Section 1.

5: Elementary matrices and a

method for finding A−1

University of Johannesburg: Linear Algebra A

(MAT2A20/MAT02A2)
Dr Wilmari Morton

Elementary matrices and a method for finding A−1

Row equivalence

Definition
Matrices A and B are said to be row equivalent if they can be
obtained from each other by a sequence of elementary row
operations.

Example
Ri +kRj Ri ↔Rj cRj
A −→ B −→ C −→ D.
Matrices A, B, C, D are all row equivalent.

Elementary matrices and a method for finding A−1

Elementary matrices
Definition
An n × n elementary matrix is a matrix obtained by performing
one elementary row operation on the identity matrix In .

Example
     
0 0 1 1 0 0 1 0 0
E1 = 0 1 0 E2 = 0 8 0 E3 = 0 1 0
1 0 0 0 0 1 0 −4 1
   
1 0 0 0 0 1
R1 ↔R3
0 1 0 −→ 0 1 0  = E1
0 0 1 1 0 0
   
1 0 0 1 0 0
8R2
0 1 0 −→ 0 8 0  = E2
0 0 1 0 0 1
   
1 0 0 1 0 0
−4R2
0 1 0 R3−→ 0 1 0 = E3
0 0 1 0 −4 1
Elementary matrices and a method for finding A−1
Examples

Example
Which of the following matrices are elementary matrices?
 
0 0 1
1 1 0 0
0 1 0
Not an elementary matrix.
 
1 0 1
2 0 1 1
0 0 1
Not an elementary matrix.
 
2 0 2
3 0 1 0
0 0 1
Not an elementary matrix.

Elementary matrices and a method for finding A−1

Multiplication with elementary matrices
Example
   
1 0 0 1 2
1 Let E1 = 0 0 1 and A = −3 0. Find E1 A.
0 1 0 0 5
 
1 2
E1 A =  0 5
−3 0
   
1 0 2 1 2
2 Let E2 = 0 1 0 and A = −3 0. Find E2 A.
0 0 1 0 5
 
1 12
E2 A = −3 0
0 5

Theorem (Theorem 1.5.1)

Let A be an arbitrary m × n matrix. If the elementary matrix E results from
performing a certain row operation on Im ; then EA is the matrix that results
when the same operation is performed on A.
Elementary matrices and a method for finding A−1
Reversing row operations

Row operations
Recall the elementary row operations:
1 Multiply by a nonzero constant c: cRi
2 Interchange two rows: Ri ↔ Rj
3 Add a multiple of one to another: Ri + kRj

Reversing row operations

The above row operations can be reversed:
1 Multiply by 1/c: 1/cRi
2 Interchange the two rows: Ri ↔ Rj
3 Add the negative multiple: Ri − kRj

Elementary matrices and a method for finding A−1

Inverses of row operations

Operation Inverse Operation

1
cRi c Ri
Ri ↔ Rj Ri ↔ Rj
Ri + kRj Ri − kRj

0 12 R2 1 0
     
1 0 2R2 1
−→ −→
0 1 0 2 0 1
     
1 0 R1 ↔R2 0 1 R1 ↔R2 1 0
−→ −→
0 1 1 0 0 1
     
1 0 R2 −3R1 1 0 R2 +3R1 1 0
−→ −→
0 1 −3 1 0 1

Elementary matrices and a method for finding A−1

Inverses of elementary matrices

 −1  
1 0 0 1 0 0
0 0 1  = 0
 0 1
0 1 0 0 1 0
 −1  
1 0 0 1 0 0
0 1
c 0  = 0

c 0

0 0 1 0 0 1
 −1  
1 0 0 1 0 0
0 1 k  = 0 1 −k 
0 0 1 0 0 1

Theorem (Theorem 1.5.2)

Each elementary matrix is invertible, and its inverse is an
elementary matrix of the same type.

Elementary matrices and a method for finding A−1

Examples

Example
Invert the following matrices:
 
1 0 0 1

0 0

1 E1 = 0 0 1 E1−1 = 0 0 1
0 1 0 0 1 0
 
1 0 0 
1 0 0

2 E2 = 0 1 0  E2−1 = 0 1 0 
0 0 −3 0 0 − 31
 
1 0 0 1

0 0

3 E3 = 0 1 −4 E3−1 = 0 1 4
0 0 1 0 0 1

Elementary matrices and a method for finding A−1

Equivalence Theorem

Theorem (Equivalence Theorem)

Let A be an n × n matrix. The following statements are
equivalent:
1 A is invertible.
2 Ax = 0 has only the trivial solution.
3 The reduced row echelon form of A is In .
4 A is expressible as a product of elementary matrices.

Elementary matrices and a method for finding A−1

A Method for Inverting Matrices

If A is invertible
Ek . . . E2 E1 A = I
then
A = E1−1 E2−1 . . . Ek−1 .
Thus
A−1 = [E1−1 E2−1 . . . Ek−1 ]−1 = Ek . . . E2 E1 I.

Inversion Algorithm
Let A be an invertible matrix. Take the sequence of elementary
row operations that reduce A to I; and perform the same
sequence on I to transform I into A−1 .

Elementary matrices and a method for finding A−1

The procedure:

Write the system as the adjoined matrix [A|I]

Perform row reduction on the WHOLE system, until A is in
reduced row echelon form
The system will then be in the form [I|A−1 ]
If A cannot be reduced to I it is not invertible by the Equivalence
Theorem.
Example
Find the inverse of  
3 0 −2
A = 0 0 1 
0 12 0
1 2

3 3
0
−1
if it exists. A = 0 0 2
0 1 0

Elementary matrices and a method for finding A−1

Examples

Example
Find the inverse of the following if they exist
 
1 0 1
1 A = −1 1 0 A is not invertible.
2 0 2
 
1 0 1 
−1 0 1

−1
2 A = −1 1 0 A = −1 1 1
2 0 1 2 0 −1
 
2 1 1 
3 1 −3

−1
3 A = −1 0 3 1
A = 2 −5 −3 7
1 1 2 1 1 −1

Elementary matrices and a method for finding A−1

Extra Exercises

Example
1 Is the transpose of an elementary matrix elementary?
2 Show that if B is square matrix satisfying BA = I then A is
invertible and B = A−1 .
3 Use the Equivalence Theorem to prove that if A contains a
row or column
 of zeros
 then A is not invertible.
1 0 −2
4 Let A = 0 0 1 
0 1 0
1 Find elementary matrices E1 and E2 such that E2 E1 A = I.
2 Hence, express A as a product of elementary matrices.

Elementary matrices and a method for finding A−1