- -

Contents__I

Chapters Pcic.
Addition and Multiplication 1-60

II Some Types of Matrices 6)-102

III Equivalence 103-121

IV- Determinants 122-204

I Rank and Adjoint of Matrix 1-104

11 Solution of Linear Equations 105-159

III Characteristic Equation of a Matrix 160-209

IV Linear Dependence of Vectors 2)0-'l9

V Quadratic Forms 220-228

Note-Please note that important questions are marked as* and very
important questions as' in this book.


CHAPTER 1
Addition and Multiplication

§ 1.01. Introduction.
The matrices (formal definition is given in § 1.02 Page 2) were invented
about a century ago in connection with the study of simple changes and
movements of geometric figures in coordinate geometry.
J. J. Sylvester was the fitst to use the word "matrix" in 1850 and later on
in 1858 Arthur Cayley developed the theory of matrices in a systematic way.
"Matrices" is a powreful tool of modern mathematics and its study is
becoming important day by day due to its wide applications in almost every
branch of science and especially in physics (atomic) and engineering. These are
used by Sociologists in the study of dominance within a group, by
Demographers in the itudy of births and deaths, mobility and class structure
etc., by Economist in the study of inter-industry economics, by Statisticians in
the study of 'design of experiments' and 'multivariate analysis', by Engineers
in the study of 'net work analysis' which is used in electrical . and
communication engineering.
Rectangular Array.
While defining matrix (see § 1.02 Page 2) we use the word 'rectangular
array', which should be understood clearly before we come to the formal
definition of 'matrices' and to understand the same we consider the following
example
In an inter-university debate, a student can speak either of the five
languages Hindi, English, Bangla, Marathi and Tamil. A certain university
(say A) sent 25 students of which 8 offered to speak in Hindi, 7 in English, 5 in
Bangla. 2 in Marathi and rest in Tajnil. Another university (say B) sent 20
students of which 10 spoke in Hindi, 7 in English and 3 in Marathi. Out of 25
students from the third university (say C), 5 spoke in Hindi, 10 in English, 6 in
Bangla and 4 in Tamil.'
The informatiorr given in the above example can be put in a compact
way if we give them in a tabular form as follows

1
University -
Number of speakers in
Hindi English Bangla Marathi Tamil
A 8 7 5 2
B 10 7 0 3 0
C 5 10 6 0 4
i
192/1/1


2 Matrices
The numbers in the above arrangement form what is known as a
rectangular array. In this array the lines down the page are called
colunun whereas those across the page are called rows. Any particular number
in this arrangement is known as an entry or an element. Thus in the
above arrangement we find that there are 3 rows and 5 columns and also
we observe that there are 5 elements in each row and so total number of
elements = 3 x 5 i.e. 15.
If the data given in the above arrangement is written without lines and
enclosed by a pair of square brackets i.e. in the form
8 7 5 2 3
10 7 0 30,
5 10 6 0 4
then this is called a matrix.
§ 1.02. Definition of a Matrix.
A system of any mn numbers arranged in a rectangular array of rn rows
and ii columns is called a matrix of order n x n or an rn x n matrix (which is
r"ad as m by n matrix).
Or
A set of mn elements of a set S arranged in a rectangular array of n TOWS
and n columns is called an n x n matrix over S.
For example [ 2 1 3 1 is a 2 x 3 matrix.
[3 —2
and ajj a7 a13 at,, isan,nxnmatrix.
aj 6122 a23 (12n

aml am2 0m3 am,,

where the symbols ay represent any numbers (ay lies in the ith row and jth
c "imn).
Note 1. A matrix may be represented by the symbols taijl, (a,j), II a, U or
by a single capital letter A, say.
Generally the first system is adopted.
Note 2. Each of the mn numbers constituting an m x n matrix is knov
as an element of the matrix.
The elements of matrix may be scalar or vector quantities.
Note 3. The plural of 'matrix' is 'matrices'.
Solved Examples on § 1.02.
Ex. 1. Find 323 inA[ 1 3 2 4
0 3 1 S
I 5 0 3 6
L

Sol. 021 = clement in the 2nd row and 3rd column

=1 Ans.


Order of Malt ices 3

Ex. 2. Write down the orders of the matrices

(a) 2 3 51 (h) [2 ; (c) [3 4 51; (d) [II.
1 0 3] [3
Sot. (a) 2 x 3 (b) 2 x I (c) I x 3 (d) I x I. Axis.
Ex. 3. How man y elements are there in a 5 x 4 matrix ?
Sot. ftc required number of elements in 5 X 4 matrix is S x -
c. 20.
Ans.
Ex. 4. The results of it music competition are given in the following
matrix
3210
0 32 4
5030
2143

Here the rows represent the teams A, B, C, 1) in that order and the
columns represent the number of wins, first place, second place, third plack
and fourth place scored by the teams.
From the above matrix find (a) How many events did the team A win
? (b) How man y first places (lid the team B win ? (c) How many third
places did the team C in ? What does 0 rcprsent in second row?
Sot. (a) ': the first row represents the team A, so the required number
Sum of the elements of first row
=t2+l=6 Axis.
(b) As first elements of - second row (which represents the team B) is zero,
so the team B did not win any first place. Axis.
(c) The third row represents the team C and third column represents the
third place scored by the learns, so fhe number of third places won by the team
Cis3. Ans.
(d) The second row represents the leant B and the first column represents
the first place scored by teams. So 0 in the second row represents that the team
B did not score any first place. Ans.
Ex. 5. The order of the matrixI 2 3 1
4 5 6]]

(I) 2 x 3, (ii) 3 2, tin) 2 x 2, (iv) None of these. Arts. (i)

Exercises on § 1.02
Ex. 1. In k\ ; Implc I above, find (i) (112 . ( ii) a. Ans. ( i ) 0. (ii) 5
Ex. 2. Write the Irlcrs of the matrices
(a) 2 3 4 2 Il; (h)[ I (c) 151
3 5 5 3 4 jo
[ 4 7 6 7 1)

A (a) 3 x 5 (h) 3' x I (C) I I
4 Matrices 989
§t03. Rectangular Matrices.
The number of rows and cohins of a matrix need not be equal when
m * n i.e. the number of rows and columns of the array are not equal, then the
matrix is known as a rectangular matrix.
Classifications of rectangular matrices are as follows :---
/Sqa.e Matrix.
Tf m = n Le.1hnumber of rows and columns of a matrix are equal, then
the matrix is of order n x n and is called a square matrix of order a.
For example 2 3 1 j is a square matrix and[1 3 2 3
I 5 2 [2 5 7 9
769

is a rectangular matrix.
o!rizontalmatrü. If in a matrix the number of columns is more than
the number of rows then it is called a horizontal matrix.
For example 1 3 2 3 us a horizontal matrix.
12 5 7 9]
kow matrix : If in a matrix, there is only one row it is called a row
matrix. For example [1, 2, 3]. This is also called a row vector
Vertical matrix. If in a matrix the number of rows is more than the
numbof columns it is called a vertical matrix.
For example 2 3 is a vertical matrix.
35
46
57

'Column matrix : If there is only one column in a matrix, it is called a

column matrix.
For example 2 . This is also called column vector
3
4
' luI1( zero) Matrix If all the elements of an m x n matrix are zero,
then it is called a null or zero matrix and is denoted by Om < n or simply 0.
Foii example0 0 0] is the 2 x 3 null matrix.
10 0 0]
-- Unit ma trix A square matrix having unity for its elements in the
leading diagonal and all other elements as zero is called an unit matrix.
For example 1 0 0 0 is a four rowed unit matrix and we denote it
o 1 0 0 by 14.
0010
0001

an n-rowed square matrix [at,] is called a unit matrix provided



Equality of Matrices 5

a,1 = I whenever i j
0, whenever i * j.
Equal matrix : Two matrices are said to be equal if (a) they are of the
same type i.e. if they have same number of rows and columns and (b) the
elements in the corresponding positions of the two matrices are equal.
For example, let two matrices be A = [a,] and B [ba] then the two
matrices are said to be equal if a = by , for all values of i and j.
From the definition given above it is evident that
(i) If A = 8, then B = A (Symmetry)
(ii) A = A, where A is any matrix. (Reflexivity)
(iii) If A = B and B C, then A = C (Transitivity)
i.e. the relation of equality in the set of all matrices is an equivalence
relation. (See Author' Set Theory)
Matrices over a number field.
A matrix A is defined as 'over the field F of numbers' if all the elenents
of the matrix A belong to the field F of thc numbers.
Diagonal Element and Principal Diagonal.
Those elements ay of any matrix fajJ are called diagonal elements for
which i = j.
The line along which the above elements lie is called the Principal
diagonal o9Diagonal of the matrix.
t-jonal Matrix: A square matrix in which all elements expect those in
the main (or leading) diagonal are zero is known as a diagonal matrix.
For example 2 0 0 j is a 3—rowed diagonal matrix.
030
007

The sum of the diagonal elements of a square matrix A (say) is called the
trace of the matrix A.
Sub-matrix : A matrix which is obtained fiom a given matrix by deleting
any number of rows and number of columns is called a sub-matrix of the given
matrix.
For example Ii 21 isasub— matrixof 5 3 2
3 4j 1 1 2
L
734

Exercises on § 1.03
Ex. 1. The unit matrix is
(1) I I I (ii) 1 0 0
Ill 110
111 110


6 Matrices

(iii) 0 0 1 (jv) I 0 0
0 I 0 0 1 0
1 0 0 0 0 1 Ans. I, iv)

Ex. 2. What is the type of the matrix (a h cI

(i) column matrix: (ii) unit matrix:
(iii) square matrix: (iv) row matrix. Ans. (iv)
Ex. 3. The unit matrix is
(i) I I ); (ii) [ 0 1:
(iii) [I 11. (iv) 10 1 1
L I 1] L' oj Ans.i)

Ex. 4. The matrix of order m x it will he a unit matrix if

(i) all its elements arc unity:
(ii) tn it and all elements are unity;
(iii) in = n, and its diagonal elements are unity:
(iv) in = ,i, diagonal elements are unity and all the renaming elements are
zero. Ans. (iv)
§ 1.04. Scalar Multiple of a matrix.
If A is a matrix and X is a number then kA is defined as the matrix each
clement of which is X times the corresponding clement of the matrix A.
For example : 2 1 3 5 7]=[6 10 14
2 3 4j 6 8
L L4
or if A = [a & ]. then XA = [A]. where X is a number.
§ 1.05. Addition of matrices.
If there be two in x it given by A a,j] and B (h,,], then the
matrix A + B is defined as the matrix each element of which is the sum of the
corresponding elements of A and B i.e. A + B = lou + hi)],
where i= 1,2,3.... . in and j= 1,2,3.... . n.
For example : If A = [ a b c 1 and B = a3 b C3
L 02 h 2 ('2] L 04 bi
then A+B=[al+a3 bj+b3 cl+c3
L a 2+ a 4 b2+b4 C2+C4
§ 1.06. Subtraction of matrices.
If there be two in x it matrices given by A = ( a u] and B = [b(j], then the
matrix A - B is defined as the matrix each clement of which is ohtacd by
subtracting the clement 6f B from the corresponding clement of
A i.e. A—B=[a,—h01,
where i= 1,2.... . in and j = 1,2.... . n.
For example : If A = 5 at hi cii and B = 5 a3 b C3
a2 b C2J 1a4 b4 C4
 Addition of Matrices 7
then ar—ai bi—b3 CI—c3
A—B=^a2-614 b2—b4 C2-C4
*Note. If the two matrices A and B are of the same order, then only their'
addition and subtraction is possible and these matrices are said to he
conformable for addition or subtraction. On the other hand if the matrices A
and B are of different orders, then their addition and subtraction is not possible
and these matrices are called non-conformable for addition and subtraction.
§ 1.07, Properties of Matrix addition.
Proerty I. Addition of matrices is commutative.
i.e. [Oy] + [bj,jJ = [h j1 + [aq],
where a,] and [b,1 ) are an y two in x it i.e. matlices of the same order
(Meerut 95)
Proof : [a4 + [h,1 J [ay + b,J , by definition of addition
= [ b id + a1J] addition of numbers (elements) i
commutative
rr [bil l +IajJ
i.e. [o.} + [b] = [b] + [a,] Hunce the theorem.
Property Il. Addition of matrices is associative.
i.e. {[Otj] + [by]} + [Cij] = [ail ] + ([b] + [c,]),
where Ia,jl, [bill and trij are any three matrices of the same order in )< n, say.
Proof: { [a,,] + [bI} + [cjj]
( a + l) + [cjJ, by law of addition for matrices
= [(U,j + brj) + CqJ. by law of addition for matrices

= ia + (h,1 + c,3 )], addition of numbers is associative

1a4 + {b] + [ cij ]}. Hence the theorem.

Property HI. Addition for matrices obey the distributive law.
i.e. k ([aJ i' [h11]) = k [aj] ± k tbjJ
where 1041 and [b id ] are an y two ,natrjces of the same order in X n, say.
Proof: A' ([ a l + fb]) = k [at, + bj], by law of addition
= [k (a 1 + b11)], by law of scalar multiplication
[kay + kb11 ), by distributive law for numbers.
= tkaiJ + [kb]
= k ]ajJ + k [bij ]. Hence the theorem.
Property IV. Existence of additive identity.
If A = [a1 ] be any in x i t matrix and 0 be the in x n null matrix then
A±0=A=O+A
Proof: Here A = I OIjImxn and °=EOJ,flxfl
Then A+O=[a,j]mxn+[O]mxn


8 Matrices
= [a1j + O ]m x,t, by def. of addition
[aijjm>nA
Again O+A=[Olmxn+[aij]rnxn
= [0 + aj]m x ,, by def. of addition
=[aijJmxn=A
From (i) and (ii) we get A+O=A=O+A
Thus we observe that 0 (the null matrix) is the additive identity.
Property V. Existence of addititi've inverse.
If A = [a] be any m X matrix, there exists another m x n matrix B such
that A+B=O=B+A.
where 0 is the m x ii null matrix.
Here the matrix B is called the additive inverse of the matrix A or the
negative of A.
Also the (ij)th element of B is - aij if A = [ay]
Property VI. Cancellation Law.
If A, B, C are three matrices of the same order m x n, say such that
A+B=A+C, then B=C
Proof: Given A+B=A+C
or - A + (A + B) = - A + (A + C), adding -A from left on both sides
or (- A + A) + B = (- A + A) + C, by associative law of adition
or 0 ± B = 0 + C, by def of additive inverse
or• BC,bydef.of additive identity,
S*d,ETmp1es on § 1.04 to § 1.07.
"N,tIfA=: [2 3 •l and F1 2 -1
[o -1 [0 -1 3
evaluate 3A - 4B. (Avadh 90)
Sol 3A-4B=3[2 3 1 1- 4 1 1 2 -
[0_i 5] [o_i. 3
=16 9 3 1-1 4 8 -4
[0 -3 15] [0 -4 12
= 16-4 9-8 3-(-4)
[0-0 -3-(-4) 15-12
=12 1 71
[0 1 3] Ans.
Ex. 2.IfA=[ 1 5 6landB=[1 -5 7
[-6 7 0] [8 -7 7
then 0 13 1, A - B= 1 0 10 -1
[2 0 7j [_14 14 -7
Sol. Do yourself as Ex. 1 above.
Ex. 3. Determine the matrix A, where


Addition of Matrices 9
=2 1 2 3 +3 3 3 —1
324 22 3
1 4 5 —1 3 1
Sot. A= - 2 4 6 + 1
9 9 _3
64 8 66 9
2 8 . 10 —3 9 3
= 2+9 4+9 6+(-3) = II 13 3
6+6 4+6 8+9 12 10 17
2+(-3) 8+9 10+3 —I 17 13 Ans.
Ex. 4. Given A = 12 —3 and B =' 3 -1 2,
502 42 5
1—I 1 20 3
find the matrix C, such that A + 2C = B.
Sot. Given that A+2C=B or 2C=B—A
or 2C= 3 —1 21 - [l2 —3
4 25 5 0 2
2 0 3 1 —I 1
= 3-1 —1-2 2-(-3)1=r2 —3 5
4-5 2-0 5-2 [-1 2 3
2-1 0—(-1) 3_i] 1 12
or C = (1/2) 1 2 - 3 5 = 1 -(3/2) (5/2)
2 3 -(1/2) 1 (3/2)
1 1 2 (1/2) (1/2) 1 Ans.
x. Solve the _foliowng equations for A and B;
2—B=J —3 0 J 2B + A=T 4 1 Tif fl
<-Th
L 3 3 2] Hi 4 -4j
Sot. Given 2A - B =1 3 -3 0
32]
L 3
Multiplying both sides by 2, we get
4A-2B=2[3 —3 o] = 1 6 —6 0
[3 3 2] [6 6 4 (i)
Also given that 2B+A=[ 4 15
[-1 4 —4 ...(ii)
Adding (i) and (ii) we get
5A=[6 —'6 01+1 4 1 s
[6 6 4] [_i 4 —4
=[6+4 —6+1 0+51=110 — . 5 5
[6_I 6+4 4-4] [ 5 10 0
1
or A=(l/5)110 - 5 1 = 1 2 —1 1 ]
[ 5 10 0] [1 2 0 A ns.


Matrices
Again from (ii) we get
2B=1 4 1 51-A
L_ 1 4 --4j
or 2B=[ 4 1 5 1-[ 2 —1 1
4 _4j [i 2 0

4 4-2 1+1 5-I]=[_ 2 2 4

1—I 4-2 4_O 2 2 —4
or B=(1/2) 2 2 4 1 I 2
Ans.

Exercises on § 1.04 tel 1.07.

* If X, Y are two matrices given by the equations
x +- y =1 I —2 and x--y=[ 3 21. IindX,Y.
L I 0]
Ans. X o1 y [- t —2
2] L. 2 2
L'
Ex2%—IfA I 2 3 - B = 2 0 3 evahiaic 2A - 3B.
• 0 5 7 3 0 5
6 8 9 5 7 0
Ans. 1, —4 4 —3
—9 10 --I
—3 —5 18]
x7 (IfA=[i 2 3 then 2A equals
3

(i) 2 4 6 (ii) 2 2 3
213 223
321 322
((iii)
iv) 246 1 2 3
426 213
642 6 4 2 Ans. (iii)

Ex. *.-ff—k sin 2 0 and B —tan 2 8 cos2O

e0sec 2 0] /3- 12 - cot 2 0
then A + B is
(i)[l 0 0
[0 I [0 0
(iji)[ i I (iv)[0 11
LI I [1 0] Ans. (in)


Multiplication of Matrices II
Ex. 5. If A 3 and A + B = 0, then B equals

I
Ans. B -1 -2 -3
-2 --3 -}
-3 -1 -2

1.08. Multiplication of matrices. (Gorak/ipur 95)

If A and B he two matrices such that the number at columns in A is
equal to the number of lOWS in B i.e. if A = [a,I and B = IbjA 1 then the product
'7

of A and B denoted by AR is defined as matrix [erk]. where c, = aj hfk or

j=t
in other words the product AR is defined as the matrix whose clement in the ith
row and Atli column is ot bt + a,, /' 2k + a1 /k + -. + s,,, bk -
The product matrix will have i rows and k columns.
l]tis we conclude that
A is an ,n x n matrix and B is an ,s x '\iiiatrix then the product matrix
AR is an M W k matrix. (Remember)
As an example, consider the matrices
A = r I 2 3] andB= 7 8
[4 5 oj 9 10
Il 12
Here the niiinhei of columns in A = 3 = the number of rows in It
thus we can evaluate ,B.
Let Alt ]c]. where [c,) l is 2 x 2 matrix.
Now to write u, we take the element of the first row of A viz- I. 2, 3 in
this order and the elements of the first column of B ci, 7, 9, II in this order
and lrm the products I 7, 2 9. 3 II and finally add them.
cii = 17+29+3 . 11 =58
Similarl y ci = 1 . 8 + 210 + 312 = 64;
C21 = 47+59+611 139
and c22=4'8+5•I0+612= 154
Hence AR Cj[ =[ e tt 12 58 64
[r21 C22
1=1
139 154
Note. The product Alt he calculated only if the number of columns
in A he equal to the number of rows in B. The two matrices A and B satisfying
this condition are called conformable to multiplication.
Post-multiplication and Pre-multiplication of matrices.
The matrix All is the rnalnx A post-multiplied by B whereas the matrix
BA is the nl;ttrlx A pre-niultiplied by B.
12 Matrices
In the product AB, the matrix A is known as the pre-factor and the
matrix B is known as the post-factor.
The product in both the above cases viz. AB and BA may or may not
exist and may be equal or different,
i.e. we say AB * BA in general. (Bundelkhand 93; Gorakhpur 90)
The same is discussed below
Case I. If the matrix A is m x n and the matrix B. is n x k, then the
product AB exists whereas BA does not exist, since we know that AB can he
calculated only if the numbers of columns in A is equal to the number of rows
in B.
Case I!. If the matrix A is ni x n and the matrix B is n x rn, then both AB
and BA exist, but the matrix AB is m x m while the matrix BA is n x n. (Note)
Hence AB * BA though AB and BA exist.
Case III. If both A and B are square matrices of the same order, then AB
as well as B Ut are not necessarily equal

i.er] and B=[3 1]

then AB =fl 2 1 x [ 3 11=11-3+2.4 11+2-7

[3 4] [. 7] [3-3+44 3•l+4-7
=1 11 15
[25 31j
and BA -[3 l i x I l 21=13.1+13 32+14
[4 7] 4] [4 . 1+7 . 3 4-2^7-4

=[ 6
[25 36]
'°l
AB*BA.
( ButifA'[l O l andB J I 0
[0 -2] o4
then =r o1[i
o1=[ii+oo 10+04
- - [o -2] [o 4J [o . 1_2 . o 00-24
Il 0
o -8
and BA =fl o l x [ l ol=rl.l+o.o 10+0(-2)
[0 4] [0 -2] [0 . 1 +40 00-i-4(-2)
=11 0
[0 -8
AB BA.
Hence in general AB # BA. (Gorakhpur 95, 90)
Note 1. If AB = BA, then matrices A and B are said to commute. If
AB = - BA, the matriccc A and B are said to anticommute.
 Multiplication of Matrices 13
**Note 2. The product of Iwo non-zero matrices can also be a zero (or
null) matrix. (Avadh 93; Goraklzpur 91; Meerut 96P)
Let A=Il l]andfl=[ 1 0
I I] [-1 0
then ABI I I x[ j_ I 0 11 1 i(- 1) 10+1.01=10
0 01=0
[I i 0] [1.1+1.(_1) 10+1 . 0 0]
i.e. AB is zero matrix (or null matrix) where neither A nor B
is a zero
matrix.
AB=O does not imply that either A=O or B=O.
Here BA=[ 1 o]x[l I
[1 I
i o]

=[ 11+01 1 , 1+0-1 I I
[ . 1 . l+o. 1 . -1 . I+0 . 1J [_i -1
i.e. BA * 0
Another Example,
If A=J4 4] and B=[- I 1J, then
[3 3] [1
AB=[4 4]x[-I
[3 3] [1 -J

=[4(-l)+4(1) 4(1)+4(-1)]=[,) 01=0

[3(-1)+3(l) 3(1)+3(- 1)] [o oj
i.e. the product of two non-zero square matrices can be a zero manLi.
and BA=[-i 1]xf4 4
[1 -1] [3 3
=1 (- 1)4 + 13 (- 1)4+ 1 1-1l - i] ;t0
[1 . 4 +(- 1)3 14 +(- 1)3] [ 1 1]
*Note 3. The multiplication of matrices generally does ,uf obri f/it Imi
of cancellation
Let A=[0 i]. B 1 a 0 1. C=[b 0
[0 0] [o oj [o 0
where a # h
ThenAJJ=[0 I][a 0 1 = 1 0 01=0
Lo o][o 0] [o 0]
and AC=[0 I][b 0]=[o
[o o][o 0] [o
TO
It is evident that here AB = AC but B * C.
Law of cancellation is not obeyed in general.
-2


4 Matrices

Solved Examples on § 1.08

F+ 1 JAA A is any m x n matrix such that AB and BA are both
defined, at is the order of 11 7
(Sec Ca se II Page 2)
Sol. The rcqjd order of B n x m.
.jy1u1tipIyI3_ 141 and —21
()
3
Aiis.
Sol. 13—I 4I)<1_2=13(2)4(_1301
6

ExiifA = 1 2 0 , 11=10 0 0 find AB.

1 1 0 000
—i 4 0 L'
Sot. AB = - I 2(1 x 0 () 0
1 1 0 0 0 0
—I 4 0 1 4 9
1.0+2()+() I() + 20 + 09
10+20+0-1
10+10+04 10+10±09
1.0+10+01
10+40+09
_i . 0+40+01 - 1 ()+40+04
0 0 0 = 0, where 0 is the null tnatnX of order 3. Ans.
000
0 1) 0
Ex. 2. If A = 3 1 2 and B = 1 4
0 1 I 22
1 2 0] LI oj
(13urv2nchal 89)
then find AB. Whether BA exists 7 Give reason.
Sol. AB 3 1 2x I
0 1 I 22
I 2 0 • 1 ()

1t4+l2+2()
= 3.1+12+21
01+12+11 0.4+12+10
11 +2-2+0 1 1 4 + 22+00
Anc.
7 14
32
58
I sanlatrl\ oi orde' 3 1.
Here AV, a matrix of order 3 > 3 and B
s iii 13 is not equal I the
Hence BA does nol exist as number of column
number of rows in A.


II
Mult i plication of Mairice IS
* '3 aJ.If = 1 -2 3 and B= 2. 3

find AB and show th (Rohilkhand 97)

^7V Xf
2

= V j+4 32 13+(-2)5+ I
'L -42+4+5 -43+25+5 I
=10 -41
[10 3] \
and BA= (2 3 x[ Cl` -2 3 1
4 5 [-4 2

= (j3(-4) 2(-2+3(2) 2(3)4 1 (5)

41-±.5(-4) 4(-2)+5(2) 4(3)+(5)
[ 2 . 1 +i(-4) 2(-2)+ 1(2) 2(3)+ 1(5)
=-I0 2 211
-16 2 371
-2-211]
Hence (_.
Ex. 3(b). If A = 2 3 4 and B = 1 3 0
1 2 3 -1 2 1
-1 1 2 0 0 2
then pro ve that AR BA. (Meerut 97)
Sol. AB= 23 4x 130
1 2 3-1 2 1
-1 1 2 0 0 2
21 +3(- 1)+40 23+32+4.0 20+31+42
II +2(- 1)+30 13 +22+3 . 0 10+21+32
(-1)1 + I (- 1)+20 (- 1)3+12+20 (- 1)0+11+2.2
= 2-3+0 6+6+0 0+3+8 = -1 12 Ii
1-2+0 3+4+0 0+2+6 -1 7 8
-1-1+0 -3+2+0 0+1+4 -2 -1 5
and BA = 1 3 0 x 2 3 4
-1 2 I 1 2 3
0 0 2 -1 I 2
= 12+31 +0(-1) 13+3240.1 14+33+02
(-1)2+21+l(_I) (-1)3+22+1 . 1 (-1)423+I2
02+01 +2(- 1) 03+02 2 1 04+03+2.2


Matrices 182111I
16
2+3+0 3±6+0 4+9+0 = 5 9 13
=1 -1 2 4
1-2+2-I -3+4+1 -4+6+2
0+0+2 0+0+4 -2 2 4 (ii)
L 0+0-2 Hence proved.
From (i) and (ii) we find that AB * BA.
**Ex 4. If 41A = -4 8 4 , find A.
1 -1 2 1
3 -3 6 3.
Sol. From § 1.08 Page ii we know that if X is an m x n matrix, Y is an
n x k matrix, then the product XV is an ni x k matrix.
1
Here 4 is 3 x 1 matrix and - 4 8 4
-12.1
1
3 -3 6 3
is 3 x 3 matrix, so A must be a I x 3 matrix i.e. a row mairix. (Note)
:. Let Atabc]
ThenI4Xta b c]- -4 8 4
ii -121
-3 6 3
which gives 4a 4b 4c = -4 8 4
a b c -1 2 1
[3a 3b 3c -3 6 3
Comparing corresponding elements we have
1, 3c = 3.
4a = -4, a =- 1, 3a = - 3, 4b= 8, b =2, 3b = 6 and 4c = 4, c =
All these are satisfied by a = - 1, b = 2, c = 1.
Ans.
Hence fronr(i) we have A = [a b cJ = - 1, 2, 1).
Ex. 5. If A 1 1 3 , show that A2=0
226
-1 -I -3

Sol. A2= 1 1 3 x 1 I 3
22 6 2 2 6
-1 -1 -3 -1 -1 -3
= 1 . 1±I2+3(-1) 1-1+I2+3(-1) 13+l6+3(-3)
21 +22+6(- 1) 21 +2•2+6(- 1) 23+26+6(-3)
- 1-1 - I2-3(- 1) - - 12-3(- I) -13- l•6-3(-3)
= 0 0 0 = 0, where 0 1 3 x 3 null matrix.
000
0 0 0 Hence proved.

Ex. 6. Find the square of the matrix

-1 1 1-1
1 1 1 1
1 1 -. 1 1
1 1 1 -1


192/112 Multiplication of Matrices I?

2
SoL —I 1 1
1 —1 1 1
1 1 —1 1
1 1 1-1
—1 1 1 1 x —I 1 1 1
1 —1 I 1 1 —1 1 1
1 1 —1 I 1 I —1 I
1 1 1 —1 1 I 1 —1
= +Il+ll+ll (—l)l+l(—l)+ll+l-1
1 (-1) + (-. 1)1 + I-I + 11 ii + (-1) (-1) + 11 + i-i
1 (— l)+ Ii +(— I) I + II II + 1 (-1)+(— 1) I + 11
1 (-1)+ 11+ 1-1+(-1) 1 11 + I(--1)+1•I +(—l) 1

(- I) I + 11 + I (- I)+ 11 (-1)1 + 1 . 1 + 11 + 1 (- 1)1

1 . 1 +(— 1)1+1 (-1) + 1 . 1 1 . 1 + (-1)1 + 1 . 1 + 1 (-1)
1•1+l•l+(—l)(—l)+11 11+l1+(-1)1+I(--1)
11 + l•1 + I (- 1)+ (-1)1 1 . 1 + 1 . 1 + 11 ±(— 1) (-1)
= 4 0 0 0 =4 1 0 0 0 =41.
o 4 0 0 0 1 0 0
0040 0010
0 0 0 4 0 0 0 1 Ans.
*EL 7.
A= 1 1 —1 ; B= —1 —2 —1 , C= —1 —1 1
2 —3 4 6 12 6 2 2 —2
3 —2 3 S 10 5 —3 —3 3
show that AB and CA are null matrices but BA * 0, AC # 0.
Sol. AB= 1 1 —1 x —1 —2 —1
2-3 4 6 12 6
3 —2 3 5 L 5
= I (-1)+l6+(-1)5 I(-2)+112+(-1)10
2(-1)-36+45 2(-2)-312+4-10
3 (-1) —26+35 3 (-2) —212+310
l(-1)+16+(-1)5
2(— 1)-36+45
3(— 1)-26+35
= 0 0 0 J, which is a null matrix.
000
0 0 0 (See 1.03 Page 4)
This is known as 'unusual property' of Mnrix Multiplication
CA= —I —1 1 x 1 1 —1
2 2 —22 —3 4
3 3 3 3 —2 3


Matrices
= -1-1-12+1-3 -I-l-I(-3)+1(-2) -l(-l)-1-4+I-3
21+2-2-23 2-l+2(--3)-2(-2) 2(-1)+2-4-2-3
-31-32+33 -3-1--3(-3)4-3(-2) -3(-1)-3-4+3-3
O 0 0 , which is a null matrix,
000
0 0 0 Hence proved.
=[
We can prove in a simiUar way that BA 0 and AC # 0.
Ex. 8. Find the product of the following two matrices
o c -b x a 2 ab ac
-c 0 a ab b2 be
b -a 0 ac be c2 (Bundelkhand 93; Kanpur 94)

Sot. The required product

= 0 c -b x a 2 ab ac
-c 0 a ab b2 bc
b -a 0 ac bc c2

= 0-a 2 + c-ab - b-ac Oab + c-b2 - b-bc 0-ac + c-bc - b-c2

- c-a2 + 0-ab + a-ac - c-ab + Ob 2 + a-bc - c-ac + 0-bc + a-c2
ba2 - a-ab + 0-ac b-cb - ab 2 + 0-bc b-ac - a-bc + 0-c2

=000
000
000 Ans.
**Ex. 9. Prove that the product of two matrices
1co 2 o cos o sin el and [c0s24, Cos 4) sin 4

LCOS 8 sin 9 sin 2 8 J [cos 4) sin 0 sin2 4)

is zero when 8 and 4) differ by an odd multiple of it.

(Bundelkhand 92,- Meerut 91 .5)

Sot. The required product
=[cos2 cos2 O +cosO sin 9 cos 4, Sifl 4,
[cos o sin 0 cos 4, + sin 9 cos 4, sin 4)
cos 2 9 cos 4, sin 4) + .o.c 9 sin 0 sin 4)
cos 8 sin 8 cos 4) sin 4, + sin2 8 sin 4)
=1 cos 9 cos 4, (cos 9 cos 4, + sin sin 4,)
[sin O cos 4)(cos 8 cos4)+sin 8 sin 4,) -
cos 0 sin 4,(cos 8 cos 4)+ sin 9 sin 4))
sin 0 sin 4) (cos 9 cos 4, + sin 0 sin 4))
=[cos8cos4)cos(0-4)) cos8sin4)cos(0-4))
sin 0 cos $ cos (8-4)) sin 9 sin 4) cos (9-4))


Multiplication of Matrices 19

If 0 4) = an odd multiple of 1 it, then cos (9 - 4)) = 0 and consequently

the above product is zero (i.e. the null matrix of order 2>< 2).
S EL1O.IfA =Cos I e — sin el,B=Fcos4) -sin4)
[sinO coseJ [sin4) cos 4)
show that AB = BA. (Gorakhpur 90)
Sol. AB=[ .-- sine l x l cos 4) —sin4)
[sin 0 cos 0 j [sin 4) cos 4)
=[cosOcos4)_sin9sin4) —cos0sin$—sin9cos4)
[sin0cos4)+cos0sin4) —sin0sin$+cos4)cos4)
=[cos(0+4)) _sin(0+4))]
[sin (9+4)) cos (0+4))j
AndBA=Icos4) —sin4)1>[cose —sine
sin4) cos 4)j [sinO cosO
=[cos4)cos0_sin4)sin8 —cos4)sinO—sin4)cos0
[sincos0+cos4)sin9 —sin4)sin0+cos4)cos0
=[cos(0+4)) —sin(9+4))1
[sin (0 + 4)) cos (ti + 4))
From (i) and (ii) vet AB = BA. Hence proved.
**Ex. 11. H A, B, C are three matrices such that
A = [x, y, zi, B = a h g , C = x evaluate ABC. I
h b f y
g f c z
(Goralthpur 94; Kanpur 93; Kumaun 94; Purvanchal 90)
Sot. AB=[x, y, z] x a h g
h b f
gfc

=[x.a+y.h+z.g x.h+y.b+z.f x.g+y.f+z.c]

or ABC=[iu+hy+gz hx+by+fz gx+j5+cz]x x
V

= [x (ax + hy + gz) + y (hx + by +ft) + z (gx + fy + cz)] (Note)

= [ax 2 + by 2 + cz 2 + 2h.ty + 2gz + 2hz] . Ans.
Ex. 12 IIA=12 3 i] and B=rl 2 -6
[0 -1 5] [0 -i 3
evaluate (a) A2 - B2 and (b) AB and BA.
Sot. (a) A 2 =I2 3 1 ]X[2 3 I
which does not exist as
0 - 5 0 - 1 5 j
number of columns in the first matrix is not equal to number of rows in the
second matrix.


20 Matrices 999
Similarly B 2 does not exit.
(b) AB and BA both do not exist, the reason being the same as in part (a)
above
*Ex 13. Evalute A 3 if A [ cosh 0 sinh 9
L
sh0 cosh
F
Sol. A2 = [cosh 0 sinh 0 lx cosh 0 sinh 0
Lsmh 0 cosh 0] [sinh 0 cosh 8
= [± sinh 2 0 cosh 0 sinh 0 + sinh B cosh 0
[sinh 0 cosh 0 + cosh 0 sin B sinh2 0 + cosh 0
r s,nh 20 I1 - - 2
= cosh 20 cosh 0 + s,nh 0 = cosh2 0,
L
sinh 20 cosh 20 j 2 sinh 0 cosh 0 = sinh 29
A3=A2A= [ cosh20 sinh2ol[cosho sinhO
L sinh2O cosh 2OjL sinhO cosh
= cosh 20 cosh 9 cosh 29 sinh 8
+sinh20 sin 9 +sjnh29 cosh 9
sinh 20 cosh-0 sinh 29 sinh 0
+ cosh 20 sinh 0 + cosh 29 cosh 9
= cosh (20 ± 0) sinh (20 + 8)
sinh(20+0) cosh (29+0)
sinh(A+B)=sjnhAcoshB+coshA sinhB
cosh (A + B) = cosh A cosh B + sinh A sinh B
= [ cosh 30 sinh 38 1
L sinh 30 cosh 30] Ans.
Ex. 14. if A 2 - 1 1 , evaluate A3.
0 12
1 01
Sol. A2 =A= 2 —1 1j)< 2 11
0 12 0 12
1 01 1 01
2-2-1-0±11 2(-1)-11+l0 2.1_1.2+1.11
02+ 10+2 . 1 0(-1)+1-1-s-2•0 01+12+2-1
12 *01 + Ii 1 (- 1)+0-1 + 1 .0 1-1 +0 . 2+ 1.1
= 4+0+1 —2-1+0 2-2+1 = 5 —3 1
0+0+2 0+1+0 0+2+2 2 14
2+0+1 —1+0+0 1+0+1 3 —1 2
A3 =A 2 A= 5 —3 1 JX[2 —1 1
2 1 4 10 12
3 —1 2
H


Multiplication of Matrices 21
52-30+11 5(-1)-31+1•0 51-32+11
22+10+41 2(-1y11+40 21+12+41
32-10+21 3(- 1)-11 +20 31-12+21
= 10-0+1 -5-3+0 5-6+1 = II -8 0
4+0+4 -2+1+0 2+2+4 8 -1 8
6-0+2 -3-1+0 3-2+2 8 -4 3 Ans.
Ex. 15. lfA=[i 0 1, B=[0 - 1 1, C=[O
[0 -i] [1 0] L' 0]
Prove that
A 2 =B 2 =C 2 =-I and AB=_C=_BA,whereI=11 0
I
'Kumaun 92)
Sol. A2=[i OlxIi 0
[ -i] [0 -i
=[i. i+O. O i0+0(-i) I 0
[O . i - jO 0 . 0+(-i) (- i)] [ 0 -I
=-[l 0
[o I § 1.04 Page 6
or
and B2=[o - il x [ o -11=10.0-1.1 0(- l)+(- 1)0
[I 0] [1 0] [1.0+0.1 I (- 1)-+-00
=1-i 0 1=-l i 01=-i
[ 0 -1] [o iJ
Sinitlar]y we can prove that C2=_ I. Hence A 2 = B 2 = C 2 =_ L
Again AB=[i 01x[0 -1
[0 _ij [1 0
=[i0+0 . 1 i(-1)+00
[0 . 0-1(l) 0(-1)-i0
- o -i1=-[ o i
1[_i 0] [1 0]
and BA=[0 _i]x[e 0
[1 0] [o -i
=[0.i_1.0 0 . 0_1(_i)IT il=c.
[1-1+00 i . 0+0(_i) 0]
Hence AB=-C=-BA.
*Ex 16. If A= 1 2 3 ; B= xand AB= 6
012 y 3
0 0 1 z 1


22 Matrices
find the values of x, y, z.
Sol. AB 1 2 3 x x
012 y
001 z
or 6 = 1x+2y+3z
3 O'x+ly+2Z
I 0x+0y+l-z
6=x+2v+3z,3Y+2Z,lZ, (Note)
or
comparing the corresponding elements of the matrices on both sides
Solving these we get x= 1, y= I. z = 1. Ans.
Ex. 17. Find the values of x, y, z in the following equation
1 2 3 x x 4 —2 42
3 1 2 y 0 —6 LI
2 3 1 z —1 2

Sol. 1 2 3 x .t = 1x+2-y+3z
3 1 2 y 3x+1y+2-z
2 3 1 z 2x + 3y + 1 z (i)

And 4 —2 421= 42+(-2)11=[61

0 —6 Lii o•2+(-6)1 —6
—1 2 —12+ 2•1 0 ...(ii)

With the help of (i) and (ii), the given equati on. reduces to
x+2y+3z = 6
3x+ y+2z —6
2x+3y+ z 0

From this on comparing the corresponding elements on both sides we get

x+2y+3z6; 3x+y+2z=-6 and 2x+3y+O.
Solving these we get x = - 4, y = 2, z 2. Ans.
Ex. 18. Given At 0 0 0 1 ; A2 = 0 0 0 I
0 0 1 0 0 0 —i 0
0100 0 00
1 0 0 0 —i 0 0 0

A3= 0. 0 1 0 and A4 = 1 0 0 0
0 00-1 01 00
1 00 0 00-1 0
0 —1 0 0 0 —1
L0 0
Show that AI A k + AkAI = 21 or 0 according as i = k or i k and I is the
unit matrix of order 4 and I and k take the values 1, 2, 3 and 4.
Sol. Let i = k I (say). Then AiAkAIAi =AkAi


Multiplication of Matrices 23
AA=AA— 0 0 0 1 x 0 0 0 I
0010 0010
0100 0100
1000 1'O 00
= 0+0+0+1 0+0+0+0 0+0+0i-0 0+0+0+0
0+0+0+0 0+0+1+0 0+0+0+0 0+0+0+0
0+0+0+0 0+0+0+0 0+1+0+0•-0+O+0+0
0+0+0+0 0+0+0+0 0+0+0+0 1+0+0+0
= 1 0 0 0 =1
01 00
0010
0001
A1Ai + AkA1 = I + I = 21 Hence proved.
If it-k, let i=3 and k=2
Then AIAk = A3A 2 = 0 0 1 0 x 0 0 0
0 0 0 —1 0 0 —i 0
1 00 0 0 i 00
0 —1 0 0 —i 0 0 0
= 0+0+0+0 0+0+j-4-O 0+0+0+0 0+0+0+0
0+0+0+i 0+0+0+0 0+0+0+0 0+0+0+0
0+0+0+0 0+0+0+0 0+0+0+0 i+0+0+0
0+0+0+0 0+0+0+0 0-f-i+0+0 0+0+0+0
= 0 i 0 0 =i 0 I 0 0
i000 1000
000i 0001
GOjO 0010
And AkA=A2A3 = 0 0 0 x[0 0 1 0
0 0 —i 0 1 0 0 0 —1
0 1 00 I 00 0
—i 0 0 0 Lo —1 0 0
= F 0 —1 0 0 multiplying in the usual way
—i 0 0 0
0 0 0 —i
0 0 -1 0
=—i 0 1 0. 0
1000
0001
0010
AAk+AJ4 1 =j 0 1 0 0 —i[0 1 0 0
1000 11000
0 0 0 1 0001
L
0 0 I 0 [0 0 1 0


24 Matrices
= 0 Hence proved.
We can in a similar way prove the above result by giving i anri k other
values also.
**Ex. 19. If A0 = cos a sin a then prove that
L_sina cos a]
(a).AAa+ P and (b)Ac.A_aiS unit matrix.
Sol. (a) A0 A = cos a sin a xi cos
I I
_sinct cosa [_sin
sin 3
cos3

=1 cos a cos 3-sin asin P cos a sin + sin a cos I

-sin asin+cosacos3
L _sac0s_c05asin P
[ cos (a + 3) sin (a + ) = A 0 +
Hence proved.
L - sin (a + cos (a + n.)]

(b) Here A..0 = cos (- ci) sin (- a) = cos a - sin a

sin (- a) cos (- a) j 1. sin a cos a

cosa sin l x [ cosa -sincx

A0 . A._ 0
=I
cos2 a+ sin a
_sina cos a] Lsina cosa
-cos asin ct+ sin acos a
=[
L_sin a cos a+ cos a sin a sin2ci+cos2a

=I ' °1 which is an unit matrix.

0 lj Hence proved.
Exercises on § 1.08.
Ex. 1. Multiply [4 5 6] and 2
3
-1 Ans. [17]

Ex. 2. Multiply [1 2 3] and 4 -6 9 6

0-7 10 7
5 8 -11 -8
Ans. [ 19 4 -4 -4j
Ex. 3. If A= [ 1 - 1 1 and B = 1 1 1 show that AB is a null
[-i i J [1 1]
matrix.
Ex. 4. Show that [-5 2 3 1 x
1 4]
i 1101
1 o


Multiplication of Matrices 25
Ex. S. Show that 0 0 1
010
Ix 0 1 0 1 = 1 0 0
001 001
100 100 010
Ex. 6. IfA=[0 11 and B=[ 1]. find AB and BA if they exist.
[1 2] [2]

Ex. 7. If A= I 1 —1 1, B= —1 —2 -11
—2 3 —4 6 12 6
3 —2 3 5 10 5
then prove that AB = 0 but BA * 0.
Ex. 8. If k = I -2 3 and B = 1 0 2
2 3 —1 0 1 2
—3 1 2 120
then prove that AB BA.
Ex. 9.IfA= —2 3 —1 ,B= 1 3 —1
—1 2 —1 2 2 —1
—6 9 —4 3 0 —1
then show that AB = r0 0 - BA 1 0 01.
0 1 0 0 1 0
0 0 1 0 0 1 (Meerut 94)
Ex. 10. Show that - I I I x 0 (1/2) (1/2) = 1 0 0
1 —1 1 (1/2) 0 (1/2) 0 1 0
I I - I (1/2) (1/2) C) 0 01
L
Ex. 11. Form the products AB and BA, when
A 234] and B= 5
4
3
2 Ans. AB = [30]
Ex. 12. If A= 1 4 0 , B= 3 2 1 1 and C= 3 2 1
250 123 123
360 456 789
then prove that AB - AC = 0.
Ex. 13. Show that [2 2 lix 2 —I
[ii 1 1] 0 4
—2 1
1 —3
(Bundelkhwid 94)
Ex. 14. If A = 1 2 , evaluate A 2 . Ans. r 9 -
14 _3j [_8 17


26 Matrices
Ex. 15. If A = 0 1 2 and B = I -21, find AR or BA
123 -I 0
2 2 4 I -I
whichever exists. Ans. AB = I - 2 and BA does not exist.
2 -5
3 -8
Ex. 16. If A= 1 -23 1 and B= 1 0 2
2 3 -I 0 1 2
3 1 2 120
then prove that AR * BA.
t Ex. 17. If X, Y are two matrices given by the equations
x+Y=1i 21 and X-Y=[ 3 2 1, findXY. Ans. XY=1-2 -
2
L 4] 0] L-' [ 3
Ex. 18. In Ex. 11 Page 19 of this chapter, evaluate A (BC).
(Purvanchal 90)
Ex. 19. If order of A is m x n and that of C is m X I and A x B C then
order of B will be (i) Ix n, (ii) n xl, (iii) I x 3, (iv) 3 x 1. Ans. (ii)
Ex. 20. If A is ,n x n matrix, B is n x I matrix and C is Ix k matrix, then
t heorderof(AB)C will be(a)mXI,(b)nXP,(c)m x k,(d)k>< m Ans.(c)
§ 1.09. Properties of Multiplication of Matrices.
**Property I Multiplication of matrices is associative.
(Agra 96; Avadh 94, 92, 90. Garhwal 91; Gorakhpur 91; Rohilkhand 94)
Let A=[a]; B [ bjk), C = [ ckrj be three m x n, n x p and p Xi matrices
s
re pectively then (AR) . C = A. (BC).
/
...Joof. Let AB '= ( d1 kJ, where dk = Y. ay bk (i)

Then (AB) . C = [d,k] x [Ckr] = [e,r],

P
where e,r = Y dk. Ckr
k=1

= aijbj,t .ck,. from (i)

P 'I
i.e. (i. r)th element of (AB) . C = E av b3 k Ckr
k=1 1=1
P
And let BC = ( girl , 93 bjk Ckr (iii
k=1
Then A . BC = [a] x (girl [hjrl,
n
where hjr a).gjr
j=I


Multiplication of Matrices 27
a (p
bjkckrl,
from (iii)
j =l tk=I )
P a
i.e. (i, r)th element of A. (BC) = Y, Y a, hJk Ckr, ...(iv)
k=I j=I
since the summation can be interchanged.
From (iii) and (iv) we can conclude that the i. r)th elements of
(AB). C and A. (BC) are the same and their orders are also m x /.
Hence (AB). C = A . (BC).
"Property II. Multiplication of matrices is distributive with respect
to matrix addition. (Bundelkhand 96, 92)
(a) Let A = [a,), B = [ bfk] and C = [ cjkJ he three rn x n, n x p and ii x p
matrices respectively, then A (B + C) = AB + AC
(Avadh 93; Gorakhpur 93; Rohilkhand 93. 92)
-Proof. A(B+C)=
v/ fail ] x ([bjk]+[cjk])
= [ai1 [bk + c1 kJ = [d1kj, say,
a
where djk Z a ij (bfk + cj)
j=1

or
aa
(i, k)th element of A (B + C) = Z arj bj k +I a, Cjj
1=1 j=1
Again AB = [ aij] [bj kj = [elk], say,
a a
where e = Z a, hk i.e. (i, k)th element of AB = aij bj k ...(ii)
j=l j=I
Similarly we can prove
a
(i, k)th element of AC = aij cjk
j=l
From (ii) and (iii) we have
a fl

(i, k)th element of AB + AC = 1 a 1 bfk + Z aj Cjk

J=1
Hence from (i) and (iv) we conclude "that A (B + C) AB + AC.
(h) Let A = [ Ow), B (bjk l and C [ck] te three a Xp. ,n x n and rn X a
matrices respectively.
Then (B + C) A = BA + CA.
(Note. If A and B be ,n x n and a x p matrices then BA can not exist
whereas AB exists).
Proof. Its proof is simiLu to that of part (a) above.
§ 1.10. Positive integral power of a square matrix.
From § 1.09 we find that if A is a square matrix, then only the product
AA is defined and we write A 2 for AA


28 Matrices
Also by associative law
A 2 A = (AA) A = A (AA) = AA
So A2 or AA 2 is written as A 3 . -
In general AAA ... A is denoted by A" if there are n factors.
Definition. If A he a square matrix, then AA .... n times =
and Am + 1 A m . A, where m is a positive integer.
Theorem 1. If A he a square matrix (n x n say). then
A1' . = A" ", for any pair of positive integers p and q.
Proof. We shall prove this by the method of induction.
From definition we know that A 1' . A = A1' + , where p is any positive
integer.
A 1' A q = A1' + q holds when q I, whatever p may be.
We shall now prove that if it holds for a partiular value in say of q for
all values of p, then it must hold for the value in + 1 of q for all values of P.
Now A1' A1 + = A1' . ( A m . A) , by definition given-above
= (A" . Atm). A, by associative law
= (A p4-rn ) A, by hypothesis
p+m+1 , by definition given above
= A
= AP + + ', by associative law of addition of numbers.
i.e. A1'.Aq = 41' + q holds for the value m + I of q, whatever p may he if
it holds for q in.
Hence the proof by mathematical induction.
Theorem H. If A be a square matrix, then
= A1', for every pair of positive integers p and q
Proof is similar to that of Theorem I above.
Solved Examples on § 1.09 - § 1.10.
*Ex. 1, Evaluate A2 -4A-5I, where
1
A= 1 2 2 and 1=11 0 0
212 010
2 2 1 0 0 1 (Garhwa'I 90)

Sot. A2 = I 2 2 x I 2 2
212 212
221 221

1 . 1+22+22 1-2+21+2-2 1-2+22+21

'2-1 + 1-2+22 2-2+ 1-1 +2-2 2-2+1.2+21
2-1 ± 2-2 + 1-2 2 . 2+2 1 ± 1-2 2 . 2 + 2 . 2 + II


Multiplication of Matrices 29
1+4+4 2-2+4 2+4+2 = 9 8 8
2+2+4 4+1+4 4+2+2 8 9 8
2+4+2 4+2+2 4+4+1 8 8 9

A2-4A-5I
=988-41 22-5I 00
8 9 8 2 1 2 0 I 0
889 221 001
= 9 8 8'+ -4 -8 78 + -5 0 0
S 9 8 -8 -4 -8 ( -5 0
8 8 9 -8 -8 -4 0 0 -5
= 9-4-5 8-8+0 8-8+0= 0 0 0 =0, Ans.
8-8+0 9-4-5 8-8+0 0 0 0
8-8+0 8-8+0 9-4-5 0 0 0
where 0 is null matrix.
the

Ex. 2. Let f (x) = x2 - 5x + 6, find f (A) if A = 2 01

2 13
1 -1 (1
Sol. f(A) = A 2 - 5A +6
= A 2 -5A+61,wherel= 1 0 0
010
001
Now proceed is in Ex. 1 above. Ans. I - I - 3
-1 -1 -10
-5 4 4
*Ex. 3. If A =[ 2 - 1 1 and B =[ 1 0 1, show that
0 i L—' -1
(A+B)2=A2+AB+BA+B2^A2+2AB+B2
So!. A2=[2 - 1 1>[ 2 -.1
L° 'i L° '
=[2.2+(-1)o 2(-1)-1 . 1]=[4 -3
L 02 + 1 -0 0-1+i.lj [0 1
AB=[2 -ilx 1 0
Lo 1]
1_
i -1
f2 . 1-1(_1) 20-I(-l)]=[ 3 1

BA=[ I -1
[-1 -iJ [0
olxr 2
1

-3

A
30 Matrices
=1 12+0-0 l(-1)+0.II=[-2 —1
[-1 .2-10 _1(_1)_1-1 2 0
B2=[ i o l x I 1 0
L-' —1
L-' -'J
=[1.1+0(—I) 1-0+0(— 1 0
1 . 1_1(_I) —1.0-1(-1)
1)1= 1 °
A+B[2 -11+1 10
[o 1] [-1 —i
=[2+1 -i+ol=1 3 —I
[0-1 1—I] [_i o
(A+B)2 =[ 3 —i]x 3 —1
[—I oJ [-'
=[ 3•3-1(—I) 3(-1)-1-011 10 —3]
1 . 3+0e-1 —1(-1)+O-Oj L- 3 ij
Now A2+AB+BA+B2
=I4 -31+1 3 i]^[ 2 -]1+1 1 0
[o 1] [_i _ij [-2 oj [o 1
14+3+2+1 —3+l-1+0]=[ 10-3
1[o_1_2+0 i-i+o+ij [-3 1
( A + B)2 from (I) Hence proved.
Also A2 + 2AB + B2
=1 4 -31+21 3 il+Ii 0
[o Ij [-1 _ij [0 i
14 -3 ] + [ _ 6 2 1 + 1 1 01
[0 1 —2] [0 I]' See 1.04 Page
2
=f4+6+1 -3+2+01=1 11 1*(A+B)2
1-2+1j 2
L- Oj Hence proved.
L°-2°
Ex. 4.lfA4O llandB10 _1],sbowthat
[iij [1 OJ

(A-4-B)(A—B) A2—B2
SoL A+B= 0 il+ro -il=Io+o 10
i ij [i oj [i+i i+oj [2 1]
I
A—B=ro 11-1 0 -11=10-0 1-(- I )] = 1 0 2
L' 1] [' o] L'-' 1-0 j [01
(A+B)(A—B)40 O l x I O 2
t 2 iI [o


Multiplication of Matrices 31

= 1 00+0.0 02+0.1 1 = 1 0 01
[20+10 2 . 2+1 . 1] [0 5j
A2=10 l l x l o 1 1 = 1 0- 0+1 - I 0.1+1.11=11 1
[I 1] [1 . 0+1 . 1 1 . 1+1 . 1 1 2
'i [I
B 2 =1 0 -1140 - 1 1 = 1 0- 0 - il - 0 - 1 - 1 - 0 1 = 1- 1 0
[i oj [i 0] [1 . 0+0 . 1 _i-1+0.Oj [ 0 -1

A2 _B2 =1l il-I-i 0 1 = 1 1+1 1 - 0 1 = 1 2 Ii

L' 2J [ 0 _1J [1-_o 2+1] [1 3j

Hence from (1) and (ii), (A + B) (A - B) * A2 - B2.

