Introduction to Determinant UAF

Introduction to Determinant

In algebra, the determinant is a special number associated with any square matrix.
As we have studied in earlier classes, that the determinant of 2x2 matrix is defined as the
product of the entries on the main diagonal minus the product of the entries off the main
diagonal.. The determinant of a matrix A is denoted by det (A) or |A|
a b 
For example: A 
c d 
Then det (A) = ad-bc.
or |A| = ad – bc
1 2 
Example: Find the determinant of the matrix A   
3 4 
1 2
| A|   1 4  2  3  4  6   2
3 4

To extend the definition of the det(A) to matrices of higher order, we will use subscripted
entries for A.
a a 
A   11 12 
b21 b22 

a11 a12
det (A) =  a11b22  a12b21
b21 b22

This is called a 2x2 determinant.

The determinant of a 3x3 matrix is also called a 3x3 determinant is defined by the
following formula.
a11 a12 a13
a21 a22 a23
(1)
a31 a32 a33

For finding the determinant of the 3x3 matrix, we look at the following diagram.

a11 a12 a13 a11 a12

a21 a22 a23 a21 a22
a31 a32 a33 a31 a32

1
Introduction to Determinant UAF

We write 1st and 2nd columns again beside the determinant. The first arrow goes from
a11 to a33, which gives us product: a11a22 a33 . The second arrow goes from a12 to a31,
which gives us product: a12 a23a31 . The third arrow goes from a13 to a32, which gives us
the product: a13a21a32 . These values are taken with positive signs.

The same method is used for the next three arrows that go from right to left downwards,
but these product are taken as negative signs.

= a11a22a33+a12a23a31+a13a21a32 - a13a22a31 - a11a23a32 - a12a21a33

 1 2 3
Example 2: Find the determinant of the matrix A   4 5 6 
 7 8 9 
1 2 3
detA = -4 5 6
7 -8 9

1 2 3 1 2
= -4 5 6 -4 5
7 -8 9 7 -8
= 1 5  9 + 2  6 7 + 3  (-4)  (-8)- 3  5 7 - 1 6  (-8)- 2  (-4)  9
= 45+84+96 - 105+48+72
= 240
We saw earlier that a 2 x2 matrix is invertible if and only if its determinant is nonzero.
In simple words, a matrix has its inverse if its determinant is nonzero. To extend this
useful fact to larger matrices, we need a definition for the determinant of the n  n matrix.
We can discover the definition for the 3  3 case by watching what happens when an
invertible 3  3 matrix A is row reduced.

Gauss’ algorithm for evaluation of determinants:

1) Firstly we apply it for 2  2 matrix say

 2 3
A 
 4 3
R2  R2  2 R1 (Multiplying 1st row by 2 and then subtracting from 2nd row)
 2 3  2 3 
 4  2(2) 3  2(3)   0 3
   
Now the determinant of this upper triangular matrix is the product of its entries on main
diagonal that is
Det ( A)  2(3)  0  3  6  0   6

2
Introduction to Determinant UAF

2) For 3  3 matrix say

 2 2 3
B   1 1 3 
 2 0 1
By R12  R21 (Interchanging of 1st and 2nd rows)
 1 1 3 
 2 2 3
 
 2 0 1

R2  R2  2 R1 (Multiplying 1st row by ‘-2’ and then adding in the 2nd row)
R3  R3  2 R1 (Multiplying 1st row by ‘2’ and then adding in the 3rd row)
 1 1 3 
 0 0 9 
 
 0 2 5 
  R32 (Interchanging of 2nd and 3rd rows)
By R23
 1 1 3 
0 2 5
 
 0 0 9 
Now the determinant of this upper triangular matrix is the product of its entries on main
diagonal that is
Det ( B)  (1)  2  (9)  18
So in general,

For a 1 1 matrix:
say, A  [aij ] - we define det A  a11 .

