal matrix is orthogonal.
c. Product of orthogonal matrices is orthogonal.
d. The determinant of any orthogonal matrix is either +1 or 1.

Solved Problems:
2 𝑥 𝑧
1. Find 𝑥, 𝑦 𝑎𝑛𝑑 𝑧 such that the matrix 𝐴 = 4 5 𝑦 is symmetric.
𝑧 1 7
2 𝑥 3
Solution: Given 𝐴 = 4 5 𝑦 is a symmetric matrix implying that entries of the matrix
𝑧 1 7
are symmetric along the diagonals. Thus we must have 𝑥 = 4, 𝑦 = 1 and since we have
𝑧 = 𝑧 comparing the entries, z can take any value.
2. Show that for any square matrix A, the matrix 𝐴 − 𝐴 is skew-symmetric matrix.
Solution: Let 𝐴 be any square matrix. Put 𝐵 = 𝐴 − 𝐴 , a square matrix. Now we shall
check B for the condition:
We have 𝐵 = (𝐴 − 𝐴 ) = 𝐴 − 𝐴 = −(𝐴 − 𝐴 ) = −𝐵. Hence, we get 𝐵 = −𝐵,
proving that B is a skew-symmetric matrix.

3. If A is a symmetric matrix, prove that 𝐴 is also symmetric.

Solution: Let A be any symmetric matrix. By definition, we shall have 𝐴 = 𝐴. Put
𝐵 = 𝐴 , then 𝐵 = (𝐴𝐴) = (𝐴 ) = 𝐴 = 𝐵. Therefore, we have 𝐵 = 𝐵 and so B is a
symmetric matrix.
7 −6 2𝑥
4. Find 𝑥, 𝑦 𝑎𝑛𝑑 𝑧 such that the matrix 𝐴 = 𝑦 𝑧 −2 is skew-symmetric.
𝑥 −2 5
7 −6 2𝑥
Solution: Given 𝐴 = 𝑦 𝑧 −2 is a symmetric matrix implying that entries of the
𝑥 −2 5
matrix are symmetric along the diagonals. Thus we must have 𝑦 = −6, 2𝑥 = 𝑥, which is
true only if 𝑥 = 0. Since we have 𝑧 = 𝑧 comparing the entries, z can take any value.

3 0 −1
5. Express 𝐴 = 1 2 1 as a sum of symmetric and skew symmetric matrix.
2 −1 3
3 0 −1 3 1 2
Solution: Consider the given matrix 𝐴 = 1 2 1 , then 𝐴 = 0 2 −1
2 −1 3 −1 1 3
⎡3 ⎤
⎢ ⎥
Put 𝐶 = , then 𝐶 = =⎢ 2 0⎥ is a symmetric matrix and the matrix
⎢ ⎥
⎣ 0 3⎦

⎡0 − − ⎤
⎢ ⎥
𝐷= =⎢ 0 ⎥ is skew symmetric matrix. Further, we have
⎢ ⎥
⎣ − 0⎦

⎡3 ⎤ ⎡0 − − ⎤
3 0 −1
⎢ ⎥ ⎢ ⎥
𝐶+𝐷 = + = ⎢ 2 0⎥ + ⎢ 0 ⎥= 1 2 1 = 𝐴represents 𝐴 as
⎢ ⎥ ⎢ ⎥ 2 −1 3
⎣ 0 3⎦ ⎣ − 0⎦
a sum of symmetric and skew–symmetric matrices.