Ex.S(a.IfAdenot The matrix Ia hi, prove that
c dj
A 2 -(a+d)A+ (ad -bc)I=O.
Sol, A 2 =I a b i x i a b
[c dJ [c d

= a-a-+-b-c a . b+bd]=1a 2 +bc b(a+d)

c-a-4-d•c cb+d . dj lc(a+d) cb+d2

A2-(a+d)A+(ad-bc)J

=[a2+bc b (a + d )1-( a -4- d)1 a b l + (ad - bc) [ I O

[c(a+d) cbd2 j [c dJ
L°
=[a2+bc b(a+d)1+1-a(a+d) -b(a+t
[c(a+d) cb+d2 j [_c(a+d) -d(a+d)

+Iad-bc 0
L
0 ad-bc
=[a2+bc_a(a++ai_,c b(a+d)-b(a+d)+0
[c(a-ci)-c(a±1)+O cb + d2 - d (a + d) + ad - bc
-[o 01=O, where Ois the 2x2 null matrix.
[0 0 j Hence proved.

Ex. S (b). if A = 1 2 3 , evaluate 6A 2 - 25A + 421.
2 3-1
3 1 2 (Agra 94)
Sol. Here A 2 -= 1 -2 3 X 1 -2 3
2 3 -1 2 3 -1
-3 1 2 --3 .1 2


32 Matrices 821112
= 11-22-33 - 12-23+31 13±21 +32
21 + 32 + 13 - 22 + 33- II 2 . 3 3 - 1 - 12 -
-31 + 12-23 32+ 13+21 -33--1I +22
= 1-4-9 -2-6+3 3+2+6 -12 - 5 II
2+6+3 -4+9-1 6-3-2 1 II 4 1
-3+2-6 6+3+2 -9-1+4 - 7 11 - 6
6A2-25A+421
=6-12-5 11-25 1 --2 3+42 I 00
4
11 1 2 3-1 010
-7 11-6 -3 1 2 001
= -72 -30 66 - 25 -50 75 + 42 0 0
66 24 6 50 75 -25 0 42 0
-42 66 -36 -75 25 50 0 0 42
' .
= -72-25+42 -30+50+ 0 66-75+ 0 = -55 20 -9
16 -9 31
66-50+ 0 24-75+42 6+25+ 0
-42+75+ 0 66-25+ 0 -36-50+42 33 41 -44

*Ex. 6. If A = r 1 - 1 ], then sbow that A2 = 2A and A3 = 4A.

[_1 1

SoL Given A= 1 -1
I_- II .( i)

A2 =A . A=[ 1 -1]x( 1 -1
[-1 1] [-1 1

=[1-l^(-1)(-1) 1(-1)+(-1)11=F 2-2

L-
1)1 + 1 (-1) (— 1)(- 1)+ [-2 2 i-ij
=2[ 1 -1]=2A,from(i)
[-1 ij (ii)
Hence proved.

Again A3 = A. A2 = A. (2A), from (ii)

=2A A- 2=2(2A),from(ii)
=4 Hence proved.
Ex. 7 a). I = F' 0] and E = F 0 1 , prove that
[0 0]
L° 'i
=a 3
! + 3a2bE. (Avadh 91; Gw-Iiwa 96)
Vt Soi.iaT[I J.1.b[O ]

=I a4*j'O b 1= [ a +0
[o a+OO] [0+0
0-+b1= a bl=B(say)
u+0 a
10
 182/1/3 Multiplication of Matrices 33

(al +bE) 2 =B2 =IablXIa b

[0 a] [0 a
= [ a-a + b0 a-b + b-a 1= a2 2ab
[0a + a-0 U-b + a.a] o a2
(al + hE)3 = B3 13213 (Note)
=[a2 2ab 4a b
[0 a2 [o a

=Ia2,a+2ab,0 a2. b+2ahal= a3 2b1

[øa+a2. 0-b+a2-a j 0 a3 ] ...(i)
Now a 3! 4 3a2 bE = a 3 11 0 1 + 3a2h 0 1
1]
L°
0 + [ 0 3a2b
[o a3 ] [ o 0
3 +0 0+3a2b1=[a3 32b1
=.[a

[o+o
a3-+O j [o a ] ...(ii)
From (i) and (ii) we get (a! + bE) 3 =a 3 14 3a2bE.
Ex 7(h). =[ 1 01 and E= 1O I
[0 1] [0 0

Th(j 'rove)hat (21 3E)3 81+ 36E 4- (Rohilkhand 95)

Do act1 y as Ex. 7 (a) above. Here a' = 2 and b = 3.
N *E. .lfA=[0 .-tan ( a/2)1, and lisa unit matrix,then
[tan(ct/2) 0]
prove that 1+A=(1-A)[ Cos cX sin (x
- [sina cosa
So!. !+A=[ t °1[° -tan (a/2)
0 lj [tan(a/2) 0

=[ 1+0 I -tan (w'2)

0-tan (2)1=ltan(ot/2)
0±tan (a/2) I +Oj I

• 1_A=[l 0 1_ l0 - tan (cX/2)

L0 1j [iancz/2) 0
=[ I-fl O+ tan ((L/fl 1=r I tan (a/2)
L° - tan (/2) 1-0] [_ tan (CX/2)

!-A)[cns - Sin (x
Ifl CX Cos CX


34 Matrices
= 1 tan (a12)] cos a - sin a
L-tan (a/2) I iL sin a co a
=[ Icos a + tan (a/2)sin a 1 (- sin a) + tan (a/2) cos a
L - tan (a/2)-cos a + 1sin a (sin (X)•tan (a/2) + ICOS a
-
= { 1 -2 sin2 (a/2)} + 2 sin (a/2) - 2 sin (c/2) cos ((x/2)
+ tan ((X/2) cos a
- tan ((x/2) cos a + 2 sin (cx./2) cos (cc/2) 2 sin (a/2'
+ {1 -2 sin - ((x/2))
writing cos a = 1 -2 sin2 a
I - 2 tan (a/2) cos2 (a/2)
+ tan (cx/2) cos a
- tan ((X/2) cog a I
+ 2 tan (a/2) cos2 (cx./2)
wrting sin 1 a as tan acos
= I - tan (a/2) [2 cos 2 (a/2)
-(2 cos 2(a/2)-1}]
tan (a/2)[-{2 cos; 2 (a12)-1} I
+2 cos 2(a/2))
writing cos a=2cos2(u/2)-1
=11 -tan (a/2)1=I+A, from (i)
tan (a/2) Ij Hence proved.
*tEx9(a)IfA-[3 1, show that A=11+2n -4n
Li - 1 J L n 1-2n
(Agra '96; Avadh 92; Gathwa! 91; Kanpur 95, Kunwun 95, 93, Meerut 90)
SOLA2 =A.A=[ 3 .-4]x[3 -4
L -'i L
=r33-4( ' ) 3(-4)--4(-1)11 =5 -8
L 1.3 - 1 u 1(._4)_l(_1) 2 -3
=11+2(2) -.4(2)1
[ (2) 1-2(2)]
An , when n=2 (Note)
An 1 +2,, 4nlholds when n=2
[ n 1-2nj

Now An+1 A". A ..Seedef. § 1.10 Page 28.



Multiplication of Matrices 35

=[ l-t-2n --41 [3 -4
n l_2n]
[
I(' +2u)3-4n(1) (I +2n) (-4)-4n (-1)
n(-.4)+(l-2n)(-l)
=[3+2n 4 -"1=[ I +2(n+1) -4(n+1)
- I - 2n j [ (n ± 1) 1 --2 (n + I)
L
i.e. A" r[ I + 2n
-
1 holds for 'n' = n + 1
n 1-2,'rj
Also we have shown above that it holds for n = 2.
Hence by mathcmatical induction it is true for all positive integral values
of n. . Hence proved.
Ex. 9(h). If [i 11, provcthat Atlr[1
1,
[0 1] L° 1i
where in is positive intgcr. (Kanpur 97, 93)
.Sol. A 2 =A . A=[ 1 il x [i 1
[0 ij [0 i

+ 1.0 - i.ii=[i 2
[0.1 + 1.0 (1.1 i.ij [o i

rA n,whcren2.
it holds when n 2
i.e., A n = [l
[0 U
,n-1-1
= An ... See def. § 1.10 Page 27
=[l nix 1 ii =[i.i +n.0 1.1 +n.l
1. 0 1] o 1] [0.1 flU 0.1 1.1

= I n+1

n
A =1 1 °1 holds for 'n' = n + I
[0 1]
Also we have showsn ahoe that it holds icr a 2. Hence by
mathe niatical induction it i s irue for all positive integral values of n.
Hence proved.
Ex. JO. Let A =b]
[a j! 0. Show that for
[ci i]'

h(i" - 1)
n ^ 0, A" 4'
(a - I)
1 0 1
 36 Matrices 989
Sol. A2 =A .A=I a b i x ia b
[ 1] [0 1
=faa+b.0 a.b+b.l1=1a 2 b(a+l)
[0.a+1.0 0.b+ L1] [o

= a2 b(a2_I)Aflh2
(a — I)
(Note)
A"=d b (an _1)/(a_l)1 holds when n=2.
[o i j
Now AIi = A.A, See def. § 110 Page 27
an b(a" — 1)/(a— 1 )14 0 b
o 1 j [0 i
a.a+0 ab+ I.{b (an — l)/(a— 1))
1[0+0 0+1
=1a' b {a(a— l)+(a— I))/(a— 1)
[0 1
[aP*1 b(d' - l)/(a— 1)
10
A[d b(a'7_1)/(a_l)11IOIdSfOr'flfl+l.
[0 1 ]
Also we have shown above that it holds for ii = 2.
Hence by mathematical iriduettoit it s true for all positive integral values
of n 2! 0. Hence proved.
*Ex. 11. (a) Show that [cos 0 -• sü ]fl =[cos nO - sin
[qin 0 cos 0] [sin nO cos no]
when n is a positive integer. (AL'a(Ih 95, Gorakhpur 90)
Sol. L.t A Tos 0 -. sin 01
inO Cos0]

Then (A2) = A.A = cosE — si no] x [coso —sinO]

i[s n 0 Cos 0] [sin 0 cos 0]

0—sin20 - 0 cos 0 - sin cos

[Si
ll 9 cos 0 +Sjfl 0 cos 9 _ sin 2.O+ Cos 29
J
or (A)2 [scos 20 -sin 20
in 20 cos 20 'ii)
Similarly tA) 3 = (,t)2 -A See def. § 1.1(1 Page 27


Multiplication of Matrices 37
= Fcos 20 — sin 201 x [cos 0 --sin 0]. from(i)and(ii)
[sin 20 cos 20] [sin 0 cos 01
=[cos20cos0—sin20sir)0 — Cos 20 sin 0— sin 28 cos O
[sin 20 COS 0 + cos 20sin 0 - sin 28 sin 0 + cos 20 cos 0
=[cos(20+0) -rsin (28+0)
[sin (20+0) Cos (20+0)
or (A)3 = cos 30 — sin
[sin 30 cos 30] ...(iii)
In the light of (i), (ii) and (iii) let us assume that
(A)n [cos nO — sin nOl
[sin nO cos
no] (iv)
Now (A)' = (A). (A)
Cos no sinne]x[coso
— — sinG
[sin nO cos nO] [sin 0 cos 0
[
=Icos ti0cos 8— sin nOsinG -ces nO sinO—sinnOcosOl
[sin nO cos 0 ± cos nO sin 0 -. sin nO sin 0 + cos nO cos oj
=[cos(nO+0) —sin(nO+0)l=[cos(n+I)8 —sin(n+1)0
1sin 0+0) Cos (nO+0)j [sin (n + 1)0 Cos (. 11 1)0
i.e. (iv) holds for n + I if is truc for n.
We hvc already proved in (ii) and (iii) that (iv) hoLs for it = 2 and 3.
Hence (iv) holds for all positive integral values of ,i.
i.e. ( = [cos 0 — sin °T=[cos nO -sin,iOl
sin 0 cos 0sin nO :os no ] Hence proved.
­ Ex. 11. (b) IfA=f cosO smOl.evau!ateAnl.
sin 0 cos 0] (Garhwal 94, 92, Meerut 97)

Sot. A 2 = A.A = F o01 x [ COS 0 Si 0

sin 0 cos 0] [- sin 0 cos 0

Cos 2 O-- Sir, 2O cos Osin 8+ sin 0 Cos O

=I—sin0cos 0 — cos 0 sin 8 - sin 0 + cos2 0
or (A)2 = [ cos 20 sin 201
[—sin 20 cos 20] .
Similarly (A)3 = ( A)2. A
r cos 28 sin 20
cos 8 sin 0 i
[— sin 28 cns 20j[_ sin 0 cos oj
r
cos 28 Cos 0— sin 20 sin O Cos 20 sin 0+ sin 20cosO
—sin20cos0—cos26sinfl —sin 2Osin0±cos2OcosO


38 Matrices

=[ cos (20-t-0) sin (20+0)

sin (20 + 0) cos (20 + 0)

=1 cos 30 sin 30
sin 30 cos 39 ME
In the light of (i), (ii), let us assume (hat
(A) n = [ cos nO sin nO]
sin nO cos nO] (iii)

Now (A)" = (A)'. A

cos nO sin nol x [ cos sin
sin nO cos nO] sn 0 cos 0
L

cos nO cos 0— sin nO sin 0 cos nO sin 0+ sin n0os 0

n n0cos9—cosnOsin0 —sinn0sin0+cosn9cosO
=[—si
= cos (nO + 0) sin (nO + 0)1 = [ cos (n + !)0 sin ( n + 1) 0
L — sin (ne + 0) cos (n0+0)] 1)0 cos(n+ 1)0

.. (iii) holds for n + 1 if it is true foi n.

We have already prtwed in (I) and (ii) that (iii) holds for n = 2 and 3.
Hence by mathematical induction (iii) holds for all +vc integral values of n and
value of A'2 is given by (iii).
Ex 12. Show that ifAl Cosh e sinhO1
coshoj
L sit0
then A'7 = c osh nO sinh nO ]
Isinh nO cosh no] (Agra 93)

Sol. Here A 2 = A.A

= I coshO sinhoixlcoshfl sinhO
9 cosh 0] Lsmnh 0 cosh 0
L sinh
cosh e + sinh o cosh 0 sinh0
+ sinh 0 cosh 0
sinh 0 cosh 0
+ cosh 0 sinh 0 sinh 2 6 + cosh 0

or A2 [cosh 20 sinh 20
Lsinh 20 cosh 20

Similarly A3 = A2. A
= Icosh 28 sigh 201 x 1coh 0 sinli 01, from (i)
Ls1 29 cosh 201 [sinh 9 cosh 0]
= Icosh 20 cosh 0 + sinh 20 sinh 0 cosh 20 sinh U + sinh 20 coh 8
[sinh 20 cosh 0 4 cosh 20 sinh 0 sinh 20 sinli 0 + cosh 20 cash 0

Multiplication of Matrices 39
cosh (2(3 +8) sinh (20 + 0)
sinh (2(3 + 0) cosh (20+0)
[
or A3 =[cosh 30 sinh 30
[sih 30 cosh 30 ...(ii)
In the light of (i), (ii) and the given value of A, let us assume that
=I
A n cosh riO sinh nO
si t0 cosh nO] (iii)

Now A'1
= {.cosla rIOsinh °1 xfcosti 0 sinh 0
nO cosh nO] [sinh 0 cosh 0
Ic o sh nO osh 0 + sinh nO sinh 0 cosh nO sinh 0 + sinh nO cosh 0
[sihh nO cosh 0 + cosh nO sinh 0 sinh nO sinh 0 + cosh nO cosh 0
=fcosh(ne+0) sinh(nO+0)1=[cosh(n+ 1)0 sinh(n+ 1)0
[inh (nO + 0) cosh (nO + 0)] [sinh (a + 1) 0 cosh (n + 1)0
i.e. (iii) holds for n + 1 if it is true for a.
Also from (1) anti (ii) we know that iii) holds for a = 2 and n = 3. Hence
(iii) holds for all positive integ r al values of a.
i.e. A' 4cosh nO sinh nOl
Isiah nO cosh no] Hence proved,
[Note. See Ex. 13 Page 20 also].
Exercises on § 109 - § 110
EL 1. Show that the matrix A = 1 21 satisfies the equation
13 1]
A2 - 2A -51 0, where 0 is the 2 x 2 null matrix.
Ex. 2. Evaluate A 2 - 3A - 131, where I is the 2 x 2 unit matrix and
A=[2 51A ns. 01.10
[3 ] Lo Oj
Ex. 3. Show that matrix A = I 0 0
2J0
321

satisfies the equation A 3 - 3A 2 + 3A - I = 0, where I is the unit matrix and 0

the null matrix of order 3.
Ex 4. lfA=12 3 1, B 13 —2],C41 2
[4 _iJ [2 1] [3 4
verify (i) (AD) C A (BC): (iii (A +B) C =AC+ BC.


40 Matrices
EL S. If A = 1 2 1, B = [2 ii; C = [i i i, show that
[3 4] [4 2] [7 4]
A(B+C)=AB-tAC.
Ex. 6. If A=1I 2] and B = [ 1 0], show that
0 L'
(A + B) (A + B) A2 + 2AB + B2.
Lx. 8. Show that I I I T = 1 n n (n + I)
0 1 I 0 1
001] 00 1
for all natural numbers n.
(Hint. See Ex. 10 Page 351
SOME TYPICAL SOLVED EXAMPLES
•SLx. 1. A manufacturer produces three products A, B, C which he
ses In the market. Annual sale volumes are indicated as follows:
Markets Products
A B C
I 8,000 10,000 15,000
II 10,000 2,000 20,000
(i) If unit sale prices of A. B and Care Rs. 2-25, Rs. 150 and Rs. 125
respectively, find the total revenue in each market with the help of
matrices, (ii) if the units costs of the above three products are Rs. 160, Rs.
1-20 and Rs. 090 respectively, find the gross profit with the help of
matrices.
Sot. (i) The total revenue in each market is given by the products matrix.
[2-25 1-50 1-251x 8,000 10,000
10,000 2,000
15.000 20,000 (Note)
= [(2-25 x 8,000) + (150 x l0,000)+ (1-25 x 15,000)
(225 x I0,000)+(1-50x 2,000)+(l . 25 x 20,000)J
= [ 18 ,000 + 15.000+ 18,750 22.500+3000+ 25,0001
= [51750 505001
Total revenue from the market I = Rs. 51,750.
and total revenue from the market H = Rs. 50,500.
(ii) Similarly the total cost of products with the marmufactujer sells in the
markets are:
 Typical Solved Examples 41
^1-60 1 . 20 0-90IX 8,000 10,0001
10,000 2,000
1 15,000 20,000

=[(1-60x 8,000)+(1-20 x 10,000)+(0-90 x 15,000)

(1-60 x l0,000)+ (120 + 2,000) + (0-90 + 20,000)]
=[12,800+ 12,000+ 13,500 16,000+2,400+ 18,000]
= (38,300 36,400)
Total cost of products which the manufacturer sells in the market I and
II are Rs, 38,300 and Rs. 36,400 respectively.
Required gross profit = (Total revenue received from both the
markets) -- (Total costs of product which the manufactuerr sold in both the
markets)
= (Rs. 51,750 + Rs, 50,500) - (Rs. 38,300 - Rs. 36,400).
= Rs. 102,250 - Rs. 74.700 = Rs. 27,550. Ans.
Ex. 2. A man buys 8 dozens of mangoes, 10 dozens of apples and 4
dozens of bananas. Mangoes cost Rs. 18 per dozen, apples Rs. 9 per dozen
and bananas Rs. 6 per dozen. Represent the quantities bought by a row
matrix and the prices by a column matrix and hence obtain the total cost
(1. C. W A. Final)
Sol. The quantities bought are represented by 3 x 1 row matrix (8 10 4)
and the prices are represented by 3 x I column matrix
18
9
6
The cost of fruits is.a single nuniber i.e. I x I matrix given by the
product matrix [8 10 4] x 18
9
6

i.e. [(8x 18)+(10x9)+(4x6)) i.e. [144+90+24] i.e [258]

The required total cost = Rs. 258. Ans.
EX. 3. A store has in stock 30 dozen shirts, 15 dozen trousers and
25 dozen pairs of socks. If the selling prices are Rs. 50 per shirt, Rs. 90 per
trouser and Rs. 12 per pair of socks, then find the total amount the store
owner will get after selling all the items in the stock.
Sol. The stock in the store can be written in the form of a row matrix A
given byA=[20x12 15x12 25x12]
or A=[240 180 300), which isa lx3rnatrix.
42 Matrices
The prices can he wr,ttcn in the form of a column ma;rix B given by
B=I 50 . which is 3 x matrix.
QI)
12
L
The required amount is a single number i.e. a matrix of order I x I and
so the same can be obtained by multiplying the matrices A and B, since their
product would be a 1 x I matrix (Note)
Now AB= [240 ISO 300]x150
90
12
= [(240 x 50) + (180 x 90) + (30)x 2)]
= [12000 ± 16200 + 360001 = [31800]
The required amount received by the store owner
=Rs, 31,800. Ans.
Ex. 4. A trust fund has Rs. 50,000 that is to be invested int3 two types
of bonds. The first bond pays 5% interest per year and the second bond
pays 6% interest per year. Using matrix multipkation, determine how to
divide Its. 50,000 among two types of bonds so as to obtain an annual total
interest of Rs 2780.
Sol. Let Rs. 50.000 hc divided into two parts Rs. x and Rs. (50.000 -
out of which first part is inve s ted in first type of bonds and the second pr.rI is
invested in second type of bonds.
The values of these bonds can be written in the form of a row matrix A
given by A = k 50,000-4 which is a ! x 2 matrix.
And the amounts received a interest per rupee annually from these two
types of bonds can be written in the form of a column matrix B given by
B =[5/100 1, which is a 2 x I matrix.
iiooj
L
Here the interest has been calculated per-rupee annually.
Now the interest , to he obtained annually is a single number i.e. a matrix
of order I x 1 and the same can be obtained by the product matrix AR, since
this product matrix would be a I x I matrix, (Note)
Here AB Ix 50,000—x)x5/100
16/100

[x.__+(50.O00_x): i]

=[30001]

Also we are given that the annual interest 2,780.



Typical Solved Examples 43

We must have [3000 - = 127801

(Note)

or 300O-j- 2780 or x=(30-2780)x

00 100
or x=220x100=22,000
Hence the required amounts are
Rs. 22,0(X) and Rs. (50,000 - 22,000) i e Rs. 22.000 and Rs. 28,000 Ans.
Ex. 5. A finance company has offices located in every division, every
district and every taluka in a certain state in India. Assume that there are
five divisions, thirty districts and 200 talukas in the state. Each office has
one headclerk, one cashier, one clerk and one peon. A divisional office has,
in addition, (PflC office superintendent, two clerks, one typist and one peon.
A district office, has in addition, one clerk and one peon, The basic
monthly salaries are as follows office superilendent Its. 500. Head clerk
Rs. 200, cashier Rs. 175, clerks and typists Rs. 150 and peon Rs. 100. Using
matrix notation find
(I) The total number of posts of each kind in all the offices taken
together, (ii) the total basic monthly salary bill of each kind of office and
(iii) the total basic monthly salary bill of all the offices taken together.
(C. A. Irnermediaze)
Sol. Let us use the symbols Div. Dis, Thl for division, district, taluka
iespectivcly and 0, H, C, Cl, T and P for office superintendent, Head clerk,
cashier, clerk, typist and peon respectively.
Then the number of offices can be arranged as elements of a row matrix
A (say) given by
Div. Dis. Tal.
A=(5 30 200)
The composition of staff in various offices can he arranged in a 3 x 6
matrix B (say) given by
O H C Cl T P
111= 1 1 1 2+1 I 1+1
0 1 I 1+1 0 1+1
0 1 1 1 0 1

The basic monthly salaries of various types of employees of these offices

correspond to the elements of the column matrix C (say) given by
C=O 500
H200
C 175
Cl 150
1 150
P 100
44 Matrices
(1) Total number of posts of each kind in all the offices are the elements
of the product matrix AB.
i.e. [5 30 2001 I I I 3 1 2
011202
0 1 1 1 0 1 (Note)

i.e. [5+0+0, 5+30+200, 5+30+200, -15+60+200.

5+0+0, 10+60+2(X)]
0 H C CIT P
5 235 235 275 5 270
i.e. Required number of posts in all the offices taken together are 5 offices supdts.,
235 Head clerks, 235 cashiers, 275 clerks, 5 typists and 270 peons. Ans.
(ii) Total basic monthly salary bill of each kind of office are the elements
of the product matrix BC
i.e. 1 1 I 31 2x500
0 1 1 2 0 2 200
0 I 1 1 0 1 175
150
ISO
[ 100
= (I x500)4 (I x200)+(l x 175)+(3x 150) -+ (lx 150)+(2.x 100)
(0x500)+(l x200)+(1 x 175)+(2x 150)+(0x 150) -s-(2x 'Y))
(0x500)+(1 x200) -(1 x 175)+(l x 150)+(0x 150)+(l x 100)

1-500 200+175 + 450+150+200 = 1675

0+200+175+300+ 0 +.2 00 875
0+200+175+150+ 0 +100 625 Ans.

i.e. The total basic monthly salary bill of each divisional, district and taluka
offices are Rs. 1675, Rs. 875 and Rs. 625 respectively. Ans.
(iii) Total basic monthly salary bill of all the officers (i.e. of five
divisional, 30 district and 200 taluka offices) is the element of the product
matrix ABC
i.e. [5 30 200] x 175
875
625 (Note)

i.e. [(5 x 1675) + (30 x 875) + (200 x 625)1

i.e. [8375+2650+1250001 i.e. [159625]
i.e. total basic monthly salary bill of all the offices taken together is
Rs. 1,325. Ans.


Miscellaneous Solved Examples 45

"Ex. 6. In a development plan of a city, a contractor has taken a
contract to construct certain houses for which he needs building materials
like stones, sand etc. There are three firms A, B, C that can supply him
these materials. At one time these firms A B,C, supplied him 40, 35 and 25
truck loads of stones and 10, 5 and S truck loads of sand respectively. If the
cost of one truck load of stone and sand are Rs. 1,200 and Rs. 500
respectively, then find the total amount paid by the contractor to each of
these firms, A, B, C separately.
Sol. The truck-loads of stone and sand supplied by the firms A, B and C
can he written in the from of a matrix A (say) given by
A BC
A = Stone[ 40 35 251 which is a 2 x 3 matrix
Sand [l0 5 8]

And the cost per truck of stone and sand can be given in the form of a
matrix B (say) gtvcn by
Stone Sand
B=(1200 5001
The required total amount paid to each of the firms .4, B and C are given
by the product matrix BA. (Note AB can not be calculated).
Now BA = [1200 500] x40 35 25
110 5 8
= [(12(8) x 40) + (500 x 10) (1200 x 35) + (500 x 5)
(1200x 25) 4 (SOOx 8)]
148000+ 5000 42000 + 2500 30000 + 4000J
= (53.00C) 44,500 34,000]
Tl:c amount paid to the firms A, B and C by the contractor are Rs,
53.000, R.c. 44,500 and Rs. 34,001) respectively. Arm.
Exercises
Ex. 1. A hua seller has in stock 20 dmen rnangocs, 16 dozen apples and
32 docn haaarias. S'ippose the selling prices are Rs. 0.35, Rs. 0.75 and Rs.
0.08 per rn:rngo, apple and banana respectively. Find the total amount the fruit
seller will get by saIling his whole stock. Ans. Rs. 25872
Ex. Z. In Ex. -1 Pagc 3 write down (0 the row m:itix which represents
team B's result: rt) the clunin matrix which rcprc:;cnt the results of first places
of vari.us tcams. (0 3 2 41 and 1 31
0
5
L
—4


46 Matrices

MISCELLANEOUS SOLVED EXAMPLES

*Ex. L If A= [l - 1], i= [a 1
[2 —1] [b —1

and (A + B) 2 = A 2 + B2, find a and b. (Kanpur 96)

Sol. Here we have
A2 =[1 —1]x[l —i
[2 -] [2 —1
=[I_2 _i+il=[-i
[2-2 —2+1] [ 0 —1
0

B2=Ia i ] X ^ I]=[2+1 a-
[bi 1 ab—b b -fl

A2 +B2 [_ 1 O1+[a 2 +b a — Il
[ 0 1] [/_h b+1]

l+a2 +b 0+a_114a 2 +b_1 a-il

0+ab—b _1+b+ij lab — b b j
[
Also A+B=[l — il+[a I
[2 —1] [b —1

=[I +a —l+l][I+a 0
[2+b - I - I] [2+b —2

(A+B)2=[1+(2 O l x[I -fa 0

• - [2+b -2j [2+b —2

=[(I+a) 2 +o 0+0
[(2--b)(I +a)-2(2+b) 0+4

[(2 + b) (a —1) 4] (ii)

Now it is given that (A + B)2 = A 2 + B2.

+ )2 [U2 a -- I from (i) and (ii)
or (1 01 h -
ah — b /3
1 (2 + b) (a — 1) 4]
or 0=a - 1 and 4 = b, comparing :hc elements of second column or both
sides-
or a= I and b=4.
Ex. 2. If A = 1 2 and B = - 3 - 2
34 1-5
56 4 3


Miscellaneous Solved Examples 47

find 1) p q , such that A + B - D =0.
rs
t U

Sol. A+B- D=0 or D=A+B

or D= 1 2 + -3 -21
3 4 I -5
5 6 4 3]
11-3 2-21 = [-2 0 = 11' q given
3+1 4-5 4 -1 Jr s
9 9 1U
L54 6+3 L
We have p -2, q--O, r = 4, 5 - I, t = 9, u = 9 which gives D.
If A = 0 1 0 and I is the unit matrix ororder 3, show that

pqr

= p1 + qA + rA2.
Sol. Here A2 =A.A= 01 0 x 0 1 0
001 001
pqr pqr

0+0+0 0+0+0 0+l+0 0 0

0+0 4 p 0+0+q 0+0+r p q r
0+0+rp p+0+rq 0+q+rl 2
rp p+rq q + r

A S= A 2 *A =[O0 1 x 0 I 0
P q r 00 1
rp p+rq q+r2 p q r

[0+0+p O+O+q 0+0+r

0+q+ r2
[0+0+pq+p, rp+0+q 2 +r2 q 0+p+rq+n7+r

=1 rp p q
p+rq
pq ip+q2 +r2q p+2rq+r ]
[+pr2

And pI -,- qA+rA2

=P 1 0 0 4-q 0 1 0 +r 0 0 1
010 001 p q r
0 0 lj p q r i7 p±rq q+?



49 Matrices 182/1/3

0 r
=p00^0 q 0+0
o p 0 •0 0 q rp. qr
0 0 p pq q2 rq r2 p pr+q? qr+r3

0+q+O 0+0+r
= p±O+O
0+0+rp p+0+pr O+q+r2
0+pq+?p O+q 2 + pr + qr2 p+2rq+

= 43
from (ft Hence proved.

Ex. 4. Show that E 2 F + OE = E, where

E=0 0 1,F=1 00
001 010
000 001

So!. E 2 0 0 1 >( 0 0 1
001 001
000 000
= 00+00+ 10 0-0+00+ 10 01 +0-1 + 1.01
00+0-0+1-0 0-0+0-0+1-0 01+0.1 + 1-0
0 . 0+0 . 0+00 00+00+00 01+0 1+00 .
1

or E2= 0 0 0
000
0 0 0

E2 = 0 0 OxI 00=000
0 00 0 1 0 0 0 0
0 0 0 0 0 1 0 0 0 (j)

Again F2 41 0 0 x 1 0 0
0 1 0 0 1 0
0 1 0 0 1
= i . 0+0-0+00 10+01+00 1.0+00+1.01
10+10*00 00+1-1+00 00+10+0-1
01+00+10 00+01+1-0 0-0+0.0+T-I-
, 00
010 -
001

F2E= 1 0 Ox[() 0 I
0 1 o! 0 0 I
0 0 .0 0 0
IL J -


18211/4 Miscellaneous Solved Examples 49

= 10+0-0+00 10+00+00 11+0-1+00

00+1-0+00 00+1-0+00 0-1+11+00
00+0-0+ 10 00+00+1-0 01+0-1+10

= 0 0 I
001
.(ii)
000
From (i) and (ii) we get

0 0 0 I
1+
E 2F+F2 E= 0 0 0 10 0 1 = 0 0 1 =E
0 1 0 0 1
0 0 0 .0 0 Lo
0 0 0 Hence proved.

Ex. 5. If A = 12 3 , find the matrix X such that A + X + 1 0,

3 —2 1
4 21
where I and 0 are unit and zero 3 x 3 matrices respectively.
Sol. Given that A+X+IOorX=O—A--I
0 00-1 23-100,
000 3 —2l 010
000 4 21 001
subsutuing values of A, I and 0
= 0—I - I 0-2-0 0-3-0 = —2 —2 —3
0-3-0 0+2-1 0-1-0 —3 1 —1
0-4-0 0-2-0 0-1-1 —4 —2 2 Ans

**Ex 6. Show that

1cose510=r -_tan0l[i
][_ tan 0
tan ! ()
F
[sinG cosoj [tan.O 1
Sol. We have
[cos 0 - sinol x 1 (nfl 9
sill cosoj —tan9

= cosO+sinOianO cosOtanO—sinO
sin O— cos 9 tan 0 sin Btan 0+ cos G
- cosOcosO+siriGsin0 cosO sin 9—sin0cosO
cos LO Cos 10
sin 0 cos --G—cos Osin . O sin Osin 0+cos Bcos 0
cos -8 cos0
 50 Matrices
1 fcoso cos . 9+ sin o sin! 0 cos 0 sin -O- sin 0 cos 0
cos0[sin9cosO-cosOsin8 cos9cos-O+sin0sin9
=(sc0) COS (00 ) -sin(0-10)
1 sin (G- . 6) cos(0-0)
=(Sec . e)[ cos 8 - sinO
[sin . 0 cos .0
=[cos.0sec0 _ sinOsecO 1_1 1 tan 10
[sin f0sec .. e Cos iO sec .ej L tan 0 i

rcoso —sinel=I I _ tan ! o1r 1 tan!2 0]

or I . 0 cOS OJ [taM
i i
1 i L U 1 j Hence proved.
L51

Ex. 7. if A and B be n-rowed square matrices, then show that

(1) (A + B)2 =A 2 ±AB + BA + B2
(li)(A+B)(A-B)=A2-AB+BA-B2;
(iii)(A-B)(A+B)A2+AB-BA-B2;
and (iv)(A-B)2=A2-AB-BA+02.
Sol. As A and B are n-rowed square matrices therefore A + B and A — B
are also n-rowed square matrices and as such distributive law is true.
(i). (A + B)2 = (A + B) x (A + B)
= (A + B) A + (A + B) B, by distributive law
= AA + BA + AB + BB, by distributive law
A2 + BA + AB + B2
=.
(il)(A+B)(A-B)
= (A + B) A + (A + B) (- B), by distributive law
= AA + BA + A (— B) + B (— B), by distributive law
=A2+BA-AB-&
(iii) (A - B) (A + B) = (A — B) A + (A - B) B, by distributive law
= AA - BA + AB — BB, by distributive law
= A 2 - BA + AB - B2.
(iv) (A — B)' = (A - B) • (A - B)
= AA + A (— B) + (— B) A + (— B) (— B), by distributive law
= A2 — AB — BA + B 2. Hence proved.
*Ex. 8. 11 A, B are two n x n matrices and if
C= A+ B, AB = BA, B 2 =0
then show that for every integer m, = Am tA + (in + 1) BJ.
 Miscellaneous Solved Examples 51

shall prove that 1. = A [A + (m + 1) B],

by mathematical induction.
Form I, from (I) we get C 2 A[A+2B] (ii)
Also C = A + B, given
:. C2=(A+B)2=(A+B)(A+B)
=A 2 +BA+AB+B 2 , asin Ex. 7(i) Page 5O.
= A 2 + 2AB, since AB = BA, B 2 = 0 (given)
or C2 = A (A + 2B), which in the same as (ii).
rn = I.
Hence (i) is true for
Let us now assume that (i) holds when m k
i.e. Ck+1 = A (A + (k + 1) B) ...(iii)
Now C + 2 = Ck ' C, by def. § I 10 Page 27.
= A [A + (k + 1) B]. (A + B), from (iii) and C A + B(given)
or CZAkIA(A+B)+(k+1)B(A+B)]
= A [A 2 + 103 + ( k + 1) BA + (k + 1) B2]

A [A
24
1 R + (k + 1) AB ], . BA = AB, B 2 = 0
=Ak[A2+(1+k+ 1)AB]
=Ak.A[A+{(k+ 1)+ 1}B]
or Ck4.2=A!LA+{(k+ 1)+1}B).
Hence (i) is true for , = k + I provided (iii) is true i.e. for rn = k. Also we
have shown that (i) is true for m = I, so it is true for rn = 1 + 1 i.e. m = 2 and so
on. Hence by induction (i) is true for all positive integral values of m.
Hence proved.
Ex. 9. If A = 2 0 0and B = xi yl Zj
o 2 0 x yz Z2
o o 2 X3 13 Z3

then prove that AB = 2B.

Sol. 1B = 2 0 x 01 Al Vj Zl
o 2 0 X2 Y2 Z2
() 0 2 .r Y .3

= 2x1+0+0 2v'+0+0 2zi+0+0

0+2x2+0 0+2 y2+() 0+2z2+0
0+0+2x 0I-0+2V3 0+0+2z

A 19


52 Matrices 989

= 2.ri 2y j 2zi =2 x i y i Zi =2B

2x2 2y2 2Z2 X2 '2 Z2
2.x3 2y3 2Z3 X3 )'I Z3 Hence proved.
Ex. 10. If A, B are two matrices given below, which of the two
statements is true AB BA or AB *-BA.
A= 23 4;B= 130
1 2 3 -1 2 1
-1 1 2 0 0 2
Sn. Do yourself. Ans. AB*BA.
Ex. 11. Find n if [a 4 1] x 2 i 01 a=O,
02 4
0 2 4 -1
where 0 is I x I null matrix.
So!. [a 4 l]x 2 1 0
102
024
= [2a+ 4 +0 a+0+2 0+8+4] (Note)
=[2a+4 a+2 12]
[a 4 11x2 1 Ox a
102 4
0 2 4 -1
=[2a+4 a+2 12]x a

-
=[((2a+4)xa}+(a+2)4+ 12(- 1)] Ans.
[2 + 4a + 4a + 8- 12] = [ 2a 2 + Sa -4] = 0 = [01, given
2a +8a-4=0 or a2+4a-2=O
or a=[-4±'I(6+8)]=--2±'i6. A.
**EX 12. Show that if A, B, C are matrices, such that A (BC) is
defined, then (AB) C is also defined and A (BC) = (AB) C.
So!. Since A (BC) is defined so the matrices A, B, C are conformable to
multipictions and we can take A = [ a 1], B = [b] and C [ckJ], where A, B. C
are m x n, n x p. p X q matrices.
Then AB = [aij ] [ b1kJ is an m x p matrix

i.e. (i, k)th element of the product AB = Z a, b1k (Note)

• j=I
P
Sirnrnarity (1. 1)Lh element of the product BC = h. ei (Note)
 Miscellaneous Solved Examples 53
Also (AR) C is the product of an rnx p and a p x q matrices and so is
conformable to multiplication, hence defined.
(z, !)th element in the product of (AB) and C
= sum of products of corresponding elements in the ith
row of AB and lth column of C with k common
P
= a1h1t kkt (Note)

k=l J11 )
P a
= L (iij bj k CkI
k= lj = I
Again (i. l)h element in the product of A and (BC).
= sum of products of corresponding elements in the ith
row of A and lth column of (BC)
a p
= Z aO L bjk cj (Note)
j=t k=I
P a
= L Uij bjk Ckl ...(ii)
k = [j= I
From (1) and (ii) we conclude that (AB) C = A (BC)
*Ex. 13, If A and B are two matrices such that AB and A + B are
both defined, then prove that A and B are square matrices.
Sol. Let A be an m x n matrix.
Since A + B is defined i.e. A and B art conformable to addition, so B
must also be an m x n matrix.
Again AB is defined i.e. A and B are conformable to multiplication and
hence the number of columns in A must be equal to number of rows in B i.e.
ri=
Hence A and B are m x m matrices i.e. square matrices.
**Ex. 14. If AB = BA then prove that (AB) = A"B".
Sol. We shall prove this by mathematical induction.
If n = I, then (AB)' = A' B" = (AR)' AR, which is true.
If n = 21 , then
(AB) = ( AB)' = (AB) (AB)
= (ABA) B, by associative jaw
= (AAB) B, BA = AB, given
=A2B2.
n
Hence (AB)" = AB is true for n = 2.
'a = AmBm
Now suppose that it is true for n = rn i.e. (AD)
or (A B) tm (AB) = (A m B tm ) (AB)


54 Matrices

or (AB) m+ = Atm (Btm A) B, by associative law

= Am (Bm BA) B, . Btm = Bml B
= Am ( Bml AB) B, . BA = AB, given
= Am (Bm2 BAB) B, Bm1 = Bm2 B.

= A m (Bm-2 ABB) B, BA = AB, given

= A m (Bm2 AB2) B
m
= Atm (AB m-2 B 2) B = (A A) ( B m_2 B2B)
or (AB)'' = A m Bm

i.e. if (AB) = A B n is true for n = rn, it is true for n = m + I.

Also we have proved that it is true for n = I and 2.
Hence by mathematical induction it is true for all +ve integral values of n.
*Ex. 15. If A = 0 1, prove that A" = 12, A, - I, - A according
L° ]
asn=4p, 4p+I, 4p+2 and 4p+3 respectively.
SoL Given A= i 0
o i ( 1)
A 2 =A.A= i 0 x Ii 0
o i [0 i

= I i.i + 0.0 iO+0.il=1i2 0 1

[0.1 + i.0 0.0+ ii] [o i2 j (ii)
A3=A2SA=Fi2 o lxri 0
[o i2] L°
=1i2 .i+0.0 2 . o+o . j = [3 0
1
[o j3j
Lo.i + i2 .0 0.0 + i2 .i (iii)
3=
From (ii) and (iii) we get A2 =[i 2 0],A 1i 01
[0 12 [O i3j

Let us assume that A"= [ii' 0]

0 ...(iv)
and also assume that (iv) is true when n = k.
i.e. Ak=[i 01
jkj
[o
i O l x [j 0], from(v)and(i)
AI1=AklA=[1*
o jkj [


Miscellaneous Solved Examples 55

=1i+o.o k0+0 =[
Lo.^.o o . ø +i* jo i

(i\) is true for n = k + 1 provided (v) is true.

Also we have shown in (ii) and (iii) that (iv) is true for n = 2 and 3. So it
is true for 3 + 1 i.e. 4 and so on.
Hence (iv) is true for all positive integral values of n.
Also if n = 4p then from (iv) we et
A= [ i 0 1= 1 1 0l,sincei4=(j)P=(1)P=1,
[o i'] [0 1] where i='J(—l)

or A = 12. Hence proved

Ifn=4p+1, then t1=i41=(i)4h).i=1.i=i
From (iv),we get Afl=I' °1=1' o]A.
0M in [0 ij Hence proved.

Ifn=4p+2, then
=(l)(—l), since i4 '=1, i2=-1
From (iv), we get

A= I
°11 0 1=- I l 0]=-12
10
,n in] [ 0 - 1] [0 1] . Hence proved.

Ifn=4p+3, then in= 3=(142).i=(— 1)1, as above

From (iv), we get

An= in °1_1-i .01=—A
o in] JO
1-10i i] Hence proved.

Ex. 16. Eva1uat[cos 0 + sin 0 42 sin 01

sin0 Cos O_ sin O]

Fcos9+sin8 42 sine
Sot. Let A= ,
L
'2 sin 0 cos 0— sin 0
Then A 2 = A.A
r [cos9+sin0 l2 sin 0][cos0+ sine '12 sin 0
[- 42 sin 9 cos 0—sin 9] [—'12 sin 9 cos 0— sin 0

(cos 9+
= sin 0)2 -2 sin 2O (cos 0-s- sin O)'12 sin 9
+'12sin eçcoso — sin 0)
- '12 sin 0 (cos 0 + sin 0) —42 sin 8 '12 sin 8
- 42 sin 0 (cos 0 - sin 0) + (cos 0 - sin 9)2


)t) Matrices

= 1(Cos 2 o — sin 0) + sin 0 cos 0 2 42 sin 0 cos 0

2 42 sin 8 cos 8 (cos2 0 - sin 2 0) - 2 cos 0 sin 0 ]
(Note)
or 2 [Cos 20 + sin 20 '12 sin 201
—'12 sin 20 cös 20—sin 20] ...(ii)

, Looking at (i) and (ii) let us assume that

A n [cos nO + sin nO 42 sin nO
[_'12 sin nO cosnO — sinnO
Let (iii) be true for n = k
i.e. A [cos k8 + sin kO 12 sin
42 sin kO cos k13 - j fl kO] (iv)

A=Ak.A
= [cos kO + sin k8 '12 sin kO l x s 0 + sin 0 '12 sin 0
cos kO - sin kO] [co s[— 42 sin 0 cos 0 - sin 0
— 12 sin kO
= (cos kO + sin kO) ( cos 0 + sin 0) (cos kO + sin kO) (42 sin 0)
+ (12 3in kO) (—'12 sin 0) + (42 sin k8) (cos 0— sin 0)
—I2 sin kO (cos 0 + sin 9) (—'12 sin k8) ('12 sin 0) +
+ (cos kf3 — sin kO) (— 42 sin 8) (cos kG — sin kG) (cos 0—sin 0)
'12 (sin k8 Cos O
= Cos kO Cos O+ Cos kO sin O+ sin kO cos 9 + cos k0 sin O)
-i-sinkOstnO-2 sin kOsin0
—2 sin kG sin 0 + cos kG cos 0
— Cos k8 sin 0— sin k8 Cos O
(sin + sin kG sin 0
= [cos (kG + 9) + sin (kG + 0) 42 sin (kG + 0)
—'I2sin(kG+O) cos(kG+9)—sin(kO+0)
L
=[Cos (k+ 1)0 +sin (k+ 1)0 '12 sin (k+ 1)0
—'12 sin (k-s-1)0 Cos (k+1)0— sin (k+ 1)0
[
(iii) is true for n = k + 1 provided (iv) is true.
Also we have shown in (ii) that (iii) is true. for n = 2.
Hence it is true for n = 2 + I i.e. 3 and so on.
Hence (iii) is true for all positive integral values of n.
Hence A' =[cos nO + sin nO 42 sin nol
[ —12 sin nO cos nO — sin noj Ans.

*Ex. 17. II P(x)=[ cosx sin x l, then show that

sin cos x]
P (x) .P (y) P (x + y) P (y).P (x)
 Miscellaneous Solved Examples 57

Sol. P (). P (y)

= cos x sin x lxi cos v sin y

sinx cosxj [—sillY cosy
I—
= [cos X COSY — Sin X Sin V cos X Sin V + Sin X cos V
—SInxsinv+COSXCOS

cos (x +y) sin (x+y)l=r(x4-y)

sin (x + v) cos ( + v)j

Similarly we can prove (to be proved in the exam) that

P (y) • P (x) = P (x + y)
Hence P (x). P (y) = P (x + y) = P (y). P (x) Hence proved.

Ex. 18. If 0 1 1, find number a,bso that (al+bA)2 =A

[-1 0]

Sol. aI±bAal O l^ b [ 0 1
10 ] [-1

=ia 0] +[ a h b1[
[0 a] [_b 0] [_b a

(al +bA) 2 =[ a b]x[ a b

[—h a] [—b a
2ab
= I a 2 — b2 ab + bal = 1a2 — b2
[-ab-ab _b2 -+-a2 ] 2ab a2—b2

If (a! + bA)2 = A. then we have

a 2 — b 2 2ah l = [ 0
—2ab a2 — h2j L I C)

Equating the corresponding elements, we have

61 2 —b 2 =0,2ab=la=b=1/2 Aris.

*Ex. 19. If e 1 is defined as I + A + (A 2 /2 ) + (A 3 /3 !) + ..., then show

that e = e Icosh x sinh xl , where A = r xl
sinh x cosh xj L xj (Budelkhand 95)

Sol. Given that A [x x

[x x

A 2 [ It xl= [•- +LX LX+ LX

x x I I x_ri- x.x X.X + t.X
L J L


58 Matrices
=2[x2 2
12 x
2
LX
A 3 =A 2 .A=2 [XX2 x2 [
x ] =2 x 2 .x+x2 x x2.x-*x2.x
x2 2 [x xj x2.x+x2.A x2..x+x2.x

=22[X3 .
3 I
I
3
In a similar way we can prove that
[X4
A4 = 2A 5 = 2 1x x 5 etc.
4 41 15 I 5
1 X J LX
n
In general A = 2" [ x,- I'
VI" (i)
Now we are given that
• eA=!+A+(A2/2!)+(A3/3!)+...
or eA [i o1+[x x]+ 2 I 2 2] 22 + 2fl_1 1I'2 x'I
[O I] L i
L1 2] +
Lx x3j T xn
+ ...... . =Iu v1,
• Lv u] ... (ii)

22 2213
1
where u=I+x+— + +...+
2! 3! n!
21 2 23x3 2l_Ix?
v=O+x+ + +...+
3'
or u=.[2+21+ + 1
'I- n!

=[i+{1^21+L++
1r
++..J1 2' 3! Jj
= - [I + e2J

and 7
2 R 2 n

=[e-1j

From (ii), we get

e2x+ I
e A_I[
e—1
2 e — I e2'+I


Miscellaneous Solved Examples 59

- I1 (e + e -x ) e' (eX - e)1 = ex 1(ex + e_X)12 (el - e_X)12

- 2 Lex (e - e) e t (eX + e - e_X)12 (e x +

ex cosh sinh x1
[sjh x cosh j Hence proved.

EXERCISES ON CHAPTER I
Ex. 1. Given A = 12 - 3 and B = 3 - 1 2
5 0 2 4 25
1-1 1 2 03
find the matrix C, such that A+C=B. Ans. 2 —3 5
—1 2 3
1 12
Ex. 2. If A= 1 1 2 hand B=
[4 0 2j
—2 2

find AB and show that AB * BA.

Ex. 3. Find AB and BA if
A= 3 4 —2 and B= —1 —1 —1
—2 —I —1 2 2 2
—1 —3 —1 1 1 1
Ans. AB= 7 7 7 , BA= [O 0 0
—1 —1 —1 0 0 0
—6 —6 —6 0 0 0
Ex. 4. If A= 2 —3 —5 and B= 2 —2 —4
—1 4 5 —1 3 4
1 —3 —4 1 —2 —3

verify that AB = A and BA = B.

Ex. 4. Find A and B, where
A+2B= I 2 0, 2A—B= 2 —1 5
6 —3 0 2 —1 6
—5 31 0 12
Ans.A= 1 0 2,B= 0 1 —1
2 —1 3 2 —1 0
—1 1 1 —2 1 0
Ex. 6. If A= I 1 —1 ,B= 1 3
2 0 3 02
3-1 2 14
and C=[1 2 3
[2 0 — 2
-11]
prove that A(BC)=(AB)C
bu Matrices

Ex. 7. If A 1 0 0 0 show that A 2 = Lt, where

1 —2 0 0
1 —1 I 0
I —3 3 —1

14 is 4 x 4 identity matrix.
Ex. 8. For two matrices A and B, state the conditions under which (i)
A = B; (ii) AB exists and (iii) (A + B) 2 = A 2 + 2AB + B2.
Ex. 9. State true or false in the case of the following statement. Justify
your answer.
If A and B are conformable for addition, then
(A+ B)2=A2+2AB+B2.
EL 10. If A = 3 —4 , B = 1[ 2 1 21, then find [Au2.
I I i 2 4j
2 0
Ex. 11. What is the difference between zero matrix and a unit matrix 7
[Hint: Sec § 1 . 03 Page 41
Ex. 12. Find non-zero matrices A and B of order 3 x 3 such that AB = 0.
where 0 is the zero matrix of order 3 x 3.
[Hint: See Ex. I (c) Page 14 or Ex. 7 Page 171


Chapter II
Some Types of Matrices
§ 201. Triangular Matrices. (Bundelkhand 94)
(a) Upper Triangular Matrix. A square matrix A whose elements aij = 0
for i >j is called an upper triangular matrix.
Forexample a a12 a13 11 in
o a22 n 23 (l7
o 0 a- ......

o 0 0 ......
(b) Lower Triangular Matrix. A square matrix A whose elements au
for i<j is called a lower triangular matrix.
Forexample all 0 0 0
021 a22 t) 0
a3l 1232 (133 0
0
ant a.2 (l3 ......