For 2  2 matrix:
 a11 a12 
a 
 21 a22 
a 
By R2  R2   21  R1 provided that a11  0
 a11 
 a11 a12 
 
 0 a22  a21 a12 
 a11 
   det A = product of the diagonal entries

3
Introduction to Determinant UAF

 a 
 a11  a22  21 a12   a11a22  a12 a21
 a11 

For 3  3 matrix say:

 a11 a12 a13 

a a22 a23 
 21
 a31 a32 a33 
a  a 
By R2  R2   21  R1 , R3  R3   31  R1 provided that a11  0
 a11   a11 
 
 
 a11 a12 a13 
 a22 a11  a12 a21 a23 a11  a13 a21 
0 
 a11 a11 
 a32 a11  a12 a31 a11a33  a13a31 
0 
 a11 a11 
 a32 a11  a12 a31 
 a11  a a a a
By R3  R3    R2 provided that 22 11 12 21  0
 a22 a11  a12 a21  a11
 
 a11 
 
a a12 
 11 a13 
 
 
 
 
 
0 a22 a11  a12 a21 a23 a11  a13 a21  a11  0
 a11 a11 
 
  a32 a11  a12 a31  
 a11a33  a13a31  a23a11  a13a21   a 11 
0 0   a a  a a 

 a11  a11   22 11 12 21  
  a 11 
 
 
 

Which is in echelon form.Now,

  det A = product of the diagonal entries

4
Introduction to Determinant UAF

 a a  a a  a a  a a  a a  a a  a a  a a  
 a 11  22 11 12 21   11 33 13 31   23 11 13 21  32 11 12 31  
 a 11  a11  a11  a22 a11  a12 a21  
a a a a   a a  a a  a a  a a 
  a22 a11  a12 a21   11 33 13 31    a22 a11  a12 a21   23 11 13 21   32 11 12 31 
 a11   a11   a22 a11  a12 a21 

1
a11
 a22a11  a12a21  a11a33  a13a31    a23a11  a13a21  a32a11  a12a31 

1 2
a11
a11a22 a33  a11a22 a13a31  a12 a21a11a33  a12 a21a13a31  a23a112 a32  a23a11a12a31  a13a21a32a11  a12a21a13a31


1 2
a11
a11a22 a33  a11a22 a13a31  a12 a21a11a33  a23a112 a32  a23a11a12 a31  a13a21a32 a11

a11
 a11a22 a33  a22 a13a31  a12 a21a33  a23a11a32  a23a12a31  a13a21a32 
a11
 a11a22 a33  a12 a23a31  a13a21a32  a12 a21a33  a11a23a32  a13a22 a31

Since A is invertible,  must be nonzero. The converse is true as well.

To generalize the definition of the determinant to larger matrices, we will use

2  2 determinants to rewrite the 3  3 determinant  described above. Since the
terms in  can be grouped as:
  (a11a22 a33  a 11 a23a32 )  (a12 a23a31  a12 a21a33 )  (a13a21a32  a13a22 a31 )
,
 a11 (a22 a33  a23a32 )  a12 (a21a33  a23a31 )  a13 (a21a32  a22 a31 )
a a23  a a23  a a22 
  a11  det  22   a12  det  21   a13  det  21 
 a32 a33   a31 a33   a31 a32 
a a23 a a23 a a22
  a11  22  a12 21  a13  21
a32 a33 a31 a33 a31 a32
For brevity, we write   a11  det A11  a12  det A12  a13  det A13 (3)

a22 a23 a21 a23 a21 a22

det( A11 )  , det( A12 )  and det( A13 ) 
a32 a33 a31 a33 a31 a32
where
A11 is obtained from A by deleting the first row and first column.
A12 is obtained from A by deleting the first row and second column.
A13 is obtained from A by deleting the first row and third column.
So in general, for any square matrix A, let Aij denote the sub-matrix formed by deleting
the ith row and jth column of A.

Let’s understand it with the help of an example.

5
Introduction to Determinant UAF

Example3:
1 4 3 
Find the determinant of the matrix A  5 2 4 
3 6 3 
1 4 3 
Solution: Given A  5 2 4 
3 6 3 

1 4 3
| A | 5 2 4
3 6 3
2 4 5 4 5 2
1 4 3
6 3 3 3 3 6
 1(2  3  4  6)  4 ( 5  3  4  3)  3( 5  6  2  3)
 1(6  24)  4(15 12)  3(30  6)
 1(18)  4(3)  3(24)
 18 12  72
 42

1 -2 5 0 
 
For instance, if 2 0 4 -1
A= 
3 1 0 7
 
 0 4 -2 0 
then A32 is obtained by crossing out row 3 and column 2,

1 2 5 0
1 5 0
2 1 A32   2 1

0 4

so that 4
3 1 0 7  
   0 2 0 
0 4 2 0

We can now give a recursive definition of a determinant.