1 2 3
6. Express 𝐴 = 4 5 6 as a sum of symmetric and skew symmetric matrix.
7 0 0
Solution: Same as above
 1 5 3
7. Express the matrix 𝐴 = 7 9 0 as a sum of a symmetric and a skew-symmetric
3 1 1
matrix.
Solution: Same as problem 5.
⎡ − ⎤
⎢ ⎥
8. Show that 𝐴 = ⎢ − − ⎥ is orthogonal.
⎢ ⎥
⎣ ⎦
⎡ − ⎤
⎢ ⎥
Solution: Given 𝐴 = ⎢ − − ⎥ and we need to prove 𝐴 is orthogonal matrix. That is
⎢ ⎥
⎣ ⎦
we need to show that 𝐴 = 𝐴 . In fact we may use the hint that 𝐴𝐴 = 𝐼 where 𝐼
⎡ ⎤
⎢ ⎥
represents an identity matrix of order 3 × 3. Now, 𝐴 = ⎢ − ⎥.
⎢ ⎥
⎣ − ⎦
Let us consider the product of 𝐴 with its transpose as follows
1 8 4 1 4 8
⎡ − ⎤⎡ ⎤
⎢9 9 9⎥ ⎢ 9 9 9⎥
4 4 7 8 4 1⎥
𝐴𝐴 = ⎢ − − ⎥⎢ −
⎢9 9 9⎥ ⎢ 9 9 9⎥
⎢ 8 1 4 ⎥⎢ −4 7 4⎥
⎣9 9 −
9 ⎦⎣ 9 9 9⎦
1 81 0 0 1 0 0
⟹ 𝐴𝐴 = 0 81 0 = 0 1 0 = 𝐼
81
0 0 81 0 0 1
Therefore, 𝐴𝐴 = 𝐼 and hence A is an orthogonal matrix.

1 0 −1
9. Prove that 𝐴 = 0 1 2 is a normal matrix.
1 −2 1
1 0 −1
Solution: Consider the given matrix 𝐴 = 0 1 2 .
1 −2 1
To prove that A is normal, we need to show that A commutes with its transpose. In
symbols, 𝐴𝐴 = 𝐴 𝐴.
1 0 1
The transpose of A is obtained A as 𝐴 = 0 1 −2 . Then we have
−1 2 1
1 0 −1 1 0 1 2 −2 0
𝐴𝐴 = 0 1 2 0 1 −2 = −2 5 0
1 −2 1 −1 2 1 0 0 6
And
 1 0 1 1 0 −1 2 −2 0
𝐴 𝐴 = 0 1 −2 0 1 2 = −2 5 0
−1 2 1 1 −2 1 0 0 6
Therefore A and 𝐴 commute each other and so the matrix A is normal.

Complex matrices
Let A be any square matrix. We say that A is a complex matrix if all its entries are
complex numbers. That is, entries of 𝐴 are of the form 𝑎 + 𝑖𝑏 where a, b are real numbers.
The conjugate of 𝑎 + 𝑖𝑏 is obtained by replacing 𝑖 by −𝑖 that is the number 𝑎 − 𝑖𝑏.
A matrix denoted by 𝐴̅ is called a conjugate matrix of 𝐴 if entries of 𝐴̅ are obtained by
replacing entries of A by their conjugates.

2+𝑖 7−𝑖 6 + 5𝑖 2−𝑖 7+𝑖 6 − 5𝑖

Example: 𝑨 = 1+𝑖 0 8𝑖 𝑩= 1−𝑖 0 −8𝑖
5 + 2𝑖 2𝑖 1 − 5𝑖 5 − 2𝑖 −2𝑖 1 + 5𝑖
Here A and B are complex matrices and B is obtained from A by replacing entries of A by
their conjugates. Hence B is a conjugate matrix of A.

The matrix operations like conjugate and transpose are commutative and so the conjugate
transpose of a matrix may be defined. For any square matrix A the matrix obtained by
taking the transpose of a conjugate matrix is called conjugate transpose of A, denoted by
𝐴∗ or 𝐴 .

1−𝑖 4 + 3𝑖 2 − 9𝑖 1+𝑖 −4𝑖 2 − 3𝑖

Example: Let 𝐴 = 4𝑖 5 + 3𝑖 8𝑖 then 𝐴∗
= 4 − 3𝑖 5 − 3𝑖 7+𝑖
2 + 3𝑖 7−𝑖 8+𝑖 2 + 9𝑖 −8𝑖 8−𝑖
Hermitian and Unitary Matrices:

Definition: A Hermitian matrix (or self-adjoint matrix) is a square matrix which is equal to
its own conjugate transpose. If the conjugate transpose of a matrix A is denoted by 𝐴∗ ,
called ’A star’, then the Hermitian property can be written concisely as A = 𝐴∗ .