§ 202. Diagonal Matrix.

Definition. A square matrix which is both upper and lower triangular is
alled a dia g onal matrix.
c (Bundelkhand 94)
Ior exam p le all 0 0 0 (Sec § 1 03 Page 4 also)
o (122 0 0
o 0 (133 0

0 0 0 .....
1 heorem 1. A nv two diagonal matrices oJ the .SQO1C order commute under
niuliuplucatton. (Ru,idelkhwid 95, (,,4)
Proof. Let any two diagonal matrices he
A= a l 0 0 . . . 0 and B = b 1 0 0 0
0 (12 0 ... 0 0 h 0 0

0 () 0 ..... . b,

Then we have
AB at 0 0 . . 0 x b1 0
0 0 ... 0 0 1) 2 0 ... 0
0
0 () 0 ... an 0 0 0 .. b


62 Matrices

or AB=aibi 0 0 .. 0
o a2b2 0 0

. 0 0 0 ... ..( i)

and BA = In 0 0 ... 0 x at 0 0 0
0 0 0 02 t) ... 0
o 1)2
• ..... . 0
() 0 0 ... bn 0 0 0 ...

biai 0 0 ... 0
o 1)202 0 ... 0

00 ... bnan (ii)

O
From (i) and (ii), we find that AB = BA and each one o them is a
(Note)
diagonal matrix of order n.
Hence proved.
Theorem 11. Product of an y two diagonI matrices of order a is a
diagonal matrix of order n.
Proof. The same as of Theorem I above.
Theorem III. Sum of an y two diagonal matrices of orth', a is a diagonal
matrix of order n and commute under addition.
Proof. Let any two diagonal matrices be
00 .0
A= at 00... O and Bbi
0 b 2 0 ... 0
0 02 0 ... 0
.0
0 ... 0 0 C) ...
0 0
A+B= a 1 +bi C) 0 ... 0
0 a2+b2 0
0
0 0 0 ... tin +brj

and B + A bi+at C) 0 ... ()

0 b2+02 0
0
0 0 () h, + 0n

From (i) and (ii). we get A + B = B + A and each one of them is a

diagonal matrix of order a.
§ 203. Scalar matrix.
Definition. If in a square mat-rix A all the diagonal elements are equal to
it is
a (where a * 0) and all the remaining elements are equal to i.eia then
called a scalar matrix.


Commutative Matrices 63
For example a 0 0 0 is a scalar matrix of order 4 x 4.
OaOO
OOaO
000a
Commutative Matrices
Definition. If A and B are two square matrices such that AB = BA, then
A and B are called commutative matrices or are said to commute.
If AB = - BA, the matrices A and B are said to anti-commute.
Solved Examples on § 203.
Ex. 1. tfA={a 0 0and B= all aiz
o a 0 a a a
[0 0 a1 a31 32 33

Then prove that AB ­ BA = all.
So!. AB= a 0 0x Oh ai au
o a 0 022 023
o o aa3l a32 233

0011 0012 aa13 =a a 012 (2J3

0021 (2022 2023 (221 022 023
aa3l a1232 aa33 '31 a32 a33
=aB.
Similarly BA= all ai a x a 0 0
021 022 023 0 a 0
03! 032 033 0 0 a

[(aa hl aai j3 a a ,, 012 0 13 =aB

aa21 (2022 (2023a 022 023
031 0033 0033 aj 032 033
Hence All = BA = aB.
Ex. 2. Show that the matrices A and B anti-commute, where
A=[1 - ii and B=f1 I
[2 —1] [4 —1
Sol. Here AB=[I —l l x [ l l
[2 _lj [4 —1

- 14 II +(— l).(— l )1 = [- 3 21
[2 . 1 - 21 + (— l).(— I )j [-2 3j
, And BA =[I I
114]
[4 —1] [2 —1]

=1 1.1 + 12 l.(— I) + l.(— I) 1=[3 —2

[4 . 1 +(— 1).2 4.(— l)+(— 1 )(— l)j L2
64 Matrices 182/1/4

=-1- 3 21
[-2 3j
From (i) and (ii) we find thatAB = - BA.
Hence A and B anti-commute.
Exercise on § 2-03
Ex. 1. Show that the matrices I0 ii and[1 0]anti-comrnute.
i oj [0 -ij

EL 2. Show that the matricesi 2] and 1 5 7 1 commute.

2 1] [7 5 ]]

§ 2-04. Unit Matrix or Identity Matrix.

Definition. If in a scalar matrix the diagonal clement a = 1, then the
matrix is called the unit matrix or identity matrix and is denoted by In in the
case of n x n matrix.
For example 1 0 0 0
0100
0010
0001
Solved Examples on § 2-04.
*Ex. 1. 11 A be any a x n matrix and In is the identity matrix of order
n x a, then prove that A In = In A A
Sol. Let us suppose that
A= all a12 ... Qln and In 1 0 ... 0

a2l a2 --. a2, - 0 I ... 0

0

ant at 12 .- - 0 0 ...

A . I = all a 12 ... at,t x I 0".0

021 a22 ... a2n 0 0

an2 - - - 0 :j: 1

alI.l+a12.0+...+aln.O ajI.0+a12.1+...+aln.0
a:i.1 +a220+ ... +a2,.O a21.0+a22.1 + ...

ant-I +an2.O+ ... al-0 + 0n2-1 + - - + a,.0

a,l.0+a2.O+...+al,.l
(12I.0+U22,0+ ...

1.0 + a,12.0 -+- -. - + a,1 ,. I



1871115 Idempotent Matrices 65

all a at,, =A
a2J 'fl2 ... a2,

an a, •..ann
Similarly we can show that I .A A.
Hence we have A .I = I .A = A.
*EX. 2. Prove that 1m = 1 m-I = -. = 12=1, where in any positive
integer and In is the unit matrix of order n x n.
Sol. Let A be any n x n matrix and I be the unit matrix of order n x n i.e.
I = In.
Now we know that Al.= IA = A (See Ex. 1 above)
But 1=I.
All =IA=A
Taking A=I,we have l.I=I or 12=1
Again front (i), taking A= 12, where 12 = I (proved), we get
12.1 = 2 or 1 = 1 = 1, from (ii).
Proceeding in this way, we can prove that
Im = 1 m-I ...
= = I, where m is any positive integer.
Exercise on * 204
Ex. If A 1 0 0 0 show that A2 = 1. where I is the unit matrix.
1 —1 0 0
I —2 1 0
1 —3 3 .1
§ 205. Periodic Matrix.
Definition. A square matrix A is called periodic, if A" = A, where k is
a positive integer.
If k is the least positive integer for which Ak+t = A, then A is said to be
of period k.
ldernpotent matrix.
Definition. A square matrix A is called idcrnpotent provided it satisfies
the relation A 2 = A.
Sy mmetric Idempotent Matrix.
Definition. A square matrix A is called symmetric ideinpoteni if A A'
and A 2 = A, where A' is the transposed matrix of A, (See § 208 Page 69).
Solved Examples on § 205.
Ex. 1 (a) Show that the matrix A = 2 -2 —4 is idempotent.
—1 3 4
1 —2 —3
(Rohilkhand 96)


66 1Matrices

Sot. A 2 = A.A1 2 -2 —4 2 -2 -4
I-i 3 4
--2 _-
---I 3
i 1 -2 -3j
4.
[1
= 2•2-2(-- 1) -41 2(-2)-23 - 4 (- . 2) 2 (-4)-24 -4 (--3)
-1 . 2+3-1)+41 _1(-2)+3.3+4-2)-1(-4)+34+4(-3)
1-2-2 (- 1)- 31 1 (-2)- 23-3 - 2) 1 (-4)- 24-3 - 3)
2 -2 -4=A
1 3 4
1 -2 -3
=[
Hence the matrix A is idempoteni.
Ex. 1. (b) Show that the matrix A 2 - - 51 is iden'ipotent.
-1 4 5
1 -3 -4 (Avadh9l)

Sot A2 =A . A= 2 -3 -5 X 2 -3 -5I
-1 4 5 -1 4 5
1 -3 -4 1 -3 -4
= 4±3-5 -6-12+15 -10-15+20
-2-4+5 3+16-15 5+20-20
2+3-4 -3-12+12 -5-15+16
= 2-3 -5=A
-1 4 5
1 -3 -4 - -
Hence the matrix A is idempo(ent.
Ex. 2. If A and B are idempotent matrices, then show that AR is
deinpotent if A and B commute.
SoL If A is the idcmpotcnt, then A 2 = A and if B i s iiernpotciit then
B2 -B .Ji)
And if A and B commute, then AB = BA Ui)
Now (AB)2 (AB'.(AB)
= A (BA) B, b y as,.ociafive law
= A (AB) B. from (ii)
(AA) (BR), by associative law
=A2B2
- AB by (i)
Hence AB is idrmpctciI
Ex. 3. If A is .in idem potent matrix, then the matz-i 14 - I -A i
idempotent and AR - 0 - BA.
Sol We know IA = Al = A. (See x 1 Pgc (34)


Idcnip!ent Matiees 67

Also A being an idumpntcnt matrix, we have A 2 = A.

Sincc I and A are square matncc. so 1-A is also a square matrix and
lhLrcforc we
(l-A)2 I--A)(1-A)
= (I - A) I (1 - A) A, by distributive law
Al IA + A2
I - A -- A + A, from (I). (ii)and 12= 1
or (I AY = I -- f, i e I - A or B is an idempotent matrix by definition.
Again AU = A (1 - A) = Al -. A 2 , by distributive law
= A -A, honi ) and (ii)
i.e AB=0.
And BA = (I - A) A = IA - A 2 , by distributive law
= A - A =0.
Ex. 4. Show that if A and B are matrices of order n x n and sucL' that
AB = A and BA = B, then A and B are idempotent matrices.
Sol We have ABA = (AB) A = (A) A, . AB = A (given)
or ABA =A 2 (i)
Also A B A = A P.A)=A(B), .' BA=B (given)
=AB=A . AB=A (given)
or ABA =A ..(ii)
From (i and (ii). we have A 2 = A i.e. A is idempotent.
In a similar manner, we can prove that
HAD B (AB) = B (A), '. AB = A (given)
BA B, . BA = B (given)
01 BAB=B ...(iii)
Also BAR (BA) B = (B) B, .' BA = B (given)
or BAB = B 2 (iv)
Fn,m ( i ii) I1v, we have B 2 = B i.e. B is idempoLent. Hence proved.
Exercises on § 205
Ex. It /i aid H are iderr.poent, then A + B will be idempotent if
AR BA 0, wherc 0 is the p ull matrix.
lint (4 R) . A 2 AR 4 BA + B . A -r 04 0 # RJ
§ 21 06. 1.n ,, o!u1O " Matrix.
D!'iriio'i. A square ruatno, A is i! Involutory provided it satisfies the
relaten .'\ where I is tie idi ntits mar ix,
c npTe. 'h- matrix A = I 1 ,rvo1utoiy matrix,
i,&.
68 Matrices 989

since A2=1l o]4i 0

L0 —1] [0 —1

=ri . i+ oo 1•0±0.(-1) 1 = 1 1 01=!.

0-1+(- 1).0 0-0+(—1) ' (-1)] [0 1]
Solved Examples on § 206.
Ex. 1. Show that the matrix A —5 —8 0 i.sinvolutory.
3 5 0
1 2 —1 (Rohilkhand9l)

Sol. A= 5 —80x-5 8 0
3 5 0 3 5 0
1 2 —1 1 2 —1
= (-5)(-5)+(-8).3+01 (-5)(-8)+(-8)5+02
3 . (-5)+53+0l 3•(-8)+55+0•2
l.(-5)+23+(— 1)•1 1(-8)+25+(— 1)2
(-5)0+(-8)0+O(— 1)
3•+50-i-0(— 1)
- l•0+20+(-1)(-1)
= 25-24+0 40-40+0 0+0+0 = 1 0 0 =1
—15+15+0 —24+25+0 0+0+0 0 1 0
—5+6-1 —8+10-2 0+0+1 0 0 1
Hence the given matrix A is involutory.
Ex. 2. If A is any square matrix of order n and In is the identity
matrix of order n, such that (J — A) (I + A) = 0, then show that A is
invotoy matrix.
SoL Given that (I — A) (1 + A) =0
or I+I0•A—A•Ia--A2=O
or 10+A—A—A2=0,
(See Ex. 1. Page 64)
or In - A2 = 0 or A2 In i.e. A is involutory by definition..
* 207. Nilpotent Matrix. (Avadh 93)
Dthnition. A square matrix A is called Nilpotent matrix of order m,
provided i satisfies the relation A'= 0 and Am.l * 0, where in is a positive
integer and 0 is the null matrix.
For exile, the matrix A = 10 ii is a nilpotent matrix,
[0 Oj
since A=[0 ii ^O,
[o 0]
A2 =[ 0 l l x IO 1 1 = 10-0+ 10 O•1 + 1-0
OJ [0 0] [0.0+0-0 O•1+00
L0
 Nilpotent Matrices 69

=fo 01=0,
[o oj
A3 = A2 • A = 0.A 0.
i.e A is a matrix which is not itself a zero matjix though its powers are zero
matriccs and so it is a nilpotent matrix (Another definition of nilpotent
matrix).
Solved Examples on § 207.
Ex. Show that A= 1 2 31
is a nilpotent matrix of order 2..
1 2 3
-1 -2 -3
So!. Given A 12 3 0
1 2 3
-1 -2 -3

A2 = 1 2 3 x 1 2 3
1 2 311 23
-1 -2 -3 H -2 -3
= 11+21+3(-l) ] . 222+3(-2) l•3+2-3+3(-3)
1 . 1 + 2-1 + 3 (-1) 1-2 + 22 + 3 (-2) 1-3 + 2-3 + 3 (-3)
-1 . 1-2I-3(-1) -l2-22-3(-2) -l3-2-3-3(-3)
= 0 ()= 0, where 0 is the null matrx of order 3.
o 00
030
L

i.e. A 2 = 0 but A * 0. Hence A is a nilpotent matrix of order 2.

Exercises on § 2-07
Ex. 1. Show that the matrix(is nilpotent.
1-7 a' -abJ

Ex. 2. Show that I I31 is a nilpotent matrix of order 3.

52 6
- 2 -1 - (Avadh 93. 90)

[JJmt Po'e that A3 = 0, A2 * 01:

** 20. Tram-posed Matrix. (Agra 94)
Definition. The matrix of order n x m obtained by interchanging the rows
and columns of a matrix A of order ii x n is called the transposed matrix of A
f the maz:u A and is denoted by A' or At (read as A transpose).
or Iranspc
Another Definition. If A = (aj) be a matrix of order i'i x n, then the
matrix B [h,] of order n x m, such that by ajg is I'ncwn as transposed matrix
f A or the trw-i cpnse of the matrix A and is denoted by A'or A. '


70 Matrices
=r1
For example: If A 5] then A' = 1 21
2 4 sJ 34
56
Note 1. The element a ij in the ith row and jth column of A stards in jh
row arid ith column of A'.
Note 2. The tianspose of an in x ii matrix is an n "< m matrix.
2 09. Some Important Theormns on Transposed Matrices.
Theorem I. The transpose of '1e sum of two itarlces is the sum of their
transpose i.e. (A + B)' = A' +- B'.
Proof. Let A = [a,,) and 1i thU]
Then A+B=1a0 +h0 ]. tC:jJ, say
thcn aij-t-
(A -B)'r-[d,j], w here d11 =c for all 1 iç,n . I 5 jn
i.e. d,=a,J+!.'.J. for all I ^i'm, I :5j<n
or (A+B)'=[c}=[a,+h,j]
Also A'= [,1, whcrefj,a,. for all I:51m. 'j5n
and B'=fgfi). where , 1 1=h 1 for all I !^i:5rn, I
A+B'=EJ)1- [gjil=[fj,-i-gj,
=[aj+b,] ...(ii)
From (i) and (ii) we get (A - B)'= A'+ B'
'Theorem H. The transpose of the transpose of a matrix is the pratris
itself i.e. (A')' = A. (Meerur 95, 94)
Proof. Let A = [a,j ] be an ir x n matrix. Then A' i.e. the trsnpose of A is
n x m matrix and (A')' i.e. the tra1pusc of A' (Or the transpose of A) is an
in X n matrix.
Therefore the matrices A and (A) are both rn x n matrices and hence
comparable. .
Also, the element in tne jib row and jth column of (A')'.
= the element in the jth row and ith column of A'
the element in the ith row and jib column of A
i.e. the corresponding elements of (A')' and A we equal
From (1) and (ii), we conciude that (A')' = A. Hence roved.
Theorem 111. If A is any in x ii r,ratrir, then (kA)' kA', where k is any
number
Proof. Let A = [a,j] be any p't x n matrix. Then kA is a!ao in x n inatnx
and therefore (kA)' i.e. the transpose 'f the matrix kA is an n x m nutrix.
Also A', the transpose of the matrix A, is n x. in matrix and kA' is also an
n X in matrix.
Thus wc find that the matrices (kA)' and kA' are both n x in matrices ar.d
hence comparable. . (i)
Again the clement in ith row and jib col'irnn of (kA)'
 Transposed Matrices 71

= the element in jth row and ith column of kA

(Note)
= k times the clement in jdi row and iLh column of A
(Note)
= k times the element in ith row and jth column of A'
(Note)
= ka1
the element in ith tow and jth column of kA'
kA' are equal (ii)
i.e. the corresponding elements of (kA)' and
Hence proved.
From (I) and (ii), we conclude that (kA)' = kA'.
Theorem IV. The transpose of the froduct of two matrices is the
product in reverse order of their transpose i.e. (AB) = B'A'.
(Garhwa! 95. 93.' Gorakhpur 96, Rohilkhand 94)
n x n and
Proof. Let A = [ atj and B = tht1 be the two matrices of orders i
n X p respectively.
Let C AB (gik] x [bkjl = [cv], say
where C is a matrix of order m x P.
'I
a1k bAlçj.
The element in the ith row and jth column of AB is c 9 =
k1

(i)
This is also the element in the ith row and jththlumn of (AB)'.
and elements in
The elements in the jth row of B' are hi1. h21, b3j ..., hnj
the ith column of A' are aI. at 2. GO, a,,,. Then the element in the jth row and

ith column of B'A' is

(ii)
b*j aik E akhk=ciJ
k=i k-i
Hence from (i) and (ii) we conclude that (AB)' = WA'.
reversal rule for the
Note. The statement of theorem IV is called the
transpose of a product.
Solved Examples on § 208 to § 209.
Ex. 1. Write down the transpose of the matrix Ai 2 4
16 8 1
Sol. Let A' be the required transpose of the matrix A. Then A' = matrix
obtained by interchanging the rows and columns of the matrix A 1 6
28
4 1 Arts.

Ex. 2. Verify that (B) t (A)' (AB) t , when'

(a) Al2 11, B' 11 2 01
[4 5 - 3,] (Budenkhand 91)

(b) A=F1 2 3],B= 1 2

—2 ij 2 0
L
—1 1


72 Matrices
(c) A=[ 2 4 - 1 1, 5= 3 4 5
[-1 0
01
2] —1 2 7
2 1 (Avadh 92)

Sot (a) Here At = [ 2], B t =[_ 1 4

2 5
0 —3
1 Bt
4 x[2]= 1-2+4-I = 6
Ar
—2 5 i —22+5-1
0 —3 02-3-1 —3
Also AB=[2 1 ] x ri —2 0
[4 5 —3
[2. 1 + 1-4 2 (-2) + 1-5 2-0 + 1 (-3)]
=[6 1 —3].
(AB) transposed matrix of AS
= from (i)
=
31 Hence proved.
(b) Here A= 3 , B = [ l 2 —1
2 —2 [2 0 1
3 1
BA = [1 2 —lix 1 3
[2 0 ij 2 —2
3 1
= 1 1-1+2.2-- 1.3 1•3+2(-2)-1-1
[2-1+0-2+1 . 3 2-3+0(-2)+1.1
= 1 2 -21
[5 7] .(ii)
Also AB= [ 1 2 31x 1 2
[3 —2 ij 20
—1 1
411+22+3(—l) 1-2+2.0+3-11=12 1
[3•1-2 .2+1(-1) 3-2-2-0+1-111-2
[-2 7]
(AB)' = transposed matrix of AS
[2 21 = BY, from (ii).
L5 7] - Hence proved.


Transposed Matrices 73

(c) Here A t = 2 -1 and B t 3 -1 2

4 0 4 21
—1 2 5 70

BA r3 1 2)( 2 —1
4 21 4 0
5 7 0 —1 2
= 3•2—l•4+2(-1) 3(-1)—l0+22 = 0
4 .2+24+1(—l) 4(-1)+2-0+12 15 —2
5 2 + 74 + 0 (— 1) 5 (— 1) + 7-0 + 02 38 - 5 •.. (iii)

Also AB = 2 4 -1 x 3 4 s]
[-1 0 2] —1 2 7
210
24+4-2— 11 25+47— 1-0
=1 2-3+4(— 1)— 12
[- 1-3+0(— 1)+2-2 - 14+02+21 - 1.5+0-7+2-0

=10 15 38
Li
—2 —5
... (AB) = transposed matrix of AB
0 1 =B t At ,from (iii).
15 —2
38 - 5 Hence proved.

*FL 3 If A= 1 —1 0 and B= 4 1 0
2 13 2-3 1
4 18 1 1-1

then verify (AB)' = I3 A t. (Meerut 93, 9))

Sot. AB = I - 1 01 x 4 1 0
2 1 3j 2-3 1
4 1 81 1. 1 —1
=Fl-4- 12+0-1 1-1 + 1-3+0-1 10— 1-1 —01
24-4- 1-2+3-1 2•1 - 1-3+3-I 20+ 1 . 1 —3-1
4-1-13+8-i 4-0+1-1-8-1

C, r ABr 2 4 —1 andso(AB) t [ 2 13 26
132-2 4 2 9
26 9 - --2 —7

Again At=r1 2 41 and B t 42 i

—1 1 1 1 --3 1
0 1 —i


74 Matrices
B t A! =4 2 lx 124
1 -3 1 -L I 1
o 1 -1 0 3 8
= 41-21+10 42+21+1-3 44+21 + 18
11+3-1+10 12-31+13 1-4-3-I + 18
01-11-10 02+1-I- 13 044 1-1 -1.8

= 2 13 26 =(AB)t. from (i)

4 2 9
-1 -2 -7 Hence proved.

Ex. 4. If A =[ cosa sin C( verify that AA' 12 = A'A.

sin a cos ci]

Sol. Here A' =F,cos a - sin cx

a cos a

AA' =1 COS Ci sin ci i [cos cx - sin cx

sin a cc's a] [sin a cos a

cos 2 a+ sin a
-cos asin cx+ sin acos ci
[
sin a cos a + COS a sin a sin a + cos 2 a

=[i 01=12.
[0 ij
Sintilarfly we can prove that
A'A[cosu _sinaix[ cos cc sina
[sin a cos a] [-sin a cos ci
cos a sin a - sin a cos a
[ cos2 a+ sin a
k in a ccs ci - cos cx sin u sin a + cos 2 a

= 1 1 01=12.
[U 1]
Hence AA' - 12 A'A.
Exeriises on § 208 -- 209
E. 1. If A = 1? 3 ] , [3 0], verify that (AB)' = B'A'. where
[1 2] [ 2]
A', B' are transposes of A and B.
Ex. 2. If A = I - 1 1 and B = 4 1 0
2 13 2-3 1
4 1 8 .. 1 1-1

then verify that (AB)' = WA'.

 Cmplcx Conj ugate of Matrices 75
Ex. 3. If Ar I 2 4 il and B=[ •
0 2]
H' H'
prove that (AB)' and WA' arc cual.
Ex.4. 11, A=2 31 . B= [3 41, then verify that JAB I' = B'At
10 ij L 2 ']
Ex. 5. if A = 1 I 1 and B {) I
223
1 32 -
2 4 9] 1 1 0

then verily that (AB) = BIAt.
210. Complex conjugate (or conjugate) of a Matrix.
Definition. The matrix obtained from any given matrix A of order pit ( n
with complex elements r by replacing its elements by the corresponding
conjugate J,MPICX numbers is called the complex conjugate or conjugate of A
denoted by A and is read as 'A conjugate.'
or if A = [a,,] and d is the complcx conjugate of the clement aj then
A rz[], for all I I in, 1 5j Sn.

For c.xarnp: If A 1 + i 2 + 3i1

L2 3ij'

[i—i 2_3]
then

Real Matrix. (.1vadh 93)

Dcfnition. A matrix A is called real provided it satisflcc the relation
A=A
Imaginary Matrix. (A'adh 93)
Definition. A. inatni A is called imaginary provided it satisfies the
relation A = - A
211. 'rheorerrLs on complex conjugate of a matrix.
Theorem I. It A = [a,,] be an" 'ri x n malrix with complex elements cijj,
then the coniples conjugate of A is the matris A itself -
Proof By definition (given in § 210 above) we know that A = [, for
all 15 i!5 in, I j :5 n and is the complex conjugate of aj.
i.e. the clement in the Ith row and jth column of complcx cuniugate of A i.e. A.
the complex corugate of element in ith row aiidjth column of A.
The clement in the ith row and jth column of the complex conjugate
of A i.e. A.

= the complex conjugate of the clement in oh row and jth column of A

the complex conj ugate of (Note)
76 Matrices
= au i.e. the clement in the ith row and jth column of A. (Note)
i.e. the corresponding elements of A and the complex conjugate of A are equal.
(i)
Also it is evident that A, A and its conmplex conjugate are in x n

matrices and hence comparable. .. (ii)

From (I) and (ii), we conclude that the complex conjugate of A is

equal to A or A = A.
Theorem II. If A = [ay] he any rn)< a matrix with complex elements ajj,
then )A=XA.
Proof: By definition, we know
A = [], for all I 15 I !^ m, I 5j !^ a and aij is the complex conjugate of
ay.
Also ?cA = [Xat1 ], for all I i:5 m. I !^j ,5 n.
:.A=[Aajj]=Rajj], for all l!^i5m,l!^-i!^n (i)
and we know that 1E = j. j. where z, Z2
are any two complex numbers,
Again ) A = [bij], where b = X aij for all I :5 i!5 m, I :!^j n

for all 1 5i5m, 1 !^j5n. ...(ii)

From (i) and (ii) we conclude that the corresponding eleme of A.A
and A. A are equal. Also it is evident that A.A and A. A are matrices of the same
order. Hence we conclude that X.A = A.
Theorem M. If A and B are two matrices conformable to addition, then
A+B=A+B.
Proof : Let A = [a 3) and K = [b,] be any two matrices of order in X n.
Then as these matrices are'given as conformable to addition, so we have
A+ B= [a j+ b1j] for all I i5m, 1 j5n. '(I)
Also A = and B = [!j, by definition.
...(ii)
for all 1 :^i:5rn, I 15j!5n
and also as F1 + = Zl+Z2, where zi, Z2
are any two complex numbers.
Again from (i), we have
A + B = complex conjugate of {ay + b]
= complex conjugate of (c 0), where Cif = aij + b ij
[ .]—[a+bI for all I:5m1<j<fl
i.e. A + B = Fa,1 + b01, for all 1 :5 i!5 m, 1 Sf 15 a (iii)


Transposed Conjugate of a Matrices 77

from (ii) and (iii) we conclude that the corresponding elements of
A + B and A + B are equal. Also it is evident that both A + U and A + B are
matrices of order m x n as A and B are given as conformable to addition.
Hence we conclude that A + B = A + B
Theorem IV. If A = [a,1] be any m x n matrix and B = [bj k] he any n x p
matrix i.e. if A and B are conformable to the product AR then AU=Aff.
Proof : Since A and B are conformable to the product AB, so
AB [a] x [h3 k] = EcL where Cik = aij bk, for all 1 :5 i !^ m, I :5 k p and there
is summation on j, where j= 1,2,3........ n.
Also A= [:j]. for all 1!^im, I :5j!^,r
and B=1b1A]forl:5j:^n,1:5k:5p
A B is defined and we have A H = x [ j k] [a1kl
where d, k =a ij l jk for 311 I ^;i:^m, I !^k !^p and j= 1,2........n.
Again All = complex conjugate of AB i.e. [ck]
or AU [CzkI, where Ck a,3 b3k
17 k) = 1,jJk1, ...
for any complex numbers zi and Z2
= [dikl, since (-I,k = ay bjk for all I 5i !^m.
l5kp and j=l,2..... n ...(ii)
From (i) and (ii), we conclude the All = AR.
§ 212. Tr n.spoed Conjugate of a Matrix.
Definition. The transpose of corijugage of a matrix A i.e. (A7 is defined
as transposed conjugagc or tranjugate A and is denoted by A i . e. A t = (A
For example : If A=[l +1 2+311,
2 3i

then A=[' 2-3i]

A 0 = tran spose of A = (A'

=[1—i 2
[2-31 —31j1
** 213. Theorems on Transposed conjugate of a matrix.
Theorem 1. Fo r any matrix A, (A)' = (A')
i.e. the transposed conjugate of a matrix is equal to conjugate of its transpose.
Proof: Let A = (o ) l he any m X n matrix
Then by definition, A = [i 11 1, for all I i < and I: ii.
(A)' = transpose of A,
i.e. (X)' = [bp] , where [b1 is n x ,n inrix and h11
for all I -Si!^m, I Sjn
Again A transpose of A i.e. [ay]
—6


78 Matriccs
= ( i ii. where i Ojj aiRl Lcs ax no matrix toi all
I in:, jn
coiiiplcx ( ' ii j ugak' 01 A'
(A)'
= 1i1. by deflniti.n.
= T111 SIflCC ( 'ji Ui1
01 = ! hjL sncc 5. where [hp] is n x in matrix br
all I K to, I j it
From (i) and (ii), we conclude [hat A' (V).
Theorem II. For wiv lIiOlfl A, ( A O) C =
Proof Let A 0 B i.e. B = (A
Then B' = transpose of B
transl)osc of (A)'
A, since we kiiw (A')' A Sec Ili. II Page 70
= i. oinplex Conjugate of B .. Sec'Ili. I above
complex conjugate Of A (Note)
= A, since sc know A A ... See Tb. I Page 75
i.e. BC = A. since B 0 = ( B') = (B)' See Th. I above
C
i.e. (A") = A. since A = B. Hence proved.
Theorem Ill. (a). I'or an y rnatrz A, (kA) 0 = kA 0 , where k iv a sea/a,
Proof : By definition, we know that
(kA)0 =
= (-k c) '. by Tb. II Page 76
= (A)', by Ill. 1!! Page 70
=kA 0. since A is a scalar. Hence proved.
Theorem Ill h}. [fir wiv Fnatrix A, tk A) A0, aijere k is cri
roniplek FlUfn/'er
Proof B y, definition, we know that

(kA)', h lii. ii Page 76

= (kA)'. by Tb. Ill Vace 7))
= TA (s ) ' AC. by definition. lkncc proved.
Theorem IV. if A aid B arc no 'natn(-es rnnJr'nab1e to avldiiu'n, flier
A + 11) A C s it ':rut 90
Proof : B y definition. v e havL
(A + B)" — (A+B' (A })' '' \ + R = A + B
Se 'th Ill i',e' 70
= (A)' + (B) , l 1 h. I P:•icc 70
= A0 it'. b y dcl nition. n N1' l


Sy.,i i u ':tric and Skew-syn ia'ric Matrices 79

\. If A wrd B art, two MOO roo's o'onfo,mahc lo the pmdrui

)E)
AR. then ( A it) °o A0
I'r of (AB', by du fin t i)l
H
r. (\ if), by Iii. 1V Page 7
b y Th. IV Page 7 (Note)
-= B Ae, ,y definition. I1CnCC proved

Fxampie: Find t A , A' and () for the matrix

=[+i 3-ii
[21 S
j
Solution, A =[i - 1 3 + 51
2i 5 - See § 2 10 Page 75

(A)' ii flS0SC of A [i —' —. 21]

[3 -4 5 i S ,j See § 208 Page (

A' Tr:nispose of A = i + i 211

1See
3 - Si 5 j § 208 Page oc

A' - conjugate of A' = [t -- - 1 = (A

[3+si 5 j

A° = con ugac lronspoSC of A I -- - 211 = A'

5 j

9
and A = conjugate tsajlsr rlse of A

1+1
[21 5
214. S y mmetric and skew-s ymmetric matrices.
tat Symmetric Matrix. (Agra 94; Avadh 92)
[)viurtition. A square ir.atrix A = kI is caflcd syrmnrtnc provided
= far all vaIi.c. of and
For cxunp1c A 1 --3 5
—3 27
[5 73
N,te. 1.\ is also syrnrnrtrlc, if k :s scalar.
(b) Sicw-vmiiietric Matrix. A'ro1 4 tn/i 92
I)if'niion. A 'narc' matrix A = la ' I is :aOcJ skew rnmt'tric providc
tb ap. tar a1 VOjll',-S of i and j.
aOl t'xau p lc A '[ 0 1—3
—1 0 5
—5 0
80 Matrices 1821115
Note. kA is also skew-symmetric, if k is scalar.
§ 215. Theorems on Symmetric and Skew-symmetric matrices.
Theorem I. A square matrix A is symmetric if A = A'. (Kanpur 90)
Proof: Let A be an a x a square matrix i.e. A = [a,,,), for all I 5 i !^ a and
I :5j5n.
If A is symmetric matrix, then by definition, we have
[ a I = [aj ,J, for all I SiSnand 15j!^n
Also, by definition,
A' = [h] such that b y for all 1 !^i5n, i:^jn ...Scc § 208 Page 69
or A'=[aj], for a]l I:5i n, 1:5j!^n
= [a,j ], from (i).
Hence A' = A.
Conversely ifA=A'. Then A must be a- square matrix
Also A = A' = [a,j ] = [aji], for all I 15 i 5 a, 1 15j !^ a
= a,1 =a11 , for all I i!^n, 1:^j:5n
A is a symmetric matrix. Hence proved.
Theorem 11. A square matrix A is skew-symmetric tff A' = - A.
Proof: Let A be an axa square matrix i.e A=[aq] for all I !^in and
I j!^n.
If A is a skew-symmetric matrix, then by definition, we have
[a,,]=[—a,), for all I !^i!^n, I !^jn
Also, by definition. A' = [b,1 ], such that bq = aji.
for all 1 f:-i:^n, I 5j!^n. ...Sce § 208 Page 69
or A'=[aji] for all I!^i_'n, l!^j:^n
= - [- a1 ] = - [d,j ], from (i).
Hence A' = - A.
Conversely if A' = -A, then A must be a square matrix.
AIo A' = - A = (a1 - [a,1 ], for all I :5 i :^ a, I 15j 5 a
=5 a, = - a,1
for all I !^i!^n, l !^j!^n
A is a skew-symmetric matrix. Hence proved.
***Theorem HI. Every square matrix can be uniquel y expressed as the
sum of a symmetric and a skew-s-vnimetric matrices.
(Atudh 94, 92, 90; Bundelkhand 95; Meerut 93)
Proof: Let A he a square matrix, then we can write
(i)
since 1 A, - A' are conformable to addition, A being a square niatrix. (Note)
Now (. (A + A'))' = transpose of 1 (A + A')
= (A+A')' ...by § 209Th. III Page 70
= (A' + (A')') ...by § 209 Th. I Page 70


Symmetric and Skew-symmetric Matrices 81

1821116
(A'+ A) ... by2 .O9Th.11 Page 7O
as matrix addition is commutative.
or (. (A + A'))' = (A + A'),
Therefore, by definition, 1 (A + A) is a symmetric matrix
...(ii)

Again {. (A — A'))' = (A — A')' by § 209 Th. III Page 70

(Note)
= (A'+{(— 1)(A')') ...by § 209Th. I Page 70
= (A'+(— 1) (A')') ...by § 209Th. III Page 70
= A'+ (— 1) A) by § 209 Th. Ii Page 70

=(A'—A)=((— 1)2A'+(-1)A} (Note)

A'+ A)
or {. (A — A'))' = — 1 (A — A'), as maIne addition is commutative.
Therefore, by definition, (A — A') is a skew-symmetric matnc.
...(iii)
Hence from (i), (Ii) and (iii), we find that the matrix A can be expressed
as the sum of a sy mmetric and a skew symmetric matrices.
To prove that the representation (i) is unique, let
AA1+A2. (iv)

hcre At is symmetric and At is skew-symmetric.

'Ihen Ai = Ai'
A = - A2' ...(vi)
and
From (Iv), we have A' = (At + A2)'
byTh.I209 Page 7O
Or A' = A — A2 from (v). (vi) (vii)
Adding and subtracting (iv) and (vii), we get
A + A' = 2A1 and A — A' = 2A2
or At = (A + A') and A2 = (A — A')
Iron (iv). we get A = (A + A') + (A — A'), which is the same as (I).
Hence the representation (i) is unique. Hence pioved.
Solved Examples on § 214 and § 2 15
Ex. I. Show that the inatric A = 9 6 7 is skew-symmetric.
—6 08
-7 — 8 01 (Meerul 94)

Soi. in too gi Yen inatnx, we find that

all = 0, (2 = 6 = — 6121, 013 = 7 = — 031, 022 = 0. a23 = S = — 032, 033 = 0
i.e. for all I :^i!^3. 1 <j< 3.
a 1 =—a,1
Hence by definition [Sec § 2 . 14 (b) Page 791 the given matrix A is
skew-symmetric.


82 Matrices
Ex. 2. If A =13 1 - 11, then show that AA' and A'A are both
1 2j
symmetric matrices.
Sol. Here A'= 3 0
11
—1 2
AA'=[3 I — lix 3 0
[o 1 2] 1 1
—1 2
=[3-3+I-1—I(-1) 30+1-1-1.2
[0. 3+1 . 1+2(-1) 00+11+22
=[ ii - i], which is a symmetric matrix.
[-1 5j [See § 214 (a) Page 79]
Similarly A'A = 3 0 x [3 1 -
i i [o 1 .2
—1 2
= 33+00 31+0-1 3(—l)+0-2
1-3+1-0 1-1+1-1 1(-1)+1-2
—13+20 —1-1+2-I —l(—I)+2-2
= 9 3 - 3 which is a symmetric matrix
32 1
—3 1 5 [See § 2-14 (a) Page 791
*Ex. 3. (a). If A and B are both skew-symmetric matrices of same
order such that AB = BA, then show that AB is symmetric.
Sot. If A and B are both skew-symmetric matrices.
then A=-A' and B=-B'
Also given that AB = BA
= (— B') (— A'), from (i)
= B'A'= (AB)' See Th. IV § 209 Page 71
or AB = (AB)' i.e. All is a symmetric matrix. Hence proved.
Ex. 3 (b). If A is a symmetric matrix, then sh'w that kA is also
symmetric for any scalar k.
Sol. Here (kA)' = kA', Sec. § 2-09 Th. Ill Page 70
= kA, -.- A' = A, A being symmetric
Hence kA is symmetric, if A is so.
**EX. 4 (a). Find the symmetric and skew-symmetric parts of the matrix
A= 1 2 4
681
357


Symmetric and Skew-symmetric Matrices 83

Sol. (Refer Theorem ill § 215 Pages 80-8 L)

Here A' = transpose of A
=163
285
417

The symmctflC part of A = (A + A')

1 2 I 6 3)
6 8 1 2 8 5
3 s 7] 4 1 7
1
1+1 2+6 4+3 28 7
6+2 8+8 1+5 8 16 6
3+4 5+1 7+7 7 6 14

= 1 .4
48
37 Ans.

And the skew-synimetn . rt of A (A - A')

=( I 2 1- I
6 8 1 2 5
5 7 4 1 7

1—i 2-6 4-3 =-[ 0 —4 I

6-2 8-8 1-5 4 0 —4
[—i 4 0

[
1
Ans.
L
x Ex . 4 (li) Express given matrix A as sum of a symmetric and
skew-symmetric matrices. A = 6 8 5
423
7 ij 1 (Agra 93)

Sol. From Theorem III § 2.15 Pages 80 - 81 we find that the symmetric
and skcw- mnicric prt o a matrix A are (A + A) and (A - A')

respectivciy hs sum is evedent1: A. (Note)

(1)
A (A + A') 4 (A - A')
I'NI'W A ti .inspose of A = 6 1 I


84 Matrices 989
A+A685+641
423 827
1 .7 1 5 3 1
= 6+6 8+4 5+1 12 126
4+8 2+2 3+7 12 4 10
1+5 7+3 1+1 6 10 2
• ! (A+ A)'= 12 12 6 = 6 6 3
2 2 12 4 10 6 2 5
6 10 2 3 5 1
which is evidently a symmetric matrix as a1 = aji for all values of i and
And A - A'= 6 8 5j - 6 4 I
423 8 27
171 531
=6-6 8-4 5-1= 04 4
4-8 2-2 3-7 -4 0 -4
1-5 7-3 1-1 -4 4 0
0 44 0 2 2
-4 0 -4 -2 0 -2
1L -4 A 0 -2 2 0
which is evidently a skew-symmetric matrix as
From (i), we get
=-a for all values of i, I
A=6 63+ 02 2
6 2 5 -2 0 -2
3 5 1 -2 2 0
= sum of a symmetric and skew-symmetric matrices, as proved above.
**Ex. S. If A is any square matrix, show that AA' is a symmetric
matrix.
Sol. (AA')' = transzxsc of AA'
= (A')' A' See Th. IV § 209 Page 71
= AA' .. See Di. II § 209 Page 7(
i.e. AA' = (AA')'. Hence AA' is a symmetric matrix by definition.
*Ex. 6. If A be a square matrix, show that A + A' is symmetric anc
A - A' is a skew-symmetric matrix. (Meerut 99
Sol. If A is a square matrix, then
See § 2.09Th. [Page 7(
... see 2.09Th.1t Page 7
= A + A' , by commutative law Of addition
Hence by definition A + A' is symmetric.
 Symmetric and Skew-symmetric Matrices 85

Again (A - A') '= A' - (A')', ...See § 2.09 Th. I Page 70
=A'—A ... See 2.09 Th. ll Page 7O
= - (A - A')
Hence by definition A - A' is skew-symmetric.
'EX. 7. If A is skew-symmetric matrix, then show that AA' = A'A and A2
is symmetric.
Sol. If A is a skew-symmetric matrix, then we know that

Pre-multiplying both sides of (i) by A, we get

AA'=—A.A=—A2 ...(ii)
Post-multiplying both sides of by A, we get
A'A=—AA=—A 2 ...(iu)
From (ii) and (iii) we conclude that AA' A'A
Further we can prove (as in Ex. 5 Page 84) that AA' and A'A arc
symmetric matrices. Hence from (ii) and (iii) we find that - A 2 is a symmetric
matrix or A 2 is a symmetric matrix, as we know that kA is also symmetric if k
is scalar and A is symmetric. Hence proved.

Exercises on § 2.14 - * 2.15

'EL I. If A and B are symmetric (or skew-symmetric) matrices, then so
is A + B.
Ex. 2. If A and B are symmetric matrices, then prove that AB + BA is
symmetric and AB - BA is skew-symmetric.
Ex. 3. Show that all pqsitive integral powers of a symmetric matrix are
symmetric.
Ex. 4. If A is any matrix, then show that A'A is a symmetric matrix.
(Hint : See Ex. 5 Page 84)
Ex. 5. If A is a symmetric matrix, then show that A.A' = A'A and A 2 is
symmetric.
(Hint: See EL 7 above)
Ex. 6. What is the main diagonal of a skew symmetric matrix?
(Kanpur 90)
[Hint : See § 2.14 (b) Page 19. Each element is zero].
Ex. 7. What is the transpose of a symmetric matrix? (Kanpur 90)
[Hint : See Th. I § 2.15 Page 80]. Ans. The matrix itself.
Ex. 8. A is a skew symmetric matrix. How will be An ? n is any positive
integer.
'Ex. 9. Prove that every diagonal elementbf a skew-symmetric matrix is
necessarily zero. (Garhwal 91; Kanpur 94)
(Hint: In the case of skew-symmetric matrix, we know
a ) =— aJE for all values of i and j
If i=j, then Cjj=C 1 j for all i
i.e. Cu + ajj = 0 or 2aj = 0 or a =0
i.e. all diagonal element of a skew symmetric matrix are necessarily zero.)


86 . Matrices
***216 Hermitian and Skew-Hermitian Matrices.
(a) Hermitian Matrix. (Avcidh 95, 91, 90)
Definition. A square matrix A such that A = A is called Hermitian i.e.
the matrix [a,) is Hermitian provided a j =_aji, for all values of i andj.
For example: A= I cs-i3 y±i
cL-43 ni x+iy
Y— i8 x — iy n

(b) Skew-Hermitian Matrix. - (Avadh 91, 90)

Definition. A square matrix A such that A' = - A is called
skew-Hermitian i.e. the matrix [aij ) is skéw-Hermitian provided ajj =- aji for
all values of i and j.
For example: A= 21 —cx—i —3+i
a—in —i —y+i
3+1 7 + i8 0

§2.17. Theorems on Hermitian and Skew-Hermitian Matrices.

*Theorem I. The diagonal elements of a Hermitian matrix are
necessarily real. (Avadh 95)
Proof: Let [a(j) be a n x n Hermitian matrix, then according to definition
[as given in § 2.16 (a) above], we have
a11 =ja, for all 1 5 i S n, 1 Sf S n

Now the diagonal elements are an, where 1 5 i S n.

From (i), we have a1=a 11 , for all I S i S n .(ii)
Ifaa=+i where a and 3 are real,
then iiji =a - if
F?öm(ii),weget a+i— a— i
or 2i,=O or (30
a,
.. au = cx + i (0) = which is purely real. -
Hence the diagonal elements of a Hermitian matrix are necessarily real.
Hence proved.
Theorem U. The diagonal elements of a skew-Hermitian matrix are
either purely imaginary or zero. (Avadh 90)
Proof Let [a0 ] be an a x a skew-Hermtian matrix, then according to
definition [as given in § 2.16 (b) above] we have
a0 =—aj , for all 1 5 1 5 a, I a.
Now the diagonal elements are aj, where I S i S n.
From (i), we have aj = -. , for all 1 5 i S a.
 Hermitian and Skew-syninetric Matrices 87

If a, = a + i, where a and 3 are real,

hen
From (ii), we get a + i = - (a - iJ)
or a+i—a+i3 or 2cx=0 or cx=O
a ii = 0 + i(3 = i3, which i purely imaginary and can be flero if0,
=
Hence the diagonal elements of a skew-Hermitian matrix are eiit
purely imaginary or zero.
**Theorem III. Every square matrix (with complex elements) can be
uniquely expressed as the sum of a Hermitian and a skew-Hermitian matrices.
(Garhwal 92)
Proof. Let A be a square matrix. Then we can write
(i)

Now (A+A u )=A+A ... See 2.11 Th. III Page 76

A+A°) ={A+} =(A)'+ (A 0 )', ...See.2.O9 Th. I Page 7O

F
= AE) (A0)', by def. (A)' = A0 , See § 2.12 Page 77
= AE) (A0).
Now (A0)' = transposed conjugate of A0
= transposed conjugate of (A)', ...See § 2.12 P. 77
= transposed matrix of (A)',
since conjugate of A is A ...See Th. I Page 75
=A, '. (A')'=A ...See2.O9 Th. I Page 7O

From (ii) we get, {(A + A0 ) = A0 + A = A + A0,

as addition of matrices obey commutative law.
E)
By definition (See § 2.16 (a) Page 86) we find that A+A is a
Hermitian matrix.

Again {(_A0)}=(A_A0)(A)_ç)

= AE) A, as above

By definition (Sec § 2.16 (b) Page 86) we find that A - A 0 is a skew-

Hermitian matrix.


88 Matrices
From (I) we conclude that the square matrix A is the sum of a
Hermitian and a skew-Hermitian matrices.
Solved Examples on § 2.16 - § 2.17.
Ex. I (a). Is A= 3 7-41 —2+51
7+41 —2 3—i
—2-51 3+1 5 j
a hermitian matrix?
SoLA'= 3 7+ 4i —2-51
7-41 —2 3+i
—2+51 3—I 5
AF 3 7 - 41 —2+51 =A
7+4i —2 3—i
—2.—Si 3+1 5

Hence by definition [See § 2.16 (a) Page 861, the given matrix A is
hermitian.
Ex. 1(b). Prove that the matrix A= I 1—i 2
1+1 3 I
2 —i .0
is Hermitian. (Avadh 91; Rohilkhand 97)
SoLA'= 1 1+i 2
• 1—i 3 —z
2 i 0
... I I — i 2 =A
1+1 3 V

2 —i 0
A is Hermitian. ...See § 2,16 () Page 86
Ex. 2.IfA= 3 2-31 3-i-51
2+ 3 5
3-si —i 7
then prove that A Is Hermitian. (Meerut 96)

SoLA= 3 2+ 3i 35j =B(say)
2-31 5
3+51 7

Then B' 3 2-31 3 + 5i
2+31 5
3—Si
7


Hermitian and Skew-symmetric Matrices 89

B'= 3 2+31 3-51 =B
2-31 5 —i
3+5i 1 7

B i.e. A is Hermitian. ...See § 2.16 (a) Page 86

Ex. 3. Show that A= 1 3+21 -2-il,
—3+21 0 3-41
2—i —3-41 —21
is skew-Hermitian Matrix. (Rohilkhand 95)
Sol. Here A'= i —3+21 2—i
3+21 0 —3-41
—2—i 3-41 —21

A'= —i —3-2i 2+i].

3-2i 0 —3+4i
—2+1 3+41 2i (Note)
I 3+2i-2--i =—A
—3+21 0 3-41
2—i —3-4i —21

Hence by definition [See § 2.16 (b) Page 861, the given matrix A is
skew-Hermition.
Ex. 4. If A and B are Hermitian, then show that AB is Hermitian if
and only if A and B commute.
Sol. If A and B are Hermitian matrices, then we have
and B(B)'=B0
E
Then (AB) ' = B8Ae ,by § 2.13 Th. V Page 79
= BA, by (1) above
= AB, if A and B commute
i.e. (AB) =AB or (AB)'=AB, . A0=(A)
Hence by definition AB is Hermitian.
Converse of this can be proved to be true by reversing the above
calculations.
Ex. 5 (a). If A is a Hermitian matrix, then show that IA is
skew-Hermitian. (Kanpur 90)
Sol. If A is a Hermitian matrix, then
we have A=A ...See § 2.16 (a) Page 86
Also A=Ae ... see 2.l2 Page 77
..Here A=A'=AE)


90 Matrices
.0 , .
.0 =—:A
(tA)
NOW
See § 2.13Th. 111 (a) Page 78
=-
(iAO)
or (iA)° = - (IA) , from (i) ...(ii)
Also from § 2.16 (h) Pagc 86 we know that if A is a skew-Hermitian
matrix, then A' = - A = A°, from (i)
And from (ii), we find that - (IA) = (iA)°, hence (iA) is a skew-
Hermitian-matrix.
Ex, S (b). If A is a skewHermitian matrix, then show that iA is
Hermitian.
Sol. If A is a skew-Hermitian matrix, then we have
—A =A' See § 2.16 (b) Page 86
Also
A'=A0 ... See 2.l2 Page 77
E)
—A=A'=A (i)
Now . (iA)9 - iA 0 , 7= -
See § 2.13 Th, 111(a) Page 78
=--i(—A), from
or (iA)° = iA (ii)
Also from § 2.16 (a) Page 86 we know that if A is a Hermitian matrix,
then A' A = A°, from (i).
And from (ii) we find that (iA) = (iA) 9, hence IA is a Hermitian matrix.
Ex. 6. If A is any square matrix, show that AA0 and A°A are
Hermitian.
Sol. (AA) =(A ®)0 A0 ...by 2.13 'M. V Page 79
=AA0 ... by2.13Th.Il Page 78
By definition (See § 216 (a) Page 86). AA 0 is Hermitian.
Similarly (A° A) 0 A0 (A0)0 ...by § 2.13Th. V Page 79
=A0 A ...hy2.13Th.11 Page 78
By definition (See § 2:16 (a) Page 86), A0 A i Hermitian.
Ex. 7. Show that A is Hermitian 1ff AU Hermitian.
So!. Let be Hermitian; then A = A 0 (i)
Now (A) ® = transposed conjugate, of A

= transposed matrix of A, since (A) A

Sec § 2.l1 Th. I Page 75


Inverse of Matrices 91

= A' = (A ® )', by (i)

= transpose of transposed conguate of A
= conguate of A, (B')' = B
i.e., (A) ® = A
Hence by definition, A is a Hermitian matrix
Again if A is Hermitian, then we have

= transposed conjuage of A
= transponsc of A by § 211Th.! Page 75
or A=A' ...(ii)
Now A® = (A)', by definition
= (A')', by (ii)
i.e., A e =A, by §209 Th. II Page 70
Hence by definition A is Hermitian. Hence proved
Exercises on § 2.16 - § 2.17
Ex.1. If A = r I + 1 2 - 3 , then slow that A is skew—Hermitian
— I+i 21 1
—2-3i —1 0
Ex.2. Show that A = 0 2— 31 - 2 1 is skew—Hermitian.
—2-31 0 —3+41
2-1 3+41 0
Ex. 3. Show that A is skew-Hermitian iff A is skew-Hermitian.
[hint See Ex. 7. Page 901
Ex. 4. Give an example of matrix which is skew symmetric but not
skew- Herm itian.
Ex. S. If A and B are Hermitian matrices, show that AB+BA is
Hermitian and AB - BA is skew-Hermitian.
Ex. 6. Show that every square matrix can be uniquely expressed as
P + IQ. where P. Q are Hermitian. (GcirhwI 95; Rohilkhand 91)
[[lint See Th. 111 Page 87, Ex 5(a) Page 891.
* 2.18. The inverse of a matrix.
(Avadh 91; Th4ndelkhand 93; Garhwal 91)
If for a given square matrix A. there exists a matrix B such that
AB = BA = I where I is an unit matrix, then A is called non-singular or
Invertible and B is called inverse of A and we write B = A' (read as B equals
A inverse).
Here A is the inverse of B and we can write A = B'


92 Matrices

A 1 does not exist, then A is called singular.