When n = 3, det A is defined using determinants of the 2  2 submatrices A 1 j .

When n = 4, det A uses determinants of the 3  3 submatrices A1 j
In general, an nxn determinant is defined by determinants of (n  1)  (n  1) sub matrices.

6
Introduction to Determinant UAF

Definition:
For n  2, the determinant of n  n matrix A  [aij ] is the sum of n terms of the form
a1 j  (det A1 j ) , with plus and minus signs alternating, where the entries a11 , a12 , , a1n
are from the first row of A.
Here for aij ,
i  1, 2,3, , n (1  i  n)
j  1, 2,3, , n (1  j  n)
n
In symbols, det A  a11 det A11  a12 det A12  ....  (1) a1n det A1n   (1)1 j a1 j det A1 j
1n

j 1

Example 4:
1 5 0 
Compute the determinant of A   2 4 1
 0 2 0 

Solution:
Here A is n  n  3  3 matrix such that
i  1, 2,3
j  1, 2,3
n
det( A)   (1)1 j a1 j det A1 j and here j  1, 2,3
j 1
3
 det( A)   ( 1)1 j a1 j det A1 j  ( 1)11 a11 det A11  ( 1)1 2 a12 det A12  (1)13 a13 det A13
j 1

= a11 det A11  a12 det A12  a13 det A13

 4 1  2 1 2 4 
det A  1.det    5.det 0 0   0.det  0 2
 2 0     

4 1 2 1 2 4
det A  1.  5.  0.
2 0 0 0 0 2

= 1 [4(0) – (-1)(-2)] -5 [ 2(0) – 0(-1)] +0[2(-2) – 4(0)]

 1(0  2)  5(0  0)  0(4  0)  2

Minor of an element:
If A is a square matrix, then the Minor of entry aij (called the ijth minor of A) is
denoted by Mij and is defined to be the determinant of the sub matrix that remains when
the ith row and jth column of A are deleted.
7
Introduction to Determinant UAF

In the above example, Minors are the followings:

4 1 2 1 2 4
M 11  , M 12  , M 13 
2 0 0 0 0 2

Cofactor of an element:
The number Cij=(-1)i+jMij is called the cofactor of entry aij(or the ijth cofactor of A).
When the + or – sign is attached to the Minor, then Minor becomes a cofactor.

In the above example, Cofactors are the followings:

C11  (1)11 M 11 , C12  (1)1 2 M12 , C13  (1)13 M13

4 1 2 1 2 4
C11  (1)11 , C12  (1)1 2 , C13  (1)13
2 0 0 0 0 2

 3 1 4 
Example 5: Find the minor and cofactor of the matrix A   2 5 6 
1 4 8 
 3 1 4 
Solution: Here A   2 5 6 
1 4 8 
The minor of entry a11 is
3 1 4
5 6
M 11  2 5 6   5  8  6  4  40  24  16
4 8
1 4 8
and the corresponding cofactor is
C11  (1)11 M11  M11  16
The minor of entry a32 is
3 1 4
3 4
M 32  2 5 6   26
2 6
1 4 8
and the corresponding cofactor is
3 4
C32  (1)3 2 M 32   M 32    26
2 6

8
Introduction to Determinant UAF

Alternate Definition:
Given A  [ai j ] , the (i, j)-cofactor of A is the number Ci j given by
Ci j  ( 1)i  j det Aij (4)
Then det A  a11C11  a12C12  ....  a1nC1n

This formula is called the cofactor expansion across the first row of A.