1 𝑖 1+𝑖 1 −𝑖 1−𝑖
Example: Let 𝐴 = −𝑖 −5 2 − 𝑖 then 𝐴 = 𝑖 −5 2+𝑖
1−𝑖 2+𝑖 3 1+𝑖 2−𝑖 3
1 𝑖 1+𝑖
Taking transpose of conjugate matrix, we get 𝐴 = −𝑖 ∗
−5 2 − 𝑖 which coincides
1−𝑖 2+𝑖 3
with the given matrix A. That is 𝐴 = 𝐴 and so the matrix is Hermitian. Same is true for
∗

matrices mentioned below.

3 3 − 2𝑖 7
2 1+𝑖 0 𝑖
a) 𝐵 = b) 𝐶 = c) 𝐴 = 3 + 2𝑖 −2 1+𝑖
1−𝑖 3 𝑖 2
7 1 − 5𝑖 6
Note: From the definition it is clear that 𝐴 is Hermitian if 𝑎 = 𝑎 . In particular, 𝑎 = 𝑎
which is true only when 𝑎 ′𝑠 are all real number. In other word, diagonal entries of an
Hermitian matrix are real numbers.

Definition: A skew-Hermitian matrix is a square matrix which is equal to negative of its

own conjugate transpose. That is, a square matrix 𝐴 is skew-Hermitian if 𝐴 = −𝐴∗ .

0 3 4 + 5𝑖
𝑖 1−𝑖
Example: a) 𝐴 = −3 0 −2𝑖 b) 𝐵 =
−1 − 𝑖 0
−4 + 5𝑖 −2𝑖 0
Note: From the definition it is clear that 𝐴 is skew-Hermitian if 𝑎 = −𝑎 . In particular,
𝑎 = −𝑎 which is true only when 𝑎 ′𝑠 are all zeros. In other word, diagonal entries of a
skew - Hermitian matrix are zeros.

Properties:

1. The sum of a square matrix and its conjugate transpose 𝐴 + 𝐴 is Hermitian.

2. If A is any Hermitian matrix then 𝑖𝐴 is skew-Hermitian and vice-versa.
3. The Eigen values of any Hermitian matrix are real and those of a skew-Hermitian
matrix are purely imaginary or zero. Eigen values of orthogonal and unitary
matrices have absolute value 1.
4. If 𝐴 is any square matrix if 𝐴 + 𝐴∗ , 𝐴∗ 𝐴, 𝐴𝐴∗ are all Hermitian and 𝐴 − 𝐴∗ is skew-
Hermitian.
5. 𝐴 , 𝐴 , 𝑘𝐴 are all unitary matrices if A is unitary.
6. Product of two unitary matrices is unitary.
7. Any square matrix can be expressed as 𝑃 + 𝑖𝑄 where P and Q are symmetric
matrices.
8. The difference of a square matrix and its conjugate transpose 𝐴 − 𝐴 is skew-
Hermitian (also called anti Hermitian,𝐴 = −𝐴∗ )
9. An arbitrary square matrix C can be written as the sum of a Hermitian matrix A
and a skew-Hermitian matrix B: C = A + B with 𝐴 = and𝐵 = .

Theorem: Every Hermitian matrix A can be written as 𝐴 = 𝑃 + 𝑖𝑄 where P is real

symmetric and Q is real skew-symmetric.

Proof: Let A be any Hermitian matrix, then by definition 𝐴 = 𝐴̅ . Equivalently, taking

transpose on both sides, we get 𝐴 = 𝐴̅. Now, we let 𝐵 = (𝐴 + 𝐴 ) and 𝐶 = (𝐴 − 𝐴̅ ).

We claim that B is real symmetric and C is real skew-symmetric. We first prove that
𝐵 = 𝐵 and 𝐶 = 𝐶̅ :

Consider 𝐵 = (𝐴 + 𝐴̅ ) = ( 𝐴̅ + 𝐴) = 𝐵, therefore B is real matrix, similarly C is real.