If B i.e.,
Note 1. If AB and BA are both defined and equal then the matrices A
and B should both be square matrices of the same order.
Note 2. Non-square matrix has no inverse.
For example : 1 1 211_ 2 1 1=[ 1 °1=
4][ j [° I]
L3
Each matrix in the product is the inverse of the other.
§ 2.19. Theorems on Inverse of a matrix.
'Theorem I. If a given square matrix A has an inverse, then it is
unique or there exists one and only one inverse matrix to a given matrix.
(Bundelkhand 93, 91)
Proof. Let us suppse that B and C are two possible inverses of A. Then
we must have (See § 2. 18 above).
AB=BA=I
and AC=CAI ..(ii)
From (1) and (ii), we get AB = AC, each being equal to 1
or B(AB)=B(AC)
or (BA)B=(BA)C ... See § 109 Prop. IPage 26
or LB IC, from (i)
or B C ...See E. Page 64
Hence there cannot be two inverses of A.
**Theorem IL if A and B be two non-sin gluar or invertible matrices of
the same order then AB is also non-singular and
(AB)- ' = B' A1
(Avadh 91; Bundelhkand 95. Garhwal 92; Gorakhpur 97; Purvanchal 97, 94)
Or
The inverse of a product is the product of the inverse taken in the reverse
order
This is also known as the Reciprocal Law for the inverse of a product.
Proof. A' and B 1 exist since A and B are non-singluar.
(AR) (B 1 A) = A (BB') A', by associative law
=AIA'AA', ...See Ex. l. Page 64
=1 ... See 2l8 Page 9l
And (B 1 A 1) (AB) B 1 (A 1 A) B, by associative law
=B 1 (1)B, A1A=I
...See Ex. 1. Page 64
=1. ... See 2l8,Page9l
(B' A' ) (AB) = (AB) (B' A1)


Orthogonal 01 Matrices 93
= A 1 and as such AB is
i.e., B' A' is the inverse of AR or (AB)'
also non-singluar.
Note : For more details on inverse of matrices see chapter V of this
book.
** 2 . 20. Orthogonal Matrix.
Definition. A square matrix A is called an orthogonal matrix if AA' = I,
where I is an identity matrix and A' is the transposed matrix of A. (Kanpur 97)
Theorems on Orthogonal Matrices.
Theorem I. For any square matrix A, rf AA' I, then A'A = I.
Proof : Since AA' = I, so A is invertible (i.e. A possesses an inverses)
and there exists another matrix Is such that
AB=BA=I
(See § 218 Page 91)
Now B = RI = B (AA'), AA' = I (given)
= (BA) A' = IA', from (i)
i.e. B=A'
From (i), we gel AA' = A'A = 1. Hence proved.
Theorem II. If A is on orthogonal matrix, then A' is also orthog'onal.
Proof: By definition if A is an orthogonat matrix, then
AA' = A'A = I
or (AA')' = (A'A)' = I, transposing and remembering I' I
or (A')' A' = A' (A')' = I. by Th. IV § 2 .09 Page 71
or A' is orthogonal by definition. Hence proved.
i.e. Transpose of an orthogonal matrix is also orthogonal.
Theorem Ill. If A is an orthogonal matrix, then A 1 is also orthogonal.
Proof: By definition if A is orthogonal, then
AA' = A'A = I
or (AA')-' = (A' A) 1 = I,
taking inverse and remembering I - = I
or (A')1 A 1 = A 1 (A'it = I. by Th. II § 219 Page 92
or (A-!)' A 1 A 1 (A)' I (Note)
Ci A' is orthogonal by definition. Hence proved.
i.e. Inverse of an orthogonal matrix is also orthogonal.
Theorem IV. For any orthogonal matrices, A and B, show that AB is an
orthogonal matrix.
Proof: If A and B are orthogonal matrices, then by definition we have
AA'=A'A=I
and BB'=B'B=I (ii)
(AB) (AB)' = (AB) (WA') by Th. IV § 209 Page 71


94 Matrices
= AR B'A' A (BB') A' (Note)
= AIA' from (ii).
= AA' = I, from (i).
Similarly, we can prove that
(AB)' (AB) = B'A' AB, by Th. IV § 209 Page 71
=B'IB, from (i).
= B'B = I, from (ii).
Hence AB is an orthogonal matrix by definition.
§ 221. Unitary Matrix.
Definition. A square matrix A is called an unitary matrix if A 0 = I,
where I is an identity matrix and A 0 is the transposed conjuage ofA.
Theorems on Unitary matrices.
Theorem I. For any square matrix, if AA° I, then A 0 A = I.
Proof : Since .AA° = 1, where I is the Unit matrix, so we find that A is
invertible and there exists another matrix B such that
AB=BA=I
Now B = RI = B (AA 8), AA8 =I (given)
= (BA) A8 = LAO, from W.
i.e. B=A0
From (i), we get AA O=A 8 A=1 Hence proved.
Theorem II. If A is an unitary matrix, then A' is also unitary.
Proof: By definition if A is an unitary matrix, then
AA O =A 0A=I
or (AA°)8 = (A9 A)8 I, taking transposed conguate and
remeribering that 1 0 = I (Note)

or (A8)0 A0 = A8 (A (3)0 = 1, using § 209 Th. IV Page 71

or AA = A8 A = I, since (A 0)0 = A
(AA e ) (A0 A)' 1, taking transpose of each side
or

or (A°)' A' A' (A8 )' = 1, using § 209 Th. IV Page 71
i (Note)
or (A')0 A'= A' (A')0
A' is an unitary matrix. Hence proved.
o r
Theorem Ill. If A is an unitary matrix then A 1 is also unitary.
Proof: B y definition if A is an unitary matrix, then
AA O=A 0 A = I

or (AA0Y' = (A0 A)- ' = 1, taking inverse

or ( A81 A- 1 = A' ( A°r' = 1, by Th. II § 219 Page 9


Unitary Matrices 95
or (A_1)8 A 1 = A 1 (A')8 = I (Note)
or A' is an unitary matrix by definition. Hence proved.
Theorem IV. For any two unitary matrices A and B show that AB is an
unitary matrix. (Bundelkhand 91)
Proof: If A an B are unitary matrices then by definition we have
4eAeA=I
and BB8 = B8 B = 1 ...(ii)
(AB) (AB) e = (All) (Be A8); by Th. V § 213 Page 79
= A (EU8) A8 = AlA8, from (ii)
= AA8 = I, from (i)
Similarly (All) 8 (AB) = Be A 8 AB, by Th. V. § 213 Page 79
= B8 LU, from (1)
=Be B v I, from(ii)
Hence AB is an unitary matrix. Hence proved.
Solved Examples on §220 and § 221.
Ex. I. Show that the matrix 1- 1 2 2 is orthogonal.
2-1 2
2 2 1 (BundeLkhand 95)
Sot. Let A= ! —12 2
2 —1 2
2 2-1
Then A'= —1 2 2
2 —1 2
2 2 —1
A'A=—1 22x-1 2 2
2-1 2 2-1 2
2 2 —1 2 2 —1
= (-1).(-1)+2.2+2.2 (-1).2+2.(-1)+2.2
2.(—l)+(-1).2+2.2 2.2+(— 1).(—l)+2.2
2(— 1)+2,2+(— 1)2 22+2.(— 1)+(— 1).2
(- l).2+2.2+2(— 1)
2.2 *(— 1) 2+2.(— 1)
2.2+2.2+(— l)(— 1)
9 0 0 = 1 0 0 =1
090 010
099 001
Hence the given matrix A is orthogonal.


96 Matrices 18211/6

Ex. 2. Verify that the matrix

A= 1/'J3 I/46 — I1 q 2j is orthogonal.
1/3 —2/46 0
1/43 1/I6 1/42

Sol. Here A'= 1/43 1143 1113

1/46 —2/I6 1/6
—1/I2 0 . 1/12
A'A= 1/43 I/43 1/431 x F 1t43 1/46 —1/I2
1/46 —2/46 ii'I€ 1t43 —2/46 0
—1/I2 0 1/12 1/43 1t46 1/'12
=11 1 1 2 1 -1 1

1 2 1 141 —1 1

+0+

I 0 0 = I. Hence A is orthogonal.
o i c
001
=[
**Ex. 3. Show that the matrix A = [ cos a sin is orthogonal.
sin (x cos a]
- (Bundelkhand 91; Kanpur 97)
Sol. A'=[cos a — sina
[sin a cos a
A'A = Icos a - sin al[ cos a sin a
[sin a cosa][—
sina
a+sin 2 ct cosasina—sinacosa
2

cos U - cos a sin a

[sinacos sin2 U + cos 2 cx
= 11 01 = I Hence A is orthogonal.

i±i] unita.
Ex. 4. Prove that the matx 1 [ 1

(Meerut 96)
iF 1 I+i
Sot. Let A=[ i i -

ei1 I 1+i
Then
II_j —1


1821117 partitioning of Matrices 97

ii I 14-il ii i l+i
A OA= [11 - 1 ]X[ 1 i
—I

=[ LI +(I +i).(I—i) 1(1 ++(1+i)(— I)

+(— I) (I—i) (1 —i)(I -i-i)+(- 1) (- I)

=i +
0
0
i—i + i
1
]=1[3 O=[ 01=1
0 3] [0 IJ

Hence A is an unitary matrix.

Exercises on § 2-20 - * 221

Ex. I. Show that the matrix A = I i is unitary.

EL 2. Show that the matrix 1 2 1 is orthogonal.

. i 3
2 2 _1
3 3 3

For any two orthogonal matrices A and B, show that BA is an

EL 3.

orthogonal matrix.
EL 4. For any two unitary matrices A and B, show that BA is an unitary
matrix.
Ex. S. Prove that the following matrix is unitary
'+o (—l+i)
13
+i)
Ex. 6. Prove that a real matrix is unitary if it is orthogonal.
(Rohilkhand 93)
§ 2-22. Partitioning of Matrices.
Submatm.
Definitioa. A matrix obtained by striking off some of the rows and
columns of another matrix A is defined as a sub-matrix of A.
For example ifA=12 3 I,then
5 7]
[2), [3), [5) etc.
12 31,[3 1]ctc. are all sub— matrices ofA
1 iL 7]
It is sometimes found useful to subdivide a matrix into sub-matrices by
drawing lines parallel to its rows and columns and to consider these
sub-matrices as the elements of the originzll matrix.


98 Matrices
Consider the matrix
A= xiyI1:ajI
X2 Y2 Z2 :a2 52
X3 Y3 Z3 : a3 03

P1 qi rj :al bi
P2 q2 r2 a2 b2

Let An = x Yi ZI ;Al2=al
X2y2Z2 cX2 1'12
X3 )'3 Z3 tX3 33

Azi=pi qi ru; Au=1ai bi

Ip2q2rj L02
Then we may writeA= I A n Al2
Azi A
The matrix A is then said to have been partitioned and the dotted lines
indicate the partitions. Here it is obvious that a matrix can be partitioned in
several ways. The elements An, Al2, A 21 and Azi are themselves matrices and
are the sub-matrices of A.
Identically partitioned matrices.
Two matrices of the same size are known as identically partitioned
matrices if when expressed as matrices of matrices (i.e. when partitioned) they
are of the same order and the corresponding submatrices (or elements) are also
of the same size. Such matrices are said to be additively coherent.
For example
123:75 and 124:30
456:98 205:46

234:23 102:12
567:45 254:34
458:67 262:56
Two matrices A and B, which are conformable to the product AB, are
called multiplicative coherent if A and B are partioned in such a way that
columns of A are partitioned in the same way as the rows of B are partitioned.
Here the rows of A and columns of B can be partitioned in any way.
For example
LetA= 1 0 0 0 and B= 2 3 5
2 0 0 01 371
2 3 1 1 402
251
 Partitioning of Matrices 99
1cre A is a 3 x 4 matrix and B is a 4 x 3 matrix, so these are
conformable to the product AB (i.e. the product AB exits). Now if write
A= 1 00:0 and B= 2:3 5
210:0 3:71
4:02
231:1
2:5 I
then the partitioning of the columns of A is in the same way as the partitioning
of the rows of B. (Here we note that after third column in A the partitioning
has been done and in B the partitioning has been done after third row). Thus
according to definition given above the matrices A and B are called
multiplicative coherent.
Exercise on § 2•22
Ex. Compute AR using partitioning
A= 1 0 0 1 B= 1 0 01
0102 010
0013 000
312

MISCELLANEOUS SOLVED EXAMPLES

Ex. 1. Show that 12 3 is the inverse of 3 -2 - 1
2 5 7 -4 1-1
-2 -4 -5 2 0 1
So!. 12 ) 3-2-1
2 5 7 -4 1 -1
-2 -4 -5 2 0 1
= 1.3 + 2 (-4) + 3.2 1 (-2) + 2.1 + 3.0 1 (- 1) + 2 (- 1) + 3.1
2.3 + 5 (-4) +7.2 2 (-2) + 5.1 + 7.0 2 (- 1) + 5 (- 1) + 7.1
-2.3-4(-4)-5.2 -2(--2)-4.l--5.0 -2(-1)-4(-1)-5.1
1 0 0] =I, where I is an unit matrix.
010
o 0 1
Hence 12 is the inverse of 3 -2 1
2 5 7 -4 1-1
-2 -4 -5 2 0 1
*Ex. 2. If A is a non-singular matrix, then prove that
All = AC -> B = C, where B and C are square matrkcs of the same .rder.
(Kanpur 96)

Sol. Since A is non-singular matrix, so A exists.

100 Matrices 989

Now AB = AC = A 1 (AR) = A 1 (AC),

prcmultiplying both sides by A'
(A 1 A) B = (A 1 A) C,
by associative law of multiplication
=W=IC,
B = C, lB = B etc. Hence proved.
**Ex 3. If product of two non-zero square matrices Is a zero matrix,
then prove hat both of them are singular matrices
Sot. Let A and B be two non-zero n x ,z matrices.
Given that AS 0, where 0 is the a x a null matrix.
Let us suppose that B is non-singluar matrix then l' exists.
Then AS =0 =a (AR) W = OB_1 post multiplying both sides by B"
= A (BB 1) =0, by associative law of multiplication.
(Note)
Al 0,

which is against hypothesis as A is a non-zero matrix.

Hence 0 is not a non-singular matrix i.e. B is a singlu.ar matrix.
Similarly we can prove that A is also a singluar matrix.
Ex. 4. Expresi the following matrix as the sum of a hermitian and
a skew hermitian matrix:
A= 2+31 1—i 2+1
3 4+31 S (Kumaun 92)
1 1+i 21
Sot. From § 217 Theorem m Page 87 we know that
A=(A+A0)+(A—A8)
i.e. the hermitian and skew-herinitian parts of the matrix A are
(A + A0) and (A - A8) respectively.
Now we know that A8 (A)' ...(ii)
where X= 2-31 1+12—i
3 4-31 5
1 1—i —2.1 (Note)
From (ii) we have A (A)' = transpose of A
=2-31 31
l+i 4-3i 1—i
2—i 5 —2i (iii,


Exercises 101

A+A= 2+31 1—i 2+ 11 + 2-3i 3 1

3 4+31 5 1+1 4-3i 1—i
2iJ 2-i 5 —2i
= 2+31+2-31 1-1+ 3 2+1+1
3+1-hi 4+3i+4-3i 5 + I— i
1+1+5 21-21
4 P 4-i 3+1
4+1 8 6-1
3—i 6+1 0
Hermitian part of the given matrix A
=(A+A) 4 4—i 3-,-!
4+1 8 6—i
3—i 6-i-i 0
Again A—A6 2+31 1—i 2+1 - 2-3i 3
3 4+31 5 1+1 4-31 1—i
I 1+i 2i 2—i 5 —21

2+31-2+31 1-1-3 2+i-

3 4+31-4+31 5-1+i
1-2-i-i 1+i-5 21+2i
= 6i —2—i 1+1
2—i 61 4+1
—1+i - . 4+1 41
Skew-hermitian part of the given matrix A
61 —2—i I+i
2—i 61 4+1
—1+1 —4+1 41
Hence from (i), we have the given matrix A
4 4—i 3+1 + 61 —2—i 1+i
.1+1 8 6--i 2—i 61 4+i
6+1 0 —1+1 —4+1 41
L3
which i the sum of a hermitian and a shcw-hermitian matrix (as proved above).
EXERCISES ON CHAPTER II
Ex. 1. Show that
1 0 0 0] is the inverse of I o 00
2100 —2 1 00
4 2 1 0 0 —2 10
8
L2
(Hint. See Ex. I Page 99)
102 Matrices

Ex. 2. If A be any square matrix, then show that A + A is Hermitian.

Er. 3. If A and B are symmetric and they commute, then A 1 B and
At W 1 aré'symmetric.
Er. 4. Show that every square matrix can be expressed in one and only
one way as P + iQ. where P and Q are Hermitian.
Ex. S. If B is any square matrix, show that B'AB is symmetric or
skew-symmetric according as A is symmetric or skew-symmetric provided
B'AB is defined.
Ex.-6. If A and B are two non-singular square matrices of the same order,
which of the following-statements is true
(i)A+BB+A;
(ii) (AB)' A'B';
(iii) (AB) -' = A 1 B'
(iv)A.A'=I => A'= A-'
(v) A + A' is a symmetric matrix,
Er.. 7. If A is Hermitian,such that A 2 = 0, show that A = 0, where 0 is
the zero matrix.
Er. 8. Show that every skew-symmetric matrix of odd order is singular.
Ex. 9. When is a matrix said to be invertible?
[Hint: See * 2. 18 Page 91).
Ex. 10. If D = diag [d1 d2..., dr],

dId2 ... dn*O, what will be D1?

Er. 11. If non-singular matrices A and B commute, then
(i) A' and B and (ii) A 1 and B1
also commute.


Chapter Ii!
Equivalence
§ 3 01. Elementary Row operations
Consider the matrices
A= 1 2 3- 1, B= 4 5 61, C= 3 6 91, D= 1 2 3
458. 6912
456 1 •2 3
789 789 789 789

Here we observe that the matrices B, C, D are related to the matrix A in

as much as
(a) B can be ob'nd from A by interchanging first and secondraws of A
(b) C can be obaiind from A by multiplying the first row of A by 3 and
(c) I) can be obtained from A by adding two times the first row to the
second row of A.
Such operations on the rows of a matrix' are known as elementary row
operations. Formal definition is given below
Definition. Let Ai denote the ith row of the mgtrix A = [a jJ . then the
elementary row operations on the marix A are defined as:
(i) the interchanging of any two rows A i and A, (i.e. ith and jth rows).
The symbols R11 or R --* R art generally employed for this
operation.
(ii) the multiplication of every element of Ai by' a non-zero scalar c i.e.
replacing the ith row A i by cA1. The symbols R' (c) or R -4 cRì are
employed for this operation.
(iii) the addition to the elements of row &, of r (a scalar) tunes the
corresponding elements of the row Ak i.e. replacing the row Ai. by
A + CAk.
The symbols Rik (c) or R -4 R + cRt are used for this operation.
Note : The above operation do not change the order of the matrix.
Example:I.etA 1 2 3
345
567
The effect of the elementary row operation R2 - Ri or R21 (- 1) is to
produce the matrix
B=1 1 2 3= 1 2 3
1 3- I 4-2 5-3 2 2 2
1 6 7 567
104 Matrices
Again the effect of elementary row operation R2-+- Pt or R2j (I) is to
produce the matrix
B = 1 3 3= 1 2 3 i.e. the matrix A.
2-s-I 2+2 2+3 3 4 5
5 6 7 5 67

Thus the above two operations are the inverse elementary row operations.
34)2. Row equivalent Matrices.
Definition. If an ,n x n matrix B can be obtained from an rn x n mauix A
by a finite number of elementary row operations. then B is called the row
equivalent to A and is written as
row
B — A.
Note : Equivalent matrices have the same order.
Example: 1 3 4 7 row 2 -3 5 6
2 —35 6 - 1 3 4 7
1 0 3 2 1 0 .32
(interchanging first and seor.d rows).
303. Elementary Row Matrix.
Definition. The matrix obtained by the application of one elementary
row operation to the identity matrix I is called an elementary row matrix.
Exampk. Examples of elementary matrices obtained from 13, where
13= 1 0 0
0. 10
001
L
(i) 13— 0 I 0=E0 (say),
100
001
)btalned by interchanging first two rows.
(ii) 13 - 1 0 0 = Eb (say),
OcO
0 0 1
obtained by multiplying the elements of second row by c.
(iii) L311 2 0 =Ec(say),
010
001
obtained by adding two times the elements of second row to the corresponding
elements of first row i.e. replacing Ri by R + 2R2 i.e. R12 (2).
 Elementary Row Matrices 105

§ 304. Types of Elementary Row Matrices and their symbols.

(i) Erj denotes the elementary matrix obtained by interchanging the ith
and jth rows (or columns) of an identity (or unit) matrix.
(ii) E (c) denotes the elementary matrix obtained by multiplying the ith
row (or column) of the identity matrix by c.
(iii) Ek (c) denotes the elementary matrix obtained by adding to the
elements of the ith row of the identity matrix c limes the corresponding
elements of the kth row;
(iv) E'1 k (c) denotes the transpose of E1k (c) and can be obtained by
adding to the elements of the ith column of the identity matrix c times the
corresponding elements of the kth column.
§ 305. Theorem. Each elementary row operation on in x n matrix can
be effected by piemuttiplying it by the corresponding elemnto.iy matrix-
Example : Let A . au 0 12 0 13 0141
021 022 023 024
ra31 032 a33 034
(i) Interchanging the first and third rows, we have
A 031 032 033 034 = B (say)
a2l a22 023 024
011 012 013 OP
The corresponding elementary matrix (obtained by interchanging first
and third row of 12) is given by
E13 0 0 1
010
100
[Here students should note that as we are to premultiply Athcrcfore the
number of columns of E13 should be 3, the number of rows of A].
Now E13, A= 00 1 X au 0 12 013 014
o 1 0 021 022 (t23 (224
1 0 0 031 032 033 034

= 0 31 032 033 034 =11

021 a22 023 024
all a12 0 13 014

This shows that B can be obtained from A by pre-multiplying it by E,

the corresponding elementary matrix.
(ii) Let A= 1 2 3
456
89
Multiplying the element,, of second row by 2, we


106 Matrices
A—. 1 2 3 =B(say) 1
8 10 12
789
The corresponding elementary matrix (obtained by multilying the
elements of second row of 13 by 2) is given by
E2(2) = 1 0 0
020
001
Then E3(2)XA 1 0 0 ri 2 3
020 456
001 789
= 11+04+07 12+0'5+08 1.3+06+09
01+2 .4+07 02+25+08 03+2•6+09
01 +04+ 17 0-2+05+1-8 0•3+06+ 19
=1 2 3] = B
8 10 12
.7 8 9
i.e. B can be obtained fromA by pie multiplying it by E2 (2),
(iii) Let A=[ 1 —2 3
—3 4 5
6 —7
[ 5
Replacing Ri by R I + 2R2 i.e. adding two times the elements of second
row to the corresponding elements of first row, we get
A— —5 6 13 =B(say)
—3 4 . 5
5 6 —7
The corresponding elementary matrix (obtained by adding two times the
elements of second row of 13 to the corresponding elements of the first row) is
given by
E12 (2) 1 2 0
010
001
Then E12 (2) x = 1 2 0 X 1 -2 3
0 1 0 —3 4 5
0 0 1 5 6. —7

= 11 +2(-3)+05 1(-2)+24+06 1•3+25+0.(-7)

0. 1 + I (-3)+0'5 0(-2)+ 1•4+0•6 03+ l5+0(-7)
01 0 (-3) + 15 0(— 2) + 0-4 + 1-6 0-3 + 0-5 + 1(— 7)
 Elementary Row Matrices 107

=-5 6 13=B
—3 4 5
5 6 —7
i.e. B can he obtained from A by pre-multiplying it by Eu (2).
COROLLARY of Theorem given in 3 . 05 Page 105.
If the matrix B is row equivalent to the matrix A, then B = S.A, where S
is a product of the elementary matrices.
§ 306. Theorem. The elementary matrices E, E1 (c) , Ej k (1) are
non-singular. (See § 218 Page 91)
Proof: (i) The elementary matrix Ejj is obtained by interchanging the ith
and jth rows of I. We shall get back I if we now apply the same row operation
upon E ij which can also be effected by pre-multiplying Eij by E11
(See § 3 .05 Page 105).
Ej.E,=L
i.e. E 1 its own inverse i.e. Eij is non-singular.
(ii) The elementary matrix E (c) is obtained by multiplying the ith row
of the identity matrix by c (where c * 0). We shall get back I if we now
multiply the elements of ith row of E (c) by 1/c which can also be effected by
pre-multiplying E1 (c) with the corresponding elementary matrix 'which is
obtained from I by multiplying its ith row by 1/c, which is therefore the
inverse of E (c). -
For example, let I = 1 0 0and E3 (c) = 1 0 0
010 010
001 OOc

Then {E3 (c)} = 1 0 0 , where {E3 (c)}' is the inverse of E3 (c)

010
0 0 1/c

(iii) The elementary matrix Eik (1) obtained from I by replacing its jth
row by (jth row + kth row).
We shall get back I if we not replace the jth row of Ey (1) by (ith row
- kth row). (Note)
Hence the inverse of E3k (1) is the elementary matrix obtained from I by
replacing its Jth row by (idi row - kth row).
For example, let I 1 0 0 and Eu (1) = .1 0 1
010 010
001 001
obtained from I by replacing its 1st row by (1s( row + 3rd row).
108 Matrices

Then {E13 (l)}_1 I 0 - I obtained from I by replacing its first row

0 0 1 by (1st row - 3rd row)
00 1
§ 307. Theorem : if the matrix B is row equivalent to the matrix A
then B' = SA where S is non-singular:
From Cor. of § 305 Page 107 we know that
row
if B A, then B = SA. where S is the
product of the elementary matrices and in § 306 above we have proved that
elementary matrices are non-singular and hence their product is also non-singular.
This proves the above theorem.
§ 308. Theorem If a square matrix A of Order n is row equivalent to
the identity matrix I,, then A is non-singular
Proof: From § 307 above we know that
row
A In then A S.In wherre S is non-sinu1ar.
Now Soln being the product of two non-singular matrices is non-singular.
Therefore A is non-singular.
Note. The converse of this theorem is also true.
§ 309. Theorem : If a sequence of mw operations applied to a square
matrix A reduces it to the identit y matrix 1. then the same sequence of row
operations applied to the iieniity matrix gives the inverse of A (i.e. A1).
Proof: From Cpr: of § 305 Page 107 we know that SA 1, where S is
the product of the elementary matrices.
i.e. (Ek ... E3 E2 Ei) A = I, where E i denotes the elementar y matrices
or E3 E2. Ei) AA'
or (Ek ... E. E2, Et) I = A'. since AA -1 I and 1A1
(See * 218 Page 91 and Et. 1 Page 64)
Hence the theorem
Note. With the help of the above theorem we shall find the inverse of the
given non-singular matrix A.
In the following examples we shall show the successive matrices row
equivalent to A and I in the left hand and right hand columns respectively.
When uhimately A is reduced to I in the left hand column, I is reduced to A1
in the right hand column.
Also R1, R2, R3 ,... etc. stand for first row, second row, third row. etc.
Solved Examples on § 309.
*Ex. I Find the inverse of the matrix A = 1 -3 2
2 00
1 4


Solved Examples on inverses 109

Sol
A I
I -3 2 1 0 0
2 00 010
I 4 I 00 1
1 -3 2 - I 0 0
1 0 0 0 f 0
I 4 1 00 1
-[
(Replacing R2 by 1 R2)
-1 • 0 0-0 • 0
1 -3 2 1 o 0
I 41 001
(Interchanging Ri and R2)
-1 00-0 (1/2)0
0 -3 2 1 -(1/2) 0
0 4 1 0 -(1/2)
(Replacing R2 by R2 - R and R3 by Rj - Ri)
-100-0 0
0 -11 0 1 -2
O 4 1 0 _..!

(Replacing R2 by R2 - 2R3)
1
-Ii 0 011-I 0 01
Jo I oil I_L _. 11
U 22 II
'JI L 0 -f 1]

(Replacing R2 by - 1j R)

-Ii 0
0i
f 01
to 1 i

1 0 0 Ij -i - Ill
fl IIJ I
(Replacing R3 by R3 - 4R2)
=1 I =A
A-1 = r 0
01
I ii - i III
I -
L -i ' lj Ans.

—8


110 Matrices

*Ex. 2. A = 1 2 1 , evaulate A',

323
112
• A I
1 21 100
So!.
3 2 3 0 1 0
1 12 001

1 2 1 - 1 0 0
0 —4 0 —3 1 0
0 —1 1 —1 0 1

(Replacing R2 by R2 - 3Ri and R3 by R3— R1)

0 —1
- I 3 0 - 2
0 1 0 - 0
0 —1 1 —1 0 1

(ReplacingRi by RI - R3 and R2 by - R2)

—1
—[i 0 01-
0 i 01 - 0
0 1] - _! 1
1o 3 4

(Replacing Ri by RI -3R2 and R3 by R3 + R2)

• =1 I =A1
A=— —1
0
I Axis,

Ex.3. Find the inverxe of the matrix A = 1 2 -2

—1 3 0
0 —2
So!. A I I
• 12-2--I 00
—1 3 0 0 1 0
0 —2 1 0 0 1

- I —2 0 - 1 0 2
—1 3 0 0 1 0
0 —2 1 0 0 1

(Replacing RI by RI + 2R2)


Elementary Column Operations - 111

- 1 —2
1 0 - 1 0 2
o 112
o —2 1 0 0 1
(Replacing R2 by R2 + Ri)
—100-326
010 112
001 225
(Replacing Ri by Ri + 2R2 and R3 by R3 + 2R2)
=1 I
A'=3 26
112
225 Ans.

Exercises on § 309
Ex. 1. Find A 1 if A= 2 43 Ans. 3 - 10 - 1
01 I —2 8 2
2 2 —1 2 —4 —2
Ex. 2. Find the reciprocal matrix of 1 1 1 Ans. - - 6 5-I
2 2 3 15 —8 1
2 4 9 —6 3 0

Ex. 3. Find A, if A = 1 2 1 Ans. 0 3 —3

3 i 2 -. 6 —2 —1
002 —3 15

§ 3-10. Elementary Column Operation and Column Equivalent

Matrices.
In § 301 Page 103, if the word row is replaced by the word column we
get the definition of the elementary column operation.
Similarly in § 3 .02 Page 104 repacing the word row by the word column
we get definition of column equivalent matrices.
col -
B - A means the matrix B is column equivalent to the matrix A.
Symbols for column operations are similar to those given' for row
operations in § 301 Page' 103. Here the letter R in the symbols are to be
replaced by C e.g. Cy, C (c), C,k (c), where stands for the interchange of
ith and jth cokimns etc or C1 -- G C1 --4 cC1, C - eCk.
§ 311. Theorem. Each elemniary column operation on an rn X n matrix
A can he effected b' post multiplying A hr the n x n matrix obtained froi-n the
n )< it ith'ntii ?u?(1frjx I b y the same elen:enta co/u?:,z operation.
112 Matrices 182/1/7
cot
Proof: if B - A
row
then B' A' where B' and A' are the transposed matrices of B and A.
(Sec § 2-08 Page 69)
Since if B is obtained from A by ckmentary column operation, then B'
can be obtained from A' by an elementary row operation.
Hence B' FA', where E is the elementary matriK obtained from 4 by
an elementary row operation. (See § 3-05 Page 105)
Therefore B = AE', (Note)
where E', the transposed matrix of E. can be obtained from In by the same
elementary column operation.
Hence the theorem.
3-12. Theorem.. it there be two m x n matrices A and B, then
cot
B - A if B = AT, where T is an n x n non-singular matrix
col cot
Proof: If B - A, then B' - A', where B' and A' are the transposed
matrices of B and A respectively.
Therefore B' SA', where S is an n x n non-singular matrix.
(See § 3-07 Page 108)
Consequently B = AS', where S' is the transposed matrix of S
= AT, where T = S', an a x z singular matrix.
Hence the theorem.
* 3-13. Equivslent Matrices (General Definition). (At'adh 95)
Definition. Two , x n matrices A and U are called equivalent if one can
be obtained from the other by a finite number of row and column operations (or
elementary operations) and written as H - A.
* 3-14. ThaijWar Matrix.
Definition. A matrix [a] is called a triangular matrix if
aii =0 for 1 >1.
For Example 2 3 1 4 or 2 3 4 5
0123 0134
0057 0025
0007
Note 1. Triangular matrix need not be square. If it is square, then it is
called upper triangular matrix. (Sec § 201 (a) Page 61)
Note 2. The elements aj for which i ^j are not necessarily zero.
9 3-15. Theorem. Every matrix can be reduced to triangular fonii by
elementary row operations.
Proof: We shall prove this theorem by Mathematical intthtction.


182/118 Theorem on Triangular Matrices 113

Assume that this theorem holds for all matrices containing n - 1 rows
and let A = [ a,j ) be an n x m matrix given below -
A= all a12 a3
a 2l a22 a23

ani an2 a3

Now the following cases arise

Case L If a I I #O, then replacing Ri by (1/au) RI (i.e. by applying
elementary row operation) the matrix A reduces to an n x m matrix
B=(b,1 1= b11 b12 ... ... bi,,,
b21b22 ......

b,,1 b,, ... ... b,,,,,

where bii = 1.
Now applying elementary row-operation Rt—btjRi to R* where k= 1, 2,
n i.e. subtract bb times Rj from Rk, where k takes values from 1 to n.
This reduces the matrix B to matrix C = [cj] where cj = 0 whenever k> I and
we have
C I C12 C13 ... Cl.
o C22 C23 ...

o Cn2 C,1 3 .. C,

Now by our assumption that the theorem which we are going to r've
holds for matrices containing (n - 1) rows we find that (n - I) rowed matrix
o C22 C23 ...
o C32 C33 ... C3,,,

o Cn2 C O ...

can always be reduced to triangular form by elementary row operations and

hence from (i) the matrix C will reduce to triangular form when the same
elementary row operations are applied to C.
Case H. If au =0 but ai #0 for some value of k then interchanging Ri
and Rk the matrix A reduces to the matrix D = [di,] where dii * 0.
Then the matrix D can always be reduced to the triangular form as in
case I above.
Case III. If aki = 0 for allvalues of k then we have
A= 0 a 12 a3 ...
o a22 a 23 •.. a2m

0 2n2 a,,3 ...

114 Matrices
By hypothesis (inductive) the (n - I) rowed matrix
0 4122 '723

- 0 a,i2 a3 £2nm

as in case I above can be reduced to triangular form by elementary row

operations and the same elementary operations when applied on A will reduce
A to triangular form.
Hence the matrix A can always be reduced to triangular form and the
proof is complete by mathematical induction.
Solved Examples on § 315.
Ex. 1. Reduce the matrix 3 1 4to triangular form.
1 2 —5
01 2
SoL Let A3 14
1 2 —5
01 2
- 1 !, replacing Ri hyRi
• 1 2 —5
01 2

- 1 !replacing R3 by R2 - RI

0 _12
3 3
01 2

- 1 replacing R3 by R3 - R2
o_4
00

This is the required triangular form.

Aliter A=3 1 4
1 2 —5
01 2
1 2 - 5 interchanging R 1 and R2
31 4
01 2
-[
- 12 —5 replacing R2 by R2 - 3Ri
I 0-519
0 1 2


Solved Examples on Triangular Matrices 115

1 2 —5 replacing R3 by 5R3
0-519
o 5 10

1 2 - 5 replacing R3 by R + R2
o —5 19
o 0 29
This is also a triangular matrix.
Note. The above shows that reduction of a matrix to triangular form is
not unique.
Ex. 2. Reduce A = 5 3 14 41 to triangular form.
01 31
2 0 (Agra 95-)
—1 1

Sol. Let A= 5 3 14 4
01 31
—1 1 20
- - 1 12 0 interchanging Ri and R3
01 1
5 3 44
- I - 1 -2 0 replacing Rj by - Ri
0 1 31
5 3 14 4
J,
- I - 1 —2 0 replacing R3 by R3 - SRi
0 1 31
0 8 24 4

- I - I .-2 0 replacing R3 by R3 - 8R2

0 1 3 1
0 0 0 —4
This is a triangular matrix as here o.1j = 0 for i > j. See definition § 314
Page 112]
Exercises on § 315
I
Ex. I. Reduce the matrix I - 1 1 to the triangular form.
2 34
3 —1 4
Ans. 1 —1
0 10
0 01


116 Matrices 989

EL 2. Reduce the matrix1 2 I S to the triangular form.
2 1 —1 0
32 1 7
r
Ans. 1 —2 —1 —8
o 5 1 16
o o 12 27

MISCELLANEOUS SOLVED EXAMPLES

Ex. 1. Apply successively the row transformations (or operation) R,
R3(-2) p'41R12(4)to the B1atrix 3 1 2 1
2032
1234
3141'
SoL. (1). 4pplying R23 opertion to the given matrix we have
3 1 1 [Here we have interchanged second and third rows].
1234
2032
3141
(i) Applyiiig R3 (-2) operation to the given matrix we have
3 1 2 1 [Here we have replaced third row R3 by - 2R31
2 0 3 2
—2 —4 —6 —8
3 1 4 1

(lii) Applying R12 (4) operation to the given matrix we have'

7 9 14 17 [Here we have replaced the first row R1 by RI + 4R31
2032
1 2 3 .4
314 1
Ex. 2. Apply successively the column operation C13 ;C2 (-4) and
C (-2) to the matrix I - 1 2 3 4
2 1 —2 1 3
3 2 1-25
4 5 678

SoL (i) Applying C13 operation to the given matrix, we have

2 -1 1 2 4 [Here we have interchanged Ci and C3
—2 1 2 1 3 i.e. first and third columns.]
1 23-25
6 54 78
(ii) Applying C2 (- 4) operation to the given matrix, we have
 Miscellaneous Solved Examples 117
1 4 2 3 4 [Here we have replaced second
2 —4 —2 1 3 column C2 by - 4C2].
3 — g l
.-2 5
4 —20 6 7 8
(iii) Applying C23 (- 2) operation to the given matrix, we have
1 - 5 2 3 4 [Here we have replaced second
2 5 -2 1 3 column C2 by C2 - 2C31.
3 0 1 —2 5
4 —7 6 78
Ex. 3. Compute the following elementary matrices of order 4
E, E2 (4), E (— 2), E' (- 2). (Refer § 304 Page 1O)
Sot. The identity (or unit) matrix of order four is given by
14= 1 0 0 0
0100
0010
0001
(i) E23 = 1 0 0 0 [Interchanging R2 and
0 0 I 0 R3o rC2 and C3]
0100
0 0 0 I (Nob)

(ii) E2 (4) = 1 0 0 0 [Replacing R2 by 4R2 or C2 by 4C21.

0400
0010
0001
(iii) E34 (-2) = 1 0 0 0 [Replacing R3 by R3 - 2/?1.
010 0
0 0 1 —2
000 1.
(iv) E'34(-2) = 1 0 0 0[Replacing C3 by C3 - 2C41.
• 01 00
00 10
• 00-21
(Here students should note that E'M (-2) is nothing but the franspose
matrix of E (- 2)].
Ex. 4. Evaluate the inverse of the following elementary niMi4ces
order four: E3 (- 2), E23 (4) (Refer § 306 Page 17-1*)
Sol. The identity matrix of order four is given by
14=1000
0100
0010
0001


118 Matrices
(i) Then E3 (- 2) 1 0 0 0 replacing R3 by - 2R3
01 00
0 0 —2 0
00 01

Inverse of E3 (-2) i.e. (E3 (- 2)[ 1 0 0 0

0100
0 0 - 1 0
0001 Ans.
(obtained by replacing R3 of 1 1 by - R). (Note)
(ii) E13 (4) = 1 0 0 0 , replacing R2 of 14 by R2 + 4R3.
0140
0010
0001
Then the inverse of E (4) i.e. {E23 (4)} is given by
1 0 0 0 , replacing R2 of 14 by R2 - 4R3.
0 1 —4 0
00 10
00 01 (Note)
Ex. 5. Find the inverse of the matrix A = i - 12i
2 02
—1 01
SoL

A I
i-12i1 00
2 02 0 10
:1 0 1 0 01

—i-12i-1 00
1 01 0 1/2 0
—1 0 1 0 01
(Replacing R2 by . R2)

—i-12i—1 00
1 01 0 1/2 0
0 02 0 1/2 1

(Replacing R 3 by R3 + R2)
-
— i—I 21—I 0 0
1 0 1 0 1/2 0
0 0 1 0 1/4 1/2

(Replacing R3 by . R3)


Miscellaneous Solved Examples 119

—1—I 2i —1 0 0
1 O'O 01/4 —1/2
o 0 1 0 1/4 1/2

(Replacing R2 by R2 - R3)
—0-1 0 —1 —i H
1 0 0 0 1/4 —1/2
o 0 I 0 1/4 1/2

(Replacing Rt by Ri - iR2 - 2i R3) (Note)

10-
0 0 1/4 —1/2
0 0 1/4 1/2
r 0111
(Replacing R by - Ri)
- 1 0 0 =1 - 0 1/4 —1/2A1
o i 0 —1
0 0 1 0 1/4 1/2
(Interchanging Ri and R2)
Therefore A 1 = o 1/4 -1/2
—1 i
0 1/4 1/2 Ans.

EXERCISES ON CHAPTER III

Ex. 1. Apply the row operation R4 (- 3) and R21 (4) to the matrix
4 —1 2 3
—1 8 —3 —4
2 3 4 —1
—3 —4 —1 8

(Hint ; See Ex. 1 Page 116)

Ans. 4 - 1 2 3 and 4 -1 2 3
—1 8 —3 —4 15 4 5 8
2 3 4 —1 2 3 4 —1
9 12 3 —24 —3 —4 —1 8

Ex. 2. Apply the column operation C3 (4) and C12 (- 3) to the matrix
01234
12340
34012
20134
[Hint : See EL 2 Page 1161


120 Matrices
Aa.0 I 83 4and-3 123 4
1 2 12 4 0 —5 2 3 4 0
3 4 0 1 2 —9 4 0 1 2
2 0 4 3 4 2 0 1 3 4
Ex. 3. Compute EB, E2 (-2) and E34 (— 1) for the identity matrix of
order 4. (Hint. See EL 3 Page 117)
Ans.1 000:1 000.100 0
00100-200010 0
0 1 0 0 0 0 1 0 0 0 1 —1
00010 0 • 01 000 1
EL 4. Evaluate the inverse of the following elementary matrices of order 4:
Em, Es (3), E22 (2);
(Hint: See EL 4 Page 117).
Ans.E14, 1 0 0 0 and 1 0 00
0 1 0 0 0 1 —2 D.
001 0 00 1o
000k 00 01
Ex. S. Find the inverse of the matrix
A=1
1 1 1 Ans. J- 2 6 6 4
1 2 3 —4 22 41 —30 —1
2 3 5 —5 —10 —44 30 —2
3 —4 —5 8 4 —13 6 —1
(Hint Set EL 5 Page 118).
* Ex. 6. Find the inverse of the matrix r 1 2 —1
1 2
2-1 1
(Hint: See EL S Page 118).
3 —1 5
5 3 —1
—1 5 3
Ex. 7. Find the inverse of the matrix 1 3 3 2 1
1 4 33-1
1 3 41 1
1 1 1 1 —1
1 —2 —1 2 2
(Hint : See Ex. 5 Page 118).

Miscellaneous Solved Examples 121
Ans. 1! 30 —20 —15 25 —5
30 —11 —18 7 —8
-30 12 21 —9 6
- 15 2 6 —9 6
15 —7 —6 —1 —1

Ex. 8. Has the following matrix an inverse?

2 1 3 1
1 2 —1 4
33 2 5
1 —1 4 —1
(Hint : It can not be reduced to 14). Ans. No.


CHAPTER IV.
Determir mis

§ 4 .01. Permutations.
Def. The operation of rearranging n distinct elements of a set among
themselves is called permutation.
Let S be a set defined by
in) : im;t ik for pn;tk.
S = k02, i3 ,...,
Let P be the transformation on S, such that P(ii) = al, P(i2)=a2,
P(i3)=a3 ,..., P(i,)=a , where 01,02,..., a, is some arrangement of the
elements i1, i2 ..., in of S.
Then a two line notation for the permutation is
p = (aill 2 i3 ... in )
02 03 ... an

(Note : The order of columns in this notation is immaterial).

If in a given permutation a larger integer precedes a smaller one then
there is an inversion. For example in 452 we see that 5 precedes 2.
If in a given permutation the number of inversions is odd, the
permutation is known as odd. For example the permutation 5312 is odd as in
this permutation we observe that 5 precedes 3, 5 precedes I, 5 precedes 2, 3
precedes I and 3 precedes 2 i.e., there are five (i.e. odd) inversion in 5312
Similarly if in a given permutation the number of inversions is even the
permutation is known as even. For example the permutation 5314 is even as in
this permutation we observe that 5 precedes 3, 5 precedes 1, 5 precedes 4 and 3
precedes 1.
Note : If there is no inversion the permutation is even, for example 345.
§ 402. Determinant of a square matrix.
Let us consider a square matrix A of order n n given by
A= an 012 0]3 ......01n
(21 (222 (223 ......02n
12 31 032 (233 ......(13n

QnI 0n2 an3

The product of the elements in the principal diagonal is
au 1 022 (233 ...
This is also called the trace of the matrix.
Now obtain n! terms of the above type by operating on the
row-subscripts of the. elements of the above expression by n! permutations

p= I '2 ... 1
'3In),
where i1,i2.... are one of then! permutations of the
integers 1,2,3.... . n.


Determinant of order two 123

The sum of n signed terms thus obtained is defined as the determinants

of the matrix A and is denoted by I A I or I aq I
Therefore the determinant of thc square matrix A = of order n x it is

given by I a 1 I = Y, ± a ap, a 03 ...

where +or —sign is taken when a, , y, 3,.... k is an even or odd permutation of

1, 2, 3 .....it. and the summation extends over it permutations of the row
subscripts 1,2......
Note : The determinant of a square matrix of order it is known as a
determinant of order it.
§ 4 . 03. Determinant of order two.
Let us consider asquare matrix A = [ all 0121 of order 2-x 2.
[021 022]

Then 2 ! permutation on two symbols I and 2 are

(I 2 (1 2
'1 22 )' ' I
The product of the elements of the principal diagonal are a 11 a.
Operating on the row subscripts of au I 022 by the permutation I we get
+011 a22, prefixing + sign as the permutation I is even (See § 401 Note) and
by the permutation p we get - aj ('12, prefixing - sign as the p.rmutation is
odd.
Hence I all 012 =011 022-021 ('12
021 ('22

§ 4-04. Determinant of order three.

Let us consider a sqaure matrix A = all ('12 a13 of order 3 x 3.
02! ('22 ('23
('31 032 033

Then 3 ! permutations on three symbols 1, 2 and 3 are

1 2 3 '); p1 =(i 2 3); p2(1 2 3
2 3) 1 3 2) 2 3 1
(
(1 2 3 '); p4=(I 2 3) and p5(l 2 -3
1 3)
3 1 2) 3 2 1

Operating on the row subscripts of all 022 ('33 by the permutation

I,p1,P2 ,..- p5 we have successively
(a) + a H 022 a3, prefixing + as permutation I is even.
(b) - 01! (232 023. prefixing - as permutation pi is odd
(c) +021 ('32013, prefixing + sign as permutaton p2 is even.
(d) — 021 012 a33, prefixing - sign as permutation P3 is odd.
(e) + OU 6112 023. prefixing + sign as permutation P4 is even.


124 Matrices
(f) -a1 a22 a t3, prefixing - sign as permutation ps is odd.
Hence we have
all a12 a3 =alla22a33-a11a32a23+a21a32a13.
a2l a22 a 23 -a21a12a33+a31a12a23-a3Ia22aI3
a3l a32 G33
=all (a22a33-a23a32)-a12(a21a33-a23a31)
+-a!3 (a21 a32 - a22 613 1)
=all a22 a12 a2l a 23 +a13 a a22
a
32 331 (43k a331 (43k 32
23-
Solved Examples on § 402 to § 404.
EL 1. Evaluate -5 0
7 -2
Sol. -5 0 (-5)(-2)-0x710-010.
7 -2 Ans.
Ex. 2. Evaluate 1 2 3
4 5 6
789
Sol. 1 2 3 = 1 (5-9-68) -2(49--67)+3 (4.8-57),
4 5 6 See *4'O4 Page l23
789
=1 (45-48)-2(36-42)+3(32-35)
=-3+12--9=0 Ans.
Exercise
Ex. Show that2 5 =29
-• 3 7
* 405. Cofactor of an element.
Defluition. If in the expansion of a determinant I au I, all the terms
containing ay as a factor are collected and their sum be denoted by au Cy, then
the factor Cj is defined as the cofactor of the element aV.
From the above definition we find that if [aij ] be the n x n matrix whose
determinant is I a(/I then if from [au] the element of its ith row and fib column
are removed, the terms of Cu are then composed of elements from the
remaining (n - 1) x (n - l)sub-matrix M9 of [aijr .
Hence the determinant I aj I can be expressed as a function of the
elements of ith row, by collecting all the terms containing ail, ai2, aj3, ... , aij
and finding their sum
i.e. Iajl=a11C11+aaC12+ ... +ajCu1
R
=Za11 C11,
j- 1


- Properties of Determinants 125

which iS the expansion of the. determinant I ay I by the elements of the ith row
and their cofactors.
In a similar manner we can expand the determinant I aq I by the elements
of the kth column and their cofactors and write as

aij E OIACIk
k=1
Example : Expand the determinant h g
h
2 b f
gfc
by the elements of 1st row. - -
Sol. Elements of first row are a, h, g.
Let A. H and G denote the cofactors of a, h, g.
Then A= b f ;H=— h f and G= It b
f g Ig f
Hence a h g = aA + hH + gG See. § 404 above
h b f
g f c

b fl_ h h f I +g I It b
f c g C g f
=a(bc—/)—h(ch—fg)+g(hf_bg)
=abc+2fgh—a/2 —bg2 —ch 2 , Ans.
* 4•e6. Properties of Determinants.
Prop I. If the elements of ith row (or ith column) of the determinant I a,,,
are multiplu d by a scalar c then the resulting determinant is c I aq I.
(Gorakhpur 93)
7

Proof: We can write Ia,I= E a,C j See § 4 .05 above,

j=I
In this case, the resulting determinant (when the elements of ith row are
multiplied by c)
Pt

caC,1 =c E a(,.C,J=ca/I.
j=l j=i
Similarly we can prove that statement when elements of kth column are
multiplied by c.
Prop 11. If A = [ a jA is an a x a matrix then I A' I = I A I , where A' is the
transpose of the rnal,'ix A. (See § 208 Page 69)
Proof: If A = [aj}, then A' = [a',j ], where a',1
Now the product of elements of the principal diagonal of A'
=a'ii 0220'33
—9


126 Matrices
Operating on the row subscripts of the elements of this product by the
permutation where i1,i2,ii,... are 1, 2, 3, ... in sonic
it
order, we have ± d1 i a'22 a',33... a'j,,, as a term of I A' I plus or minus sign to
be taken according as p is even or odd.
Now as a', = aji , so we have
d1ii a'i22 a'133... a'jn,, = alit Q2r2 a3,3... an,,,
The term alil 02i2 ... a,,j,, can be obtained from the term
ajl,j a,2j2 a 3,3... ainin by opperating on its row subscripts the permutation
(ill i2 i 3 ... in' -l
2 3 ... nJP
Hence p' is even or odd according as p is even or odd. Therefore the
term ± a'a ii a'122 a' 1 33... a'jp,,, of I A' I is also a term of I A I.
We can thus prove that every one of the a ! terms of I A' I is a term of
I A I. Hence the property.
Forexample,ifIAI= a j 02 03
bi b2 h3
Cl C2 C3
Q1 b2 h3 —a bi bi +a3 b b2
C2 C3 Cl C3 CC2

=alb2c3_atb3c2_a2blc3+a2h3Cl+03h1C2-03112C1.

Then IA'I= ai bj Cl
a2 b C2
03 b3 C3

=01 b2 C2 —bi az C2 +C[ 02 b2

b3 C31 0 C31 03 b3

=(Z1b2C3 —a l b 3 c 2 —a2 h l c3 +a3b l c2+a2b3ct —a3b2e1

IA'lIAI. (Gorakhpu..r95)
Prop. III. If B is obtained from A by interchanging two rows (or
columns) then I B I = - I A I.
Proof : Let us suppose that sth and tth columns of the determinant A are
interchanged where s < t.
Let A= al I .l2 ... at,, .-. (11, , .. 01n
021 022 ... (12.ç ... 02r ...
... ... ... ... ...
O il a,i ... aix ... air . . . a,,,

ant an2 ... (1ns ... a,, tann

Then the 'trace of A i.e. the product of the elements of the principal
. . . a, ..


Properties of Determinants 127

If the sth arid tth columns are interchanged the product of the elements of
the principal diagonal of B
= ail a22 ass an ... a, (Note)
In order to have a term of I A I, let us operate on the row subscripts of the
trace of.A by the permutation
(i 2 s ... t ... n
12 ... ...
=1:1
Then we have a1 j 1 (2122 a 55 a
This terms can also he obtained from the trace of B by operating on its
row subscripts by the permutation
P,=)(I 2 .. s ... t n
i2 ... ... is in
Here we observe that p'= p ( i5 i t ), since (i5 i,) is a transposition. Therefore
p' is odd or even according as p is even or odd. Hence the term
± Oil I (122 a 55 a,t ainn is also a term of I B I but with its sign changed.