1 2 3
Example 6: Expand a 3x3 determinant using cofactor concept A  4 5 6
7 8 9
Solution: Using cofactor expansion along the first column;

1 2 3
5 6 2 3 2 3
4 5 6  (1)( 1)11  ( 4)( 1) 2 1  (7)( 1) 31
8 9 8 9 5 6
7 8 9

Now if we compare it with the formula (4),

 1C11  ( 4) C21  7 C31

5 6 2 3 2 3
 (1)( 1) 2  ( 4)( 1)3  (7)( 1) 4
8 9 8 9 5 6

5 6 2 3 2 3
 (1)(1)  ( 4)( 1)  (7)(1)
8 9 8 9 5 6

5 6 2 3 2 3
1 4 7
8 9 8 9 5 6

 1(45  ( 48))  4 (18  ( 24))  7 (12  15 )

 1( 45  48 )  4 (18  24)  7 (12 15 )

 (1)(93)  (4)(42)  (7)( 3)  240

Using cofactor expansion along the second column,

9
Introduction to Determinant UAF

1 2 3
4 6 1 3 1 3
4 5 6  (2)( 1)1 2  (5)( 1) 2 2  ( 8)( 1) 3 2
7 9 7 9 4 6
7 8 9

4 6 1 3 1 3
 (2)(1)3  (5)(1) 4  (8)( 1) 5
7 9 7 9 4 6

4 6 1 3 1 3
 (2)(1)  (5)(1)  (8)(1)
7 9 7 9 4 6

4 6 1 3 1 3
 2 5 8
7 9 7 9 4 6

  2 (  36  42 )  5( 9  21)  8( 6  (  12))

 (2)(78)  (5)(12)  (8)(18)  240

Theorem 1: The determinant of an n  n matrix A can be computed by a cofactor

expansion across any row or down any column. The expansion across the ith row using
the cofactors in (4) is
det A  ai1Ci1  ai 2Ci 2   ainCin

The cofactor expansion down the jth column is

det A  a1 j C1 j  a2 j C2 j   anj Cnj

The plus or minus sign in the (i, j)-cofactor depends on the position of aij in the matrix,
regardless of the sign of aij itself. The factor ( 1)i  j determines the following
checkerboard pattern of signs:

   ....
   
 
   
 
 

Example 7: Use a cofactor expansion across the third row to compute det A, where

10
Introduction to Determinant UAF

1 5 0 
A   2 4 1
 0 2 0 
Solution: Compute det A  a31C31  a32C32  a33C33

 (1)31 a31 det A31  (1)32 a32 det A32  (1)33 a33 det A33

5 0 1 0 1 5
0  (2) 0
4 1 2 1 2 4

 0  2(1)  0  2

Theorem 1 is helpful for computing the determinant of a matrix that contains many
zeros. For example, if a row is mostly zeros, then the cofactor expansion across that row
has many terms that are zero, and the cofactors in those terms need not be calculated.
The same approach works with a column that contains many zeros.

2 0 0 5
 1 2 4 1 
Example 8: Evaluate the determinant of A  
3 0 0 3
 
8 6 0 0

2 0 0 5
1 2 4 1
Solution: det( A) 
3 0 0 3
8 6 0 0

Expand from third column

11
Introduction to Determinant UAF

det(A)  0  C13  4  C23  0  C33  0  C43

 0  4  C23  0  0

 4  C23

2 0 5
23
 4  (1) 3 0 3
8 6 0

Expand from second column

 2 5
  4  0  0  (6) 
 3 3
2 5
 (4 ) (6)
3 3

 216

Example 9: Show that the value of the determinant is independent of 

sin  cos 0
A  cos  sin  0
cos   sin  sin   cos 1

sin  cos 0
Solution: Consider A   cos  sin  0
cos   sin  sin   cos 1

Expand the given determinant from 3rd column we have

 0  0  (1)33[sin 2   cos2  ]  1

12
Introduction to Determinant UAF

 3 7 8 9 6 
0 2 5 7 3 

Example 10: Compute det A, where A  0 0 1 5 0
 
0 0 2 4 1
0 0 0 2 0 
Solution: The cofactor expansion down the first column of A has all terms equal to zero
except the first.
2 5 7 3
0 1 5 0
Thus det A  3  0.C21  0.C31  0.C41  0.C51
0 2 4 1
0 0 2 0
Henceforth we will omit the zero terms in the cofactor expansion.

Next, expand this 4  4 determinant down the first column, in order to take advantage of
the zeros there.
1 5 0
We have det A  3  2 2 4 1
0 2 0
This 3  3 determinant was computed above and found to equal –2.

Hence det A = 3x2x(-2) = - 12.

