Download as pdf or txt
Download as pdf or txt
You are on page 1of 26

Introduction to Matrices

Rochana Meegaskumbura
Engineering Mathematics

University of Peradeniya

July 11, 2020

Rochana Meegaskumbura Matrices 1/18


Types of matrices

Definition
Zero Matrices Let n ∈ N and m ∈ N. Then, the n × m zero
matrix means the matrix (0)1≤i≤n, 1≤j≤m .

Rochana Meegaskumbura Matrices 2/18


Types of matrices

Definition
Zero Matrices Let n ∈ N and m ∈ N. Then, the n × m zero
matrix means the matrix (0)1≤i≤n, 1≤j≤m .This is the
n × m-matrix filled with zeroes. It is called 0n×m . (When no
confusion with the number 0 can arise, we will just call it 0.)

Rochana Meegaskumbura Matrices 2/18


Types of matrices

Definition
Zero Matrices Let n ∈ N and m ∈ N. Then, the n × m zero
matrix means the matrix (0)1≤i≤n, 1≤j≤m .This is the
n × m-matrix filled with zeroes. It is called 0n×m . (When no
confusion with the number 0 can arise, we will just call it 0.)
 
0 0 0
For example, the 2 × 3 zero matrix is .
0 0 0

Rochana Meegaskumbura Matrices 2/18


Zero Matrix

The zero matrix behaves very much like the number 0:

Rochana Meegaskumbura Matrices 3/18


Zero Matrix

The zero matrix behaves very much like the number 0: Let
n ∈ N and m ∈ N. Then:
(a) We have 0n×m + A = A + 0n×m = A for each n × m-matrix
A.

Rochana Meegaskumbura Matrices 3/18


Zero Matrix

The zero matrix behaves very much like the number 0: Let
n ∈ N and m ∈ N. Then:
(a) We have 0n×m + A = A + 0n×m = A for each n × m-matrix
A. (b) We have 0n×m A = 0n×p for each p ∈ N and each
m × p-matrix A.
(c) We have A0n×m = 0p×m for each p ∈ N and each
p × n-matrix A.
(d) We have 0A = 0n×m for each n × m-matrix A.
(e) We have λ0n×m = 0n×m for each number λ.

Rochana Meegaskumbura Matrices 3/18


Zero Matrix

Numbers are known to be zero-divisor-free: If a product ab of


two numbers
a and b is 0, thenone of a and b must
 be 0. This
 fails for
1 0 0 0
matrices: If A = and B = , then AB = 02×2
0 0 0 1
is the zero matrix, although neither A nor B is the zero matrix.

Rochana Meegaskumbura Matrices 4/18


Identity matrix

Definition
Let n ∈ N. The diagonal entries of an n × n-matrix A are its
entries A1,1 , A2,2 , . . . , An,n . In other words, they are the entries
Ai,j for i = j.
 
a b
For example, the diagonal entries of are a and d. The
c d
name “diagonal entries” comes from the visualization of an
n × n-matrix as a square table: When we say “diagonal”, we
always mean the diagonal of the square that connects the
upper-left corner with the lower-right corner1 ; the diagonal
entries are simply the entries along this diagonal. (The other
diagonal is called the “antidiagonal” in linear algebra.)

1
Often, this diagonal is also called the “main diagonal”.
Rochana Meegaskumbura Matrices 5/18
Identity Matrix

Definition
If i and j are two objects (for example, numbers or sets or
functions), then we set
(
1, if i = j;
δi,j = . (1)
0, if i 6= j

For example, δ3,3 = 1 (since 3 = 3) but δ1,2 = 0 (since 1 6= 2).


For another example, δ(1,2),(1,3) = 0 (because (1, 2) 6= (1, 3));
here we are using the notation δi,j in a situation where i and j
are pairs of numbers.

Rochana Meegaskumbura Matrices 6/18


The notation δi,j defined in (1) is called the Kronecker delta; it
is extremely simple and yet highly useful. It has the property
that δi,j = δj,i for any i and j (because i = j holds if and only if
j = i).

Rochana Meegaskumbura Matrices 7/18


Identity Matrix
Definition
Let n ∈ N. Then, the n × n identity matrix means the matrix
(δi,j )1≤i≤n, 1≤j≤n . This is the n × n-matrix whose diagonal
entries all equal 1, and whose all other entries equal 0. It is
denoted by In . (Other people call it I or E or En .)
The n × n identity matrix In looks as follows:
 
1 0 0 ··· 0
 0 1 0 ··· 0 
 
In =  0 0 1 · · · 0 
 
 .. .. .. . . . 
 . . . . .. 
0 0 0 ··· 1
(with n rows and
 n  For example, the 3 × 3 identity
columns).
1 0 0
matrix is I3 =  0 1 0 .
0 0 1
Rochana Meegaskumbura Matrices 8/18
The identity matrix behaves very much like the number 1:
Let n ∈ N and m ∈ N.
(a) We have In A = A for each n × m-matrix A.
(b) We have AIm = A for each n × m-matrix A.
Proposition says that multiplying a matrix A by an identity
matrix (from either side) does not change A. Thus, identity
matrices have no effect inside a product, and so can be
“cancelled” (or, more precisely, dropped). For example, if A, B,
C and D are four n × n-matrices, then
In ABIn In CIn D = ABCD. (Of course, this is similar to
dropping 1’s from products of numbers: 1ab · 1 · 1c · 1d = abcd.)