Thus every one of the n terms of I A I is a term of I B I but with sign changed
Hepce IBI= - IAI.
Example: Let I A I = at a2 a3
bi b2 b3
C 1 C2 C3

Let the determinant I B I be formed by interchanging second and third

columns.
Then IBI= ai at a2
bi b3 b2
Cl C3 C2

Exapanding i B I by the elements of its first row, we get

IBI = al b3 b2 — 03 b 1 b2 +a2 bi b3
C C2 Icj C2 Cj C3

=ai (blc2-b2C3)-a3(hlc2-h2c1)+02(blc3-b3ct)
= - [a I (bci - b 3 c2 ) - a (bid - bId) + 613 (h t c2 - bci)}
h 1) 3 -a2 lb b +O .b1½ c
=-
C2 (3 Cl Cl C2
[
(li 02 (t IAI.
=-
bi 1)2 h3
Cl C2 (3

Prop. IV. if two rows or two columns of a determinant I A I are identical

then I A I = C)


128 Matrices 182/1/8

Proof In Prop. 111 above we have proved that if any two rows or
columns of a determina't are interchanged then the value of the determinant
changes in sign only.
Thus if the two identical columns (or rows) of a determinant are
interchanged, then the sterminant does not change but its sign only changes.
Hence IA l=—IAIj.e. IA 1+ IA =0 i.e. IA 1= 0.
Example: Evaluate al a2 a
hi b2 b1
C 1 C2 Cj
Expanding the determinant with respect to the first row we have the
given determinant
a1 h2 b1 a2 I b, b1 +a j bi 12
C2 Cl Cj Cl Cl C

'=aI(b2cI-bIc2)—a2(bjcl--blcI)+l(blC2_b2c1)
= alb2cI —aIblc2—a2(0)+albjc2—aIb2c10.
407 Minor of an element
Definition : If Mj be the (n— l)x(n — 1) sub-matrix of the matrix
A = [Oij] obtained by removing the ith row and jth column, then the determinant
I My I is defined as the minor of the element aq in the determinant I a,j I of
order n.
§ 408. Theorem. The cofactor Cij of the element aq in the determinant
IaijIisgivn byCij=(_l)"ulMjI.
Proof : Let us first of all prove the case C1 = (_ 1)2 I Mit L i.e.
Cii=I Mil l.
• Thç tems in Cii are composed of elements taken from the
(n—l)X(n— 1) sub-matrix M11 -of A.
The general term of a I I C11 =±a 1 0122 aj33 ... aj,, where i2, 13, .in
are 2,3, ..., n in some order.
This term can also be obtained from the trace of matrix A i.e. the product
Cif the elements of the diagonal of the matrix A i.e. a t I a22 033 .. . ann, by

operating on its row subscripts by the permutation p = (1 where

are defined as above.
Thus the permutation p may be regarded as a permutation on the symbols
2 3.... . n. Hence all the terms of a 11 C1 can be obtained by running p through
the (n - I) ! permutations of the symbols 2, 3..... n keeping I fixed.
Thus the terms of Ci i can be obtained by operating on the row subscripts
of the elements of the product a22 033... a, which is the puoduct of the
elements of the diagonal of Mn i.e. the trace of Mil.


821119 Minors and Cofactors 129

Hence ClMiil.
)l4
Now let us prove C,1 = I I M,, I.
Move the ph column of the matrix A to the first column by performing
- 1) successive interchanges of adjacent columns and move the ith row of the
matrix A to the first row by performing (i - I) successive interchanges of
adjacent rows. 'I'hcn the clement a,j is in th first row and first column of the
resulting matrix B. say. The sub-matrix of B obtained by removing the first row
and first column is the sub-matrix M 1 , of the matrix A. Hence aq I Mq I is the
term of I B I containing a:j.
Also we know that if two rows or two columns of determinant I A I are
interchanged, the new determinant - I A I.
We have IBl=(—l)'''IAI (Note)
1)-2 1 A
= ( I)' J (-
=(- l)' 1 lAl. . (- lY 2= I
or lAI=(—l)'IBI (Note)
Equating the coefficients of u, from both sides, we have
Cj1=(—l)"1lM,jl.
' 4 09. Theorem. If C,1 is the cofactor of a, in the determinant
A I = I aq I of order n, then (i) the sum of the products of the elements of the
ith row with the cofactors of the corresponding elements of the kth row is zero
provided i * k.
(ii) Also the sum of the products of the elements of the jth column with
the cofactors of the corresponding elements of the kth column is zero provided
,j # k
?1

(i) E a vCè .= O,ifi k.

1=I
11

and (ii) Z a11 Ct=O,ifj k.

Proof: (i) The given determinant I A I E

A Oki C.

Now eplace the kth row by ith row, then we have the new determinant
'1
= L avC.
j=1
But the Ath and éth rows of the new determinant are identical, hence its
value is zero.

aij
130 Matrices
'I
(ii) The given determinant I A I = T. aik C&.
1=1
Now replace the kth column by jth column, then we have the new
'I

determinant = L aikCj.
But the Ath and jth columns of the new determinant are identical, hence
its value is zero.
R

al) C1k0.
1=1
at bici prove that
aib2c2
a3b3c3

ajAá+bi+c,C2=O, bjCj+b2C2+b3C3=O and cB+c2B2-e.c3B3=O,

wbeie r2i letter dentoe the cofactors of the corresponding smafl

AI,o prove that

aAi + biBi + ciC = t A I = 22A2 + b2B2 + c2C2 23A3 + b3B3 + c3C3.
Sol. In thedet. IA I = at bi ci we have
- a2 b2 C2
a3b3c3.

At = b C2 ;A2— b1 c ;Bi— a C2 ;B2= a c

l3 C3 b3 03 a3 C3 Q3 C)

B=— acj ;Ci= ab ;C2=— a bi I A3= bj ci

a2 C2 al b3 a3 b3 b2 C2

C3 = al bi
a2 b2

alA2+bIB2+c,C2 -
t
—a b1c +b1 at Cl --et Jai& bj
1b3 C)
I a3 C3 1
C3
I
= -al (b1c3-!7c1)+b1 (alc3-a3cl)--cI (aIb3-a3bt).
= O on simplifying.
b1C1 +b2C2+b3C3=bt a2 b2 -b2laibt +b3 at bt
a3 b31 1a3 b3 a2 b2

=0, on simplifying.
In a similar way the remaining pt can also be proved.


Minors and Cofactors 131

Also IAI al bt c
02 b2 Cl
03 b3 C3
= ai fri - bi 02 C2 + C 02 b2
173 C31 03 C3 03 b
expanding with respect to first row.
Again aiAi +biBi + c1Ci
02
C2j+C1 02
b3 C3 t23 C3J 03 t3I
= at b2 c2- bj 02 C2 + cl 1 02 172
b3c3 03C3 a3 b3

=tAIfron(i)
Similarly a2A2 +17282+ C2C2
22J_ bi ci l + b2 01 a3 Cl +C2S_ l ai bi
1173c3J a3 b3
I
=-02 bj Cl +ba cj -Cl jai bi
b3 C3 iO3 03173
Also IAI 01 hi C
02 1? ,. Cl
03 173 C3
02 b1 Cl +172 01 Cl 2 al hi *
173 C31 03 C3 03 173 ...(iii)
expanding with respect to second row.
From (ii) and (iii), we get a2A2 +17282+ c2C2 = I A I.
In a similar way we can prove that a3A3 +17383+ c3C3 = I A I, by
expanding I A I with respect to third row.
4. 10. An Important Property Of theDetermkant.
(a) if A 1 . the jilt mw of a determinant I At I ay I of order n, be replaced
by A + cAj. where c is a scalar and Ak denotes the kh row of the deermirant
I A I, then the value of the determinant remains unaltered.
(Gorakhpur 94)

Proof : The determinant I A I = L ay C4.

j1
Replacing Ai by AI+cAk we got the new determinant IBI 'say.

Thcr.101 £ (aij+ca&j)Ci
j=I
4
£ a j C4j +c £ aijC,j
J.I i-I


132. Matrices 989

=1 A I+c.O ... See § 409 Pages I2.131
i.e. IBI=Iki.
(b) If C. the ith column of dererrninw:t I A I = I a il I of order n be replaced
by Cr + XC'k where X is a scalar and Ck deuoes the kth column
of I A I then
the value c/ the determinant remains unaltered
Proof is similar to part (a) above.
Solved Examples on the Evaluation of Determinants:
In the following examples R 1 , R2, R .. stand for first, second, third,
rows and C1 , C2 , C3 , ... stand for first, second, third.., columns.
Ex. I. V " ithout expanding the determinants, prove that
a b c bq = x y z
xyz x a p p q r
p q r z C
a tic
Sol. a b c = - x y , interchanging R 1 and R2

1p
X y z . a b c
q r p q r

x y z interchanging R2 and Rj
p q r
a b c Hence proved.
I
Again a b c = a x pinterchanging rows and colurnnc
X y Z b v q
p q r C z

a p , interchanging C 1 and C2
Y b q
z c
y b q , interchanging R 1 and R2
x a p
Z C r Hence proved.
Ex. 2 (a). Evaluate" b + c
Ibc+a
ca+b
Sol. The given determinant
= I a a + b + c replacing C3 by C1+C2
I b a+b+c
P o+b+c
(a ' b t- c) I 1 o I , liking out (a + b + c) common from
C3
' 1 b I
C
IJ
since two columns are identical, Ans.
Ex. 2. b). Evaluate 13 16 19
14 17 20
15 18 211
 S. Examples on Evaluation of Determinants 133

Sol. The given determinant

= 13 16 3 .replacing C3 by C3 - ('2
14 17 3
15 18 3
= 13 3 3 replacing C2 by C2 - Ci
14 3 3
15 3 3
= 0, since two columns are idcnticaF. Ans.
Ex. 3. Evaluate 265 240 219
240 225 198
219 198 181 (Kanpur 90)
Sol. Given determinant
25 21 219 .replacing Ct, C2 by CI - C2,C2—Clrespectively
15 27 198
21 17 181
4 21 9 , replacing Ci, C3 by Ci - C2,
- 12 27 —72 C3 - 10C2 respectively
4 17 11
= 0 4 - 2 . replacing RI, R2 by RI - R3, R2 + 3 R3 respectively
o 78 —39
4 17 II
= 4 4 -2 . expanding with respect Co C1
178 —39
=4 0 —2 .replacing Ci by C1+2C2.
0 —39
=4x0=0 Ans.
Ex. 4. Evaluate a -a -a -a
b —b —b —b
c —c —c —c
d —d —d —d
Sot. Since three columns of the given determinant are identical, so the
value of the determinant is zero. Ans.
Ex. 5. (a) Evaluate 1 2 3 4
2341
3412.
4123
&L-The given determinant
= 10 2 3 4 , replacing ClbyC)+C2+C3+C4
10 3 4 1
10 4 1 2
10 1 2 3


134 Matrices
=10 1 2 3 4 ,taking out l0 common fromCi
1341
1412
1 t23
= 10 1 2 3 4 , replacing Rz. R3, R4 by R2 - Ri,
0 1 1 -3 R3—RIandR4—R1
0 2 —2 —2 respectively
0 —1 —1 —1
=10 1 1 -3 expanding with respect to Ct
2 —2 —2
—1 —1 —1
=-20 1 1 —3 ,taking out 2 common from R2 and —1 from R3

1 11
=-20 1 • 1 —3 ,replacing R2 and R3byR2—RI and
0 - 2 2 R3 - Ri respectively
0 0 4
--20- 2 2 , expanding with respect to Ct
04
= - 20 [(-- 2) 4 - 0•2 ] = - 20 [- 8] =160. Aiis.
Ex. 5(b) Evaluate —4 1 1 1 1
1 —4 1 1 1
1 1 —4 1 1
1 1 1 —4 1
1 1 1 1 —4
Sol. The given determinant
= 0 1 1 1 1 ,replacing ClbyCl+C2+C3+C4+C5
0 —4 1 1 1
0 1-4 l 1.
0 1 1 —4 1
0 1 1 1 —4 (Note)
=0, expanding with respect to elements of first column. Axis.
*Ex. 6. Evaluate cosO — sinO 0
sin cos9 0
0 0 1 (Meerut 9o)
Sol. The given detrLninant
= - cos 0 0 - sin 0 , interchanging C2 and C3
sin 0 cosO
0 1 0


Evaluation of Determinants 135

= 0 cos 0 - sin 0 interchanging Ci and C2

o sin cosO
1 0 0
= cos 8 - sin 9 expanding with respect to Ci
sin cosO
Ans.
= cos 8. cos 9— (— sin 0) sin o = cos 2 8 + in 0 =
Ex. 7.Evakiate 1 1 1
1 1±x I
1 1 l+y
Sol. The given determinant
I I I ,replacing R2byR2_R1 and R3bYR3R1
OxO
OQy
= I X x 0 expanding with respect to the first column,
0
= xy_
s*Ex. 8. Evaluate 1 2 22 32 42
22 32 42 2
32 42 52 62
42 2 62 72 (Meerut 97; Punanchal 97)

W. The given determinant

= 1 4 9 16
4 9 16 25
9 16 25 36
16 25 36 49
= 1 0 0 0 , replacing C2. C3, C4 by C2 — 4Ci,
4 —7 —20 —39 C 3 _9C 1 and C4_16Cirespectively.
9 —20 —56 —108
6 —39 —108 —207

= (— 1) 7 20 39 1, expanding with respect to first row

20 56 108
39 108 207
7 —1 —1 , replacing c2 and C3byC2-3dIand
20 —4 - 4 C — 2C2 repsectively.
39 —9 —9
1
= 0, as two columns are identical Ans.
Ex. 9. Evaluate 5 7 10 14
2376
3369
5 6 11 20


136 Matrices
Sol. The given determinant
0 1 - 1 —6 , replacing R and R3 by R - R4 and
2 3 7 6 R - R2 rcpscctivcly.
l 0 —I 3
5 6 11 20
= 0 1 0 Q replicing C3 and ('4 by C2 + C3 and
2 3 - 10 24 ('4 + 6C2 respectively.
1 0 —1 3
5 6 17 56

Now expand with respect to 1st row and proceed as in Ex. 8 Page 135
Ans. 96
Ex. 10. Show that 1 1 1 1 = 0
a 13 y 8
13+ y y+6 6+a a+13
13
Sol. The given determinant
= I I I
a S
13 y
'y+8 5+a
a+13+y+8 a+13+y+8 ct+13+y+6 a+13+y+6
replacing R4 by R2 + R3 + R4
=(a+J3+y+&) I I I I
a 13 Y 8
13+' y+8 6+a cz+13
I I I I
Liking out (a + 13 + y+ 5) common from R4.
0, since two rows are identical. Hence proved.
Ex. 11.Evaluate 1 1 1 1 1
12345
1 3 6 10 15
1 4 10 20 35
1 5 15 35 69
SoL The given determinant
= 1 0 0 0 0 , replacing C2 ,C3 ,C4 and c5byc2_c1,
1 1 2 3 4 C3 — CI.C4 — ' 1 and C5—CJ respectively.
I 2 5 9 14
1 3 9 19 34
1 4 14 34 68
= 1 2 3 4 , expanding wih respect to first row.
2 5 9 14
3 9 19 34
4 14 34 68


Evaluation of Determinants 137

= 1 0 0 , replacirtgC2,C3 and C 4 hYC22Cl C3-3C

0
2 1 3 6 and C4 - 4C1 repsectively.
3 3 10 22
4 6 22 52
1 3 6 expanding with respect to first row
3 10 22
6 22 52
= 1 00 replacing C2 and C3 by C2 - 3Ci
3 1 4 and ('3 - 6C1 respectively.
6 4 16
= 1 4 1, expandin with respect to first row.
4 16
=1.(16)-4.4=0. Ans.

Ex. 12. Show that a + b + 2c a b = 2 (a + b + c)3

C b+c+2a b
C a c+a+2b
(GarhwQl 90; Meerut 92)
Sol The given determinant
= 2i + 2b + 2c a b . replacing Ci by Ct + C2 * C3
2a+2b+2c b+c+2a b
2a+2b+2c a c+a+2b

(2a+2b+2e) I a b .taking out 2+2b+2c

I b+c+2a b common.
1 a c+a+2b

= 2 (a + b + c) 1 a b replacing R2 and R3
o b + c + a 0 byR2—RI and R3—R1
0 0 c+a+b respectively

=2(a+b+ . c) b+c+a 0 ,expanding with respect to

0 c+a+b lstcolumn.

=2(a+b+c)[(b+c+a)(c+a+b)12(a+b+c)3
*Ex- 13. Prove that b±c c + a a + b 2 a b c
q + r r + p p + q p q r
Y +z z + x x + y x y z (Meerut 90)

Sol. [,.H.S.= b+c '+a a + b

q + r r+p p+q
y+z z+x x+y
b c+a a+b + c c+a a+b
q r+p p+q r r+p p+q
Y z+x x+y z z+x x+y (Note)
=1


138 Matrices
= b c+a a + b c+a b 4-cc a+b + C aa1-b
q r+p p q r'-p q r r p+q r p p+q
Y z+x x y z+x y z z x+y z x x+y
= b c a + b a a a a + +1c cab,
q r p q p p r p p r p q
Y z x y X z X z x y
second and third determinants vanish as two columns in each are identical.
= b c a + c a b , second and third determinants vanish
q r p r p q as two columns in each are identical.
Y z x z x y
= - b a c - a c b , interchanging C2 and C3 in first and
q p r p r q Ci and 02 in second determinant
Y x z xzy
= a b c + a b c , interchanging Ci and C2 in first and
p q r p q r C2 and C3 in second determinant
X y z x y z
=2 a b c =RH.S.
p q r
xyz Hence proved.
Ex. 14. Evaluate y + z x y
z+x z x
x+y y z
Hint: Add all the rows to first, take out (x+y+z) common from first
row and then subtract sum of second and third columns from first. Then
expand. Ans. (x+y+z)(x—z)2
I
Ex. 15. Evaluate b + c a + b a
c+a b-+-c b
a+b c+a C
Sol The given determinant
= 2a+2b+2c 2a+2b+2c a+b+c , replacing Riby
c + a b+c b Ri+R2+R
a±b c+a
=(a+b+c) 2 2 1 ,taking out (a+bi-c)common
c+a b+ b
a+b c+a c
=(a-i-b+c) 0 2 1 ,replacing ClhyCl—C2
a — b b+c b
b — c c+a c
= (a ± b + c) 0 0 1replacing C2 by C2 - 2C3
a—b c—h b
b—b a—c c


Evaluation of Determinants 139

= (a+b+c) [(a—b) (a — c) — (b - -c) (c —b)]

=(a+b+c)[a2_aC_ba+bc+b2+C2_21d]
-(a ±b+c)(a 2 +b2 +c 2 _abbC _ ca) =a 3 +b3 +c 3_ 3
_ Ans.
*Ex. 16. a—b—c 2a Zn
Zb b — c — a 2b
2c 2c c—a—b
(Kanpur 96; Kurnaun 93, Meerut 95)
Sol. The given determinant
= a+b+c b+c+a a+b+c ,replacing RlbyRl+R2+R3
2b b — c — a 2.13
2c 2c c — a --b

=(a+b+c) I I I ,taking out (a+b+c)

2b b—c—a 2b common
2c 2c c—a—b

=(a+b+c) 1 0 0 ,replacing C2 and C3

2b —a—b—c 0 byC2—Cland
2c 0 - a - b - c C3 - Ci respectively.
=(a±b+c) a
—— b — c 0 , expanding with respect
0 —a—b to first row.
(— a — b — c) (—a—b—c)=(a+bC)3. Ans.
E. 17. Evaluate a 1 1 1
lal I
hi
111 a
Sol. The given determinant
= a+3 1 1 1 , replacing ClbyCt+C2+C3+C4
a+3 a I I
a±3 I .a I
a+3 1 1 a

Now proceed as in Ex. 5 (a) Page 133. Arts. (a - 1) 3(a + 3)

Ex. 1K Evaluate I 1 + a 1 1 1
I 1+a 1 1
1 1 1+a 1
1 1 1 1+a
Sol. The given determinant
= 4+a I I I ,replacing C1yC1+Cz+C3+C4
4+a 1-s-a 1 I
4 I 1+a I
4+-a 1 1 1±a


140 Matrices
= (4 + a) I ' 1 1 1 taking out (4 + a) common
I I + a 1 1 from first column.
I I 1-a I
I I I 1+a

= (4 + a) I 0 0 0 replacing C2, C3 and C4 by C2—CI,

I a 0 0 C3—ClandC4—CI
1 GaO
lOOa
= (4+a) a 0 0expanding with respect to R1
GaO
OOa
= (4 ± a) a a 0expanding with respect to RI
0

=(4+a)a(a2)=(4+a)a3, Ans.
Ex. 19. Prove that
I+ai a2 a =1-fal+a2+a3+a4
at 1+a2 a3 34
at a2 1+a3 34
al a2 a 1+a4
So!. The given determinant
= I +aj +a2+a3+a4 02 03 a4 replacing C1 by
l+aI+02+03+a4 1+02 03 04 CI+C2+C3+C4
l+al+a2+a3+a4 02 1+03 04
l+aI+a+a3+a4 a2 03 1+04
Nuw proceed further as in Ex. 18 above.
*Ex 20. Evaluate x a a a
ax a a
a ax a
a a ax
So!. The given determinant
= x+3a a a a , replacing C1byCI+C+C3+C4
x3a x a a
x+3a a . a
x+3a a x
= (x + 3a) I a a a , taking out (x + 3a) common fromCi
I x a a
lax a
I a ax


Evaluation of I)creniijnants 141

= (x + 3a) I a a a replacing R. R i and R 4 by
o x—a 0 0 R2—Rj,R3—RI and R4—k1
() 0 .V d 0 respectively.
o o 0
= (x + 3u) x - a 0 0 Iexpanding with respect to Ci
I
o X - O (J
o o I - a
+ 3a) (r - a) .r - a 0 expanding with respect to Ct
o

( + I(,) (t - a) (x - a) (x (z) = (x - 3(j) (x - a) 3 . Ans.

Ex. 21. Evaluate x + a 1 c d
• a x*b c d
a h x+- c d
a b c x + d (Meerut 96)
Sol The given determinant
A + a + h + c + d b c t replacing Ct by
x+a+b+c-s-d x+h C d CI+C2+C3+C4
x+a4-b+c+d b x+c d
x+a+h+c-d b c
=(x+a+h+c-i-1) I b c d taking out
I x + b c d (x -u+b++J)
1 b x + c d common tromC1
I b c x+d
b c d , replacing k2,R3,k4hyR2—Rt.
o x 0 0 R3 - Pt and R4 - Rt respectively.
OOxO
000x
=(+a+b-4-c+d) x 0 0 expanding with respect coCj
OxO
o 0
x(x+(i + b + c 4-d) X 0 .expanding with rcspe oC1
0

Ans.
EN. 22. Evaluate b c a a
b c + a b
C C a F b (Garhwal 90)
Sol. The given 1ctrmii1ant
= h - c 0 a . replacing C2 by C2 -
b c+a — b b V

C —a—h a+b


142 Matrices

=(b+c) c+a—b b +a Ib c+a—b

c—a—b o+b c c—a—b
expanding with respect to R1
=(b+c) c+a—b b I +a I b c+a—b
—2a a c—b —2a
replacing R2 by R2 - i in each determinant-
= (b + c) (a(c+a_b)—b(-2a)J+a[b(-a)—(C—b)(C+a—b)
=(b +c) (ac 4a2+ab)+a(_2ab_c2_Ca+bC+bC+ab_b2)
=abc +a2b+ab2 +ac2 +a2c+ac_2a2 b_ ac' _Ca2+2abC+02bb2
=4abc. Ans.
Ex. 23.Evalnate b + c a—c a—b
b—c c+a b—a
c—b c—a a+b
Sal. The given determinant -
= 21' 2a 0 . replacing Rt and R3 by Ri + R2 and
I
b - c c + a b - a R3 + R2 respectively.
o 2c 21
4 b a 0taking out 2 common from R and R3
b—c c+a b—a
o c b

4 b a 0 ,replacingR2byR2—R1
—c c b—a
o c b
=4[b c b - a + c a 0 1 1 expanding with respect Ic Ci
C b c

=4 [ b c b—a I +c I a OJ,
replacing R2byR2—RI
0 a c bj 1n first determinant.

=4[b(ca)+c(ab)]=4(2abC)=8abC. Ans
Ex. 24.Evaluate x a b c I
dxi h 1
dexkl
deg x
dcgml

Sol. The given detenninart

x—d a—i b—f c—h 0
0 x—e f-x h—k 0
0 0 x—g k—x 0
- - 0 0 0 x—nz 0
d e 9 nil
 Evaluation of Determinants 143
replacing Ri, R2,R3 and R4 by RI-R2, R2-R3, R3-R4 and R4-R5
rcsplactively.
x-d a-x b-f 0 'exapanding with respect toR4.
o .-e f-x 0
o 0 x-g 0
d e g I (Note)
=(x-m) x-d a- x b-f .expanding with respect toC4
o .z-e f-x
o
0 ,-g
= ( - m) (x - d) x - e f- x , expanding with respect to Ci
0 A-
=(x-nt)(x-d)(x-e)(x-g). Ai
Ei,25.EvaIuak I . I y
xIyI
lyix
Y I x-1
Sol. The given determinant
= x+y+2 x+y+2 x+y+2 x+y+2 ,replacingRiby
X 1 y 1 R1+R2+R3+R4
1 y I x
Y I I
= (x + y -i- 2) 1 1 1 1 taking out (x + y + 2) common from Ri
xlyl
lylx
Y I x I
= (x + y + 2) 1 0 0 0 replacing C2, C3 and C4 by
x 1-x y-x l-x C2-C1,C3--Clafld
I y-1 0 X -i C4-CI
I
Y I - y . t y 1 y respectively.
(x+y+2) 1 -x y -x 1 -x ,expanding with respect to RI
y - 1 0 x-l
I
'-Y X-Y 1
0 y-x I -x •replacing CjbyC-C3
y -x 0 x-1
0 x - y t_y
= - (x + y + 2) (v - 'X) - x I - x , expasdmg with respect to C1
t_Y I4
=-(x+v+2)(y--x)(y-x) I 1-x
-1 I - y -
taking out (y - x) common from Ci


144 Matrices 182/1/9

(x +),+ 2) x)2 ((1 _) -(-1) (I )]

= (x + y + 2) (x - ),) 2 (x + y - 2). Ans.
Ex. 26. Evaluate a x y a
xOOy
yOOx
a y x a
So!. Then given determinant
= 0 x +y a replacing C i and C: by Cj - C4 and
X- y 0 0 y " C2 + C4 respectively.
y- 0 0 x
0 )+X x a
r(x..y)(x+y) C) I v a ,takingoutx - vandx+y
I 0 0 y common from Ci and C2
- I 0 0 x respectic1y.
0 I .r a
( 2-y 0 1 y a , resplacing RI and R4 by R2 + Ri
2)
1 0 0 y and R4 - rspecvely.
ti
o 0 0 x+y
o o x-y 0

(x2
_ ).2) I
y a expanding with respect to Ci
o o
o x-y 0
=-(x2_ y 2) C X+) , , expanding with respect to Ci
x - y 0
=-(x2 - y 2 )[-(x+y)(x-y)]=(x2 -y 22 Ans
EL 27 (a) Evaluate 1 bc a (b + c)
1 ca b(c+a)
1 ab c(a+b)
Sol. The given determinant
bc a(b+c) , replacing R2 and R3byR2-Ri
o c(a-b) c(b-a) and R3-RI respectively.
o b(a-c) b(c - a)
c (a - b) c (b - a) ,expanding with respect to Ci
b(a-(-) b(c-a)

= (a - b) (a -c) c - c ,taking out(a - b) and (a - c)

b -b common from Ri and R2 respectively
= - (a - b) (a - c) c c = 0, two columns being identical.
bb


182/1/10 Evaluation of Determinants 145

Ex. 27 (b) Evaluate I/a a 2 bc

I/b b2 Ca
1/c c2 ab (Kanpur 92)
Sol. The given determinant
_...±.. 1 a 3 abc ,multiplying RI, R2 and R3bya,b
abc 3
b bca and c respectively.
I c3 cab (Note)

= ---- abc 1 a 3 I , taking out abc common from C3

abc

I C i

= 0, since Ci and C3 are identical. Aits.

Ex. 28. Evaluate1 1 1
abc
a2 b2 c 2 (Meerut 96P)
Sot. The given determinant
= 1 0 0 , replacing C2 and C3 by C2 - Ct
a b - a c - a and C3 - Ci respectively.
2 2 2 2 2
a b-a c -a

= b - a c- a expanding with respect to RI

b 2 -a 2 c.2-a2
= (b - a) (c - a) 1 1 . taking common factors out
b+a c+a
=(b-a) (c - a) [(c+a)-(b+a)]--(a-b) (c - a) (b-c). Ans.
*Ex. 29. Show that 1 a be =(b-c)(c---a)(a-b)
1 b ca
1 c ab (Meerut 96P; Purvanchal 96)
Sol. The given determinant
I a bc replacing R2 and R3 by R2 - R
o b - a ca - bc and R3 - R1 respectively.
o c-a ab-bc

= b - a - c(b - a) expanding with respect to Ct

c-a -b(c-a)

=(b-a)(c-a) I -c takingcornmon factors out

1 -b

= (b - a) (c - a) (- b + c) = (a - b) (b - c) (c - a). Hence proved.

Ex. 30. Show that ab C =abc(a-b)(b-c)(c-a)
2
b2 c 2
3 :3
a b c3
146 Matrices
So!. The given determinant
—abc 1 1 1 ,taking out a, b and ccommon from Ci, C2
a b c and C3 respectively.
2 2
abc2
Now proceed as in Ex. 28 Page 145.
Ex. 31. Evaluate 1 11
a b c
a 3 b 3 c3
So!. The given determinant
= 1(3 0 replacing C2, C3 by C2 - C1
a b - a c - a and C3 - Ci respectively.
a 3 b3 .—a3 c3—a3
= b—a c—a ,expanding with respect to Ri
3 3 3
h—a c—a3
=(b—a)(c—a) I
b2 +ab+a 2 c2+ac+a2
taking common factors out from C 1 and C2
=(b a) (c — a) [(c2+ac+a2)—(b2+ab+a2)j
=(b—a)(c—a)[(c2—b2)+a(c—b)J
=(b—a)(c—a)(c—b)[(c+b)+a]
=(a—b)(b—c)(c—a)(a+b+c), . Ans.
*EL 32. If a, b, care all different and
a a2 1±a3 =0, prove that abc=-1.
- b b 2 1+b3
C c2 I + c3 (Meerut 91 S)
So!. The given determinant -
= a a2 1 + a a2 a3 breaking into two determinants.
b b2 1 b b2 b3
C c2 i c c2 c3 (Note)
= a b c + a b C
interchanging rows and columns
a2 b2 c2 a2 b2 c2 in each determinant,
1 1 1 a3 b 3 c3 (Note)
Now proceed as in Ex. 28 Page 145 and E. 30 above.
The value of given determinant = (1 + abc) [(a - b) (b - c) (c - a)]
Ex. 33 (a). Evaluate 1 1 1
a 2 b 2 c2
a3 b3 c3
 Evaluation of Determinants 147

Sol. The given determinant

= 1 0 0 replacing C2 and C3 by C2 - CI
a2 b 2 - a2 (2 - a 2 and C3 - Ci respectively.
3 3 3 3 3
b-a c-a
' 2 2 '
= b - a c - a . expanding with respect to Ri
3 3 3 3
b-a c -a
=(h-a)(c-a) b + a c+a
2 1
b +ab+a c+ac+a
taking out common factors of Ci, C2
=(b-a)(c-a) b - c c+a
b2 +ab-c 2 -ca c2+ca+a2
replacing Ci by CI - C2
=(b-a)(c-a) b - c c+a
(b2-c2)+a(b-c) c2+ca±a2
=(b-a)(c-a)(b-C) 1 c+a
b+c+a c2+ca+a2

1 (c2 +ca+a2 ) -(c+a) (b+c+a)1

=-(a - b) (b-c) (c-
=-(a-b)(b-c)(c- 2+ca+a2cbc2_ac__ac_a2]
ab - bc - cal
Ans.
Ex. 33. (b) a a 3 a4 - 1 =0, then prove that
b b 3 b4-1
C c3 c4-1
abc (ab + bc + Ca) =a+ b± c.
Sol. The given equation is
a a 3 a4 - a a 3 1 =0
b b 3 b4 b b3 I

C c 3 c 4 c c 3 I

2 3
or abc I a a 3 - I a a =0
1 b2 h3 1 h b3
2 3 3
1 c C icc

or abc [(a - b) (b - c) (c - a) (ab + bc + ca)]

[(a-b) (b - c) (c - a) (a+b+c)] =0
interchairig rows and columns of the determinant and evaluating
determinants as in Ex. 31 Page 146 and Ex. 33 (a) above
or (a - b) (b - c) (c - a) [abc (ab+bc+ca) -(a+b+ c)] =0

or abc(ab+bc+ca)-(a+b+c)O. a#b#c
or abc (ab + bc 4- ca) = a + h + c. Hence preyed.


148 Matrices 989

Ex.' 34. Prove that
1 a 2 +bc a 3 —(b—c)(c—a)(a--b)(a2+b2+c2)
I b2 +ca b3
1 c2 +ab c3
Sol. The given determinant
= I a2 a3
b 2 b 3
+ Ii bc a 3 breaking into two determinants (Note)
1 ca b3
1 c2 1 ab c3
1 1 1 + i ' I interchanging rows
2 b (Note)
2 c2 bc Ca ab and columns.
a 3 b3 c3 a3 b3 c3
.
The first determinant = (a - b) (b - c) (c - a) (ab + bc + ca). ...(ii)
[See Ex. 33 (a) Page 1461
The second determinant
0 0 replacing C2 and C3 by C2 - Ci
=1 bc ca - bc ab - bc and C3 - C1
a3 6 3 —a3 c3—a3
= c(a — b) b(a — c)
(b—a)(b2 +ab+a 2) ( c—a) (c2+acfa2)

expanding with respect to R1

(a--b)(a—c) C b
—(b2 +ab+a2) _(c2+ac+a2)

taking out common factors from C1 and C2

=(a—b)(a--c) (c — b) b
—(b2 —c2 +ab — ac) _(C2+ac+a2)

replacing Ci by C - C2
(c—b) b
(c—b)(c+b+a) —(c2+ca-f-a2)
(a—b)(a—c)(C—h) I b
a+b+c _(c2+ca+a2)

taking (c - b) common from C.

= (a - 14(b— c) —a) {- (c2 + ca + a 2) - b (a + b + c)]
_(a_b) (bc) (ca) (a2+b2+C2+ab+bc+a)
.Substituting values from (ii) and (iii) in (i), we find the given determinant
=(a—b)(b—c)(c_a)(ab+bc+ca)

(a_b) (b — c) (c — a) (a 2 +b2+c2+ab+bC+ca)


Evaluation of Determinants 149

=(a_b)(b_0(c_a)[(th+bC+Ca)_(a2+b2

= - (a - b) (b - c) (C - a) (a 2 +b 2 + c 2 ). Hence proved.
Ex. 35. Show that
Y z = 1 1 1 =(y—z)(z—x)(X—Y)(YZ+ZX+XY).
2 2 2 2 2 2
X y z Y Z
(
yz zx Xy X3 y 3

(Gorakhpur. 93; Kuniaun 96; Meerut 94)

Sot.
X y z = _ x2 y 2 z 2 L multiplying Ci, C2, C3
X2 y 2 z2 XYZ x
3y
3 z3 by x, y, z respectively.
YZ ZK xy xyz rz xyz (Note)
taking out .zyz common from R3
= xyz x2 y2 z2
.yz 3
y3
1 1 I
1 1 1 interchanging R2 and R3 and then R and R2
2 2
3 3 3
Xy z
For the 2nd part do as Ex. 33 (a) Pages 146-47. (Note)
Ex. 36.Evaluate 1 1 1 1
1 1+x 1 1
1 1 l+y 1
1 1 1 1+z

Sot. The given determinant

1 1 1 1 replacing R2, R3 and R4 by R2 - Ri, R3 - Ri
o x 0 0 and R4 - R respectively
OOyO
000z
= x 0 0 , expanding with respect to Ci
o y 0
OOz
x y 0 , expanding with respect to Ri
Oz
= xyz. Ans.
Ex. 37. Evaluate a 3 3a2 3a 1
a2 a2 +2a 2a-+-1 I
a 2a+1 +2 1
1 3 3 1


150 Matrices
Sol. The given determinant
a 3 -3a 2 +3a-1 3a 2 3a 1 replacing Ci by ,
o a2+24 2a+1 1 Cj—C2+C3—C4
o 2a+1 a+2
o 3 3 l (Note)
a
= ( a 3_ 3a 2 +3a_1)a 2 +2 2a+l1Iexpandingwith
2a + 1 a + 2 1 respect to Ci
3 3 II
= (a 1)3 (a2 + 2a) - (2a + 1) + I 2a + 1 1 replacing Ci by
(2a+I)—(a+2)+I a+2 I C1—C2+C3
3-3+1 3 1
= (a 1)3 a 2 2a + 1 1 . on simplying
a a+2 I
1 31

(a-1)3 a 2 -1 2a-2 0 ,replacingR andR2byRi—R3 i

a - I a - 1 0 and R2 - R3 respectively.
1 3 1

- 1) a2 — 1 2(a— 1) ,expanding with respect to C3

a—i a—i
- 1) 3(a - 1) (a - I) a + 1 2taking out common factors.
1 1

Ans,
Ex.38. Evaluate 1 +a 1 1 1
1 1+b 1 1
1 1 1+c 1
1 1 '1 1+d
(Gorakhpur 90; Kanpur 97; Meerut 91; Pu)vanchal 93)
SoL The given determinant
abcd I + I J taking out a, b, c and d
a a a a common from RI R2, R3

b
I + i and R4 respectively.
b b b
- I I I
C-
- — I+1 -
C C C
1 1 1 1
d d d


151
Evaluation of Determinants
1 1
"i 1 1 i 1
=abcdI—+ b c d 1 i 1 1
-+1 -
) -
11 1
-- —+1-
C C C C

11 1

Teplacing. R I by R1 + R + R3 + R4 antaking
1 + COofl

from RI

( i i i
=ab cdI+i+b c d
1 1 0 0 0
)
100

001

1 and C4 - Ci respectively.
replacing C2. C3, and C4 by C2 - Ci. C3 - C
1 0 0 , expanding with respect t0
(i 1+1 I 0 1 0
001
1 0 , expanding with respect t0
=abcd(+i+i+_+
b c d 01

(i
=abcd+i +'J i
)
Ans

-(b-
Ex. 39. Prove that a b b b =-(b a)4
- a)4
a b a a
a a b a
bbba

Sol. The given determinant

, and C4bYC2C1
= a b—a b—a b—a replacing C2,C3 Ci respectively.
a b - a 0 0 C3 - Ci and C4 -
a 0 b—a 0
b 0 0 a—b
=—(b—a) a b—a 0 +(a—b) a b—a b—a
a 0 b — a ab—a 0
b 0 0 a 0 b—a
expanding with respect to Ci


152 Matrices
— (b—a)b b—a 0 +(a—b) 0 0 b — a
0 b—a ab—a 0
a 0 b—a
expanding first determinant with respect to R3 and in the second determinant
replacing R by RI —R2.
--b(b—a)3+(a—b)(b—a) a b—a
0 la
expanding the second determinant with respect to R1

=b(b—a)3+a(b_a)_(b_a)3(b...a)
- (b - a)4
Hence proved
**Ex40Showtbat (b+c)2 a2 =2abc(a+b+c)3
12
b2 (c-fa) 2 b2
c2 c2 (a+b)2
(Kurnaun 94; Meerut 93; Purvanchal 95)
SoL The given determinant
= (b + c)2 - a2 0 a2 replacing Ci and C2 by
0 (c+a)2—b2 b2 Ci—C3&C2—C3
?_(a+b)2 c2—(a+b)2 (a+b)2 respectively.

= (b + c + a) (b + c — a) 0 a2
0 (c+a+b)(c+a—b)
b2
(c+a+b)(c—a—b) (c+a+b)(c—a—b) (a+b)2

—=(a+b+c) 2 b+c—a 0 a2 takingoutthe

o c + a - b b2 common factors from
c—a—b c—a—b (a+b)2 Cl and C2,
—(a+b+c)2 b+c—a 0 a2 , replacing R3by
o c+a—b b 2 R3—Ri—R2.
—2b —2a 2ab

= (a + b + c)2 b + c a2/b a2 . replacing Ci and C2 by C1 +

C3
b21a c + a b2
0 0 2ab and C2 + - C3 respectively.
b
(Note)
2ab (a + b + c)2 b + a , expanding with respect to R3
b2/a c+a
= 2ab (a + b + c)2 [(b + c) (c + a) - (b 2/a) (a2/b)J


Evaluation of Determinants 153

=2ab(a+b+c)2Ebc+bo+C2+Ca—abl
+ c)3 .
= Zab (a + b + c) 2 c [b + a + cJ = 2abc (a + b Hence proved.
Ex. 41. Evaluate a b ax+by
b c bx+cy
ax+by bc+cy 0 (Kanpur 90)
Sol, The given determinant
= a b 0 ,replacing C3hyC3—xCl—yC2-
b c 0
a.r+by bx+cy —x(ax+by)
—y(bx+cy) (Note)

- (a? + Thx)' + cy2) a b expanding with respect to C3

b

Ans.
= - (a? + 2bxy + ,2) (ac - b2).
Lx. 42. Prove that
1 0 x 0 x =(x-1)2(x+1)(1+2x—x2)
01 0 x
x 0 x+1 0 x
Ox 0 10
x x 01
So I The given determinat
10 0 0 0 replacing C3 and C5 by C3 - xCi
01 0 x 0 and C - C3 respectively.
X 0 x+1—x2 0 —1

Or 0 1 0
X 0 0 l—x
=1 0 x 0 expanding with respect to R1
0 x+l—x2 0
X 0 1 0
0 .-? 0 1 —x
1 0 0 ,replacing C3byC3—xCI
O x+l —x 2 0 —1
X 0 i-2
0 -? 0 1—x

= x +I -? 0- 1 expanding wi th respect to RI
0 1—? 0
x(l — x) 0 1
154 Matrices
(1 —x) o i 0 . interchanging R 1 and R2 an
2
. + 1 -X 0 - 1 taking out(1 - x) common from R3
0 1
=0 —x)(1 —x 2 ) x+ i —x2 - Iexpanding with respect to RI
X I
2
= (1 —X) (1 + X) [(X + 1 - x 2 ). 1) .x]
= (x 1) 2 (1 +x) (Zx+ 1 —x2). Hence proved.
Ex. 43. Show that 1 1 1
bc(b+c) ca(c+a) ab(a+b)
b2c2 c2a2 a2b2
=abc (a—b)(b--c)(c—a) (a+b+c)
Sol. The given determinant
= I I I
b2c+bc2 c2a+ca2 a2b+ab2
b2c2 c2a2 a2b2
= 1 0 0
b 2c+bc2 c(a—b) (a+b+c) b (a—c) (a+b+ c)
b22 c2(a—b)(a+b) b2(a—c)(a+c)
replacing C2. C3 by C2 — Ci, C3 - Ci
-= c(a—b)(a+b+c) b(a—c)(a+b+c) , expanding with
c2(a—b)(a+b) b2(a—c)(a+c) respect toRi
=c(a—b)b(a—c) a+b+c a+b+c ,takingoutcommon
c(a-t-b) b(a+c) .factoi-sfromC1,C2
=be (a—b)(a—c)(a+b+c) I I ,taking out a+b+c
ca+cb ba+bc common from Ri
= be (a — b) (a — c) (a + b + c) 1 0 , replacing C2 by
ca+cb ba—ca C2—CI
= be (a — b) (a — c) (a-+-b+ C) (ab — ca)
= — abc (a - b) (Li - c) (c - a) (a + Li + c). Hence proved.
*EL 44. Evaluate 1. bc + ad b2c2 + a2d2
1 ca+bd c2a2+b2d2
1 ab+cd a2b2+c2d2
Sol. The given determinant
= I bc+ad bc2+a2d'
0 ëa+bd—bc—ad c2a2+b2d_b2c2_a2d2
0 ab + cd - bc — ad a2b2 + C2d -b 2 C 2 — a2d2


Evaluation of Determinants 155

replacing R2 and R3 by R2 - R I and R3 - R 1 respectively
= ca—bc+bd—ad c 2a2 —b2c2 + h2d2 —a2d2
ah—bc+cd—ad a2b2_b2c2+c2d2_ad2

expanding with respect to Ci

= (c - d) (a - b) (c 2 - d2 ) (a 2 - b2) factorising the elements
(b—d)(a--c) (b2—d2)(a2—c2)

=(c—d)(a—b)(b—d)(a—C) I (c+d)(a+b)
I (b+d)(a+c)
taking out the common factors
=(c—d)(a—b)(b—(aC) 1 ca+bc+d+db
0 ba+dc — ca — db

replacing R2 by R2—Ri
= (c — d) (a — b) (b — d) (a — c) (ba+dc—ca—db)
=(cd) (a — b) (b — d) (a — c) (a — d) (b—c). Ans.
*Ex. 45. Prove that
a2 +X ab ac ad =X3(a2+b2+c2+d2+X)
ba b2 +?. be Ix)
Ca cb c2 +. cd
2
de d + X (Kanpur 90)
da db

So). The given determinant

= aba! a + a a a , taking out a, b, c, d

common from Ci, C2, C3
b b + b b and C4 respectively.
A.
C C C± C

C
d d d d+

= abc a+ ---- -- replacing (22, C3 and C4 by (22 - Ct,

a a a a
C3 - C1 and C4 - C1 respectively.
b 0 0

C 0 0
C
A.
d 0 0 -
d
156 Matrices

zabcd a 0 0 - ,replacing Ci, C2 an C3 by
C 1 +C4,C2-C4 and C-C4
0 respectively.

d±
d

= a2bc
0 0 +Abcd 0

o : !

d+ - -
d d d

expanding with respect to RI

2 I A
= abcd -Oi+Abcd.bI
I - 0 -
bt c I I c
A XI I x A

I
b c
X
d+ -A

expanding each determinant with respect to R1

= A a cd + X b2cd (2-2)_ A2cd1- --

cd) cd d c cd

( 2 2
.3a2 ±A3 b2 +A2c diL+ 11 +X
cd

= A3a2 + A3b 2 + A3 (c2 + d2 + A) = A (a 2 + b2 + c 2 + d2 + )L) Hence proved.

Ex. 46. Evaluate a2 a2_ (b - c)2 be
b2 b 2 -(c-a) 2 ca
c2 c2 -(a-h)2 abi
Sol. The given determinant
= a 2 - ( b - c)2 bc replacing C2 by C2 - Ci
b2 -(c-a) 2 ca
c2 -(a-b)2 ab
 ft
Evaluation of Determinants 157

= - a 2 (b 2 + c 2 ) - 2bc bc
b 2 (c 2 +a 2 )-2ca ca
c2 (a 2 ± b 2 ) — 2ab ab (Note)

= — a 2 (b 2 + c2 ) bc Ireplacing C2 by C2 + 2C3
b2 (c2 + a 2 ) ca
c2 (a 2 + b 2) ab

— a2 b2+ c 2 + a2 bc . replacing C2 by C2 + Ci
b 2 c 2 +a 2 +b 2 ca
c 2 a 2 +b 2 +c 2 ab

= — (a2 + b + c 2 ) a 2 1 bc taking out the common factor from C2

b2lca
c 2 1 at,
2 2 2
= - (a + b + c ) a 2 I bc replacing R2 and R3 by
b2 —a 2 0 ca—bc R2—RI and R3--Rt
c2 — a 2 0 ab—bc respectively

=(a 2 +b 2 +c 2 ) b 2 —a 2 c(a — b) ,expanding with respect toC2

c2 — a 2 b(a—c)

2
= (a — b) (a - c) (2 + L +C - ( b + a) c taking out the
- (c + a) b common factors

= (a — b) (a — c) (a2+b2+c2)[—b(b+a)+c(c+a)J
=(a—b)(a—c)(a2-1-b2+c2)[—b2—ab+c2+ac]
=(a—b)(a—c)(a2+b24-c2)[(c2—b2)+a(c—b)]
=(a—b)(a—c)(a2+b2+c2)(c—b)(a-4-b+c).
2 2
=(ab) (b—c) (c—a) (a+b+c) (a 2 +b +C ). Ans.
**Ex 47. Prove that
0 —c h —I =(aI+bm+cn) (ax +by-fcy)
C 0 —a —m
—b a 0 —n
X y z 0

Sol. The given determinant

1 0 —ac ab —al , taking 1/a common from Ri
a c 0 —a —'n
—b a 0—n
x y z 0 (Note)
'158 Matrices
1 0 0 0 —a!—brn—cn replacing RI by
a c 0 —a —m Ri+bR2+cR3
• —ba 0 -Ti
X z 0

- (al + bm + cn) c 0 -a expanding with respect to R1

- a —b 0
X y z I

(al +bm+cn) ac 0 . — a . taking l/acommon from C1

-
a2 —ab a 0
ax y z (Note)

- (a! + bm + cu) 0 0 - a replacing Cj by

- a2 a 0 C1+bC2+cC3
ax+by+cz y z

(al + bm + cn) (ax + by + cz) 0 —a

- a2 ' a 0

= (a! + bm + cu) (ax + by +cz).. Hence proved.

Exercises on Evaluation of Determinants
Ex. 1. Evaluate 1 2 22 32
22 3 2 42
32 42 52 Ans.-8
Ex. 2. Show that 29 26 22 = 132
25 31 27
65 54 46
Ex. 3-.Evaluate I a b+c
I b c+a
I c a+b
[Hint: Replace C3 by C3 + C21. Ans. 0.
Ex. 4. Show that a - b b - c c - a = 0.
b—c c—a a—b
c—a a—b h—c
[Hint: Replace C 1 by C1 + C2 + C3],
Ex. 5. Evaluate 1 to to2 . where to is one of the
2
imaginary cube
to2 1 to roots of unity. Ans. 0
[Hint: Replace Ci by Ci + Cz + C3 remembering I +co + to2 =01
 Product of Determinants 159

Ex. 6. Prove that —4 1 I 1 1=0

1-4 1 1 1
1 1 —4 1 1
1 1 1 —4 1
1 1 1 . 1
—4
[Hint: Replace Ci by C1 + C2 + C3 + C4 + C51.
Ex. 7. Show that 21 17 7 10 -0
24 22 6 10
6 82 3
5 71 2
Ex. 8. Calculate the value of the determinant 7 13 10 6
5974
8 12 11 7
4 10 6 3 Ans.0

Ex. 9. Evaluate x +a x + 2a x + 3a
x+2a x+3a x+4a
x4-4a x+5a x+6a
[Hint: Replace C2 and C3 by C2 - Ct and C3 - C21, Ans. 0
Ex. 10. Prove that
a+b b+c c+a 2 a b C =2(3abc—a3—b3—C3).
b+c c+a a+b b c a
c+a a+b b+c c a b
[Hint : See Ex. 13 Page 1371.
'Ex.11. Prove that —2a a+b c+a = 4 (a+ b) (b + c) (c + a)
b+a —2b c+b
c+a c+b —2c

Products of Determinants
§ 411. Theorem. If A is an n x n matrix and E is an elementary matrix
obtained from the identity matrx In, then
IEAI=IAEI=IEI.IAI=IAkIEI (Purvancha194)

Proof: I n is the n x n identity matrix.

t1nI1.
Let Ea, Eb, Ec be three elementary matrices as defined in § 303 Page 104.
Then lEa 1-1L11 ...Sce §406 Prop. III Page 126.
From (i) we get I Ea I - I ...(ii)
Again IEbI=clLI ..See § 406 Prop. 11, Page 125.
From (i) we get I Eb I = c ...(iii)
Similarly IEl = IIl ... See § 410 Page 131.
From (i) we get I Ec I = I ..,(iv)


160 Matrices 18211110

Now as the matrix Ea A can be obtained by interchanging two rows of
the matrix A, so we have from (ii)
lEaAl=—IAl=IEj.lAt
Similarly the product EbA can be obtained by applying second
elementary row operation (as given in § 301 Page 103) on the matrix A, so we
have from (iii)
IEbAI=cIM=lEbJ.JAI
Again the product Ec A can be obtained by applying third elementary row
operation (as given in § 3•01 Page 103) on the matrix A, so we have from (iv)
EcAl = lAI=IEHAI ...(vii)
Henëe from (v), (vi) and (vii? we conclude that
EAt = I E I • I A I,
where E is any one of the elementary matrices as defined in § 301 Page 103.
In a similar way we can prove that
IAEI=IAI.IEI. Hence the theorem.
* 412. Canonical Form (or Normal Form)' of a matrix.
Every non-zero in x n matrix A can be reduced by means of elementary
transformations (i.e. elementary row and column operations) to the form
h1r 0],
L°°]
where Ar is the rx r identity matrix and the remaining sub-matrices are zero
matrices.
The above form is called the canonical form or orthogonal form or
normal form of the matrix A.
Solved Examples on * 411
Ex. 1. Reduce A= 1 2 —1 4' to the canonical form.
• 2 4 35
1 2 3 4
—1 —2 6 —7
Sol. A - 1 2 -1 4 , replacing R2, R3 and R 4 by R2 -
0 0 -3 R3 - Ri and R4 + R1 respectively.
5
00 4 0
0 0 5 —3
J
—1 0 0 0 " , replacing C2, C3 and C4 by C2 — 2C I , C3 + C,
0 0 5 -3 and C4 - 4C1 respectively,
0 0 4 0
0 0 5 —3
—1 0 0 , replacing R3 and R4by . R3 and R4_2
0 5 3 respectively.
0 1 0
0 00


18211111 Reduction to Canonical Form 161

- 1 0 0 0 interchanging C2 and C4
o —3 5 0
o 010
0 000
- 1 0 0 0 replacing R2 by R2 - 5R2.
0 —3 0 0
0 0 10
0 0 00
- 1 0 0 0 replacing R2 by - . R.
0100
0010
0000

-1 b 0], which is the required canonical form.