Next, we shall prove that B is symmetric and C is skew-symmetric. We have

𝐵 = (𝐴 + 𝐴 ) = (𝐴 + 𝐴̅ ) = (𝐴 + 𝐴 ) = 𝐵, (Since A is Hermitian, 𝐴 = 𝐴̅ ).
Therefore, B is symmetric. Similarly, we have the following:

𝐶 = (𝐴 − 𝐴̅ ) = (𝐴̅ − 𝐴) = − (𝐴 − 𝐴̅) = −𝐶. This establishes that C is skew-

symmetric matrix. Therefore, every Hermitian matrix can be expressed as a sum of real
symmetric and skew-symmetric matrix.

Unitary matrices: A complex square matrix A is unitary if 𝐴∗ 𝐴 = 𝐴𝐴∗ = 𝐼 or 𝐴 =

𝐴∗ where 𝐼 denotes an identity matrix is and 𝐴∗ is the conjugate transpose of A.

− −
√ √
Example: a) 𝐴 = b) 𝐵 = c) 𝐴 =
− − √ √
√ √

Normal Matrix:

2 + 2𝑖 𝑖 1−𝑖
Example: 𝐴 = 𝑖 −2𝑖 1 − 3𝑖
1−𝑖 1 − 3𝑖 −3 + 8𝑖

Solved Problems on Hermitian and skew-Hermitian matrices:

3 𝑥 + 2𝑖 𝑦𝑖
1. Find x, y, z such that 𝐴 = 3 − 2𝑖 0 1 + 𝑧𝑖 is Hermitian matrix.
𝑦𝑖 1 − 𝑥𝑖 −1
3 𝑥 + 2𝑖 𝑦𝑖
Solution: Consider the given matrix 𝐴 = 3 − 2𝑖 0 1 + 𝑧𝑖 . Since the given
𝑦𝑖 1 − 𝑥𝑖 −1
matrix is Hermitian, it has to satisfy the condition 𝑎 = 𝑎 . Comparing the entries,
we have the following: 3 − 2𝑖 = 𝑥 + 2𝚤 = 𝑥 − 2𝑖, we get 𝑥 = 3. Next, 1 − 𝑥𝑖 = 1 + 𝑧𝚤
implies that 𝑥 = 𝑧 = 3 and as we have 𝑦𝑖 = −𝑦𝑖, we must have 𝑦 = 0.

−1 2+𝑖 5 − 3𝑖
2. If 𝐴 = 2 − 𝑖 7 5𝑖 verify that 𝑖𝐴 is skew-Hermitian matrix.
5 + 3𝑖 −5𝑖 2
−1 2 + 𝑖 5 − 3𝑖
Solution: Consider the given matrix 𝐴 = 2 − 𝑖 7 5𝑖 .
5 + 3𝑖 −5𝑖 2
−𝑖 2𝑖 − 1 5𝑖 + 3
Multiplying it by 𝑖, we get 𝑖𝐴 = 2𝑖 + 1 7𝑖 −5 = 𝐵 (say).
5𝑖 − 3 5 2𝑖
 𝑖 −2𝑖 − 1 −5𝑖 + 3
The conjugate matrix is given by 𝐵 = −2𝑖 + 1 −7𝑖 −5
−5𝑖 − 3 5 −2𝑖
Taking the transpose of B, we get
𝑖 −2𝑖 + 1 −5𝑖 − 3 −𝑖 2𝑖 − 1 5𝑖 − 3
𝐵 ∗ = −2𝑖 − 1 7𝑖 5 = −1 2𝑖 − 1 7𝑖 −5 = −𝐵
−5𝑖 + 3 −5 −2𝑖 5𝑖 − 3 −5 2𝑖
Therefore, we have 𝐵 ∗ = 𝐵 implies that 𝐵 = 𝑖𝐴 is skew-Hermitian.