Rochana Meegaskumbura Matrices 9/18


Transpose of a Matrix

Definition
The transpose of an n × m-matrix A is defined to be the
m × n-matrix (Aj,i )1≤i≤m, 1≤j≤n . It is denoted by AT .
The definition says that the transpose of an n × m-matrix A is
the m × n-matrix whose (i, j)-th entry (for i ∈ {1, 2, . . . , m} and
j ∈ {1, 2, . . . , n}) is the (j, i)-th entry of A. So the transpose of
A has the very same entries as A, but in different position:
namely, the entry in position (i, j) gets moved into position
(j, i). In other words, the entry that was in row i and column j
gets moved into column i and row j. So, visually speaking, the
transpose of the matrix A is obtained by “reflecting A around
the diagonal”. Some examples should help clarify this:

Rochana Meegaskumbura Matrices 10/18


a a0
 
 T
a b c
=  b b0  ;
a0 b0 c0
c c0
T 
a a0
 
a b
= ;
a0 b0 b b0
 T
a 
 b  = a b c .
c

Rochana Meegaskumbura Matrices 11/18


Proposition
Here is a very simple fact about transposes: The transpose of
the transpose of a matrix A is the matrix A itself. In other
words:
Let n ∈ N and m ∈ N. Let A be an n × m-matrix. Then,
T
AT = A.
Proof of Proposition .
We have AT = (Aj,i )1≤i≤m, 1≤j≤n (by the definition of AT ).
Thus, AT is an m × n-matrix and satisfies

AT i,j = Aj,i

for all i ∈ {1, 2, . . . , m} and j ∈ {1, 2, . . . , n} .
(2)
Hence,

AT j,i = Ai,j

for all i ∈ {1, 2, . . . , n} and j ∈ {1, 2, . . . , m} .
(3)
(Indeed, this follows by applying (2) to j and i instead of i and
j.) Rochana Meegaskumbura Matrices 12/18
proof continued

 
 
 
T T  T 

A =  A j,i  = (Ai,j )1≤i≤n, 1≤j≤m = A.
| {z }
 
=Ai,j
(by (3)) 1≤i≤n, 1≤j≤m

This proves Proposition

Rochana Meegaskumbura Matrices 13/18


More on Transposes

How do the matrix operations we have seen above (addition,


multiplication, , etc.) behave with respect to transposes? The
answer is “fairly nicely”:
(a) Let n ∈ N. Then, (In )T = In .
(b) Let n ∈ N and m ∈ N. Then, (0n×m )T = 0m×n .
(c) Let n ∈ N and m ∈ N. Let A be an n × m-matrix. Let λ be
a number. Then, (λA)T = λAT .
(d) Let n ∈ N and m ∈ N. Let A and B be two n × m-matrices.
Then, (A + B)T = AT + B T .
(e) Let n ∈ N, m ∈ N and p ∈ N. Let A be an n × m-matrix.
Let B be an m × p-matrix. Then, (AB)T = B T AT .

Rochana Meegaskumbura Matrices 14/18


Symmetric Matrix

Definition
Symmetric Matrix
The Square matrix A is symmetric if A = AT .

Rochana Meegaskumbura Matrices 15/18


Symmetric Matrix

Definition
Symmetric Matrix
The Square matrix A is symmetric if A = AT .

Definition
Anti Symmetric Matrix The Square matrix A is
anti-symmetric if A =
-AT .NB:I have corrected this typo appearing in the recording

Rochana Meegaskumbura Matrices 15/18


Matrices with Complex Entries

In a matrix the entries could be real numbers,complex numbers


or even functions. But in this course we consider matrices
mostly with real entries (real matrices) and sometimes with
complex numbers(complex matrices). When a matrix is complex
matrix we also talk about complex conjugation of a matrix.
Definition
Conjugate of a Complex Matrix (A)
Conjugate of a complex matrix denoted(A)is the matrix
obtained by taking the conjugate of each entry.

Rochana Meegaskumbura Matrices 16/18


Matrices with Complex Entries

In a matrix the entries could be real numbers,complex numbers


or even functions. But in this course we consider matrices
mostly with real entries (real matrices) and sometimes with
complex numbers(complex matrices). When a matrix is complex
matrix we also talk about complex conjugation of a matrix.
Definition
Conjugate of a Complex Matrix (A)
Conjugate of a complex matrix denoted(A)is the matrix
obtained by taking the conjugate of each entry.
Example
 
3+i 4
A=
8i 8−i
 
3−i 4
then the complex conjugate A=
−8i 8 + i

Rochana Meegaskumbura Matrices 16/18


Conjugate Transpose of a Matrix

Given a matrix A, we can form its conjugate transpose AH by


taking the transpose and conjugating
Example
 
1 1 + i −2 − i
A=
3i 4 −3 + i

Rochana Meegaskumbura Matrices 17/18


Conjugate Transpose of a Matrix

Given a matrix A, we can form its conjugate transpose AH by


taking the transpose and conjugating
Example
 
1 1 + i −2 − i
A=
3i 4 −3 + i
 
1 3i
AT = 1 + i 4 
−2 − i −3 + i

Rochana Meegaskumbura Matrices 17/18


Conjugate Transpose of a Matrix

Given a matrix A, we can form its conjugate transpose AH by


taking the transpose and conjugating
Example
 
1 1+i −2 − i
A=
3i 4 −3 + i
 
1 3i
AT = 1 + i 4 
−2 − i −3 + i
 
1 −3i
AH = 1 − i 4 
−2 + i −3 − i

Rochana Meegaskumbura Matrices 17/18


Hermitian Matrix

A square complex matrix is Hermitian or Self-Adjoint if


AH = A

Rochana Meegaskumbura Matrices 18/18

You might also like