The matrix in this example was nearly triangular. The method in that example is easily
adapted to prove the following theorem.

Triangular Matrix:

A triangular matrix is a special kind of m x n matrix where the entries either below or
above the main diagonal are zero.

is upper triangular and is lower triangular matrices.

Determinants of Triangular Matrices:

Determinants of the triangular matrices are also easy to evaluate regardless of size.

Theorem: If A is triangular matrix, then det (A) is the product of the entries on the main
diagonal.

13
Introduction to Determinant UAF

Consider a 4x4 lower triangular matrix.

 a11 0 0 0
a a 0 0 
A   21 22
 a31 a32 a33 0
 
 a41 a42 a43 a44 

Keeping in mind that an elementary product must have exactly one factor from each row
and one factor from each column, the only elementary product that does not have one of
the six zeros as a factor is (a11a22 a33a44 ) . The column indices of this elementary product
are in natural order, so the associated signed elementary product takes a +.

Thus, det (A)= a11  a22  a33  a44

2 5 7
Example 11: 0 3 8  (2)(3)(5)  30
0 0 5

1 0 0 0
4 9 0 0
 (1)(9)(1)(2)  18
7 6 1 0
3 8 5 2

1 2 7 3
0 1 4 1
 (1)(1)(2)(3)  6
0 0 2 7
0 0 0 3

The strategy in the above Example of looking for zeros works extremely well when an
entire row or column consists of zeros. In such a case, the cofactor expansion along such
a row or column is a sum of zeros! So the determinant is zero. Unfortunately, most
cofactor expansions are not so quickly evaluated.

Numerical Note: By today’s standards, a 25x25 matrix is small. Yet it would be

impossible to calculate a 25  25 determinant by cofactor expansion. In general, a cofactor
expansion requires over n! multiplications, and 25! 1.5 x1025 .

If a supercomputer could make one trillion multiplications per second, it would have to
run for over 500,000 years to compute a 25  25 determinant by this method.
Fortunately, there are faster methods, as we’ll soon discover.

14
Introduction to Determinant UAF

5 7 2 2
0 3 0 4
Example 12: Compute
 5 8 0 3
0 5 0 6
Solution: Take advantage of the zeros. Begin with a cofactor expansion down the third
column to obtain a 3  3 matrix, which may be evaluated by an expansion down its first
column,
5 7 2 2
0 3 4
0 3 0 4 1 3
 ( 1) 2 5 8 3
5 8 0 3
0 5 6
0 5 0 6

3 4
 2  ( 1) 21 ( 5)  20
5 6

The –1 in the next-to-last calculation came from the position of the –5 in the 3  3
determinant.

15
Introduction to Determinant UAF

Exercises:
Compute the determinants in exercises 1 to 6 by cofactor expansions. At each step,
choose a row or column that involves the least amount of computation.

6 0 0 5 1 2 5 2
1 7 2 5 0 0 3 0
1. 2.
2 0 0 0 2 6 7 5
8 3 1 8 5 0 4 4

3 5 8 4 4 0 0 0
0 2 3 7 7 1 0 0
3. 4.
0 0 1 5 2 6 3 0
0 0 0 2 5 8 4 3

4 0 7 3 5 6 3 2 4 0
0 0 2 0 0 9 0 4 1 0
5. 7 3 6 4 8 6. 8 5 6 7 1
5 0 5 2 3 3 0 0 0 0
0 0 9 1 2 4 2 3 2 0

Use the method of Example 2 to compute the determinants in exercises 7 and 8. In

exercises 9 to 11, compute the determinant of elementary matrix. In exercises 12 and 13,
a b 
verify that det EA = (det E) . (det A), where E is the elementary matrix and A   .
c d 

3 0 4 2 4 3 1 0 0  k 0 0
7. 2 3 2 8. 3 1 2 9. 0 1 0  10.  0 1 0
   
0 5 1 1 4 1  0 k 1   0 0 1 
0 1 0 
1 0 0  1 k  0 1 
11.   12. 0 1  13. 1 0 
0 0 1     

3 1
14. Let A    . Write 5A. Is det 5A = 5 det A?
4 2

a b 
15. Let A    and k be a scalar. Find a formula that relates det (kA)to k and det A.
c d 

16