−𝑖 −1 + 2𝑖 3 + 5𝑖
3. If 𝐴 = 1 + 2𝑖 7𝑖 −5 verify that 𝑖𝐴 is a Hermitian matrix.
−3 + 5𝑖 5 2𝑖
Solution: Same as above

2+𝑖 3 −1 + 3𝑖
4. If 𝐴 = , verify that 𝐴∗ 𝐴 is a Hermitian matrix.
−5 𝑖 4 − 2𝑖
2 + 𝑖 3 −1 + 3𝑖
Solution: Consider the given matrix 𝐴 =
−5 𝑖 4 − 2𝑖
2+𝑖 −5
The transpose matrix is 𝐴 = 3 𝑖 .
−1 + 3𝑖 4 − 2𝑖
2−𝑖 −5
The transpose conjugate of A is given by 𝐴 = ∗
3 −𝑖
−1 − 3𝑖 4 + 2𝑖
2−𝑖 −5
2 + 𝑖 3 −1 + 3𝑖
Now 𝐴∗ 𝐴 = 3 −𝑖
−5 𝑖 4 − 2𝑖
−1 − 3𝑖 4 + 2𝑖
30 6 − 8𝑖 −19 + 17𝑖
= 6 + 8𝑖 10 −5 + 5𝑖 = 𝐵 (say)
−19 − 17𝑖 −5 − 5𝑖 30
30 6 + 8𝑖 −19 − 17𝑖
The conjugate matrix of B is 𝐵 = 6 − 8𝑖 10 −5 − 5𝑖
−19 − 17𝑖 −5 − 5𝑖 30
30 6 − 8𝑖 −19 + 17𝑖
The conjugate transpose of 𝐵 is 𝐵 ∗ = 6 + 8𝑖 10 −5 + 5𝑖 = 𝐵
−19 − 17𝑖 −5 − 5𝑖 30
Thus we have by definition, 𝐵 = 𝐴∗ 𝐴 is a Hermitian matrix.

2𝑖 2+𝑖 4−𝑖
5. Express 𝐴 = 3 + 𝑖 −𝑖 2𝑖 as sum of a Hermitian and skew-Hermitian
5−𝑖 3𝑖 0
matrix.
2𝑖 2+𝑖 4−𝑖
Solution: Consider the given matrix 𝐴 = 3 + 𝑖 −𝑖 2𝑖 .
5−𝑖 3𝑖 0
−2𝑖 3 − 𝑖 5 + 𝑖
The conjugate transpose of 𝐴 is given by 𝐴∗ = 2 − 𝑖 𝑖 −3𝑖 .
4 + 𝑖 −2𝑖 0
 0 5 9
Put 𝐵 = (𝐴 + 𝐴 ∗)
= 5 0 −𝑖 which is a Hermitian matrix. Similarly, we let
9 𝑖 0
4𝑖 −1 + 2𝑖 −1 − 2𝑖
𝐶 = (𝐴 − 𝐴 ) = 1 + 2𝑖 −2𝑖 5𝑖 is skew-Hermitian. Further, we
1 − 2𝑖 5𝑖 0
have
5 9 1 1
⎡0 ⎤ ⎡ 2𝑖 − + 𝑖 − − 𝑖⎤
⎢ 2 2⎥ ⎢ 2 2 ⎥
5 𝑖 1 5𝑖 ⎥ 2𝑖 2+𝑖 4−𝑖
𝐵+𝐶 = ⎢ 0 − + +𝑖 ⎥ ⎢ −𝑖 = 3+𝑖 −𝑖 2𝑖 = 𝐴
⎢2 2⎥ ⎢2 2 ⎥ 5−𝑖 3𝑖 0
⎢9 𝑖 ⎥ ⎢1 5𝑖 ⎥
⎣2 2 0 ⎦ ⎣2 − 𝑖 0 ⎦
2
Therefore, 𝐴 = 𝐵 + 𝐶, where B is Hermitian and C is skew-Hermitian.

−1 2 + 𝑖 5 − 3𝑖
6. Express 𝐴 = 2 − 𝑖 7 5𝑖 in the form 𝑃 + 𝑖𝑄 where P is real symmetric
5 + 3𝑖 −5𝑖 2
and Q are real skew symmetric matrix..
−1 2 + 𝑖 5 − 3𝑖
Solution: Consider the given matrix 𝐴 = 2 − 𝑖 7 5𝑖 . Conjugate of A
5 + 3𝑖 −5𝑖 2
−1 2 − 𝑖 5 + 3𝑖
is 𝐴̅ = 2 + 𝑖 7 −5𝑖 . Since we need express A as a sum of symmetric and
5 − 3𝑖 5𝑖 2
skew-symmetric matrix, we let 𝑃 = (𝐴 + 𝐴̅) given by
−1 2 + 𝑖 5 − 3𝑖 −1 2 − 𝑖 5 + 3𝑖
𝑃= 2−𝑖 7 5𝑖 + 2+𝑖 7 −5𝑖
5 + 3𝑖 −5𝑖 2 5 − 3𝑖 5𝑖 2
−1 2 5
That is, 𝑃 = 2 7 0 , which is a real and symmetric matrix.
5 0 2
Similarly, we let 𝑄 = (𝐴 − 𝐴̅) which is real skew-symmetric given by