[00]
1
Ex. 2. Reduce A = 13 16 19 to the canonical form.
.14 17 20
15 18 21
1,
SoLA- 13 16 19 replacing R2 and R3byR2—RI
1 1 1 and R3 - R2 respectively.
1 11
- 13 3 3 ,replacing C2 and C3byC2—C
1 0 0 and C3 - C2 respectively-
100
- 13 3 0 ,replacing C3byC3—C2
100
100
- 13 3 0 ,replacing R3byR3—R2
100
000
- 3 13 0 , interchanging Ci and C2.
010
000
- 3 0 0 ,replacing C2byC2—(13/3)CI.
010
000
- 1 0 0 replacing Ri by 1 Ri.
010
000
162 Matrices
Iz 0, which is the required canonical form.
t.00j
Exercise on § 412
Ex. Reduce 1 2 1 4 to the canonical form.
32 02
01.32
33 34
413. Definition. If A and B be two m '< n matrices, then B -A if and
only if B = SAT, where S is an m x m non-singular matrix and T is an n x n
singular matrix.
With the help of § 303 Page 104 and § 311 Page 111 it can be proved.
The following two properties of the above relation are fundamental.
• 1. Symmetry. If A - B, then B - A for if A = PBQ.
'then B = P 1 AQ', where P and Q are non-singular matrices.
• 2. Reflexivity. Every matrix A is equivalent to itself since we can write
A=IAI, so thatP=I=Q.
• 414. Theorem. An n x n matrix A is non-singular (or invertible) if
and only if the determinant I A I * 0.
Proof: If C be the canonical form of the matrix A, then C - A
Therefore C=SAT,
where S and T are non singular. (See § 413 above)
Hence A = S 1 C T'
or AEr... E2EICD1D2 ... Ds
where E1 and Di are elementary matrices. ...See § 109 Page 108
By the successive application of § 411 Page 159, we have
IAI=IErI...IE2II El. IICIID1IID2I ... 1D51
If lAI=0, then ICI=0,asiE,l*0andlDl*0.
If I C I = 9, then it has at least one row of zero.
The rank of matrix A is less Than n (see next chapter) i.e. the matrix A
is singular.
If the matrix A is non-singular, then
C where In i g the n x n identity matrix.
i.e. IC 1=1 In I = I

From (i) above, we have I A I * 0. Hence the theorem.

* 4 . 15. Theorem. IAIA2I=I All .1A21, where IAt and 1A21 are
two determinants: (Purvanchal 94)
Proof: Let Cj. and C2 be the canonical form of the matrices Ai and A2
i.e. Ai - Ci and A2 - C2.
If Al - Cl then from § 4 . 13 above we have
 Multiplication of DeterminantS 163

A l - S C l T, where S and T are non-singular matrices.

or At = Er .. E2 Ei Cl Di D2 ...
where Ej and Di are elementary matrices. (See § 309 Page 108)
Similarly AZ = Ft ... F2 Fj C2 Kj K2 ...
where Ft and Kt are elemenlary matrices.
AjAZ=Er...EZEIClDtD3t2F1C2K1K2
Hence by4ll Page 159, we have
lAlAZllEr...E2EJllCiD1D2DtF22l
•IK 1 K2 ... Ksl ...(i)
Now the following cases ariese.
(See § 218 Page 91)
Case I. Let Al he a singular matrix.
Then Cl has at least one row of zero.
I Cl Di Dz D5 Ft ... F2 Fi C21 0,
since the matrix C 1 Di 1)2... D 5 Ft. ..F2 Ft C2 has a row of zero.
From (i) above we have I At A2 = 0.
Case IL If Az is a singular matrix. Then Cz has at least one column of
zero, hence as in Case I aheve
Ci Di D2 ... Ds Ft ... Fi C2 1=0
From (i) above we have I At A21 = 0.
Case IlL If either Al or A2 is singular, then IA I • I Az I 0
HencclAi A2I0=IAIIIA2I.
Case IV. 11 Al and A2 are non-singular matrices . Then Ci and C2 are
identity matrices. Hence from § 411 Page 159, we have
A l A2IHEr... E2 E 1 Ci Di Dz... DI I F t ... F2 F 1 C2 Kj... Ks
= I Al 1 • I A21
From all the above cases it is clear that
IA! A I=lAile IA2 I.
Cor. I At A'2 I = I At1 • I A21, where A'z is the transpose of AZ.

Proof: IA'2I=IA2I, A'2 is the transpose

. of A
IAlA'zI=IAlI IA '2L from *4t4above.
=IA,IlA2I, IA'2I=1A21
The corolllary leads to the row by row rule of multiplication of the
determinants as given in Examples below
Solved Examples on Multiplication of Determinants.
*Ex. 1. Evaluatel 0 c b 2
c 0 a
baO
164 Matrices 989
2
Sol. 0 c b o c b x 0 c b
C a c a c a
ba0 baQ ba0
= 00+c-ci-b-b 0ci-c-O+b . a 0-b+c•a+b.O
cO + 0-c + a•b c-c + 00 + a-a cb+ 0a + a0
b0+a•c4-Q-b bc+a0+0 . a b-b+aa+O-O
ha ca
= c2+b2
ab C2 +a2 bc
ac bc
b2+a2 Ans.'
EL 2. Evaluate o cos x - sin x 2
sin 0 cosx
cos x sin 0
Sal. The required product
= 0 cos x - sin x I x 1 0 cos x - sin x
sin x. 0 cosx sin 0 cosx
cos x sin .r 0 cos x sin x 0
0-0 + cos xcos x + sin xsin x 0-sin x+ cos x-0 - sin x-cos x
sin x-O + 0-cos x - cos xsin x sin x-sin x +00 + cos xcos x
cos xO + sin xcos x + 0(— sin x) cos x-sin x + sin x-0 +.O-cos x
0-cos x + cos xsin x - sin xO
sin x•cos x + 0-sin x + cos xO
cos x•cos x + sin x•sin x + 0•0
cos 2 x+sin2 x sin XCos X Cos X sin X
— cosxsmx sin Z x+ COS 2 X sin X Cos X
SInx Cos x Cos x sin x Cos 2 x+ sin 2x
1 - X ,where?= sin x Cos x

Ans.
Ex. 3. Evaluate al b 1 q x xj yj z
a2 b2 C2 X2 Y2 Z2
a3 b3 C3 X3 Y3 Z3
Sol. The required product
= axi+biyi +cizi alx2+bly2+c1z2 01X3+bIy3+C1Z3
a2x1 + byi + czz a2x2 + b 2y2 + C2Z2 a 2.x3 + b
2y3 + CZZ3
ax1 + b 3 y l + C3ZI ax2 + b 3 + C3Z2 a
y2 3X3 + b3y3 + C3Z3
**Ex 4. Prove that
a b c 1'=1
2bc—a2 c2 b2
b C a c2 2ac—b2 a2
c a b
b2 a2 2ab—c2


Product of Determinants 165

= (a 3 + b 3 + C 3 - 3abc) 2 (Gorakhpur 91; Kanpur 95)

Sol. a b c 2' a b c X a b c
b c a b c a b c a
cab cab cab
- a b c X a c b interchanging C2, C3 of the second
b c a b a c determinant.
cab c b a (Note)
- a b .c f
X a c b multiplying C1 of first determinant by
—b c a b a c
—c a b c b a (Note)

= — aa+-bc+cb —ab+ba+cc —ac±bb+ca

— ba+cc+ab —bb+ca+ac —bc+cb+aa
—ca±ac+bb —cb+aa+bc —cc+ab+ba

2bc—a 2 c2 b2
= c2 2ac - b2 a2 Hence proved.
b2 a2 2ab - c2 (See Ex. 9 Page 167 aiso)
Also a b c =a c a —b b a +c b c I

b c a a b c b •c a
cab

=a(cb—a 2 )—b(b2 — ac) +c(ab—c2)

= abc - a 3 - b3 + abc + abc -
= - (a3 + b3 + c3 - 3abc)
a b c 2=(a3+b3+c3_3abc)2
b c a
c a b Hence proved.
tEx.5.Ifu=ax+by+cz,v=ay+bz+cx,w=az+bx+cy,
prove that a b c X x y z =u 3 + v + w3 - 3uvw
b c a yzx
cab zxy
Sol. By row-by-row multiplication, the product of the given determinants
ax+by+cz ay+bz+cx az+bx+cy
bx+cy+az by+cz+ax bz+cx+ay
cx+ay±bz cy+az+bx cz+ax+by

u v w 1, since ax + by + cz = u etc. (given)

w U V -
V W U

U V I —v w I v ^ +w W U
w U V•'U V W
166 Matrices

u(u2_ VW) _v(uw.v2)+w(w2_UV)

= a3 + V + w3 - 3uvw. Hence proved.

**Ex. 6. Express (a - x) 2 (b - x)2 (c - x)2
(a-y) 2 (b-y) 2 (c
(a_z) 2 (b-z)2 (c_z)2
as the product of two determinants. 95)
SoL The given determinant.
= a2 -2ax+x2 b2 -2bx+x2 c2-2cx+x2
a2 -2ay+y2 b2 -2by+y2 c2-2cy+y2
a2 -Zaz+z2 b2 -2bz+z2 c2-2cz+z2
The element in the first row and first column is a 2 - 2ax + x 2 , which can
be written as 1(a 2 ) + (- Zx) (a) + x2 (1). (Note)
This suggests that the first row of the required determinants are I - 2x,
and a2,a, 1.
Hence proceeding in this way we may write the given determinant
= I -2x x2 x a2 a I
1 -2y 2 b2 b I
I -2z z2 C c I
We can verify by multiplying with the help of row-by-row rule that
3bove two determinants are the required ones.
[Note: Such questions are acutally done by trial and error].
Ex. 7.Express (1+ax) 2 (1+ay) 2 (1+az)2
(1+bx)2 (1+by) 2 (1+bz)2
(1+cx) 2 (1+cy) 2 (1+cz)2
as the product of two determinants. (Gorakhpur 95; Purvanchal 96)
Sol. The given determinant
= I +2w-i-a2x2 i +2ay+a2y I + 2az+a2z2
l+2bx+b2 ? 1+2by+b2y2 1+2bz+b2z2
1+2cx+c22 1+2cy+c2y2 1+2cz+c2z2

The element in the first row and first column is 1 + 2ax + a2? which can
be written as (1) (1) +2a).(x) 4?(a2).(?)
This suggests that the first rows of the two required determinants are
1,2a,a2 and 1,x.x2
Hence the given determinant may be written as -.
 Product of Determinants 167

1 2a a 2 x
I '-1 x x 2
I 2h b2 I y y2
I 2c c2 1 I z z2 Ans.

*Ex. 8. Express b 2 + c2 Aca

Ac2+a2 be
C2 be b2+b2
as the square of a determinant.
Hence evaluate. (Purvanchal 94)

Sol. The element in first row and first column is b 2 +c 2 which can be
written as 0 . 0 + cc + bb. (Note)
So by trial and error method, we get the given determinant
=OcbXOcb
c a C a
baO b a 0

= 0 c b2
c0a
b a 0

Now 0 c b
c 0 a
boO
=-C c a + b c 0 , expanding with respect to Ri.
b ha

= - [cO - ab] + b [ca - b01 2abc.

The given determinat
= 0 c h 2 = ( 2abc) 2 = 4a2b2c2.
c 0 a
b a 0 Afts.

**Ex. 9. Express Zbc - a 2 c2 b2

c2 2ca—b2 a2
b2 a2 2ab—c2
as the product of two determiants.
Sot. The elements in the first row and first column is 2bc - a2 which can
be written as a (— a) + b (c) + c (b)
The element in the first row and second column is c2 which can be
written as a(-rb)+b(a)+c(c) ...(ii)

The element in the first row and third column is b2 which can be written
as a (— c) + b (b) + c (a) . ...(iii)


168 Matrices
(iii and (iii) suggest that the given determinant can be written
tentatively as
a b c 11 —a c b
b c a —b a c
c a b —c b a
But actually multiplying these two determinants we get the given
determinant. Hence these determinants are the required ones.
Ex. 1. Show that
Zbc - a 2 c2 . b 2 = (a 3 + 0 + c3 - 3abc)2
C! 2ac—b2 a2
b 2 a2 2ac—b2

Sol. As in Ex. 9. above, we can show that the given determinant

= a b c x —a c b
b c a —h d c
c a b —c b a
I.
= a b c I )( abc,
b Ca b c a
cab 'cab (Note)
I
taking - sign common from Ci and interchanging C2, C3 in 2nd
determinant
I,a
= c 2=(a3+b3+c3_3abc)2,
b c a
cab
on expanding the determinant. Hence proved.
Ex. 11. Find the product of determinants of different orders
or evaluate ai PI yi x al bi
2 P 2 Ti a2 b2
a3 03

SoL Here the two given determinants are of different orders, so we adopt
the following method:
(Xi Pi yi x at bi
cZ2 02 72 a2 b2
a3373
= at i TI x at b1 0
(X2272 ?2b20
(X3 53 0 0 1 (Note)
makffig the two determinats of the same order


Product of Determinants 169

a i a i + 13i b i+yt .0 a I a2 + 13I b2 +y, 0 alO+131•0+?I1
cz2aI + 132b1 + )2'O a2a2 + 132 b2 + )2-0 a2-0+P2'0+72-1
(X3aj + 133b1 + .0 (X3a2 + 133 b2 + 130 (13-0 + 133-0 + 131
= atct i+bi13i a2czI+b2131 yi
aIcx2+b1132 a2(X2+b2132 12
a I Ct3+ b 1133 a2a3+b2133 13 Ans.
Exercises on Multiplication, of Determinants
Ex. 1. Show that

i i i '= a+b+c a2+b2+c2
a b 'c a+b+c a 2 +b2 +c2 a3+b3+c3
a2 b2 C2 a2+b2+c2 a 3 +b 3 +c3 a4+b4+c4
Ex. 2. Show that
a 2 +).2 ab+c ca—bA. x X c —b
• ab—cA. b2—A.2 bc+aA. —c X a
• ac+bX bc—ak c2 + b —a A.

=X3 (A.2 +a2 +b2 +c2)

Ex. 3. 11(1) is one of the imaginary cube roots of unity, show that
2
I (0 0) 2 3 =
1 1 —2 1 =-27
2 CO
3 i 1 1 —2
0)2 (03 1 —2 1 1 1
2 1 —2 1 1
(Gorakhpur9ó, 93)
*Ex. 4. Prove that the determinant
- a2 ab ac is a perfect square and find its value.
ab —b2 bC
ac bc - c 2 (Gorakhpur 92)
Ex. 5. Express the product of the following determinats as a . single
determinant
a b and u r
c vs
4-16. Theorem. If C1 be the cofactor of a j in the n x n matrix
A=[a,j] then lCi1I= {IaJkl}"1
[Note. I C,3 I is known as reciprocal of the determinant I aj I],
Proof. If A = [aq], then A'= (a'kj], where A' is the transpose of A and
= a,k
Now A'. [c,3] = [ a'ki] [Go] = [bkj], 5Y
• n
where bkf = I a'kj Cj = E aj C, a'ki = ak
i=1 1=1


170 Matrices
Also by § 409 Page 129 we know that
n
bkj= Cij
- 1=1
=IAI,ifj=k.
From (i) we conclude that for the product A'. [C1 ] i.e. (bkf] all the
diagonal terms (for which j = k) are I A I, whereas the nondiagonal terms (for
which j#k) are zero.
i.e. A'. [C] = I A 1 0 ... 0
0 Al ... 0

0. 0 ... Al
Hence tA'.[C1j ]I=I Al 0 ... 0
0 IAI ... 0

0 0 ... Al

or lA'l.$CJl=(lAI}',.lA1.A2I=lA1l.lA21
n
or l Al. lC1l={lAlI , . IA'llAl
r lC,1I={lAIj'
§ 417. Complementary Minor of a Determinant.
Definition. If B is r x r submatrix of an n x a matrix A, then the
determinant E' of A formed by removing the rows and columns of A
containing the elements of B is called the cornpletementary minor of B.
For example : In the matrix al bi ci di
02 b2 C2 d2
a3 b3 C3 d3
04 b4 C4 d4

the completemtary minor,of the det al h i I is C3 d3

02 b2 C4 d4 (Note)
thecompleteñientary minor of bi ci di is a4
b2c2d2
b3 C3 d3
and the complementary minor of b2 C2 is at di
173 C3 (Note)
04 d.4
§ 418. Laplace's Expansion of a determinant by the minors of first r
Collumns.
If I Bi I is r x - minor of an i x n matrix A fomied by the elements of the
first r columns of A and I B'i I is the complementary minor of I B1 I, then


Laplace's Expansion of Determinants 171

1AI=±B 1 1.IB',I, -
where the summation is extended over all the possible r x r minors of A whth
o
can be formed from the elements of the first r olumns and + r —sign taken
according as an even or odd number of interchanges of adjacent rows of A is
required to bring the suhmatrix Bi into the first r rows of A.
The following solved examples explain the above theorem.
Ex. 1. Expand a x y a Laplace's expansion by the minors of
xO 0 y
y 0 0 x
a y x a
the first two columns. Hence evaluate it.
Sol. All the possible minors of the first two columns and their
complementary minors are given by
a x :IB'tI= 0 x
X 0 x a
1B21
Y
a x I;IB'21 = 0 y
X

B31= a xI IR'31 = 0 y
a v 0 x

Y
1B41= x 0 ;IB'41= y a
X

X 0 ;IB'51= y a
a y 0 x

ay
and IB6= y 0 1;]B'6 1=
0
y a

Therefore the given determinant

= ax. 1 Ox I - ax. 1 0 y I + I ax .0 y
X 0 x a y 0 x a ay Ox

+xo.ya—xo.ya+YO.Ya
y x ay Ox a)' 0 . .(i)
The submatrix B2 requires one interchange of rows viz, of second and
third rows to bring it into the first two rows therefore —sign is put before the
product I Wi I I B'2 I . Again the submatrix B3 requires two interchanges of rows
to bring fourth row to the position of second row i.e. to bring B3 into first two
rows, therefore + sign is put before the product I B3 II B'3 I.
Similarly B4 requires two interchanges, B5 requires three interchanges
and B6 requires four interchanges, hence +, - and + sign are put before
I B4 I. I B'4 I, 1 B5 I. 1 B'51 and I B6 I. I B'61 respectively.
172 Matrices
Hence from (i) we have (expanding the determinants) the given

determinants.
= (-?) ( x) - (- xy) (- xy)+ (ay - ax) (0) + (0) (ay - ax)
- (xy) (xv) + ,2)
2)

=x4—z2y2+y4= (XI —) 2 )2 - Ans.

Ex. 2. Expand a b c d by Laplace's expansion by the minors
e f g h
o 0 jk
OOlm
of the first two columns.
Sol. All the possible minors of the first two columns and their
complementary minor are given by
IBI= a b ,IB'il= j k
e I
1B21 a 1, =0, hence lB'2l need not becalculated
•00
Similarly I B3 1 = a b = 0; B4 I = e f = 0;
00 00
• I Bc I = e f = 0 and I Bs I = 0 0 1 = 0 and therefore their comple-
00 00
mentary minors need not be calculated.
Then the given determinant by Laplace's Expansion method
ab.jk
e 1 Ans.
Ex. 3. Expand 3 2 1 4 by Laplace's expansion by the
15 29 2 14
16 19 3 17
33 39 8 38
minors of the first two columns.
Sol. All the possible. minors of the first two columns and their
complementary minors are given by:
lBiI= 3 2 , IB'iI= 1 3 17
15 29 8 38
1B2l= 3 2 , IB'21= 2 14
16 1 8 38
IB3l= 3 2, IB'31= 1 2 14
33 39 • 3 17
1
lB41= 15 29 ,JB 41= 1 4
16 19 8 38


Laplace's Expansion of Determinants 173

IBI= 15 29 1B'5=
, 1 1 4
33 39 3 17
1B61= 16 19 , IB'61= .1 4
33 39 2 14

The given determinat

= 3 2 317— 3 2 214
15 29 8 38 16 19 8 38
+ 3 2 S 2 14 1 + 15 29 S
1 4
33 39 3 17 16 19 8 38
- 15 29 1 14 1 + 1 16 191 4
33 39 • 3 17 33 39 • 2 14
EL 4. Expand a 1 0 0 0 by Laplace's expansion by the
balOO
ObalO
SObal
lOOk.
minors of the first two columns. Hence evaluate it.
Sol. All the possible minors of the f'rst two columns and their
complementary minors are given by:
IB 1 I= a 1 , IBjI= a 1 0
b bal
Oba

IBzl=I a 1 , IB'21=J 1 0 0
0 b b a 1
Oba

1B31= b a , IB'31= 0 0 0 =0
0 hal
Oba
All other minors of the first two columns are equal to zero as they have
at least one row of zero.
Hence the given determinant
=al alO —al 100
ha • b a I 1 b • b a 1
0 a b 0 b a (Note)
Now a 1 0 = a I a— I a 1 .1-f- b a 0,
bal ba Oh Oh
Oba
expanding by the minors of first two columns.
:(a2_b)a_(ab)=a3_2ab.
From (I), the given determinant
-12


174 Matrices
= a I 3 0 1 - a I
(a-2ab)— Oh
ha b-a
expanding the last determinant with respect to R1.

= (a2 - b) (a 3 - 2ab) - ( oh) (a 2 - b) = ( a - b) - 3ab]

=a(a2—h)(a2-3b). Ans.
Exercises on § 418
Ex. 1. Expand at 02 a3 04 by Laplace's method of expansion by
bi b2 b3 h4
Cl C2 C3 C4
d1 d2 d3 d4

the minors of the first two columns.

Ex. 2. Use Laplace's method of expansion of a determinant by rneais of
its second minors to expand
—I 0 0 1
o —1 0 rn
o 0 —1
p q r—1
** 4-19. Solution of Linear Equations. (Cramer's Rule)
Let the n simultaneous equations in n unknown quantities XI, x2 ,..., xt, be
a li x i ± U t2X 2 + ... + U1JXj + ... + U InX n k
021X1 + a 22x2.+ .. +U2jXj+ ... +U2nXnk2
ailxl + a 32,r 2 + ... + ( 23jXj + ... + a3,X,j

ail xi + a12x2 +.. + ± ... + Oj.X =

Untxl+an2X2+...+anj.j+ ... +anflxn=kn

These equations -can be written as
n
£ a 1 x=k1 , 1=1,2 ..... n (i)

Let the determinant of the coefficients, I A I = I a,1 I ;t 0.

Multiplying (i) by the cofactor of ay in I au I vii. C,. i = 1, 2, 3.....n and
summing with respect to i, we have
n
•Z a,Cjjxj
j=l j=l
n
or IALx3 = L k 1 C,, :IAI= I aij
• jl j=l
= determinant formed by replacing jth column of the deL
I Al by the constants k1, k2.....k, . (Note)


Cramer's Rule 175

= an 012 ... a1.-1 k 1 01n

Olj-4-I ...

021 a 22 •.. (J2J 1 k2 02j+1 ... a2

Oil 0i2 .. Oq-I ki aj+1 .. am

(1n1 0n2 .. a,t,i k ... a,1,

all a12 ... aij-I ki ai1+i ... 01n

- 1 021 022 ... a2j-1 k2 02j+I ...

or
—IA
0n1 a,,2 ... anj -i k2 Onj+l ..

Solved Examples on § 419

Ex. 1 (a). Solve the following equations by Cramer's Rule.
x + y + z = 1, ax + by + cz 'k,a 2x + b2 + c2 = k2.
Sol. The given equations are x +)- + z =
ax + by + cz = k
2 2 2 2
ax+by+cz=k
By Cramer's Rule we have
X V Z

= 1 1 1
= 1.1 1 1 1
ba kca h
a2 b 2 k 2 a2 b 2 c2
Fkk h 2 c2 a2 k2 c2
or
(k b) (b c) (c k) - (a —k) (k . ) (c a)- ( a b))(h— k) (k a)
= 1
See Ex. 28 Page 145
(a — b) (b — c) (c —a)
(k—b)(c—k)
(a—b)(b--c)(c—a)(b—c)(c—a)
(a—k)(k—c) ändz=k
Similarly
(o—b)(b—c) (b—c)(c—a) Ans.
Ex. 1. (b) Solve the following equations by the method of
determinants -
aix + bjy + ciz + d1 = 0
a 2x + by + czz + d2 =0
a3x + by -f OZ + d3 = 0
Sol. Do cxacity as Ex. 1 (a) above
Ex. 2. Solve the equations
x+y+z=7; x-+-2y+3z=16; x+3y+4z=22
Sol. Solving the equations by Cramer's Rule we get
176 Matrices 182/1I11
z - 1
711171117F
'16 2 3 1 16 3 1 2 16
22 3 4 1 224 13 22 13.
Now 7 1 I = 1 7 1 interchanging C and C2.
16 23 2 16 3 -
2234 3 22 4
= - 1 7 1 , replacing R2, R3 by R2 - 2Ri and R3 - 3Ri
O 2 1 respectively.
011
= - 2 1 expanding w.r to C
11
=—(2— 1)=— 1;
1 7 1 = 17 t ,replacingR2,R3byR2—R1,R3—RI
1 16 3 0 9 2 respectively.
1 22 4 0 15 3 -
= 9 2 ,expanding w. r. to Cj.
15 3
=27 - 30=— 3;
1 1 7 = 1 1 7 , replacing R2,R3byR2—RJR3--RI
1 2 16 0 1 9 respectively.
1 3 22 0 2 15
=19
2 15
="— 18=-3
and 1 1 1 = 1 1 1 , replacing R2, R3 by R2 - R1, R3 - R2
1 2 3 0 1 2 respectively.
1 3 4 0 1 1
= 1 2 , expanding w.r.toCj.
11
=1-2=—i.
- x 1
From (1) we have --j
= =-z= —j
which gives x=1y=3,z=3. Ans.
*Ex. 3. Solve the equations (with the help of determinants)
x + y + z = 1; x + 2y + 3z = 2; x + 4y + 9z = 4.
Sot. The given equations are x + y + z = 1.
- x±2y+3z=2
x + 4y + 9z = 4


Cramer's Rule 177

18211112
By Cramer's Rule we have
x Y Z I
111
22 123122 123
149 144 149
F
111
or ==-1-'whereD 123
ODOD 149
This gives x=0. 0; y =D.(-) 1, z=O. [-] o.

x=0,y=1,z0, Ans.
i.e.
*Ex. 4. Solve the equations by determinants 3x + Sy - 7z = 13,
(Purvanchal 97)
4x + y - 12z = 6, 2x + 9y - 3z 20
Sol. The given equation are
• 3x+5y-7=13
4x+y- 12z6

i1
2x + 9y - 3z =20
By Cramer's Rule, we have
X ______ z
-:i-; - = ] 13 -7 = 3 5 -7
-12 4 6 -12 4 1 6 4 1 -12
-3 2,20 -3 2 9 20 2 9 -'3
F

Now 13 5 -7 = -17 0 53 , adding -5R2tOR1 and

6 1 -12 6 1 -12 -9R2toR3
20 9 -3 -34 0 105
= -17 53 ,expanding w.r.tOC2
-34 105
= 1 - 17 . 53 , adding - 2Ri to R2
0 -1
=(-17)(- 1)_(0)(53)=17;
3 13 -7 = 1 - 7 -4 , adding - 2R3 to R2
4 6 -12 0 -34 -6 and-R3t0RI
2 20 -3 2 20 -3
= 1 -7 -4 . adding - 2R to R3
0 -34 -6
0 34 5
- 34 - 6 •expaiiflg w.r. to Ci
34 5


178 Matrices
= 0 —1 , adding R2toRI
34 5
=05—(— 1)34=34;
3 5 13 = - 17 0 - 17 , adding - 5R2 to R1
4 1 6 4 1 6 and —9R2t0R3
2 9 20 —34 0 —34
= - 17 - 17 , expanding w.r. to C2
—34 —34
=0

And 1 3 5 —7 = —17 0 53 1, adding —5R2toR1

4 1 -12 4 1 - 12 and - 9R2 to R3 1

2 9 —3 —34 0 105
= -17 53 expanding w,r. to C2
—34 105
= - 17 53 , adding - 2-RI to R2
0 —1
=17
x
From (i), we get -j- j_z 1
=
which gives x=l,y=2,z=0 Ans.
Ex. 5. Solve the eqations+y+z=3,x+2y+3=4 x+4y+9z=6.
- (Pun'anchiil 94)
Sol Given equation are x+y+z= 3
X + 2y + 3z =4
x+4y+9z=6
.. By Cramer's Rule, we get
X z
TfI 3 1 - fT I 31]ijTT.
64 ^
F
Now 3 I I =
1
1 4 3
69 H1 2 4)
4 6
1 2 3
j14 9 1

0 1 0 replacing Cj, C3, b y Cj - 3C2,

4 2 3 —2 2 I Ci - C1 respectively.
6 4 9j l--6
—2 1 =--(- 10+6=4,
—6 5
1 3 1 1 i.replacing R2, R3 by R2 - R1,
1 4 3 0 1 2 R3 —R1 respectively.
16 9 0
1 38


Cramer's Rule 179

1 2 =8-6=2
38
1 1 3 = 1 1 3 replacing R2. R3 by R2 - Ri,

1 2 4 0 1 1 R3 - RI respectively.
146 03 3j.
1 1 =0
.(iv)
3 3
I I I = 1 0 0 , replacing C2, C3 by C2 - CI,
And
1 2 3 I 1 2 C3 - Ct respectively.
149 138
1 2 =8-6=2
38
From (i), (ii), (iii), (iv) and (v) we have
xxz I
4202 or x=2,yi,Z=O Ans.

Ex. 6 (a). Using determinants, solve the simultaneous equations:

(Pur'anchaI9O)
x+2y+3z6; 2x+:-z=7,3x+2Y+974.
SoL By Cramer's Rule e get
X z 1
1 2 3
6 2 3 TT u 3 ]T26
271 247 241
741
3 14 9 3214 329
14291
Now 6 2 3 = - 26 3 interchanging Ct and
C2

741 471
14 2 9 2 14 9
2 63 , replacing R2,R3hYR22RI
o -5 - 5 and R3 - RI respectively.
0 8 6
= - 2- 5 - 5 expanding with respect to CI
8 6
=.2'_30+4O1=-20;
., 3 = I 6 3 , replacing R2, R3 by R2 - 2Ri,
0 --5 -5R3-3R
- 3 14 9i t) -4
- 5 - 5 exp:inding with rnpccttO Ci
-4 o
=-20:
6 replacing
I. R2, R 3 by R2 - 2Ri,
I 2 6 = 1 2
2 4 7 0 0 -5 R-t-Vi.
3 2 14 0 4--4


180 Matrices 989

= 0 - 5 , expanding with respect to C1
—4 —4
= —20
1 2 3 = 1 00 ,replacingc2,Cbyc22c1
2 4 I 2 0 - 5 C3
3Ci - respectively.
3 2 9 3.-4 0
and 1
= 0 —5 , expanding with respect to Rj
—4 0

-
= —20
From (i) we get
x _ _ y
— 20 =-20 —20 =
z

or x=l,y=1,z=l Ans.
Ex. 6 (b). Solve the following equation with the help of determinants
2X+y+Z1,x-2y3z= 1,3x+2y—z5. (Purvanchal96)
Sol. By Carmer's Rule, we get

fiTiJ =12
X_______

1 —2 —3 I I —3
z
1 1 = 2 I iJ
1 —2 1
= TTTJ
1 —2 —3
5 2—i 35-1 3 25
Now I I I
1 —2 —3
= 1 0 0 , replacing C2, C 3, by C2 - C1
1 —3 —4 and C3—C 1 respectively
5 2 —1 5 —3 —6
—3 —4 1. expanding w.r.toR1
—3 —6
—3)(-6)—(-3)(-4)= 18-126;
2 1 1 = 2 1 1 , replacing R2, R3 by R2+ 3Ri
I I - 3 7 4 0 and R3 + RA-respectively
3 5 —1 5 6 0
= 7 4 , expanding w.r. to C3
56
=(7)(6)—(5)(4)422022;
2
1 1 = 2 1 1 , replacing J?2,R3byR-RI,
-2 1 - .1 - 3 0 R3 - 5R 1 respectively.
3 2 5 —7 —3 0 -

= —1 —3, expanding w.r,toc3

—7 —3

- =(- 1) (—)—(-7)(-3)=3-2I =- 18
Also 2 1 1 = 2 1 1 j,repJacingR2, R3 by R2 + 3Ri
1 —2 —3 7 1 0 and R3+Rlrespectively
3 2 —1 5 3 0


Cramer's Rule 181

= 7 1 , expandirg w.r. to C3
53 -
=73-51=21-5=16
x y . z 1
From (i)wchave = 22j-=-j-
6 3 22 11 18 9
which gives x =1 '' = -j- = --' z = - Ans.
= -
Ex. 7. Solve the equations:
'x+y+z+u=l, ax + by + cz + du k,
a2x+b2y+c 2z+d2u=k2 and a3x+b3y+c3z+d3u=k3
Sot. The given equations are

ax + by + cz + du =
2x + b 2y + c2z + d2-u = k2,
and a3x+b3y+c3z+d3u=k3
Solving these by Cramer's Rule, we get
X z
11 11 = 1 11 1 = 1 1 11
k b c a k c a b k d
k2 b2 c2 d2 a2 k2 c2 d2 a2 b2 k2 d2
k3 b3 c3 d3 a3 k3 c3 d3 a3 b3 k3 d3

- U
_ 1 11lill 1
a b c k a b c
a2 b2 c2 k2 a2 b2 c2 d2
a3 b3 c 3 k3 a3 b3 c3 d3

Now 1111 0 0 0
a b c d a b — a c—a d—a
a2 b2 c2 d2 a2 b2 —a2 c2 —a2 d2—a2
a3 b3 c3 d3 a3 b3 —a3 c3 —a3 d3—a3

replacing C2, C3 and C4 by C2 - C1, C3 - C1 and C4 - C1 respectively.

= b—a c—a d—a 'expandingwith respect toRi
2 2 2 .a2a
2
b —a c — —a 2
b3 —a3 c3 —a 3 d3 —a3 -

=(b—a)(c—a)(d—a) 1 1 1
b+a c+a d+a
b2 +ab+a 2 c2 +ac+a 2 d2+ad+a2
12 Matrices

=(b-a)(c-a)(d-a)I 1 0 0
I b+a c - b d-b
b2 +ab+a2 c2 +ac-b2 d2+ad-b2
-ab -ab
replacing C2 and C3 by C2 - Ci and* C3 - C1 respectively.
=(b- a) (c- a) (d- a) c - b d-b
(c-b)(i+b+c) (d-b)(a+b+d)
expanding with respect to Ri,
=(a-b)(a-c)(d-a)(c-b)(d-b) 1 1
a+b-*c a+b+d
(a - b) (a -c) (a - d) (b - c) (d - b) [(a +b + d) - (a + b + c)]
- (a - b) (a - c) (a - d) (b - c) (b - d) (c - d)
Similarly we can have 1 1 1 1 = (k - b) (k - :) (k - d)
k b c d (b-c)(b-i)(c-d)
k2 b2 c 2 d2
P b3 c3 d3

Frnin (I), (k - b) (k - c) (k- d) (b - c) (b - d) (c !A -

(a - b) (a - c) (a - d) (b - c) (b - d) (c - d)
(k - b) (k - c) (k - d)
of X=(a_b)(a_c)(a.d) Ans.
Similarly from (i) we can find the values of yz and u.
Exercises on * 419
Solve the equations by Cramer's Rule
Ex. 1.x-2y+z=-1,3x+y-2z=4,y-z=1. An.x=l, yl, z=0,
Ex. 2.2.x+3y-4z=2, 3x_2y+5z=z5,x-i-2y+3z= 11.
1050 33
Ans. x=-j' -j-' 19-
Rx. 3. x + 2y - z =3; 3x - y + z 8, x + y + z =0.
Ans.x=-'y=-I' z=-.
- Ex. 4.x+y+z3;2x+3y+4z=9,x+2y4z=-1.
Ans. x=1, Y= 1, z=l
EL S. 2x-y+ z=9, x+y+z='6, x-y+z=2.
Ans.x= 1, y=2, z=3.
EL 6. 3x+y+2z=3, 2x-3y-z=-3, x+2y+z=4.
Ans. x=j, y=2, z=-1
Ex. 7.xi +2.r2+3x3+5=0, 2xi + x2+x3±7 = 0. x +x2+x3=0.
Ans. xi=-7, 12 = 19, x3=-12.
Ex. &6x+y+2z=7, 3x-y+4z= 14,5x+2y-3z=-7.
(Puri'anchal 91)
Ansx= 1,y-3,z=2


Derivative of Determinants 183

*EL9x+y+z9,2X+5Y+7Z=52.2.+YZO
Ans. x=l, yrr3. z=5.
* 420. Derivative of a determinant..
If some elements of the n x n matrix A = [ aV ] are differentiable functions
of a variable x, then the derivative of I A I with respect to x i.e. dx I A I is the

sum of n determinants formed by replacing in all possible ways the elements of

one row (or column) of the del I A I by their differential coefficients with
respect to X.
The above procedure will be illustrated by the following examples -
Ex. 1. Find the derivative of the det. x 3 2x +3 *
32 x4

d x3 2.x+3 = 3x2 2 1 + x3 2x+3

d132 x 3?x I & 4x
differentiating the elements of Rj in the first dci
whereas differentiating the elements of R2 in the
second del
=[3x6-6x2]+[4x6-6x(2X+3)]
6 _ 6x2 + 4X6 _ 12x2 — I 8x = 7x6 — 182 — 18x. ADL
=U
X32
*Ex. 2. Find the derivative of X 2
2x 3x+1
0 3x-2 j;2j
Sol. The derivative of the given determinant
2 2
= 2x 3X20 + 2 + x2 X

2x 3x 4- 1 2 3 3x2 Zr 3x + i
o 3x-2 x2 +l 0 3x-2 ?-+-i 0 3 Zr

=Zx 3x2 0+? x3 2

o 3x+J-3x2 x3 2 5-3x 2x 2_
o 3x-2 x2 +1 0 3x-2 x2-i-1

+ 2

? 2 ,replacing R2oflstdet.byR2—RI
Zr 3x+l x 3 and R2 of 2nd det.byR2—R3
0 3 Zr

=2x 3x+1-3? X3 22 -1
5-3.r
3x-2 x2+I 3x-2 ?+i
—2 x3 2
3x-2 x2+l


Matrices
+ 3x + i - 2 . expanding det. w.r. to C1
3 2x 3 2.x
= 2-X [(x2 + 1) (3x+ I - 3x2 ) - x3 (3x - 2)] + x 2 [(5 - 3x) (x2 + 1)
—(3x-2)(2x2— 1)]-2[x3(x2+ l)-2(3x-2)]
+x2[2x(3x+1)-3x3]-2.x[i-6],
—0x5 + 25.x4 - 4x3 + 9x2 + 26x - 8, on sirnihfying. Ans.
Exercise on § 420
Ex. Find the derivative of X2- I X - I
4 x3 zr+s
X +
x2 x

Ans. 6x 5 - 5x4 - 28x3 + 9x2 + 20x —2

MISCELLANEOUS SOLVED EXAMPLES
4 Ex.1.Solve 1+x 2 3 4
12+x 3 4
1 2 3+x 4
1 2 3 4+x
So!. The given determinant
= x +10
2 3 4 replacing C1 by
x+10 2+x 3 4 CI+C2+C3+C4
x+l0 2 3+x 4
x+l0 2 3 4+x
=(x+10) 1 2 3 4 ,taking out (x+l0)
I 2+x 3 4 common from Cj
I 2 3+x 4
1 2 3 4+x
= (x + 10) 1 2 3 i , replacing R2, R3 and R4 by R2 -RI,
0 x 0 0 R3 - Ri and R4 - Ri respectively.
OOxO
000x
= (x + 10) x 0 0 , expanding with respect to Ci
0x0
OOx
= (x + 10) x x 0 , expanding with respect to Ri
10
= (x + 10) x (xx) =x3 (x + 10). Ans.
 Miscellaneous Solved Examples 185

Ex. 2. Show that - (a + b + c) is a root of the equation

x+a b c =0
b x +c a
C ax+b
Sol. The given equation can be written as
'x+a+b+c b c =0replacingCIbyC1+C2+C3
b+x+c-4-a x+c a jn the det.
c+ai-x+b a x+b

or (x + a + b + c) I b c = 0, taking out the common

1 x+c a factor fromCi
1 a x+b
= 0, replacing R2, R3 by
or (x + a + b + c) 1 b c
o x+c—b a—c R2— R I , R3--Rj
o a—b x+b—c
a - c =0, expanding with respect
or (x + a + b + c) x + c - b
a—b x+b—c toCt

or (x+a+b+c)[(x+C_b)(X+b—C)—(aC)(a—b)]=0
or (x+a +b+c)(x2+ab+bc+ca_a2_b2C2)=0

This gives x=_(a+b+c),±'J(a2+b2+C2abbC—ca)

Hence - (a + b + c) is a root of the given equation.
Ex. 3. Show that x I m 1 = (x - a) (x - 3) (x -
ax n 1
aI3xl

Sot. The given determinant

= x 1 m I , replacing R2, R3, R4 by
a - x x - I n - rn 0 R2 - Ri, R3 - Rt,R4 - Ri respectively.
cz — X —1 x—m 0
a—x —1 y—rn 0
a—x x-1 n—rn , expanding with respect to C4
a—x , — l x—m
a—x i-1 y—m

- ((x - x) 1 x - I n - rn , taking out (a - x) common

I —I x—rn
1 —I y—m
=(x—a) 1 x-1 n— r n .replacingR2 and R3bYR2—Ri
0 13—x x—n and R3—RI respectively.
0 1 — x •)'—fl
=(x—(X) —x x—n ,expanding with respect toCi
f3—x y—n


186 Matrices
=(x-a)(--5r) I x-n =(x-u)(-x)[(y-n)-(x-n)]
I )'-n
=(x- cc)(13 —x)(y—x)(x - (X)(x- )(x-y) Hence proved.
**Ex. 4. Evaluate 0 * y z
-x 0 c
-c 0 a
-b -a
01
Sol. The given determinant
- 1 0 ax - by + cz y z , replacing C2 by aC2 - bC3 + cC4.
- a - x 0-bc + cb c b Here (1/a) has been taken
- y -ac+0+ca 0 a common due toaC2
- z -ab+ba+0 -a 0 (Note)
0 ax-by+cz y z
a -x 0 cb
- y 0 0a
-z 0 -a 0
=-(ax -by+cz)(l/a) -xc b , expanding with respect
- y 0 a to C2.
-z -a 0
=(ax+by+cz)(11a) xc b ,takingoul- Icornmonfromci
Y Oa
z -a 0
(ax-by+ ax-by+cz ac-0-ca ba-ab+0
a.a y 0 a
Z -a 0

replacing R1 by aRj - bR2 + cR3 and taking out (I/a)

common as before
_(ax-by+cz) ax-by+cz 0 0
- 2 y Oa
z -a 0.
- (ax + by + cz) 0 a , expanding with respect
(ax - + CZ1
- a2 -a 0 to RI
- by + cz)2
[0-0 - a- a)] (ax - by+ cz)2 [2]

(ax - by + cz) 2 Axis.

Ex. 5. Show that I a a2' 3 + bed =0
1 bb2 b3+cda
1 c c2 c3-+-dab
1 d d2 d3 + abc (Gorakhpur 92; Kwnaun 95)


Miscellaneous Solved Examples 187

given that abcd *0.

Sol. The given determinant
= i a a 2 a3 + 1 a 02 bcd
1 b b2 b3 1 b b2 cda
I c c2 c3 I c c2 dab
1 dd2 d3 ldd2abc

Now I a a2 bcd'
I b b2 cda
i c c2 dab
ldd2abc
2
= _L a abcd multiplying Ri, R2, R3, R4 by a, b, c, d
abcd b b 2 b3 beda respectively and dividing the result by

c c2 c3 cdab abcd, where abcd * 0.

dd2 d3 dab c (Note)
2
- a a a3 1 , taking out abcd common from C4
abcd (abcd) b b 2 b3 1
2
3
dd2d3l

= - a a 1 interchanging C3 and C4
b b2 1 b
c C I
dd2 Id3

= ( 1) a I a2 03 interchanging C2 and C2
b I b2 b3
C I C 2 C3

dld2d3

I a a a interchanging Cl and C2, also I)=-I.

I b b2 b3
I cc 2 c 3
ldd2d3
Substituting this value in (i), the value of the given determinant is zero.
Ex. 6. Show that - 1 0 0 a = 1 - ax - by - cz
0-1 0 b
o 0 -1
x y z -1 (Gorakhpur9l)


188 Matrices
Sot. The given determinant
-= —I 0 0 —a+O+0+a ,replacingC4hy
0 —1 0 0—b+0+b aCI+bC2+cC3+C4
o 0 —1 0+0—c+c
X y z ax+bv+cz — J
=-1 0 0 0
0—i 0 0
O0-1 0
X y z ax+by+cz—i
=(ax -s-by+cz--1) —1 0 0 , expanding with respect toC4
(I —i o
0 0—i
= - (ax + by + cz - 1) - 10 , expanding with respect to Ci
0 —1
• =(1 — ax — by+cz) [(- 1)(— 1)-00J
=(l—ax—by—cz)[l]=(1—ax—by—cz). Hence proved
*Ex. 7. Prove that
2 zx—y 3 xy—z 2 2 2 2 2
yz—x = v- U U = x y a
V2 y z x
ax -? xy - a2 yz - x2 u2
,2 z x y
y—z2 yz_x 2 zx—y2 u2 U2

where 'v2 = x2 + y2 + z2, u2 = yz + zx + xy. (Goraknpur 90)

Sot. X z2xy z xxy z

- yzx yzx yzx
zx.y zxy zxy
Xy+yZ+ ZX xz+yx+zy
=
yx+zy+xz y2+z2+x2
zx+xy+yz zy+xz+y Z2 +x +y
=V
2 2 2 2
U U u=y:+z.a+xy,
2-2 2
U V U v = x2 + y2 + z, (given)
2
2 ?

Again from § 416 Page 169 we knwo that if Cij be the cofactor of aij in
the nxn matrix A=[a1],then
• ICgjI={AI}'
Here n=3,soIC,I={IAI} 2, ... ( ii)
where A= x y z
y z x
zxy


Miscellaneous Solved Examples 189

c j i= zy—x 2 xz—y 2 zy — z where C,js the cofactor
zx—y 2 xy — z 2 yz—x 2 ofajinIAl
.xy—z 2 yz—x 2 u—y22
2 2 2 2
From(ii), Z,V — X ZX — y AY — Z = x y z
2 2
U - y .xy - z yz - f y z x
2 2 z x -y
XY — Z )'Z — X 2 U -)'
From (i) and this we have the required result.
Ex. 8. Write down as a determinant the product
abc. x y z
cab z x y
b Ca y z x
Sol. Multiplying by the row-by-row' rule we get
ax+bs'+c a+hx+cy a%'+bZ+cZ
Cr + a ) ' +bz cz +ax + by cv + az + bx
hA- + cv + az bz + cx + ay b y + cz +

=(a+h+c)(x+y+z) 1 I
cx+av+bz cz +ax+hv c+az+bx
bx+cy +az bz+cx+ay hy+cz+ax
replacing R1 by R + R 2 + R3 and taking the common factor out.
Ex, 9. Expand 0 1 x y by Laplace's Expansion's by the
0 Oyx
Z wOO
w 00
minors of the Fst two columns.
Sol. All the possible minors of the frist two columns and their
complementary minors are giv'n below:
BiI = 0 I 1 B'l= y x=0
Z w 0 0
1B21= 0 I , IB'21= y x I =0
14' z 0 0
1B31= z w JB'jI= x y
wZ
Y
The given determiriat = z iv .X
W z y x
the remaing minors or complementary minors are zero.
2 2 2 2
= (z - w ) (x - y ). Ans.
**Ex 10 (a). Show that at b1 c 2 = At Di C1
22 b2 C2 A2 82 C2
a3 b3 C3 A3 B3 C3
-13


190 Matrices
where the capital letters denote the cofactors of the corresponding small
letters. (Gorakhpur 96. 92; Kanpur 93; Purvanchal 97)
SoLLetIAI aibi C1 and = Ai Bi C1
a2b2C2 A-2 B2
03 173 C3 A3 83 (23

Then x I A I
= At Ri Ct x ai hi
A2B2C2 a2b2c2
A3 B3 C3 03 fr3 C3
a3Al+b3Bt+C'1
= aAi+biB1 +ctCi a2Al+b2Bl+C2CI
b2.B2 + c2 2 aA2 + b3B2 + c3C2
aiA2 + b,B2 +C1 C2 a2A2 + C

a1A3+ b l B3 +C1 C3 a2A3+b283+C2C3 a3A3+LB3+c3C3

= IAI 00 ,sinceaiAi+blBl+C1Cl=Ietc.
o IAI 0 and alA2+b1B2+c1C2
o 0 A I - (See example on Page 117)
Note. Students are to prove these
in the examination.
=IAI . IAI 0 ,expanding with respect first row.
t0

0 IAI
[I AU3
=IAIAI.IAI_001U
or A xIAltlAl)0t2
or Al BI C, = hi c
A2 82 C2 02 b2 C2
Hence proved.
A3 83 C3 03 b3 C3
S Ex. 10 (b). Prove that
2 a2 —bc b 2 —ca c2—ab
a h
c a b c2—ab n2 —jc b2—ac
b C a b2-ca c2 - ab a 2- be (GorakiipUr 90)
SoL We know [See E. 10 (a) above] that
at c 12 =. At Bi C,
02 17-2 .'2 A2 82 C2
03 13 C3 A3 83 C
.jwll letters.
where capital letters denote the cofactors of the correspmding .
b c 2 IA 8 cI.
c a b ¶c A RI At
c a B C A
specitvciy : 11 imi .: rn'e.
where A, B, C are the cofactcrs of a, b, r
left.


Misc. Solved Examps on Determinants 191

a b = a2 —bc,8=—jc b b2—ac
C a Jb a
and C= c a =c2—ab
b c (Note)
Hence from (I) we get

a b c 2 a2 —bc b2 -ca c2—ab
c a b, a2 —bc b2.-ac
b C a
bi - ca c2 - ab a2- bc
I Hence proved.
Ex. 11. Solve the equatIon 31-8 3 3 = 0.
3 3x-8 3
3 3 3x-8 (Meerut 97)
Sol. Given that 3x —8 3 3 1 = (J
3 3x-8 3
3 3 3x-8
3x - 2 3 3 0, replacing Ci by C1 + C'2 + c3
3x-2 3x-8 3
3x-2 3 3x-8
(3x —2) 1 3 3 = 0, taking out (3x - 2)common.
I 3x-8 3
1 3 3x-81
=(3x-2) 1 3 3 I O,repacingR2,R3byR2—Rl
0 3x - Ii 0 - R 1 respectively.
O 0 3x— 11
(3x - 2) 3x- 11 0 0, expanding with respect to C
0 3x-1l
(3x-2)(3r-- 11)2=0
=> x=2/3 or 11/3. Anzc.
Ex. 12. Prove that the value of ctcznnat
x+ I x-4- 2 x -i J is!epcndntofx,
+2 X-r3 +h
1x+3 +4
Sn!. The given detem1,rFafl

x+2 1 2 respct.vc1y
jxi-3 I
- a— I t,rer)acin9P2,RhyA2 R1,
O b—a—I Th--R1 rpccthe
2 0


192 Matrices 182/1/12

- I x + I a - I interchanging Cl and C2
o i b — a —i
o 2 c—a—2
I b—a—i I =—[(c—a-2)-2(b—a— I)]
2 c—a--2
[c_a_2-2b+2a + 2 ] =—a±2bC . which is independent ofx.
Hence proved
*Ex. 13. Give correct answer to the following:
The value of the determinant - 3 1 1 1
1-3 1 1
1 1 —3 1
1 1 1 —3
is (A) - 1, (B) 1, (C) 0, (D) 4.
Sol. The correct answer is (C) i.e. 0, since replacing Ci by
CI ± C2 + C3 + G4 we find that all the ekments of Ct are zero. Hence the value
of the given determinant is zero.
Ex. 14. Show that (a + b + c) and (a2 +b 2 + c 2) are factors of
determinant a 2 (b + c)2 be and find the remaining factors.
b 2 (c+a) 2 ca
c2 (a+b)2 ab
Sot. The given determinant
= 0 2 (b2+c2+2hc)+a2 bc ,replacing C2 by C2+Ct
h2 ( c2 +a2 +2ca+b2 Ca
+ 1,2 2 ab
C2 (a2 + 2ab) + c
= 2 b2 + c 2 + a 2 bc replacing C2 by C2 - 2C3
,2 c 2 +a2 +b2 ca

c2 a 2 +b2 +c 2 ab
= (02+ b2 + c2) a2 I bc taking out (02 + b 2 + c2) common
1,2
ca from C2
c2 i a1
- a2 + h2 +c 2 a3 a abc taking I/a I/b, 1/c common from
- abc b3 b bca RI, R2 and RI respectively.
c cab (Note)
2 + C2 I I
a2+b I a 1 taking out abc common from C3.
- xabc a 3
- abc b ii
c
Now proceed as in Ex. 31 Page 146.


18211/13 Misc. Solved Examples on Determinants 193

Ex. 15. Prove that a2 be ac + c2 = 4a2b2c2.

a2+ab b2 ac
ab b2 +bc c2
Sol. The given determinant
= abc a c a + c taking out a, b, c common from
a + b b a Ci, C2, C3 respectively
b b+c c
= abc a + c c a + c . replacing Ci by C1 + C2
a-s-2b b a
2b+c b+c c
= abc 0 c a+c, replacing C1 by Ci - C3
2b b a
2b h+c c
= abc 0 c a + c , replacing R 3 by R3 — R2
2b b a
0 c c—a

= — 2ab 2c c a + c , expanding with respect to Cj

c c—a

= — 2ab 2 c 0 2a , replacing R1 by R 1 — R2
C c—a

= — 2ab 2c (— 2ac), expanding the determinant.