1 −1 2+𝑖 5 − 3𝑖 1 −1 2−𝑖 5 + 3𝑖
𝑄= 2−𝑖 7 5𝑖 − 2+𝑖 7 −5𝑖
2𝑖 2𝑖
5 + 3𝑖 −5𝑖 2 5 − 3𝑖 5𝑖 2
0 2𝑖 −6𝑖 0 1 −3
𝑄= −2𝑖 0 10𝑖 = −1 0 5 which is skew-symmetric.
6𝑖 −10𝑖 0 3 −5 0
−1 2 5 0 1 −3 −1 2 + 𝑖 5 − 3𝑖
Finally, 𝑃 + 𝑖𝑄 = 2 7 0 + 𝑖 −1 0 5 = 2−𝑖 7 5𝑖 = 𝐴,
5 0 2 3 −5 0 5 + 3𝑖 −5𝑖 2
which represents A as a sum of real symmetric and skew-symmetric matrices.

7. Show that the following matrices are unitary.

 √
− 𝑖 −
a. 𝐴 = c. 𝐶 =
√
− − − 𝑖 −
1 −𝑖 −1 + 𝑖
b. 𝐵 = 𝑖 1 1+𝑖
1+𝑖 −1 + 𝑖 0

Solution:

− 𝑖
a) Given the matrix 𝐴 =
− − − 𝑖

+ 𝑖 −
Conjugate transpose of the matrix A is given by 𝐴∗ = .
− + 𝑖

− 𝑖 + 𝑖 − 1 0
Now, 𝐴𝐴∗ = = , hence A is unitary.
− − − 𝑖 − + 𝑖 0 1

1 −𝑖 −1 + 𝑖
b) Given the matrix 𝐵 = 𝑖 1 1+𝑖
1+𝑖 −1 + 𝑖 0
1 −𝑖 −1 − 𝑖
Conjugate transpose of B is given by 𝐵 ∗ = 𝑖 1 −1 − 𝑖 .
−1 − 𝑖 1−𝑖 0
1 −𝑖 −1 + 𝑖 1 −𝑖 −1 − 𝑖 1 0 0
Now, 𝐵𝐵 = ∗
𝑖 1 1+𝑖 𝑖 1 −1 − 𝑖 = 0 1 0
1+𝑖 −1 + 𝑖 0 −1 − 𝑖 1−𝑖 0 0 0 1
Hence B is unitary.

c) Similar to above.

0 1 + 2𝑖
8. Given that 𝐴 = , show that (𝐼 − 𝐴)(𝐼 + 𝐴) is a unitary matrix.
−1 + 2𝑖 0
0 1 + 2𝑖 1 0
Solution: Consider the given matrix 𝐴 = and 𝐼 = an
−1 + 2𝑖 0 0 1
identity matrix of order 2 × 2. Then

1 −1 − 2𝑖 1 −1 − 2𝑖
(𝐼 − 𝐴) = and (𝐼 + 𝐴) = .
−1 + 2𝑖 1 1 − 2𝑖 1
1 −1 − 2𝑖 1 −1 − 2𝑖
Now, (𝐼 − 𝐴)(𝐼 + 𝐴) =
−1 + 2𝑖 1 1 − 2𝑖 1
 −4 −2 − 4𝑖
= = 𝐵 (𝑠𝑎𝑦)
2 − 4𝑖 −4
−4 2 + 4𝑖
The conjugate transpose of 𝐴 is given by 𝐵 ∗ = . To establish the
−2 + 4𝑖 −4
result, it is enough if we prove that 𝐵𝐵 ∗ = 𝐼.

−4 −2 − 4𝑖 −4 2 + 4𝑖 1 0
𝐵𝐵 ∗ = = = 𝐼, thus (𝐼 − 𝐴)(𝐼 + 𝐴) is a
2 − 4𝑖 −4 −2 + 4𝑖 −4 0 1
unitary matrix.

(1 + 𝑖) (−1 + 𝑖)
9. Prove that 𝐴 = is unitary and find 𝐴 .
(1 + 𝑖) (1 − 𝑖)

(1 + 𝑖) (−1 + 𝑖)
Solution: Given the matrix 𝐴 = , its conjugate transpose is
(1 + 𝑖) (1 − 𝑖)

(1 − 𝑖) (1 − 𝑖)
given by 𝐴∗ = . Since we need to A is unitary, consider the
(−1 − 𝑖) (1 + 𝑖)

(1 + 𝑖) (−1 + 𝑖) (1 − 𝑖) (1 − 𝑖) 4 0
product 𝐴𝐴∗ = = = 𝐼, an
(1 + 𝑖) (1 − 𝑖) (−1 − 𝑖) (1 + 𝑖) 0 4

identity matrix. Therefore, A is unitary and its inverse is given by

(1 − 𝑖) (1 − 𝑖)
𝐴 =
(−1 − 𝑖) (1 + 𝑖)
10. Verify which of the following matrices are normal.
3 + 4𝑖 1 1 0
a. 𝐴 = b. 𝐴 =
𝑖 2 + 3𝑖 1−𝑖 𝑖
3 + 4𝑖
1
Solution: a) Consider 𝐴 = . To prove that A is normal, we need to
𝑖
2 + 3𝑖
3 − 4𝑖 −𝑖
check if 𝐴𝐴∗ = 𝐴∗ 𝐴. The transpose conjugate is given by 𝐴 = . Now,
1 2 − 3𝑖
3 + 4𝑖 1 3 − 4𝑖 −𝑖 26 6 − 6𝑖
𝐴𝐴∗ = = − − − −(1) and
𝑖 2 + 3𝑖 1 2 − 3𝑖 6 + 6𝑖 14
3 − 4𝑖 −𝑖 3 + 4𝑖 1 26 6 − 6𝑖
𝐴∗ 𝐴 = = − − − −(2).
1 2 − 3𝑖 𝑖 2 + 3𝑖 6 + 6𝑖 14

From (1) and (2) it follows that 𝐴𝐴∗ = 𝐴∗ 𝐴, proving that A is normal.

1 0
b) 𝐴 = is not normal. Verify it on your own.
1−𝑖 𝑖
Involutory and Nilpotent matrices:

Let A be any square matrix. We say that A is an involutory matrix if 𝐴 = 𝐼. For example
−5 −8 0 −5 −8 0 −5 −8 0 1 0 0
𝐴= 3 5 0 is involutory as 𝐴 = 3 5 0 3 5 0 = 0 1 0 is an
1 2 1 1 2 1 1 2 1 0 0 1
identity matrix.

We say that 𝐴 is a nilpotent matrix if 𝐴 = 0 for some positive integer 𝑘. Then we say that 𝑘
is an index/order of 𝐴.

5 −3 2
0 1
Example: Let 𝐴 = and 𝐵 = 15 −9 6 are nilpotent matrices of index 2.
0 0
10 −6 4
0 2 1 6
0 0 1 2
𝐵= is a nilpotent matrix of index 4.
0 0 0 3
0 0 0 0

Note: Orthogonal matrices we have already discussed above. If you are asked to
show that Eigen vectors corresponding t o different Eigen values are orthogonal or
not follow the procedure given below.

Let 𝑿𝟏 , 𝑿𝟐 , 𝑿𝟑 be the Eigen vectors corresponding to 𝝀𝟏 , 𝝀𝟐 𝒂𝒏𝒅 𝝀𝟑 . Then A is the

matrix whose columns are 𝑿𝟏 , 𝑿𝟐 , 𝑿𝟑 . That is 𝑨 = [𝑿𝟏 , 𝑿𝟐 , 𝑿𝟑 ]. Now check 𝑨𝑨𝑻 = 𝑰, if so
then you are done.