=.
4a2 b 2c2 Hence proved.
*Ex. l& Solve x+2 2x+3 3x+4 =0
2x±3 3x+4 4x+5
1. 3x+5 5x+8 10x+17
Sol. The given equation is
x+2 —1 —2 =0, replacing C2, C3 by C2-2C1,
2x+ 3 — x - 2 - 2x —4 C3 — 3C1 respectively.
3x+5 —x-2 x+2
or x+2 — 1 — 20, 0, replacing R3 by R3 - R2
2x+3 —x-2 —2x4
--
x+2 0 3x+6
or x+2 — 1 0 = 0, replacing C3 by C3 — 2C2
2.r+3 —x-2 0
.r+2 0 3x+6
or (3x + 6) x+2 — 1 = 0, expanding with respect to C3
2x--3 —x--2 .
or (3x+6)[—(x+2)2+(2x+3)]0


194 Matrices

or (3x+6)(x2 +2x+1)0 or (3x+6)(x+I)2=0

or x=-1,-2 Ans.
*Ex. 17. Prove that CL I X x = f (x) - x r (x),
I 0 X

X Y I

xxx 8
where f (x) = (x - Ct) (x - ) (x - (x - 8) and f' (x) is the first derivative of
f (x) with respect to x.
SoL The given determinant
• cx x 0 0 , replacing C3 and C4 by C3 - C2
x - x - and Ct - C2 respectively.
X x y—x 0
X x 0 6—i

X-0 x—f —x x x-3 x— ,expandingw.r.toRi

X •f—x 0 x y—x 0
X 0 8—x x 0 8—x

=a P x - 0 1 - x' 1 x - 0 , replacing C3 by
X y—x x—y 1 y—x x—y C3—C2ineach
X 0 6 '— x 1 0 6— x determinant
=cz3 I y—x x—y I —cz(x—) x x—y—x 2 y—x x—y
o &—x 0 8—x
x 8—i
each det. w.r. to Ri
+x2 (x—) 1 x—', expanding
I 6—i
=(4—x2) (y—x)(6—x)—(x—)x(U—x) ((8—x)—(x—))]
= (cx —x2) (x - ) (x —8) - x (x —a) (x - 3) (x —6)
—x(x—a)(x—)(x—y)
=(_a—x+4—x2+x—x2+C1x)(x—y)(x-8)
_x(x—a)(x—)(x-8)—x(X—(x)(X—)(X (Note)
= [(x—u) (x—)—x(x—O) —x(x—CO](x— (x-8)
—x(x—CL)(x— )(x-8)—x(x—(Z) (x—)(x— (Note)
=(x—çz) (x-)(x— (x-8)—x[(x—)(x— (x-6)
+(x .a)(x—(x-6)+(x--a)(x—)(x-8)
+(x—cz) (x— 3) (x—y)]
=f (X) - if' (x), where f(x) = (x - a) (x - ) (x - (x - 8) Hence proved.
Ex. 1& Prove that 1 o3, where 0) is one of the imaginary
• ,
o i to
to 1
cube roots of unIty.


Misc. Solved Examples on Determinants 195

Sol. If w be one of the imaginary cube roots of unity, then
w3 =l and l+o+w2=O
Now the given determinant
= I from (i) using (03 1=
1 1(0
(02. co
i
= 0 W2 , replacing Ct by Ci - C2
0 1(1)
W2 CO (I) I -

(2 - ()) 1 2 , expanding with respect to C

lc
= (co2 _ (0 ) ( 0) _ o
c2 ) (034 _ 0)2 + J)3 o)2 +l , . coi
=2_((0+2)=2_(_l), ...
=2+1=3. Hence proved
**Ex 19. Prove that the determinant
• 1 cos (13 - a) cos (y - a)
cos (Ct - 13) 1 cos ( y— 13)
cos(cz — y) Cos Y) 1
is a perfect square (of a determinant) and find its value. (Gorakhpur 94)
So!. The given determinant
= cos a cos a+ sin Ct sin a cosa cos13+ sin a sin 13
cos 13 cos a+ sin 13 sin a cos 13cos 13+ sin 13 sin 13
cos y cos a + sin ysin a cos ycos 13 + sin y sin 13
cos a cos y+ sin a sin y
cos 13 cos y+ sin 13 sin y
cos ycos y+ sin ysin y
(Note)
The element in the first row and first column is cos a cos a + sin a sin a,
which can be written as
(cos a) (cos (X) + ( sin a) (sin a) + 0 . 0 (Note)
Similarly the element in the first row and second column is
cos a cos 13 + sin a sin 13, which can be writter as
(cos (x) (cos 13) + ( sin a) (sin 13) + 00, (Note)
Proceeding in this way we can write the given determinant
= cos a sin cx 0 X cos a sin a 0
cos 13 sin 13 0 cos 13 sin 5 0
cos y sin 7 0 cosy sin 7 0
02
= cos a sin Ct hence a perfect square of a determinant
cos 13 sin 13 0
cosy sin 0


16 Matrices 989
0, since the value of this determinant is zero as all the elements of one
of its columns are zero.
*Ex. 20. (a) Show that a + lb c + Id I )< I a + i3 y + i
-c+id a-lb -y+i a-W
can be expressed as A+IB C+ID
-C+ID A-lB
Hence prove the Euler's Theorem, 'the product of two sums of four
squares each is equal to the sum of the four squares'.
Soil .+ib c+id x a+i y+i
-c+id a-lb -y+iö a-if3
= (a+ib)(ci+i3) (-c+ id) (a+i)
+ (c + id) (y+ i8) +(a-ib)(y+i)
(a + ib) (- y -r i6) (- c + id) (- y+ i)
±(c+ id) (a-1f3) +(a-ib)(a-i)
= (aa-b+cy-d) (-ca-d+ cry +M)
±i(aj3+ba+c+dy) -I--i(-c13+da+a-by)
(-ay-bö+ca-i-d3) (aa-b+-d8)
+i(a- by- c+da) +i(-a13-ba--c6-dy)
A+iB C+iD
-C+iD A-lB
where A =aa-bj3+c'y-dö, B-a3 +ba+c+dy,
C -Y+b-ca-dt3,D=aö-fr(-c+da ...(ii)
Now a+ib c+id
-c+id a-lb
=(a-+-ib)(a-ib)-(c+ id) (-c+id)
(a2 - i2b2) - ( i 2d2 - c2) = a2 + b2 + c2 +
I
Similaryly a+i3 y+iö =cz2+j32+y2+52
-y+i (X-i5
and A+iB c+iD
I -C-4-iD A-lB
.. From (I) we have
(a2 + b2 + c2 +d2)(cz2 + 2 +y+ 2)=(A 2 +B2 +C2 +D2) (iii)
i.e. product of two sums of four squares each is equal to the sum of four
squares. Hence proved.
Ex. 20 (b). With the help of determinants express the following as a
sum of four squares
(12+2 2 +32 42)(52+62+7 2 8)
Sol. As in Ex. 20 (a) above we can show that,
(a2+b2+c2 + d2)(a2 + 2 +y+ 2)(A 2 +B2 +C.2 +D2) (i)


Misc. Solved Examples on Determinants 197

where A, B, C. D are given by Ex. 20 (a) result (ii).

Now Ieta2, b2, c3, d=4, a=5, 3=6, i=, 5=8
ThenAarx-b+C1-d6 1(5)-2(6)+3(7)-4(g)
=5-12+21-32=-18.
B=a3+htX+c+dY 1 (6)+2(5)+3(8)+4(7)
=6+10+24+28=68,
Cay+b-cU-d3 1 (7)+2(8)-3(5)-4(6)
=7+6-15-24=-- 16
D=aS-1Yy-c13+dU 1 (8)-2(7)-3(6)+4(5)
=8-l4-18+20=-4
2 +32 +4 2 ) ( 52 +6 2 + 72 +8 2
From (I) above we have (1 2 +2
= ( 18) +(68) 2 +( 16)2 +(-4) 2
(18)2 +(68) 2 +(16)2+ (4)2. Ans.
Ex. 21. Evaluate a -a - a -a
b b -b -b
C C C -c
d d d d
Sol. The given determinant
a 0 0 0 , replacing C2,C3andC4bYC2+C1.C3+C1
b 2b 0 0 and C4 + Ci respe tively
c 2c 2c 0
d2d242d
= a 2b 0 0 expanding with respect to RI
2c 2c 0
2d U 2d
= 2ab 2c 0expanding with respect to RI
2d24
=2ab[2cX2d-(2MX0]=8(thcd. Ans.
*Ex. 22. Prove that 4 5 6 x = - Zy + z)2
567y
678z
x y z 0 (Gorakhpur 94; Kanpur 94)
Sot. The given determinant
= 10 5 6 x , replacing ClbyCl+C3
12 6 7 y
14 78z
x+z y z 0
= 0 5 6 x , replacing Ci by C1 - 2C2
0 67y
0 78z
x-2y+z y z 0


198 Matrices
=—(x-2y±z 5 6 x I.exP * nding with respect to Cj
I

8
=— (x - 2y+ z)112 14 x+zI. replacing R1byR1+R2
• 167 y
8 z
= -(x-2y+z) 0 0 x+z-2y !,replacingR j byRi-2R2
• 6 7 y
7 8 z
— (x-2y+z) 2 J6 7 l,expanding with respect to R1
I 8!
= - (x - 2y + z) 2 [48 - 49}, expanding the det.
= (x - 2y +
Hence proved,
Ex. 23, Evaluate 0 a J3
jl 0 c —bj
0 a -C
n b -a 0 (Gorakhpur 95; Ka'pur 90)
I
Sol. The given determinant
.1 0 a 13 taking (1/a) common from R2
a1al 0 ac —ab1 y I
—c 0 a
b —a 0
(Note)
- I 0 a P Y
aa1+bm+cm 0 0 0 replacing R by R2 ± bR3 + cRj
M —c 0 a1 I'
n b —a 0!

=— -(al +bm+cn) a 13 yI expanding with respect fo2

•a I—c 0 aJ
b —a 0

(a! +bm+cn) aa $3 T I, taking out 1/a common fromC1

1 —ac 0 aj
ab —a 01
(a1 + brn+cn) faa+b$3+cv $3- y1,
a. 0 0 aJI
o —a

replacing C 1 by Cj + bC2 + cC3

----4 (al +bm +cn)(aa+b$3+cy) 0 a Iexpandingwithrespect
a 1—a 0! toCj


199
Misc. Solved Exs. on Determinants
= _ (l /a 2 ) (al +bm+cn)(+
Ans.
=_ (al +bm+Cfl)(aa
Ex. 24. Prove that 1 1 1 1

a3 I3

= a) (a - ) (a -6) (1 -6) (y - &)

- y) (y-
= aT + pr + +
Hence evaluate $0 SI S2 $3 'where Sr
si s2 S3 $4
S2 S3 S4 55
S3 S4 $5 S6

sa 11 1
a 5
a22282
p3 83

' replacing C2, C3, C4 by

0 0 C2 - Ci,
0
= 1
a -a y-a 5-a C3_CI and C4-Cl
• ,22 82_a2
cxl 2_a2

3 a3
a3 0 y3 -a3 53-a3
1 1
5+a
+cr1' 52+a2+cth
32 +a2 +al3 y+a2
expanding w.r. to Ri and taking
out common factors.
0 0
1 - S
+_2 62+_2
j3 2+ a2 +a (3 -a(3
replacing C2, C3 by C2 - Ci, C3 - Ct
0 0
1
2 +a2 +ap (r-(3) (ö-1)
p (a+(3+y) (a+(3+&)
1
ct+13+y a+ +5 0
^1

expanding w.r.t. to'Ri

=(a-13) (a-5) ('y-a) ((3 -((3-6) [(a+(3+S) -(a+(3+y)]
 200 Matrices 182/1113
=-(a-f3)(a-S)(y_(x) (3-7)(i3-6)(y-6).
Hence proved.
Squaring both sides of (I), we get
((x )2 (a - 6)2 (y )2 ( - )2 (3 - 6) 2 (y_ 6)2
= 1 1 I 1• j 1 1
a a
a2 2 72 62 a2 2 72 82
a3 33 y 53 a3 p3 Y3 53
(Note)
1+1+1+1
- a+±y+& a2+p2+y2+62
a2+2+y2+52 a3+3+y+6
a3++y+3 a4+t34+y+8
a2+2++2 a3+3++6
a3+3+y+6 a4+i34+7+8
a4+p4+y+6 (X5++'y+6

a5+5+y+6 a6+6++66
So S2 S3 ,
si
SI 52 S3 54
where s,=a r + r +YT+6r
S2 53 S4 S5 and so a° + 30 + +
53 54 55 S6 - 1 + 1 + I + 1,etc.
Hence proved.
*Ex- 25. 11(0 is the cube root of unity, then one root of the equation
x+1 Ci)
(02 OisO.
x+(02 1
0)2 1 x-f-o) (MNR9O)
Sol. Adding all the rows of the given determinant to first, the given
equation reduces to
x+1±(j+(02 x+U)+(02+1 (02 +1+X+(0 0
(02
1 x+o
or x xx 0, 1+(0+0)2=0
(I) X+002 I
2
Co 1 X+0)

or x i 1 1 =O=x=O
0) X-I-0)2 i
0)2
x +
Hence proved.
*Ex. 26. The value of 9 which lies between 0 = 0
satisfy the equation and 9= it/2 and


Misc. Solved Exs. on Dete ants 201

18211113
1 + sin2 B Cos 2 0 4 sin 40 =0 is
sin 2 e 1 + cos 2 B 4 sin 40
sin 2 0 cos2 0 1 + 4 sin 40
(LL?)
(a) 7it/24, (b) 5it/24, (c) I1ic/24, (d) it/24.
Sol. Given equation is
i + sin2 e + cos2 e cos2 0 4 sin 40 = 0,
4 sin 4O
sin 2 O+l+ C0s2 0 1+ Cos 20
Cos 2 0 I + 4 sin 40
sin o + Cos e
adding 2nd col. to 13t.
or 2 cos2 0 4 sin 40 = 0, Cos e + sin 2 0 = i
2 14-cos2 O 4 sin 40
I Cos 2 0 1+4 sin 4B

or 2 cos 2 O 4 sin 4O =0, replacing R2bYR2"Rl

0 1 0
I Cos 20 1+4sin419
or 2 4 sin 40 = 0, expanding w r. to R2
I 1+4 sin 4O ^

or 2(1+4 sin 4O)-4 sin 40=0 or 4 sin 4O+2O

or sin 40 = - 1/2 = sin 210 0 or sin 330°
= sin (7x/6) or sin (1 1,t/6)
or 40=7t/6 or llir/6 or 0=7it/24 or llx/24
Hence the required values of 0 are given by (a), (c). Ans.
Ex. 27. If A t = x b b and L2 = x b ahe given
ax b ax
a ax

determinants then show that dx it = 32 (M.N.R.)

S l. -4-1
o
100+ x b b + x b b I I

dx a x b 010 a x b
a ax a ax 001
=xb+xb+xb,
ax ax ax
expanding 1st, 2nd and 3rd determinants
w.r. to 1st, 2nd and 3rd row respectively.
=3 x b =3E2
la x Hence proved.


202 Matrices 989

**Ex 28. The value of determinant a b aa + b
b c bcc+c
laa+b bcx+c 0
is zero if
(a)a,b,c are an H.P.
(b) a, b, c, are an G.P.
(c)czLsa lxotof the equation ax2++Ø
(d)(x - a) is a factor of ax2 +2bx +c
(/.LT)
SoLIf a b aa+b0,
b c ba+c
aa+b ha-c 0
then a b 0=0,
b c 0
aa+b ba+c —a(acz+b)—(ba+C)
replacing C3 by C3 - aC; - C2
or —(acx2 +2ba+c) a b 0
1b
or (act+2bd+c)(ac_b2)0
i.e. either b2=aoraa2+j,+=Ø
If b2 = ac, then a, b, c are in GP
Hence result (b) is true.
If aa 2 + 2bcz + c .= 0, then Ct is a root of the equation ax2
+ 2bx + c =0 or
(x — cz) is a factor ofax2+2bx+c
Hence result (d) is true.
Ans. (b) and (d).
EXERCISES ON CHAPTER IV
*EL 1. If 2s = a + b + c, prove that
a' (s—a)2 (s —a)2s3(s—a)(s—b)(s....c)
(s—b)2 b 2 (s—b)2
(S—C) 2 (s—c)2 2
Ex. 2. If a + b + c =0, solve the equation
a—x c b =0
C b—x a
b a c—x
[Hint ,,Add all the rows or columns]
Ex. 3. Show that a a2 a3 + bc (a - b) (b - c) (c - a)
b b2 b3 +ca (ab + bc + ca + abc)
?


Exercises on Determinants 203

S EX. 4. Show that there are three values of t for which the system of
equations
az = 0, cx +av + (b - t) z = 0
(a - t) x + by + c17 = 0, bx + (c - 1) +
have a common non-zero solution. If the three values of tare fl, t2,13

show that ti (2 t = a h c
bca
cab

Ex. 5. Prove that 0 x y z = ( + b y + cz)2

-x 0 c -h
-y -c 0 a
-z b -a Di
2ab - 2b = (1 +a 2 + b2)3
EX. 6. Show that 1 + 2 - b2
2ab l - a+b
2b -2a 1+a2-b2

Ex. 7. Solve x -6 - = 0
2 -3x x-3
3x x+2 (Meerut 92)
-3
Ans. x-3,2, I
Ex. 8. Show that the roots of the following equations are all real
a+x h g =0,
h b+x f
g f c+X
where a, b, c,f. g, h are all real numbers.
Ex. 9. Solve I x -6 -I=0
2 -3x x-3
-3 Zr x+3
*Ex. 10. By the product of determinants estabilish Eulier's theorem that
the product of any two sums each of four squares is expressible as the sum of
+2 2 + 32)?
2 )(1 2
four squares. Does the theorem hold for (32+4
Hence express (92 + 2 + 32 ± 42) (52+6 2 +7 2 +82 ) as sum of four
squares.
(Hint. See Ex. 20 (a) and (b) Pages 196-97.
Also (32+42) (12+2'+3 2) (32+42+02+0 2) (1 2 +2 2 +3 2+02

Now proceed as in Ex. 20 (b) Page 197.

(32+42)(12+22+32)_52+(l0)2+92+(12)
Ans.
(92+22+32+42)(52+62+72+8)
and
= (22)2 + ( 116)2 + (40)2 + (f3)2
204 Matrices 182/1113
Four possible answers for the following questions are given. Choose the
correct answer
Ex. 11. The value of the determinant 1 2 3 1 is
357
8 15 20
(a) 20, (b) 10, (c) 2, (d) 5. (M.N.R. 9/)
Ans. (c)
Ex. 12. The value of the determinant I a b + c is
I b c+a
I c a'+-b
(a) a + b + c, (b) I, (c) 0, (d) abc Ans. (c)
Ex. 13. If = ai bi ci 'then A is equal to
a2 b2 C2
03 b3 C3
(a) - bjB1 - b2B2 - b3B3; (b) - biBi + b2B2 - b383;
(c) bB1 - b2B2 + b38 3 ; ( d) b 1 B 1 + bzB2 + b3B3 Ans. (d)
Ex. 14. The value of the determinant 3 6 12 1 is
5a 5b 5c
a b c
(a) 15, (b) 2. (c) 0, (d) 4. Ans. (c)
Ex. 15. The cofactor of a in the determinant 3 4
5 is
789
a b c,
(a) 4 5 . (b) 1 3 4 , (c) 1 3 5 1 (d) None of These
89 78 79
Ans.(a)

OBJECTIVE TYPE QUESTIONS

CH. TOIV

(A) SHORT & VERY SHORT ANSWER TYPE QUESTIONS

I. I)cfine matrix. ( A.ao,iuir 2001) Se § 1-02 Pate 21
2. Dchne a rectangular and a sivarc fiat rex. jSec § I AB
3. What are horizontal, and vertical mat rices 7 See § 1-0.1 P. 41
4. What are row a nil coluiu ii vctors ISee 1 . 03 P. 41
5. What k 1 4 7 ]See § I 03 P.
(. Dctine an unit matrix. See § 1-03 P. 4
7. Define a diagonal matrix. ISce § I (13 P. 51
S. What do you understand b y a th-mai rex 7 jScc § I -03 P 51
. Write down the properties of 11`1a0 11X additum. See § 1 07 P. ]
UI. If A, H, C he three matrices ut the same order and arc such that
A + B A ± C, I lien show thai B C. 15cc Prop. VI Page 5]
11. \Vlien are the two matrices conformable to multiplication 7
See § 1-08 Page 11]
12. Give an example to show that the product of two nun-zero matrices
can he a zero mat lix. See § I - OS Note 2 Page 131
3 12 1 4
3. If A = t) 1 1 and B = 2 2 then does BA exist 7
I 2 (1 1 0
Ans. No.
1 1 3
14. If A = 2 2 6 , then show that A 2 = 0.
—3

a I, [x
IS. Ilk = I. v , y , : 1 B It / then C = show
, that
H
f -
g J c

- BC 2
I =ax +b+cf+2ievi+2z.v+2Jvz.

41
2 3
10. it A calculate integral ro\verS of A.
= IL 2 1 1
3 4
(['arvanchal 97)
17. When are two matrices said to be equal ? ( Pwvanchal 2001)
15. Define null matrix. (P -...... . hal 2001) ]See 6 1-03 Page 4
19. Define transposed matrix. 208 Page (>)]
21l. Define Nilpotcnt matrix. Isce 5 2 ­07 651
1G2/tAZO

-14
2 Matrices

21. State reversal rule for the inverse of product of two matrices,
(Pun'an ch a! 9) jSec § 2 () Th . It Page 92
22. Show that AR B\ where
- Cosf) Sif91 - cuse, —sin6-
A
sin 01 CON - Sin ()2

(/1ecni( 2001)
23. State reversal rule for the transpose of a product.
15cc § 209 Th. IV Page
24. Describe elementary row operations on a matrix.
(Purranchal 99) ISce § 301 Page 1031
25.-Define cofactor of an element of a determinant.
jSec § 40 1241
26. Define minor of an element of a determinant.
Isce § 407 P. 1251
• I a h-f-c
27. Show that I b c + a = (1 15cc Ex. 2(u) P. 1321
I c -f-h

—2 1 1
• 28. Evaluate 1 —2 1 Ans. ft
1 1 —2
1 1 1
29. Show that 1 1 +.x 1 =xy. ISCC Ex. 7. Patc
1 1 1+)'
X a a
30. Evaluate a x a . Ans. (x + 2a) tx -
ci a x [Sec Ex. 20. P. 1401
1 1 1
31. Show that U b C = (a - h) (h - c) (c - a).
a 2 h 2 c 2 15cc Ex. 28. P. 1451
1 1 1 1
I 1 +a 1 1.
32. Evaluate • Ans. abc.
1 1 1 + h 1
- 1 1 1 1 + c [Sec Ex. 36 P. 149
2
0 c h 2 + h2 ha Cci
33. Show that c 0 a ab C2 + a2 bc
h a 0 ac hc hh+2

ISee Ex, 1, P. 163J

Dc b
34. Express c 0 a as a determinant. (Puivanclwl 97)
ha 0
 Objective Tpc Questions Ch. I to IV 3

(B) OBJECTIVE TYPE QUESTIONS

(I) MULTIPLE CHOICE TYPE
Select the correct answer of the follQwing
2 3
I. The order of the matrix 41 i s -
. 7
15
(i) 3 x 2 (ii) 2 x 3 ; (iii) 3 ; (iv) none of these.
2. A matrix A = I aj , X;) is called a square matrix, if
(1) in ii ; ( ii) 'U it ; (iii) in = it ; ( iv) none of these.
3. A matrix A = ,, is calicd it vertical matrix, if
(I) in > it ; (ii) iU = it ; ( iii) in < it ; ( iv) none of these.

matrix A = aj i,,, X ,j is called

4. A niatrix it horizontal matrix, if
III
(1)/n > it ; (ii) in = it ; ( iii) it ; ( iv) none of these.

5. If a is the number nI elements in it row and h is the number of

lements in it of it A. then the order of A is
(I) a xl; - (ii) h X a; (iii) axa; (iv) hx b-
6. A matrix A = a j Im x, is called a row matrix, if
(j)m<n; (ii)nt=n; (iii)m=1; (iv) 11
[i
. ] is called a

(i) column matrix; - (ii) row matrix;

(iii) null matrix; (iv) unit matrix.
. The matrix I a ij x, will he an unit matrix, if
(I) all its mit elements are unity;
(ii) in = it and all elements are unity;
(iii) in = ii, diagonal elements are unity and other elements are zero;
(iv) none Of ilicsc.
9. The negative of it matrix A is
(i) zero .; (ii) —A ; (iii) +A ; (iv) non-cXisttnt.
'(Kanpur 2001)
10. It A is of order 2 x 2 and B is of order 2 X 3, then BA is of
order
(i) 2 x 3 ; (ii) 3 x 2 ; (iii) 2 x 2 ; (iv) none of these.
11. A is any in x it matrix such that AB and BA are both defined,
then the order of B is
(i) in x in ; ( ii) U X it ; (iii) it X nt ; (iv) none of these.
a
12. If A = hI and B = [b ], then AB is
[
(i) null matrix ; (ii) unit matrix
(iii) vcrilcal matrix ; (ii') none of .these.


4 Matrices

13. If A = [ ani B = 0 7 then

(i) AB = BA ; (ii) AB dcs not exist but BA cxists
(iii) exists but BA does not exist
AB
(iv) AK BA.
•
14.11 A = I
H 01 2.
I. then A is it
[0 Oj
(i) diagonal matrix - (ii) unit mat rix
(iii) null mi( rix (iv) none of these.
• 1 2.2
15.IfA= 2 1 2 ,the value ofA2-4Ais
22 1
(I) I , (ii) 2 I, (iii) 4 1, (iv) 5 I.
16.* Transpose of it column 'matrix is
(1) square matrix ; (ii) column matrix
(iii) row matrix ; OvI none of these.
It I
17. If A = - I 0 5 , then A' =
2 —5 tt
(I) A (ii) - A ; (iii) 2 A ; ( is') none of these.
(kaiipur 20011'
18. If 'A is a square matrix, then A + A' will he
(i) diagonal (ii) symmetric
(iii) skew-symmetric ; (iv) identity matrix. (Kwmjmr 2001)

19. The matrix 0 1 0 ix

I) 0 1
(I) unitary ; (ii) idempotcnt
(iii) nilpotcnt ; (iv) identity.
20. 1n an upper triangular matrix, the elements a ij = 0 for
(I) 1 >; (ii) I = j ; (iii) 1<]; (iv) / j.
21. A matrix A is called involulory if
2 =0
(i) A I = -A (ii) A 2 ; (iii) A = I (iv) none of these.
22. Which of the following is a scalar matrix !
fa 0.0 a 0 0 0 a 0
(1) 0 b (I ; (ii) 0 a 0 ; (iii) a (1 1)
[0 1) b 0 I) a 0 0 a
(iv) flOflC of t licsc.


Ohjectie T\'pc Questions Ch. I to IV 5

23. 11w be the cube root of unity, then the value of the determinant
8
w
2 is

(ii) () (iii) —1; (iv) none of these.

24. Ii = 1, then the value O f the dete'rminant

3 2
I cii (I)

1 is:

W W I

(i) (I; (ii) 1 (iii) 2; (iv) 3.

I I 1
25. The determinant U I) . C is divisible by
2 2 2
(I C

(I) a — b; (ii) ü +h (iii)ab ; (iv) a/b.

x= 2, the value of x is
26. It.2

(i) 1; (ii) 2 ; (iii) 3 ; (iv) 4.

27. II two rows of i determinant ae proportional, the value of the
del erniinanl is
(i) infinite ; (ii) not zero ; (iii) negative ; (iv) zero.
2—x 3 3
28, One of the i-oots of 3 4 - x S 0 is
3 5 4—x

(ii) U ; (iii) 1; (iv) none of these.

h
29. II each element in the matrix is doubled, the value of
LC ilj
the determinant of the imitrix is
(I) doubled ; (ii) unchangc
(iii) multiplied by 4 ; (iv) none of I hc.e.

30. A = I aii I is a determinant of order three ind A' denotes the

cotactors of uj in A, then which of the following is not c i) ' reeL ?

(I) a1A 1 = A; (ii) u1 4(2 = 0;

(ii) (iv) ajA 0. (Kanpur 2001)
A - = 0 ;
31. If two rows of a determinant are interchanged, the value of the
determinant
(i) remains unchanged
(ii) Is negative of the vlue of original determinant


6 Matrices

(iii) doubles ; (iv) none of these.

32. The system of linear equations can he solved easily b y the rule
of
(i) Newton ; (ii) Bessel ; (iii) Cramer iv) none of these.
(II) TRUE AND FALSE TYPE
Write "T" or "F" according as the statement is true or false
1. The elements of matrix may he scalar or Vector q Ll;iIlt U ies
2. Matrix denotes a number.
3. -The order of the matrix 121 is 2 x I.
4. If in it matrix, the ii unihr of columns is more than I he number
of rows, then it Lc called a horizontal matrix.
S. In a row matrix there is only one column.

6.L j is an unit matrix.

• V () 0
7. 1) 2 (1 is called a diagonal matrix.
I) () —5
8. It. is not necessary For the two matrices A and B to be l the same
order so as to be conformable for addition and subtraction.
9. Commutative law holds but associative law does not for addition
of matrices.
10. Addition for matrices obeys the distributive law.
11. The product of two non-zero matrices can he a zero matrix.
12. Matrices of different orders can be subtracted,
13. Matrix iiiultiplicatioll in general is commutative.
14. Commutative law does not hold for addition of nlatrice,\.
15. The transpose of the transpose of it matrix is the matrix itself.
16. A square matrix A = aij J is symmetric it a,j = ajj for all values
of i and j.

17. If A = - J.J , then show that AA' is a symmetric matrix.

L
ISi.e Ex. 2 Page 2l
18. The inverse of a matrix is not unique. ISee § 2 Pt Th. I P. 921
19. If A is an orthogonal matrix, then A' is idso orthogonal.
20. If A is an unitary matrix, then A is not an unitary matrix.
21. A matrix I ai,, I js called a triangular matrix if a,1 = () for i >
22. Non-square matrix hit ,, no inverse.
23. If two rows of a determinant I A I are identical, then I A =
24. A square matrix A is singular if and only if 1 A 0.


Objective Type Oticstions Ch. 1 to IV 7

2 2 '
(a -(h - x (c -.v)
. t) I - 2r x u a I
- )2 (e - v) 2 = —2v b 1, I
25. ' b - v) 2
(a :) (b - z (c - 1 —2: z c c I

26. A deternunatit can be expanded using an y row or column of the

dcter,niii,tflt
27. The value iii a determinant changes it' the elements (if' a row .arc
tskled to or suht r.iied Ironi the corresponding elements ol another row,
2M. II each element of a column ol a determinant be multiplied by
s'iiiie cotistailt, rhen the determinant is mull iplied b y hint constant.
29. The value ( ') a determinant chiiigcs ii its rows and columns are
nit e rch ii ge (1.
30. The value oh a determinant changes in sign if two consecutive
rows (or colu iii is) arc i ntcrchanged.
ililt FILL IN TIlE BLANKS TYPE
liii in I lie blanks in the following
Each of. the inn 11 utithers constituting an 'ii X n matrix is known
as an ...............ol the matrix.
2. The plural of the word matrix ' is ................
3. The order of the matrix 131 i
4. A matrix which is not a square matrix is known as a.............matrix.
5. 11'171 = I in the matrix A = a 1 1,,,Ihen it is called a
iii it r ix.
6. If in > it in the matrix A =aijlui , , then it is called a ...............
matrix.
[1 ol . -
7. IS called ............... matrix.
I

is called a ...............matrix.
]
9. Two matrices are conformable for addition and .subtracl ion if they
are of the ...............order. -
10. Matrices of different orders ...............be added.
II. Additive identity ...............for addition of matrices.
12. 11 B be the additive inverse of the matrix A ajj it then

(i, j'gh element of B is ................

13. If A is an in X n matrix and B is an it x k matrix, then AB is an
....... . matrix.
14. If AB = - BA, the matrices A and B arc said to ................
iS. Multiplication of matrices is ...............with respect to matrix
addition


8. Matrices
182/1/10 0

16. Negative matrix is obtained by niultiplying it by

17. A matrix wl\cn added to its ncgal ivc gives he . nial rix.
18. A matrix when niultif)lied by . uives the null matrix.
19. 11 A is skew-H ermil ian ilial rix, the ii iA is
21). II a square matrix A is idenipotent I hen ................
21. The square matrix A = J aj1 J is skew-synimet nc, if (I,
for all VaILICS Of / and j.
22. II A is a Hermitian matrix, then íA is ............... . A/een(( 2(101)
23 11 A is an orthogonal matrix, Olen A i ................
24 a b u
2 ,2
u _ = ............... ( a - h) (b - c) (c
3 3 '3
(1 1) C

I I I
25. C
=( a - I) (b c) (c - a).
.3 1 3
ci b c
26. II any two columns of a determinant are ............... , then a Inc
of the determinant is zero.
27. The s y stcni of linear equations can easily he solved b y
the rule
of..................
28. A determinant can be expanded using ...............row of it.
Answers to Objective Questions
(I) Multiple Choice 1'pe
1. (ii) ; 2. (iii) 3. (i) 4. (ii i ; 5. (ii) ; 6. (iii) ; 7. (iv)
9. (ii) 10. (iv); 11. (iii); 12. (i) 13. (iv); 14. (iii); 15. (iv); 16. (iii);
17. (ii) IS. (ii) ; 19. (iv) ; 29. (i) 21.
; (iii) ; 22. (ii) ; 23. (ii) ; 24. (iv)
25. (i) ; 26. (iv) ; 27. (iv) ; 28. (i) 29. (iii) ; 30. (iv) ; 31. (ii) ; 32. (iii).
01) . T-ue and False Type

12.F;I3.F;14.F;I5T.I6T;17T;18F;19T;29F21T
22. T ; 23. T ; 24. F; 25. 1 ; 26. T ; 27. F ; 28. T ; 29. F; 30. T.
(III) Fill in the blanks type
1. element ; 2. matrices ; 3. 1 x 1; 4. rectangular ; 5. row; 6. vertical
7. an unit ; 8. diagonal ; 9. same ; 10. cannot ; 11. exists
12. -aj,
13, in x k ; 14. aiiticornruutc ; 15. distributive ; 16. --1 ; 17. null ; IS. zero
19. Hermitian ; 20. A 2 = A ; 21. -a11 . 22. skew- Hermitian
23. also
orthogonal ; 24. abc ; 25. (cz + h + c) ; 26. identical ; 27. Cramer ; 28. any.
Chapter V

Rank and Adjoint of a Matrix

§ 5-01. Order of a minor.
Definition. If any r rows and any r columns from an in x ii matrix A are
retained and remaining (m - r) rows and (n - r) columns removed, then the
determinant of the remaining r x r submatrix of A is called minor of A of
order r.
For example : In the matrix
I'll all a 11 014
02 1 0 22 1723 024
a 31 a 32 a 33 a

0 41 042 043 044

051 052 053 054

elements a 11 , 0i2, 413j, etc. are minors of order unity

07 a 11 013 (233 014 etc.
(2 21 022 02, °23 ' (254

are minors of order 2

a t l 012 (1 13 a2 l 0 23 024 etc.
CJ, a22 (123 . a41 (143 044
-, 0 a 53

re minors of order 3
l uh ll L7 (213 14 'l 22 '23 0 24 etc,
1. 21 ( 1 22 123 24 3i a12 a33 1134
04 042 0 43 0 44 04 , a42 (143 (144
a 5l 11 52 0 5 3 (15.1 (1 51 1152 053 (154

minors of order 4.
Note. In the UtOVC example there cannot be any minor of order higher than 4.
502. Rank of a matrix.
Consider the matrix A = 1 2
234
357
98; L,ucknow 91)
(4vadh 97,Garini'aI 90; Gorakhpur

This matrix A has only one three-rowed minor i.e. minor of' order. 3, viz.
1 2 31 and its value, can easily be calculated to be hero, by expanding with
234
3 5 7,
rcSpeCt to first row, 182/11/1


2 Matrices
This matrix A has 9 minors of order 2 (Or two-rowed minors) znd one of
them is 3 4 which has the value
.57
(3x7)—(5x4)=21 —20=1 *0.
This fact that A is a matrix whose every minor of order 3 is zero and there
is at least one minor of order 2 which is not equal to zero is also expressed as
'the rank of the matrix A is 2'.
"Definition of Rank of a Matrix:
(Avadh 92 ; Bundelkhand 6, 95, 94; Purvanchal 98, 96: Rohilkhand 92)
ij in an m X n matrix A, at least one of its r x r minors is d4fferentfro,n zero
while all the minors of order (r + I) are zero, then r is defined as the rank of the
matrix A.
A number r is defined as the rank of an m x a matrix A provided
(i) A has at least one minor of order r which does not vanish and (ii) (here
is no minor of order (r + 1) which is not equal to zero.
Note 1. The rank of a matrix A is also denoted by p (A).
*Note 2. The rank of a zero matrix by definition is 0 i.e. p(0) =0.
Note 3. The rank of a matrix remains unaltered by the application of
elementary row or column operations i.e. all equivalent matrices have the same
rank.
**Note 4. From the definition of rank of a matrix we conclude that
(a) If a matrix A does not possess any minor of order (r + 1) then p(A) 5 r.
(b) If at least one minor of order r of the matrix A is not equal to zero, then
p(A) ^ r.
Note 5. If every minor of order p of a matrix A is zero then every minor of
order higher than p is definitely zero.
Solved Examples on § 5.02.
Ex. 1 (a). Find the rank of the matrix A 1 2 3
258
4 10 18 (Gorakhpur92)
Sol. The determinant of order 3 formed by A
= 1 2 3 1 0 0 replacing
C2, C3 , by
4 10 18 C2 - 2C 1 , C3 - 3C 1 respectively.
42 6
=1 2 =6-4=2;10
26
p(A) ^! 3.
Also the matrix A does not possess any minor of order 4 i.e. 3 + 1, so
p(A):53 ...(ii)
Froln(i)and(ii)wegetp(A)3 Ans.


Rank of Matrix 3
Ex. 1 (b). Find the rank of the matrix A = 1 2 3
234
4 10 18
Sot. Do as Lx. 1 (a) above. A. 3
Ex. 1 (c). Find the rank of the matrix A = 1 3 2
223
154
Hint: Do as Ex. 1(a) above. Ans.3
Ex. 2 (a). Determine the rank of A = 6 1 8 3

O
2 1 0 2
4-1-8-3
So!. The given matrix A possesses a minor of order.3 viz.
61 = 10 0 0 replacing R 1 , R2 by
8
2 1 0 6 0 -8 R1+R3,R2+R3
4-1-8 4-1-8
=10 0 -8 =10(0-8)=-80*0
-1 -8
p(A)>3.
Also A does not possess any minor or order 4 i.e. 3 + 1, so
p(A)!^3. ...(ii)
From (i) and (ii), we get p (A) = 3. Ans..
Ex. 2 (b). Find the rank of the matrix
A=1 3 5 1
2480
3175
Hint: Do as Ex. 2 (a) above. Ans.3.
*Ex 3 (a). Find the rank of matrix A = 1 -1 3 6
1 3-3-4
5 3 3 11 (Kanpur 95)

Sot. The given matrix A possesses a minor of order 3

viz 1 3 6 - 2 0 2 replacing R 1 , R3 by
1 -3 -4 1 -3 -4 R2 + R I , R3+R2
5 3 11 6 0 7
=-3 2 2 =-3(14- 12)=-6*0
67
p(A)^!3.
Also A does not possess any minor of order 4 i.e. 3 + 1, so
p(A)<3 ...(ii)
From (i) and (ii) we get p (A) = 3. . Ans..


4 Matrices 999
Ex. 3 (b). Find the rank of the matrix A 168
253
(Purvc,nc/ia/ 96)
SoL The determinant of order 3 formed by A
= 1 6 8 = 1 1 6 8 replacing R 2 , R 1 by R 2 - 2R1,
2 5 3 0 —7 —13 R 3 - 7k 1 repectheiy
7 9 4 0 —33 —52
—7 —13 = -7 —13 replacing
1
R2 by R, - 4R1
—33 —52 —5 0
=- 65
p(A)-^!3.
Also A does not possess any minor of order 4 /.e.3 + 1, so
p(A):53 ..(ii)
From (i) and (ii), we get p (A) 3
A ns
Ex. 3(c). Find the rank of the matrix A 2 3 8
506
*
8 9 10 (Puri'anciza/ 94)
So!. The determinant of order 3 formed by the matrix
A
= 2 3 .8 = 2 3 8 replacing R 3 by R 3 -- 3R1
1,

506 50 6
8910 20-14
=- 3 1 56 , expanding w.r. to C2
2 —14
=-3(-70-- 12) --- 3x82=246O
p(A)>3 -
Also A does not possess any matrix of order 4 i.e. 3 + I and
- p(A)<3. ...(ii)
From (I) and (ii) we get p (A) = 3
An;.
E. 4 (a). Fiid the z-aiik of tile matrix A = 1 2
2 3 1
—2 —3 —1
SoL The deterniinan of order 3 formed by !his Inalrix A
= I 2 3 •= 1 2 3, replacing R 3 by R + R2

=J: - -
Also there exists a minor of order 2 of A.
9


Rank of Matrix 5
viz. I 2 =3-4=--IeO
23
}lcnce the rank of the given marix A is 2. Ans.
Fx. 4 (h) Find the rank of matrix , = 1 3 4
L2 6 8
Sol. A minor of order 2 formed by this matrix.
= 1 3 = 6 -- 6 = 0. Similarly all minors of order 2 are zero:
26
Now we are let', with minors of order t it'. elements of A which are not
equal to zero.
Hence the rink of the given matrix A is 1. Ans.
Ex. 5. Find the rank of the matrix.
r1 2 3 1
2462
1 2 3 2
(Gorakhpur 96)
Sol. In this matrix, a minor of order 3
= 1 2 3 0. R and R 3 are identical
246
123
In a similar way we prove that all the minors of order 3 are zero.
Now a minor of order 2 = 1 2 1 = 0.
2 4
But another minor of order 2 = 3 I 0,
34
Hence rank of the given matrix is 2. Ans.
Ex. 6. Find the rank of the matrix A = I -3 2
3 -9 6
-2 6-4

Sol. The determinant of order 3 formed by this matrix A.

1. - 3 2 = 1 0 replacing C2. C3 by
3 -Q 6 I 3 0 0 ,C2+3Ctan&1C3-2C3
6 -4 -2 0 0 1 respectively.

= (j_
Also there exists no minor of order 2 of A which is not equal to zero.
(Students can verify for (hemselves).
Finall y all minors of order I of the matrix A are non-zero, as no element of
the matrix A is 0.
Hence the rank of A is I. kns.


6 Matrices
Ex. 7. Find the rank of the matrix
1 3 43
3 9 12 9
-1 -3 -4 -3
Sot In this matrix, a minor of order 3
I 3 4 =3 I 3 4', taking 3 common from R2
3 9 12 1 3 4
= 1 -1 -3 -4 -1 -3 -4

=0, asR and R2 are identical.

In a similar way we can prove that all minors of order 3 are zero.
Now a minor of order 2
= 1 3 1 = 3 1 3 , taking out 3 common from R2
1 39 13
=0, as rows are identical.
Similarly all the minors of order 2 are zero.
Hence we are left with minors of order unity, viz, the elements of the given
rna*rix, which are not equal to zero.
Hence rank of the given matrix = I. Ans,
EL 8(a). Find the rank of the matrix
A= 6 1 3 8
16 4 12 15
533 4
4 .2 6 -1 (Kanpur 96)
Sol. The determinant of order 4 formed by A
= 61 3 8= 01 0 0
16 4 12 15 -8 4 0 -17
53 3 4 -13 3 -6 -20'
4 2 6 -1 -8 2 0 -17
, replacing C 1 ,C3 ,C4 by C1-6C2,C3-3C2
and C4 - 8C3 respectiiely
= - -8 0 - 17 = 0, as R 1 , R3 are indentical.
-13 -6 -20
-8 0 -17
Also one minor of order 3 viz.
i 3 SI=fI 0 0I=-6 -20 *0.
3 3 41 I -6 -201 1i
1 2 6 -i 12 0 -'l
0 -171
Hence the rank of given matrix A is 3. Axis.


Rank of Matrix 7
Ex. 8 (b). Find the rank of the matrix
A=1 2 1 2
1322
2934
3746 (Garhwal 93)
Sol. The determinant of order 4 formed by A
_Ii - I i
L _
-.1I,
4 • I -.1
L replacing R 2 , R , R by
3 4
Ii 3 2 2 01 R2-R1,R3-2R1
2 9 3 4.! 0 5 ol
3 7 4 61 10 1 1 ol R4-3R1respectively
= 0, R2 , I 4 being identical
Also one minor of order 3 viz.
1 2 1 1= 1 2 1 = 0, as above.
132 011
374 011
But all minor of order 3 are not zero.
eg.2 12= 21 2=--i -'2
3 2 2 -1 0 -2 3 -2
9 3 4 3 0 -2
=-[2+6]=--8 ;a0
Hence the rank of the given matrix A is 3. Ans.
Ex. 9 (a). Find the rank of the matrix
A= 1 -2 0 1
2 -1 1 0
3 -3 1 1

Sol. The determinant of order 4 formed by A

= 1 -2 0 1 = 0 0 0 1 replacing C1 , C2, by C1 - C4
2 -1 1 0 2 -1 1 0 and C2 + 2C4 respectively
3 -3 1 1 2 -1 11
-1 -1 -1 1 -2 1 -1 1
= - 2 -1 I , expanding w.r. to R1
2 -1 I
-2 1 -1
= 0, R 1 , R2 being identical.
Also one minor of order 3 viz.
1-21=100=0
2 -1 0 2 3 -2
3 -3 1 3 3 -2


8 Matrices
Similarly all minors of order 3 are zero
Now one minor of order 2 VZ 0 1 =- I *O
t0
Rank of given matrix A is 2. Ans.
**Ex . 9 (b). Find the rank of the matrix
A= 1 1 1 -1
123 4
3 4 5 2
(Ak'adh 92; Thrnde/k/ujnd 92;
Gorakhpur 93, Rohilkhand 98)
Hint: Do exactly as Ex. 9 (a) above. Ans. 2
Ex. 9 (c). Find the rank of the matrix
A= 1 -2 34
-2 4 -1 -3
-1 2 7 6
Hint Do as Ex. 9 (a) above. Ans. 2
Ex. 10. Find the rank of the matrix
1 -I -2 -4
3 1 3 -2
6 3 0 -7
2 3 -1 -1
So!. The determinant of order 4 formed by the given matrix
= 1 -1 -2 -4
3 1 3 -2
6 3 .0 -7
2 3 -1 -1
1 0 0 0
Jreplacing CCCbc+c
34 9 10
6 9 12 17 C3+2C1,C4+4C1respectively
2 5 3
= 4 .9 10 = 4 9 10 ,replacingR,hyR2-R1
9 12 17 5 3 7
5 3 7 5 3 7.
=0, as its two rows are identical.
A minor of order 3
= 1 - i -2 = I 00 . replacing C2 . C3 by
3 1 3 3 4 9 C2 + C 1 . C3 + 2C 1 respcctivley
6 3 0 6 9 12
I
= 4 9 =48-81=-33^,Q
9 12
 Rank of Matrix 9

Hence the rank of the given matrix is 3. Ans.

**Ex. 11. Find the rank of the matrix
A= 6 1 3 S
42 6-i
10397
16 4 12 15 (4 rod/i 90: kwiwun 90)

Sol. The deeirninant of order 4 formed b y this matrix

6 1 3
4 2 6 -1
hO 3 9 7
16 4 12 15

!o 1 3 81 replacing R 3 and R 4 by R 3 - R2
4 2 6 -1!
and R4 - R1 respectively.
1 3 81
6
6 1 3 81

= 0, as its three rows are identical

A minor of order 3
= 6 I 3 = 6 I 3 replacing R 1 by R 1 -
4 2 6 4 2 6
lU 3 9, 6 1 3I
= 0, two rows being identical-
In a similar way we can prove that all the minors of order 3 arc zero.
Now a minor oford'r2= 6 I =12-4-8()
42
Hence the rank of the given martix = 2. Ans.
Ex. 12. Find the rank of the matrix
A= 1 3 4 5
1267
5 0 1 (Gora(hpur 94)
L
Sol. One minor of order three of A
={i 4 5 Hi () 0 replacing(.-.,hv
6 7 Il 2 2 C2 - 4C 1 and C 5c'1
I 0 1 1 -. respectively.
--
-= 2 2. c pendi ng with respect to
-4 -4
2 (-4)- (-4) 0.
In a similar wa\ we can prove that all the minors of order three are ,ero.
Now a nhinor of order 2 is 2 6 = 20 -65 - 30 0
IS I)
I Ia:c die rank ot A is 2. Ans.
--Is
10 Matrices
Ex. 13. Find the rank of the matrix
A I a b 0
OedI
I a b0
Ocdl

So!. A I a b 0 =0,'.- R 1 , R are identical

0 c d I
labO
0 c d I
A minor of order 3 of
= a h 0 =0.asR 1 , R 3 are identical
c d 1
a h 0
In a similar way we can show that all the minors of order 3 are zero in value.
A minor of order 2 of A a b=ad - be 0
c
i
Hence the rank of the matrix A is 2. Ans.
Ex. 14. Find the rank of the matrix
A=3 4 5 6 7
45678
56789
10 11 12 13 14
15 16 17 18 19
(Kwnaun 91)
So!. One minor of order 3 of A
5 7 8 = 5 7 8 1 , replacing R2 , R 3 by
16 18 19 11
R2 - R 1 and R 1 - R 1 respectively.
11

= 5 7 8 ,replacing R 3 by R3 - hR2
1 11
000
=0.
In a similar way we can prove that all the minors of order 3 of A are zero.
This shows that all minors of order 4 and I A I of A are automatically zero.
(See Note 5 Page 2 of this chapter)
Now one minor of order 2 of A
= 7 8 =(7x9)-(8x8)=63-64=- I ;e0
89
Hence the rank of A is 2. Ans.


Rank of Matrix II
Ex. 15. Find the rank of A = I I I
b+c c+a a+b
be ca ab
(Kanpur 91)
Sot. IA= 1 I I
hi-c c+a a+b
be ca ab
= - (a - b) (b - c) (c— a), on evaluating.
Now following cases arise :-
Case I. a h = c.
Ifa=h=c,thenA= 1 1 1
2a2a2a
2 2 2
a a a
Therefore all minors of order 2 and 3 of A vanish.
Also A has non-zero minor of order 1, since no element of A is zero.
Hence the rank of A in this case is 1. Ans.
Case 11, Two of numbers a,b,c are equal, but are different from the
third.
Let a = h ;t c.
Then I A J = 1 1
1 = 0. as C1 , C2 are identical.
a+c c+a 2a
2
ac ca a
Also A has a minor of order 2 viz. 1 1
la+c 2a
=2a—(a+c)=a—c#O, '. a*c.
Hence the rank of A in this case is 2.
Similarly we can discuss the cases b = c # a, c =a # b. Ans.
Case M. a, b.c are all different.
In this case I A I ;^ 0, as is evident from (i) above.
i.e. A has a non-zero minor of order 3 and there exists no minor of order greater
than 3.
Hence the rank of A in this case in 3. Ans.
**Ex 16. Find the rank of the matrix
A 1 1 1 where a, b, c are all real.
a b c (Rohilkhand 97)
3
abc3 3

Sot.
A I 1 1 = I 0 0 replacing C2, C3 by
a I, c a b — a c— a
3 3 3 3 3 3 3 3 -
a b c a b--a c—a
12 Matrices
= b - a C- a , expanding with respect to R1
3 3 3 3 -
b —a C —a

(b—a)(c—a) I
2 2
Ii + ab 1- a c + ('a 1- a

taking (b - a). (c - a) common from C 1 arid C2

=(l,Lii)(c_a) 1 0
2 2
if + ab + a e + (O - / 2-- ab
replacing -2 by C 2 C1
/2 -
= (b - a) (c - a) ((c 2 + ((t- ab) -
/2) +
= (b - a) (c - a) 1Cc2 - a (c - b)J (Note)
=(b—a)(c—a) ((c—b)(c+b+a)j
or IAl=(a—b)(h—c)(c—a)i+b+c)
Now following cases arise
Case L a = b = c, then A = 1 1
a a
3 3 . 3
a a
11c',
Therefore all minors of order 3 and 2 of A are zero.
Also as no element of A is zero, so A has non-zero minors of order 1.
Hence in this case the rank of A is 1. Ans.
Case H. Two of the numbers a, b, c are equal but are different frorn the
third.
Let a=b;tc.
Then IA I = 1 1 I =0, as C and C2 are identical.
a a (2
3 3 3
a a a

Also A has a minor of order 2. viz. 1 1 = c - a 0 a c

I(I C

Hence in this case the rank of A is 2. Ans.

Similarly we can di.sus.s the cases b = c * a, c = a ;#' b.
Case HI. a, b.c are all different but a + b + c = 0.
In this case from (i). 0 is evident that = 0. IAI (Note)
Also A has a minor of order 2, viz I I = h - a :t 0. . (2 ^ h
h
Hence in this case the rank of A is 2. Us.
Case IV. a,b.,c are all different but + b + c . 0.
In this case from (I), it is evident that 0. Note)


Rank of Matrix 13

i.e. A has a non-zero minor of order 3.

Also A has no minor of order greater than 3.
Hence in this case the rank of A is 3. Ans.
**Ex. 17. Prove that the points (x 1 , y), (x 2, Y2)' (x3, y. ) are collinear if
the rank of the matrix is x 1 Yi 1 is less than 3.
X2 Y2 I
X3 Y3 1 (Agra 91; Kanpur 95, 93)
Sol. If the rank of the given matrix is less than 3, then the minor of order 3
of this matrix must be zero. (See § 502 Page 1 of this chapter)

i.e. - Yi 1 =0
X2 Y2 1
X3 Y3 I

Now the area of triangle whose vertices are (x 1 . v1). (x2 , vi), and (x3 , y)
= lxi Yi ) (See Authors Co.ordinale Geometry)
X2 Y2 I
X 1 y- I
= 0, from W.
Since tire area of this triangle is zero, so its vertices (x 1 , yi)' (x2, Y2) and
(X 1, yi) are collinear. Hence proved.
Ex. ]. Under what conditions the rank of the following matrix A is-3?
Is it possible for the rank to be 1? Why ?
A= 2 4 2
312
1 0 x (Kanpur 94)
Sol. 11 rhe r,uik of ihe rnairix A is 3, then the minor of order 3 of A should
1-t non-/ero.
i.e. 2 4 3 0. which is the required cc,riditjon.
3 i 21
1 0 VI
Ai ,, the rairkji A can not be I as at least onc minor of order 1 of A i.e., one
cierncni of A is ero.
if we are find the condition under which the rank of A is 2, then the same
is I A J 0 i x. minor of order 3 of A must he zcra
i.e 2 4 21 =0,i.c. 2 4 2 =0,repiacir.gR2byR2-Ri
3 1 21 1 -3 0
1 0 .t 1 0 x
ic. lo 10 2 0, replacing R 1 by R 1 - 2k2
Ii -3 0
0 x
14 Matrices
i.e. 10 2 -0+x 0 10 =0, expanding with respect to
1- 3 0 I -3
i.e. 6- I0x=0 i.e. x=6/10=3/5 Ans.J
Ex. 19. Are the matrices
A 12 3 and B = - 5
equivalent?
2
5 -2 -9 14
4 -2 -4 8
Sol. Since A is a 3 x 3 matrix and B is a 4 x 4 matrix i.e. their dimensions
are different, so these can not be equivalent.
Exercises on § 5.02
Find the rank of the following matrices
Ex. 1. (a) 1 2(b) 0 1 2 (c) 4 5 6
25 8 123 567
4 10 18] 3 1 1 7 8 9

Ans. (a) 3; (b) 3; (c) 2.

Ex. 2. 3 ii I 5
5 13 -1 11
-2 2 4
Ans.2
,Ex. 3. 1 2 4 3
3 9 12 9
1
3 4 1
Ans. 2
Ex. 4. 13 16 19
14 17 20
III 18 21 Ans.2
Ex. S. 1 0 -5 6
3-2 12
5 -2 -9 14
4 - 2 - 4 8
(Lucknow 90) Ans. 2
Ex. 6. 1 1 1 ], where a, b, c are all real.
abc
2 2 2
a b
(Kanpur 90)
[Hint See EL I6 Page 11 of his chapter]
Ex. 7 0 c -b al ,where a, b, c are, all positive
- 0 a numbers and aa+b+cyrO
1- : - a 0 y
[-a - -Y 0]
Ans.2
 Rank of Matrix 15

3
1
Ex. 8. 1
3 20
1 -2 -3 -3
1 .1 2 3 Ans.3
§ 503. Normal Form of a Matrix. (Agra 96)

Every non-/ero matrix A of order ru x n can be reduced by application of

elementary row and column operations into equivalent matrix of one of the
following Rrms

(ii) (iii) [J r, 0], (iv) ['ri'

[ ]
where 'r is r x r identity matrix and 0 is null matrix of any order.

These four forms are called Normal or canonical form of A.

Important Theorem (without Pr(of). (Avadlz 94)

Th. 1. If in x n matrix A is reduced to the canoncial form or normal form

[ ir 0
by the application of elementary row or column operations, then r, the
[o°]
order of the identity sub-matrix I r is the rank of the matrix A
Th. II. If a non-singular matrix of order n x n is reduced to the identity
matrix I (which is its canorHcal or normal form), then the rank of the matrix is n.
n
Solved Examples on § 503.
Ex. 1(a). Find rank of the matrix A = [ 1 2 31
3 41
5 6]
L4 (GOrakhpur 95)

Sol. A- I 2 I replacing C 3 by C 3 - C2
2 31
4 51
1 I replacing C2 by C 2 - C1
2 11
4 I I
-0 1 0 ,replacing C 3 byC 3 -C2 and ClbyCI-C2
10
3 10
-0 1 0 , replacing R 3 by R3 - R2
10
I 00
-0 1 0 . replacing R 2 by R2 - R1
00
2 00


IV Matrices 182/Il/I
-0 1 0 replacing R 1 by R 1 - 2R2
00
0 o 0
o 0 interchanging R 1 and R2
0 I0
0 00

[2 0
[0 0
Hence the rank of A is 2. Ans.
Ex. J (b). Find the rank of the matrix A =10 2 3
0 4 6
[0 6 9
Sol. A 0 2 1 repla.ing C 3 by C - C2
o 4 2
0 6

- 0 0 1 replacing C2 by C2 - 2C3
002
003

0 0 1 replacing R 2 . R 3 by R2 - 2!?, R - 3R1

000
000

1 0 0 incrct)aiIg1ng C 1 and C
0 0 0
0 0 0
-[
o
[o o
Hence the rank of A is 1. Ans.
*Ex I (c). Find the rank of the matrix A 1 1 1
222
3J
L3
Hint : Do as Ex. I (h) above. Replace C 2 . C 3 by C 2
- C. c -.CI
respectively and then from the result so obtained replace 1?, R 3 b y R2 - 2k1,
R3 - 3R 1 respectively.
Ans. I.
Ex. I (d). Find the rank of the matrix.
A=r_1
6 12 6
5 10 5


Rank of Matrix 17
182/11/2

Hint Do Ex. I (b) above. Ans. 1.

Ex. 2. Find the rank of tthe matrix

A= 1 2 3
456
212

Sot A i oa replacing C2 . C3 by C2 - 2C1,

4 - 3 - 6 C 3 - 3C 1 respectively
2 -3 -4

- 1 001 replacing R 2 , R 3 by R2 - 4R1,

0 -3 -6 R 3 - 2R 1 respectively
0 -3

- 10 ol. replacing R2 by R2 - R3
0 0 -2
o -3 -4
- 1 0 01. replacing C2 , C 3 by - C2
0 0 Ii
0 I 2] --i- C 3 respectively

- i 0 01. replacing R3 by R 3 - 2R2

001
010
- 1 0
010
01 interchanging C2 and C3

001

[131.
Hence the rank of A is 3. Ans.
Ex. 3 (a) Find the rank of the matrix A = 1 2 3
1 4 2 (Bundelkhand 94)
265

Sot. A - 1 ü replacing C2, C3 by C2 - 2C1

01.
I 2 - 1 and C3 - 3C 1 respectively
2 2 _lj

- I 0 0. replacing R 3 by R 3 - R2
1 2 -1
10 0
R2 , R 3 by R2 - R1
- i o 01. replacing
0 2 - 1 and R 3 - R 1 respectively
00
Oj


18 Matrices
o o], replacing C2 by C2 +2C3
0 o —1
0 o 0
o o]. replacing C3 by - C3
01
00
100,
o 01, interchanging C2 and C3
10
00
-100I
- [12 0
0
Hence the rank of A is 2.
Ans.
*Ex. 3 (b). Find the rank of the matrix A = [31 2 3 after reducing it
234
357
o the normal form.
t . ( Avadh 97; Garhwa/ 90; Meerut 92)
soLArl replacing C2 and C3 by C2 — C 1 and C3 - C2
21
[3 2 2j

1 1 i]. replacing R 2 and R 3 by R2 - R 1 and R3 - R2

10
11 ii
—1 1 'j, replacing R 1 by R 3 -
10 0
00 0

1 1 ol. replacing C 3 by C3 - C2
10 0
00 0
-[00
01, replacing C i by C 1 - C2
i 0 0
0 0
—10 o1, interchanging C, and C,
01 0
00 0

[ 12 0
00
Hence the rank of A is 2.
Ans.
 Rank of Matrix 19

Ex. 3 (c). Reduce matrix A to its normal form and then find its rank,
where A=1 11 —1
123 4
3 4 5 2 (Agra 93).
So!. A i 0 0 01. replacing C2 . C 3 , C4 by C2 - Cl-
1 1 2 51C3_ CI, C4+Cirespectively
3 1 2 5]
- 1 0 0 0 replacingR3byR3—R2
1125
2000
- 1 0 0 0 replacing R 3 by R 3 - 2J?
1125
0000
1 0 0 01. replacing C 1 . C3 C4 by C 1 - C2,
o i 0 0 C3 - 2C2 , C4 - 5C2 respectively
0000
[12 °1 which is in the normal form.
[oOj
Hence the rank ofA is 2. Ans.
Ex, 4. (a). Reduce the matrix A to the normal form.
where A = 1 1 2 - 3 hence find the rank of A.
410 2
030 1
010 2 (Meerul 92 P)
Sot. A 1 0 2 0 replacing C2 , C4 by
4 s 0 14 C 2 + C 1 , C4 + 3C 1 respectively.
0304
0102
1 0 1 0 replacing C3 , C4 by
45O7i i
C3 . - C respectively
0 3 0 2
010•1
--0 0 1 0 replacing C 1 . C4 by
4 5 0 2 C1—C3.C4—C2respectively
0 3 0 —1
010 0
0 0 I 0 replacing C1 by , C1
1 5 0 2
0 3 0 —I
0100


20 Matrices 999
0 0 1 0 replacing C2, C4 by
1 0 0 0 C2 - 5C 1 ; C4 - 2C respectively
0 3 0 -1
0100
- 0 0 I0 replacing C2, C4 by
1 0 0 0 C2+3C4,-C4respectively.
0001
0100
1 0 0 0 rearranging columns
0- 1 - 0 0
0010
0001
-['4].
Rank of A is 4. Ans.
EL 4 (b). Express the matrix
A= 3 -1 -1 3
-1 -4 -2 -7
2 1 3 0
-1 -2 3 0
lathe normal form and find Its rank. (Purvanc/aal 93)
Sol. A - 0 - 1 o 3], replacing C 1 . C3 by C1 - C4,
6 -4 2 - 7 1 C3-C2respectively
2 12 0
-1 -2 5 0
0 -1 0 3 1replacing C1 by C1 -. C3
4 -42 -7
0 12 0
-6 -2 5 0
- 0 - 1 0 3 replacing R2, R4 by R2 - R3
4 -5 0 -7 R4 + 2R3 respectively
0 12 0
-6 0 9 0
- 0. 0 23]. replacing R1,R4byR1+R3
4 -5 0 - 7 R + R2 respectively
0 12 0
-2 -5 9 -7
- 0 0 0 3]. replacing C3 by C3 - 2C2
4 -5 10 -7
0 10 0
-2 -5 19 -7


Rank of Matrix 21
- o 0 0 3 replacing R2 , R4 by R2 + 5R3
4 0 10 —7 R4 + 5R 3 repectively
010 0
—2 0 19 —7

0 0 0 3 replacing R4 by R 4 - R2
4 0 10 —7
01 0 0
—6 0 9 0

- 0 0 0 1 replacing R 1 , R4 by R1, -R respectively

4 0 10 —7
010 0
—2 0 3 0

- 0 0 0 1 replacing R2 by R2 + 2R4
0 0 16 —7
010 0
—2 0 3 0

0 0 0 1 replacing R 2 by R 2 + 7R1
0 0 16 0
01 00
—2 0 3 0
- 0 0 0 1 replacing R2 by (1/ 16) R2
0010
0100
—2 0 3 0
- 0 0 0 1 replacing R4 by R4 - 3R2
0010
0100
—2 0 0 0
--0 0 0 1 replacing R4 by (1/2) R4
0010
0100
1000
- i 0 0 interchanging R 1 and R4
0 1 0 0 and interchanging R2 and R3
0010
0001
or A 1141 The rank of A is 4. Ans.


22 Matrices
Ex. 5. Find the rank of the matrix
Az —2 —1 —3 —I
1 2 3—I
1 0 1_i
0 1 1 l
(Garhvu1 94)
Sol. A - 0 - - t i replacing R 1 , R2 by
o 2 2 - 2 R 1 + 2R3 , R2 - R3 respectively
I 0 I I
o i i —1
- 0 0 0 1 replacing C2 , C3 by
0 0 0 - 2 C2 + C4, C3 + C4 respectively
112 I
• 0 0 0 —1
- 0 0 0 i].. replacing R2. R3 , R4 by i2 + 2-R 1,
0 0 0 0 R3 - R 1 , R. + R respectively
1120
0000
- 0 0 0 I],replacing C2,C3by
0 0 0 0 C2 - C 1 , C3 - 2C1 respectively
1000
0000
r
0 0 0. interchanging R I , R3
0000
0001
0000
- 1 0 0 o]. interchanging C2, C4
0000
0100
0000
1 0 0 o], interchanging R2, R3
0100
0 '0 0 0
0000
—[12 0
Loo
Hence the rank of A is 2. A ns.
Ex. 6 (a). Reduce A = 1 - 1 2 -I to normal forn.
4 2 —1 2
2 2 —2 0
(Gczrhwal 91)
 Rank of Mati ix 23

Sol. A = i o o o replacing C,. C3 , C4 by

4 6 -9 6 C, + C 1 , C3 - 2C2 , C4 + C 1 respectively
2 4 -6 2

- 1 0 0 o]. replacing C2 . C 3 by
4 C) 0 6 C2 - (4. (3 -- (3/2) C2 respectively
2 2 3 2]

0 0 o]. replacing C3.C4by

4 0 0 6 C 3 - (3/2) C2 . C4 - C2 respectively
2 0 0

- 1 0 0 01 replacing R 2 , R3 by
200 3I t i
R,, R 1 respectively
I 1 0 oj

- I (1 0 o replacing R 2 . R 3 by
() () 0 3 fR3 - 2R 1 , R 3 - R 1 respectively
0 1 0 0]

- 1 0 o 01 . replacing C3 by I C3
0001
0 1 0 0
- I 0 0 0 . interchanging C2 , C4
0100
0 () 0 1
- 1 0 0 01, interchaning C3 , C4
0100
L0 0 1 0

- 11 3 01 which is the required normal form Ans.

Ex. 6 (b). Reduce the matrix A to the normal form and hence find the
rank of the matrix A, where
A= 1 -1 1 -ii
4 2 --1 2
2 2 -2 2

Hint Do as Ex. (a) above.

6 Ans. 3
Ex. 6 (c). Find the rank of the matrix
A1 3 4 71
2458
3124

Hint : Do as Ex. 6 (a) above. Ans. 3

**Ex. 7. By elementary operations, find the rank of the matrix.
24 Matrices

[bN. U (Avcid/z 95, 93, 91; Car/twa! 95,

(3nr-iL-Irn,jr 90d A',,,,n,r 07 &.1,,,,-,,v OR

Sol. A-6_3 0 —7 replacing R by R 1 - R 2 + R3

1 —1 —2 -4
3 1 3 —2
6 3 0 —7
or 6
, 3 C) —7, replacing R4byR4—R1
1 —1 —2 —4
3 1 3 —2
o 0 0 0
..-0 9 12 17 replacing R 1 and R
1 —1 - 2 —4 by R 1 - 6R 2 , R 3 - 3R 2 respectively.
o 4 9 10
0 0 0 0
—09 12 Il replacing C2. C3 and C4 by C2 + C1.
10 0 0 C3 + 3C ! and C4 + 4C 1 respectively
04 9 10
00 00
—10 o o , interchanging R 1 and R2
09 12 17
04 9 10
00 0 0•
—10 0 0 , replacing R2 by R2 - 2R3
01 —6 —3
04 9 10
00 0 0
—10 0 0 ,replacing C3 and c4byc3+6C2
01 0 0 and C4 + 3C-, respectively.
0 '4 33 22
00 00
0 0 0-1 replacing C3 and C4 by 313 C3 and
0 0 0- C2 respectively.
04 112
0 0 00
—1 0 0 0 . replacing C2 and C4 by C2 - 4C3
0 0 0 and C4 - C3 respectively.
0 0 10
0 0 00
 Rank of Matrix 25

^1 .4 ()
o°
Ans.
Hence the rank of the given matrix = 3.
Ex. 8. Find the rank of the matrix
J 0 21
0 1 -2 1
A-1 -1 40
2 2 80
Ans. 3
Hint : Do as Ex. 7. on Piges.23-24
**Ex 9 (a). Find the rank of A = 0 1 -3 -1
10 1 1
31 0 2
1 1 -2 0

(Agra 91; Bundelkhand 91, Garhwal 92; Kurnaun 96; Lucknow 92;
Meerut 90, Purvanchai 98; Rohjlkhand 91)

Sol. A - 0 i -3 - replacing C3, C4 by

1 0 0 0 C3 - C 1 and C4 - C 1 respectively
3 1 -3 -1
1 1 -3 -1
- 0 I -3 - i replacing R 3, R4 by
1 0 0 0 R3 - R 1 and R4 - R 1 respectively
30 0 0
10 0 0

- o 1 0 0 replacing C3 , C4
by
1 0 0 0 C3 - 3C2 C4 + C2 respectively
3000
1000

- 0 1 00 replacing R3 . R 4 by
1 0 0 0 R 3 - 3R 2 and R4 - R2 respectively
0000
0000

- 1 ü 0 o interchanging C i , C2
0100
0000
0000

-[12 0
[00
Hence the rank of A is 2. Ans.
Ex. 9 (b). Find the rank of the matrix
16-
 26 Matrices
A= 1 2 3 4
2345
3 4 5 6
4 5 6 7 (Purvanc/zal 95)
Sol. A - 1 2 3 4 replacing R2, R3 , R 4 by
1 1 1 I R,- R I , R3 - R I , R 4 -R, respectively
2222
3333
1 1 2 3 replacing C2 , C 3 , C4 by
1 0 0 . 0 - C,, C 3 - C,, C4 - C, respectively
2d d 0
3000
- 0 I 2' 3 ,replacing R I , R3,R4by
1 0 0 0 R 1 - R2, R 3 - 2R2 , R4 - 3R 2 respectively
0000
0000
0 1 0 0 replacing C3 . C'4 by
1 0 0 0 C 3 - 2c2 , C4 - 3C'2 respectively
•0000
0000
- 1 0 0 0 , interchanging C 1 and C2
0 100
0000
0000
_[12 0
[00
Hence the rank of A is 2. Ans.
*Ex 10. Find the rank of matrix A = 2 - 1 3 4
0 341
2 3 75.
2 5 11 6
by reducing to canonical form. Also show that it is not equivalent to the
matrix B= 1 0 -5 61
3 -21 ol
5 -2 -9 14
4 -2 -4 8
Sol. A 2 - 1- 3. 4 ,replacing R3 and R 4 by
0 '3 4 1 R3 - R, and R4 - R3 respectively
0 441
0 - 41
 Rank of Matrix 27.

2 - i 3 4 replacing R 2 and R 3 by
o 1 0 0 R2 - R4 and R - R 4 respectively
o 200
0 2 4 I

- 1 3 4 replacing C 1 and R 3 by
0 IOOi
C 1 and - R 3 respectively
o 1 0 0
0 241

- 1 —1 3 4 replacing R 3 by R 3 - R2

0 +1 0 0
0 000
0 241

- I - 1 3 4interchang R 3 and R4
0 1 0.0
0 241
0 000
1 0 0 0 .replacing C 2, C 3 and C4bYC2+CI,
0 1 0 0 C3 - 3C 1 and C4 4C 1 respectively.
0241
0000

r eplacing C2 and C 3 by
? g
C2 — 2C4 and C3 — 4C4
0 01
0 0 0 0 respectively.

1 0 0 0 interchaining C 3 and C4
0100
0010
0000

0
[13
[00
The required rank of the matrix A= 3.
Also we can prove as in Ex. 5 Page 5 Chapter V that the rank of the matrix
Bi.e. 1 0 —5 6 is 2.
3 —2 1 0
5 —2 —9 14
4 —2 —4 8
Since the ranks of the two matrices A and B are different so these are not
equivalent. (See Note 3 of § 502 Page 2 Chapter V)


28 Matrices
Ex. 11(a). Find the rank of matrix A = 3 - 2 0 -1'
o 2 2 1
1 —2 —3 2
o i 2 1
(Gorakhpur 97; Lucknow 92)
Sol. A- 0 4 9 .-7 replacing R I , R2by
0 1 0 0 R c - 3R3, R2 - R4 respectively
—2 —3 2
0 1 2

—2 —3 2] interchanging R 1 , R3
0 1 0 0 and replacing R4 by R4 —R2
0 4 9 —7
0 0 2 I
00 0 . rephcing C2 , C3 , C4 by C2 + 2C1,
0 1 0 0 C 3 + 3C 1 and C4 - 2C 1 respectively
0 49
0 02
--1 00 0 1. replacing R 3 by R3 - 4R2
0 Jo 0
0 09 —7
0 02

—1 0 .0 01,replacing R3byR3+7R4
0 1 0 oJ
0 0 23 0
0 0 2 1]
00
g, replacing R3 by R3
0 10
0 01
0 02

replacing C3 by C3 - 2C4
0
0
0 I$
-1141
Hence the rank of A is 4
Ans.
Ex. 11(b). Reduce the following matrix A to the form
1r 0
100

and hence determine its rank.

 Rank of Matrix 29

A=1 —1 2 —3
4 1 0 2
o 31 4
o 1 0 2 (Kumaun 92)

A - 1 0 2 - 1 replacing C2 , C4 by C2 + C1.
Sol.
4 5 0 2 C4 + C3 respectively
031 5
010 2
1 0 2 0 -.replacing C4byC4+C
4506
0315
0102
-.5R4
- 1 0 2 0 replacing R2, R 1 by R 2
4 0 0 —4 R 3 - 3R4 respectively
o 0 1 —1
010 2

1 0 2 0 replacing C4 by C4 - 2C2
4 . 0 0 —4
0 0 1 —1
010 0

1 0 2 3 replacing C4 by C4 + C 1 + C2 + C3
4. 0 0 0
0010
0101
- I 0 2 3 replacing R2by(1/4)R2
1000
0 01 0
0101
- 0 0 2 3 replacing R I , byR1—R2
I 0 0 0 and C4 - 2 respectively
0 0 . 1 0
0100
- 0 0 0 1 replacing R 1 , C4 by R 1 - 2R3

1 0 0 0 (1/3) C4 respectively
0010
0100

- I 0 0 0interchanging R 1 and R 2 , R3 and R4

0001
0100
0010


30 Matrices
I 0 0 0 interchanging R2 and R3
0100
0. 0 0 I
0 0 I •0
1 0 0 0 interchanging R3 and R4
0100
0 0- 1 0
0001
-1141
Hence the rank of A is 4. Ans.
*EX . 12. Is the matrix A = 1 1 2j equivalent to 13?
-1 20
2 -3 1
Sot. Here we find that the minor of order 3 of A
= 11 2 = 1 0 0 , replacing C2,C3byC2-C1,
- 1 2 0 - 1 3 2 C3 - 2C1 respectively
2 -3 1 2 -5 -3
3 2 , expanding with respect to R1
-5 -3
=3(-3)-2(-5)=-9+10=1:^t0.
Also from 5-03 Th, II Paper 15 Chapter V we know that this matrix A can
be recduced to 13 by elementary row or column operations.
Bence A is equivalent to 13.
Ex. 13. Determine by reducing to normal form the rank of the matrix
A= 8 1 36
0 3 22
-8 -1 -3 4
Sot. A I
0
1
3
3 31
211
and C4 by
C1
1
by
C4
j C1

-1 -1 - 3
- 1 0 0 0 . replacing C2, C3 and
0 . 3 2 1 C4byC2-C1,C3-3C2
- 1 0 0 5 C4 - 3C respectively
- 1 0 0 0 replacing R3 by R- + R1
0321
0005
- 1 0 0 01, replaing C2 by 13 C2 and C3 by 1 C3
0111
0 0 0 5


Rank of Matrix 31

1 0 0 replacing C3 and C4 by
C3 - C2 and C4 - C2 respectively
g
1 0 0 01. interchanging C3 and C4
0 1 0 0
oO0

- I 0 0 01. replacing C3 by C5
0100
0010

11 3 01.
(Note
Ans.
Hence the rank of A is 3.
Ex. 14. Find the rank of the matrix A = 1 2 -1 3
2 4 -4 7
-1 -2 -1 1

(Bundellkhand 95; Garhwal 96; Purvanchal 97; Rohilkhand 9

1 0 replacing C2 . C 3 , C4 by
Sol. A - o ol.
2 0 -2 11C2_2Ct.C3+Cl and
- 1 0 -2 4] C4 - 3C 1 respectively
i 0 o 01, replacing R 3 by R 3 - R,
2 1) -2 1
-3 0 0 3

- 1 0 0 0 . replacing R 2. R 3 by
0 0 - 2 1 R 2 - 2R 1 and R 3 + 3R 1 respectively
00 03 -

- 10 0 0 ,replacingC 3 hyC 3 and. R3bYR3

0011
0001

1 0 0 0, replacing C4 by C4 - C3
0 0 1 0
0001

- 1 0 0 0 interchanging C2 and C3
0100
0001
- i 0 0 0, interchanging C3 and C4
0100
0010

1 1 1 01- (Note)


32 Matrices 182/11/2
Hence the rank of A is 3.
Ans.
Ex. 15. Use elementary transformations to reduce the following matrix
to triangular form and hence find the rank of A.
A=5 3 144
0 1 21
1 -1 2 0
Sol. A - 53 8 1 replacing C3 , C4 by
o 1 0 0 C3-2C2,C4--C2
1 -1 4 1

5 8 - 12 - 4]. replacing C2. C3, C4 by

o 1 0 0 1 C2 +C
i, C 3- 4c 1,c4_c 1 respectively.
1 0 0 0]

5 0 0 4]. replacing C2. C3 by

o i 0 0 IC 2 + 2c4 , ('3 - 3C4 respectively.
-[ 1 0 0 Oj
Js 0 0 i]. replacing C1 bi - C4
10 1 0 0
1000
0 0 0 11 replacing C 1 by C1 - 5C4
0100
1000
[00 1. 0 '0 , interchanging R 1 , R 2
001
1000
O 1 0 o]. interchanging R2, P3
.1000
0001
- 1. 0 0 o]. interchanging R 1 , R2
0100
0001
1 0 0 0 , interchanging C3, C4
0 1 00•
0010
-1 13 0]
Hence the rank A is 3.
Ans.
**EX 16. Reduce the matrix A to the normal (or canonical) form and
hence obtain its rank.


182/11/3 Rank of Matrix 33

A=[ 1 2 0.—I
341 2
[-2 3 2 5

Sol. A - 1 o o a replacing C2 and C4 by C2 - 2C1

3 -2 1 5 and C4 + C 1 respectively.
—2 723
- 1 0 0 01
replacing C2 and C4 by
3 0 1 0 IC2 + 2C 3 and C4 - 5C3 respectively.
—2 11 2 -7j
- 1 0 0 o replacing C 1 by C 1 + C3
C4byC4+.j7jC2
4 0 1 0 and
0 II 2 Oj

- 1 0 0 01, replacing R2 by R 2 - 4R1

0010
0 II 2 0
- i o 0 0 1, replacing R3 by R 3 - 2R2
0010
0 11 0 0

- 1 0 0 01, interchanging ('2 and C3

01 00
0 0 11 0
- 1 0 0 01 replacing C 3 by C3
0100
0010
11 3 0] (Note)
Hence the rank of A is 3. Ans.
*Ex. 17. Is rank of A = 1 0 equal to 2? 01
010
0 0 1 (Agra9O)
Sol. Here A =13, so rank of is 3 and not 2. Ans.

Ex. 18. If A = -1 2 - 1 4 1, find its rank.

2 4 3 4
1 2 3 4
1 -2 6 -7 (Rohilkhand 92)
Sol. A ----I 2 -- 1 41 replacing R2 , R 3, R4 by
0 8 I 12 R2 +2R 1 ,R3 +R 1 ,R4 +R 1 respectively.
04 2 8
0 0 5 —3

34 Matrices

0
-J 0 01 replacing C2 , C3, C4 by
0 I 12 IC2 + 2C,, C 3 - C 1 , C4 + 4C 1 respectively.
0 2 8
0 5 -3
-10 0 0 replacing C 1 , C2 by - C 1 and
02 12 (1/4) C2 respectively.
01 2 8
00 5 -3
-10 0 0 replacing C3 , C4 by C3 - 2C2,
02 -3 - 4 C4 - 8C2 respectively.
.0l 0 0
00 5 -3
-10 0 01.
replacing R 2, R4 by
00 • -3 - - 2R3 , (- 1/3) R4 respectively
01 0
00 -7/3 1]
-10 0 0 replacing R2 by R2 + 4R4
00 -29/3 0
01 0 0
00 -5/3
-10 0 0replacing R 2 by - (3/29) R2'
00 0
01 0 0
00 -5/3
-10 0 0 interchanging R2 and R3 and replacing.
01 0 0 R4 by R4 + ( 5/3) R3
00 10
'0'0 01
['41
Hence the rank of A is 4. Ans.
Exercises on * 503
Find the rank of the following matrices by reducing these to the normal (or
canonical) form :— *
Ex. 1. [0 01; Ii 01; fi T ft ii
°9J L0 jLi [l 2] Ans. 0, 1, 1, 1
Ex.2. 2 1 3
4 7 13
4 . Ans.2
 Rank of Matrix 35

Ex. 3. I I I I
I 3 -2 I
2 0 -3 2
.33 03 Ans. 3
Ex. 4. 1 2 1 0
-2 4 3 0
1 0 2 -8 (Kumaun 93) Aid. 3
*Ex. 5 1 2 3 2
235 I
1345 Ans.2
Ex. 6. 1 -3 2
3 -9 6
-2 6 -4 Ans. I
*Ex. 7 1 2 -1 4
2 4 3 5
-I -2 6 -7 Ans. 2
Ex. 8. 1 1 1 1
1 3 -2 1
2 0 -3 2
3 3 -3 3 Aiis. 3
Ex. 9. I 4 3 2
1234
2675 (Garakhpur 99) Aus. 3
Ex. 10. 1 2 -1 3
4 1 21
3 -1 1.2
1 2 01 (Meerut 91S) Ans. 3.
Ex. 11. 1 1 2 3
1 3 0 3
1 -2 -3 -3
1 1 2 3 Ans. 3
*Ex. 12 0 1 2 -2
4 0 2 6
213 1 (Garhwal9i) Ans.2
Ex. 13. 6 1 3 S
5 2 6 -I
10 3 9 7
16 4 12 15 (Avadh 98) Ans. 3
36 Matrices 999
: Ex. 14. 3 4 5 6 7
45678
56789
10 11 12 13 14
15 16 17 18 19 (Agra 92) Ans. 2
Ex. 15. 1 3 2 0 I
9 2 —1 6 4
7 —4 —5 6 5
17 1 -4 12 7 Ans.2
Ex. 16. 9 7 3 6
5 —1 4 1
6 8 2 4
(Rohilkhand 96) Ans. 3
Ex. 17. 1 2 —1 4
32 02
01 32
(Lucknow 90) Ans. 4
§ 504. Echelon Form of a Matrix.
Definition. If in a matrix.
(1) all the non-zero rows, if any, precede the zero rows,
(ii) the number of zero prececding the first non-zero element in a row is less
than the number of such zero in the succeeding row.
(iii) the first non-zero element in a row is unity, then it is in the Echelon
form.
NOte. The number of non-zero rows of a matrix given in the Echelon form
is its rank. (Remember)
Example of q matrix in the Echelon Form
13456
01234
•
00123
00000
In this matrix we observe that
(i) the first three non-zero rows precede the fourth row which is a zero row.
(ii) the number of zero in R4 , R 3 and R2 are 5, 2 and 1 respectively which are
in descending order.
(iii) the first non-zero term in each row is unity.
Hence all the three conditions of the Echelon form are satisfied.
Also there being three non-zero rows-in this matrix, its rank is 3. This fact
can be proved by actually finding the rank of this matrix.
In this matrix, a minor of order 4


F. 'hêlon For of Matrix 37

= 1 3 4 5 = 0. one row being of zero.
0123
0 0 1 2
() () () 0,

In a similar .ay we can show that Al minors of order 4 are zero.

Now a minor of order 3 = 1 3 4
012
0 0 1
expanding w.r.toC1
= I 2,
0

= 1 :P, 0.
Hence the rink of this matrix = 3.
Ex. 1. Find the rank of the matrix.
A=[0 1 2 3
o 0 1 —1
000 0
Sol. In the given matrix we observe that
(i) the first two non-zero rows precede the third row which is a zero row,
(ii) the number of zero in R 3 , R 2 and R 1 are 4, 2 and I respectively which are
in descending order, and
(iii) the first non-zero term in each row is unity.
Hence all the three conditions of the Echelon form are satified.
Also there being two non-zero rows in this matrix, its rank is 2. Ans.
Ex. 2. Reduce the following matrix to its Echelon form and find its
rank
A= 1 3 4 5
3 9 12 9
- 1
-4 3 (Meeru: 93)
So!. A - 1 3 4 5 , replacing R2 , R3 by
0 0 0 -6 R2 - 3R 1 , R3 + R 1 respectively
000 2

F 3 4 5 1 replacing R 2, R3 by
0 0 0 0 k2 + 3R 3 , ( 1/2) R3 respectively.
[o 0 1) 1]

- I 3 4 5 . interchanging R 2 and R3
.0001
0000
In the above Matrix we observe that.
38 Matrices
(i) the first two non-zero rows precede the third row which is a zero row,
(ii) the number of zero in R3 and R2 are 4 and 3 respectively which are in
descending order, and
(iii) the first non-zero term in each row is unity.
Hence all the three conditions of the Echelon form are satisfied and then the
given matrix is reduced to its Echelon forrii.
Also there being, two non-zero rows in this matrix, its rank is 2. Ans.
5. 05. Invariante of rank under elementary operations.
Theorem. All equivalent matrices have the same ranks i.e. the rank of a
matrix remains unaltered by the application of elementary row and column
operations. (Avadh 99: Bundeikhamd 93)
Proof. Let rbe the rank of x matrix A[aj]
Case I. If fib and jth rows are interchanged (which may be written
symbOlically as R ij or (R1 ---_+R) then it does not effect the rank.
Let B denote the matrix obtained from the matrix A by the elementary
operation Ri R and let p be the rank of B.
Also if D be any (r + I) rowed square sub-matrix of B, then I D = ± IC
where C is a particular (r+ 1) rowed sub matrix of A.
As r is the rank of the matrix A so every (r + 1) rowed minor of A vanishes
and therefore p, the rank of B, cannot exceed r, the rank of A i.e. p:5 r.
Also we can obtain A from B by the elementary operation R• E-* R1
Therefore in that case interchanging the roles of A and B we shall get r <p.
-Hence r=p.
Case H. If the elements of a row are multiplied by a non-zero number A
(which may be written symbolically, as R 1 - A R1 , A 0) then it does not effect
the rank.
Let B denote the matrix obtained from the matrix A by the elementary
operion R —,XR1 and let pbe the rank of B.
Let D be any (r + 1) rowed square sub-matrix of B and let C be the
sub-matrix of A having the same position as D. Then either I D = I C I or I D
= A I C.
[Here ID I = IC I happens if the ith row of A is one of those rows which are
reiaoved to obtain D from B and I D = A IT I happens when the ith row is not
r'ed while obtaining C from A].
Also as r is the rank* of the matrix A so every (r+ 1)—rowed minor of A
vaniirs and therefore in particular I C = 0 and consequently in both the above
cues I D I =0.
p, the rank of B, cannot exceed r, the rank of A.
i.e. p5r.
Also we can obtain A from B by the elementary orw ration R R,

therefore it at case interchan . the roles of A and J' -hall get r

 Invariance of Rank of Matrix 39
Hence r=p.
Case III. If to the elements of the ith row are added the products by any
non-zero number A of the corresponding elements of jth row (which may be
written symbolically as R, —p R + A X;4 0) then it does not effect the rank.
Let B denote the matrix obtained from the matrix A by the elementary
operation R• —+ R, + AR and let p be the rank of B.
Let D be any (r + 1) rowed square submatrix of B and let C be the submatrix
of A having the same position as D.
Now three sub-cases'arise
(i) If A and B differ only in the ith row i.e. if ith row of B is one of those
rows which have been removed while obtaining C.
In this case D = C and therefore I D = I C
The rank ofA isr,soCj=0 and consequently DJ0
(ii) If ith row of 13 has not been removed but jth row has been removed
while obtaining D.
In this case I I) = C I + A I C O 1, where CO is an (r + 1) rowed matrix which
is obtained from C by replacing ak by '1jk i.e. is obtained from C by
performing the elementary operation R 1 or R1 R1 and then removing those
rows and columns of the new matrix which are removed to obtain D from B.
I C O I is negative of some (r -f. l)-rowed minor of A and as the rank of A
is r so every (r+ 1)-rowed minor of A is zero i.e. ICI=O,lCokO and
consequently I D I = 0.
(iii) If neither the ith row not the jth row of B has been removed while
obtaining D.
Here IDJ=j CL and so as before IJ)
Every (r + 1)-rowed minor of B vanishes so p, the rank of B, cannot
exceed r, the rank of A i.e. p < r.
Also we can obtain A from B by the elementary operation R, — R 1 - A
therefore in that case interchanging the roles of A and B we shall get r_-5p.
Hence r=p,
Thus we have observed that the rank of a matrix remains invariant tmder
elementary row operations. Similarly it can be shown that the rank of a n1rix
remains invariant under elementary column operations too.
Note. By the applications of the above theorem we can easily obtain the
rank of a matrix for if we can obtain a matrixB by elementary operations an a
matrix A and of the rank of B can be easily determined by inspection or simple
calculations as given in previous articles in this chapter, then we can determine
the rank of A.-
§ 506. Some Important Theorems.
Theorem I. The rank of a matrix is equal to the rank of the transposed
matrix.
40 Matrices
or p (A) = p (A'), where p (A) denotes rank of A. (Kanpur 94: Rohilkhand 92)
• ProoL Let A = [a ] be any mXnmatrix.
Then the transposed matrix A' = [ar ] is an n X m matrix.
Let the rank of A be r and let B be the r x r sub-matrix of A such that
B I *0.
Also we know that the value of a determinant remains unaltered if its rows
and columns are interchanged. (See Prop. II of Determinants)
i.e. I B' I = I B I -t- 0, where B is evidently a r x r sub-matrix of A'.
The rank of A' ^t r, (See Note 4 (h) Page 2 Ch. V)
Again if C be a (r+ I) x (r+ 1) sub-matrix of A, then by definition of rank
(See * 502 Page 1 Ch. V) we must have all I C I
Also C' is a (r+ l)x(r+ I) submatrix of A' so we have
I C' j C 0, as explained above.
We conclude that there cannot be any (r+ l)x(r+ 1) sub-matrix of A'
with non-zero determinant.
The rank of A' ^t r and it cannot be greater then r as above.
The rank of A' is r which is also the rank of A. Hence proved.
Theorem IL The rank of the product matrix AB of two matrices A and B is
less that the rank of either of the matrices A and B.
Pteol. Let r 1 and r2 be the ranks of the matrices A and B.
r1 is the rank of A therefore A - [M]l, where M is a submatrix of rank

r1 and contains r 1 rows.

Post multiplying it by B, we get
AB - [M]
[0]
But [Ml B can have r1 non-zero rows at the most which are obtained on
L°i
non-zero rows of M with columns of B.
multiplying r 1
Rank of AB = Rank of [M] < r1
[0]
i.e. Rank of AB 5 Rank of A
In a similar way we get B - [N 0], where N is a submatrix of rank r2 and
contains r2 columns.
Pre-multiplying it by A, we get
AB—A[N0]
But [N 0] can have r2 non-zero columns at the most which are obtained on
multiplying the rows-of A with r2 non-zero columns of [N 0]


Rank of Matrix 4I
Rank of AB = Rank of A [N 01 r2
i.e. Rank of AB Rank of B. ...(ii)
Hence the theorem from (i) and (ii),
Solved Examples on § 5 . 05 and § 506.
Ex. 1. Show that the rank of a matrix A does not alter by pre or post
multiplying it with any non-singular matrix R.
Sol. Let B = RA.
Then rank of B = rank of RA :5 rank A.
See § 5.06 Th II above
Also A = R B, where R _ the inverse matrix of R.
rank of A = rank of (R'B) !^ rank of B. ...(ii)
From (i) and (ii) we conclude that
rank of A = rank of B. . . Hence proved.
Ex. 2. Show that AA' has the same rank as A, where A' is the transpose
of A.
Sot. Let B = AA', then the rank of B :5 rank of A.
Also A - A', and so we have
rank of A = rank of A' !^ rank of B.
From (i) and (ii), rank of A rank of B.
Er. 3. Show that AA 0 has the same rank as A, where A is the
transpose conjugate of A.
[Hint: Do as Ex. 2 above]
Ex. 4, Prove that if A is a matrix of order n x n and if B non-singular
matrix of order n, then the product P = AR has the same rank as A.
Sol. Here A - (m x n), B - (n x n)
P=AB-(mxn)
Ifm<zn, rank ofAm but rank ofB=n
rank of A rank of B,
Now rank (P) = rank (AB) rank A
But we can write A = PB'
rank of A = rank of (PB' rank of P. ...(ii)
From (L) and (ii) we get rank of P = rank of A.
§ 5 . 07. Sweep out method of finding the rank of a matrix.
In the process of evaluation of the rank of a matrix by means of elementary
row and column transformations, if certain rows or columns are zero-rows or
zero-columns i.e. each element of these rows or columns are zero, then we can
remove these rows or columns Without any effect- on the rank of the matrix. (See
§ 505 Page 38 Ch. V). This method is general ly called the Sweep out
method.
-17


42 Matrices
Solved Examples An § 5.07.
Ex. 1. Find the rank of the matrix
=61 3 8
426-1
1039 7
16 4 12 15
Sol. A 6 1 3 8 replacing R 3 and R4 by
• 4 2 6 - I R 3 - R 1 and R4 - 2R 1 respectively.
4 2 6 -1
4 2 6 -1

.A-[6 1 3 81
[4 2 6 ij
...See § 507 abov
Now a minor of order 2 is 18 = - 1 - 16 = - 17 0.
• . 2 -1
Hence its rank is 2. Ans.
Ex. 2. Find the rank of the matrix
0.1 -3 -1
10 1 1
31 0 2
1 1 -2 0
Sol. Let A= 0 1 -3 -1
10 1 1.
31 0 2
11-2 0
Now A - [0 1 -3 - i replacing R2 and R3 by
1 1 -2 0 R 2 + R 1 and R 3 + 2R 1 respectively.
3 3 -6 0
[1 1 -2 0
- 0 1 - 3 - i replacing R 3 and R 4 by
• 1 1 -2 0 1 R3 and R4 - P2 respectively.
1 1 - .2 0
00 0 0
- 0 1 -3 - i replacing R3 by R3 - R2 and then
1 0 0 0 C2 and C3 byC2 -C 1 and C3+2C1.
• 0 0 0 0 respectively.
00 0 0
-[0. 1 -3 -11
[i 0 0 0] • ... See § 507 Page 41 Ch. V.
Now a minor of order 2 is
 Adjoint of Matrix 43
-[o i]=— I *0. Hence its rank is 2.
[1 0] .
- Ans.
Ex. 3. Find the rank of the matrix A =1 1 —3 4 71
1 2 0] (Meerut 95. 94)
[9

Sot. Here A -1 28 - 3 10 7 ] , C 1 , C3 by C 1 - 9C2,

[
0 1 0 Q] C3 - 2C2 respectively.

-10 —3 3 71 replacing C 1 , C3 by C1 - 4C4,

[0 1 0 oj C3 - C4 respectively.

—[0 0 3 i. replacing C2.C4byC2+C3,

[0 1 0 0] C4 - 2C3 respectively.
—[0 0 0 '1 replacing C3 by C3 - 3C4
[01 00]
Now aminaroforder2 is 0 1 =-1*0
10
Hence its rank is 2. Ans.
Exercises on § 5-07
Find the rank of the following matrices :-
Ex. 1. 4 3 0 2
3 4 —1 —3
—7 —7 1 5
Ans.3
Ex. 2. 3 2 —2
2 3-3
—2 4 2
5 —2 4
Ans.3

1 -2 2 - ^

i r,^^d
j i t of a Matrix. (Agra 94, 92; Rohilkhand9l, 90)
- 0 ition. If C
ii
he the cotactor of the
maix A Ea qJ. then - -
adjoini of A = C11 C21 C,1
C 12 C22 ... C,2

44 Matrices
This is also rewritten as Adj. A
or adjointofA= transposed ofC,where c= c 11 c 12 ....
C In -
C21 C22 .. C211

Cf1 C,12 .... C.

While solving problems we generally use this definition.

Here students should note carefully that the cofactors of the elements of the
first row of I aij I are the elements of the first column of Adj A.

Similarly the cofactors of the elements of the first column of Jaij I are the
elements of first ro of Adj. A.
S'ed E 7thp1es on § 508.
If A= 1 2 41, rind Adj. A.

9 10 12
(Avadh 95)
Sol. For the given matrix A, we have
C 1 = 7 8 =4; C 12 =- 5 8 =12;C 13 = 5 7 =-l3;
10 12 9 12 9 10

C21 - 2 4 =16; C22 = 1 4 =-24; C23 -1 2 8;

10 12 9 12 19 10

C31 = 2 4 =-12; C32 =- 1 4 =12.'C33 = 1 2 =-3

78 58 57

C= 4 12 -13
16 -24 8
-12 12 -3

Adj. A=C'= 4 16 -12

12 -24 12
-'13 8 -3 Ans.
*Ex. 1. (b). Find the adjoint of the matrix A = 1 0 - 1
3 4 5
0 -6 -7
Sol. Do as Ex. I (a) above. 2 -6 4
21 -7 -8
-18 6 4Ans.
ç, 4x,4. Find the adjoint of the matrix A = 1 1 1

2 - 1 3
(Kanpur 96)
Sol. For the given matrix A. we have
 Adjoint of Matrix 45

C 11 = 2 -3 =3; C 12 1 -3 -9 C 13 1 2 =-5;
-i 3 2 3 2-1

C2 = - =-4; C22 = 1 1 I; C 23 I 1 =3;

-1 3 2 3 2 -1

1 1 =-5;C32 - i I =4;C 3 1 1=1

C 31
2 -3 I - .3 1 2

C= 3 -9 -5
-4 1 3
-5 4 1
dthe3 -4 -5
Adj. A=j
9 1 4
3 1 Ans.
Ex. 34)adjoint of the matrix A, if
A=11 3
0 1 -1
20 4

Sol. For the matrix A, we have

C 11 1 -1 =4; C 12 - 0 -1 =-2; C 13 = 0 1 1 =-2;.
o 4 24 20
C21 - 1 3 =-4; C72 = 1 3 =-2; C23 =- 1 1 1 =2;
04 24 20

C31 = 1 3 =-4; C32 =- 1 3 = 1; C 3 = 1 1 = 1

I-I 0-1 .01

C= 4 -2 -2
-4-2 2
-4 1 1

..Adj. A=C'= 4 -4 -4
2 -2 1
-2 2 1 Ans.

Ex. 4. Find the adjoint of A = 5 0 0 2

1102
0021
1001

Sot. For the matrix A, we have

=2 1=2C 12 = - =-. g =2;

001 001 101

46 Matrices
C 3 = =_ i i = 1;C 4 =- 1 1 0 =0 2=-2;
101 1 00
C21 0 02=0; C22 5 02=25 2=6
0 2 1 0 2 1 1 1
jooi 101
C- 5 0 2 = 0; C24- 5 0 0 =0;
001 002
101 100.
C3 0 0 2 = 0; C32 - 5 0 2=0;
102 102
001 101
C33 5 0 2 = 5 0 2 = 5 2 = 3; C34 50 0 =0;
12 011 11 111
• 101101 100
C41 0 0 2-2 1 0 -4;
1 02 0,2
021
C42 5 0 2 =-2 5 2 =-,16
102 12
021
C43 50 2 =- 5 2 =-5;
112 01
001
C= 5 0 0 =2 5 0 =10
110 11
002
.. C= 2 2 1 -
0 6 0 0
00 3 0
-4 -16 -5 10

Adj.AC'= 2 0 0 -4
2 6 0 -16
103-5
-2 0 0 10 Ans.
EL S. Verify that the adjoint of a diagonal matrix of order 3 Is a
diagonal matrix.
• SoL Let A be a diagonal matrix of order given by
 Adjoint of Matrix 47

A=a 00
ObO
OOc
Then for the matrix A we have
C 11 = h 0 =bc; C 12 =— 0 0 =0; C 13 = 0 b =0;
10 Oc 00

C21 =— 0 0 =0; C22 — I a 0 1 =ac; C23 = a 0=0;

10 0 00

C31 = 0 0 =0; C32 — a 0 =0; C33 a 0 ab

b 00 0

C=bc 0 0
0 ca 0
0 Oab
Adj. A = C '= bc 0 0 . which is evidently a diagonal matrix
0 ca 0
0 0 ab Hence proved.
Ex, 6. If A = 1 2 31, rind A2 - 2A + Adj. A
0
2 4 3] (Agra 95)

SOLA 2 1 2 3x I 2 3
050 050
243 243

= 1+0+6 2+10+12 3+0+9 7 24 12

0+0+0 0+25+0 0+0+0 0 25 0
2+0+6 4+20+12 6+0+9 8 36 15

Also C 11 = 5 01 = 15,C, 2 =- 10 0 = 0; C 13 = 0 5 = - 10
43 23 24

C 21 — 231=6,C22=11 3 =-3; C = 1 2 =0;

43 23

C33 2 3 =-15,C32 = 1 3 =0; C33 I =5

50 00. 25,

C= 15 0 —10
6 —3. 0
—15 0 5

Adj. A=C'= 15 6 — IS
0 —3 0 ...(ii)
—10 0 5


48 Matrices 182/11/3
A 2 - 2A + Adj. A
7 24 12 -2 1 2 3 + 15 6 - 15
o 25 0 0 5 0 0 -3 (1
=[ 8 36 15 2 4 3 -10 0 5
from (i) and (ii)
= [7 24 12 - 2 4 6 + 15 6 - IS
025 0 0 10.0 0-3 0
8 36 15 4 8 6 -10 0 S
=7-2+15 24-4+6 12-6-15
0-0+0 25-10-3 0-0+0
8-4-10 36-8+0 15-6+5
= 20 26 -9
0 12 0
-6 28 14 A ns.
Exercises on § 508
Find theadjoint of the following matrices
Ex. 1. -1 -2 3
9 -11-8
-2 21 Ans.8 -14 -5
4 -5 2 2 -13 -6
Ex. 2. 2 -1 3 7 9 -10
-5 31 Ans. 12 15 - 17
-3 23
Ex. 3. -4 -3 -3
-4 -3 -3
1 0 Ans. I - 0 1
' 4 4 3
Ex. 4. [1 5 7 3 II -16
231 Ans. 0 -26 13
32 -6 17 -7
Ex. 5. 2 3 '1
1 -5 7
123 Ans. 7 1 -5
31-2 -5 7 1
Ex. 6. 3 3 4
-7 1 24
234 Ans. -2 3 •-4
011 2 -3 -15
Ex. 7. 2 1 3
1 2 -3
020 Ans. 0 -2 0
211 -2 2 2


182/11/4 Adjoint of Matrix 49

Ex.8. 0 I l' 8 -5 2
120 Ans. -4 -3
3 I 4j 3 -1
Ex. 9. 1 2 3 2 -1
2 3 4. Ans.[2 -4 2
3 4 5j 2 -1
Ex. 10. 0 1 1 0
1
I 0 1 Ans. 0 -I
1 1 -I
Fx.11. I 2 3 3 -3
456 Ans. 0 -9 6
679 5 -3
Ex. 12, 1 2
0 5 0
A •
(Agra 95; Bundeik/zand 92; Gari,wa('I')
V '
Ans. 15 6 -15
0-3 0
- 10 0 5
Ex. 1. [- i -2 31 Ans. 7 - II -5
H2 l 0 JO -5
-5 2J 14 3 -5
*Ex 14. I C) 21 Ans. I 4-2
2 I 0 -2
3 2 ij --2
L
§5 09. Theorems on Adjoint of a Matrix.
**Theorem 1. if A . ja j j I . e an n X n matrix, then
A • (Adj A) (Adj A) • A A I • I,: where tn is .-zn n X n identit y mairLr.
(Agra 94. 91: Aradhi 94, 92, 90; IIundelA-ha,zd 94, 93; Garhwa! 90,
Gorakiipur 97. 92; Kanpur 96; Mecrut 91: Purvanchal 95: R/ilIA/:and W);
Proof. We know Adj A [C ].
where Ckj is the cofactor of a, in f A I and C C.1.
Therefore A • (Adj A) jj,) [C ',I
[B, say. .. Ill
14

where B, = i a.0 j k 0J
J-r J=I
= IA I. ifi=A
if See § 405 and § 409 in CL


50 M.iriccs
From (i), (1, k)th clement of A • Adj A) A . or 0 acordng as
ik or i*k.
i.e. All diagonal terms of A • (Adj A) aie I A and non-diagonal terms are
zero,
A.(Adj.A)= A) () 0 .... 0
o A) 0 .... 0
0 0 Al... 0'

o o 0 .... A)

= A ) I 0 0 . 0 , See Chapter 1
o I 0 ... 0
o o I .. 0

() .......
=IAI . 1 (ii)
Similarly we can prove that (Adj. A) • A - A I • I (iii)
Hence from (ii) and (iii), we get
A • (Adj. A) (Adj. A) • A =JA I • I

or

or A AAL
I=A tA
JAI

e. The inverse of A = -- A
i.e. (iv)

Note : The result (iv) gives us another method of finding the inverse of 'a
given matrix.
* *Theom 11. If
A = I a J I be an /2 X fl mU! rrr, then
Adj A I = I A n - if A I r 0. (Agra 96; Gorakhpur 92; Ruliilklcand 99, 91)
Proof. We know that I A • I B I = I AR I ...,See ('h. on Determinants
)ANIAdjAI=IAAdJ.Al
= JAI 0 0 .... 0],as proved in
0 A
I j C) .... 0 Theorem I above
() 0 IA) .... 0

.0 0 0 .... Au

or I Al.lAdjAI=)IA)) (Note)
Dividing both sides by I A j. since I A I ;e ft we get
fl - .
I Ad) A = A Hence proved.
III. If A timid B we too ii x ii ,mia!rice.i. 1/1(12
Ad) (AR) = (fldj B) • (Ac/f A). (Agra 93; Ro/iiIIw,id 98; Gnokhpur 98)


Adjoint of Matrix 51
Proof. We know A. (AdjA ) = I A J .I ... Sec Th. I Page 49Ch. V
So we have (AB).(Adj AB) =JAB!.1
Now (AB) • (Aj B) • ( Adj A)
= A • B • Adj B • Adj A
=A s ( B . AdjB).(AdjA) (Note)
A . ;BI.1 . ndjA, .. B.AdjB=IBJ.I

=A 1 B 1 .A (IjA , I'fldjA=AdjA a s, I.A=A always

=BJ . A . AdjA (Note)

= I B H A I .1 A.AdjA=IAI.t.
= I A I •I B I • 1.
=JABj.1. AI.IBI=IAB! ...(ii)
From (I) and (ii) we get
(AB) • (Adj AB) = (AB). (AdjB) . (Adj A)
or Adj (AB) = ( Adj B) • (Adj A). Hence proved.
Solved Examples on § 509.
I (a). For the matrix A given in Ex. 2 Page 44 Ch. V verify the
theorem A • (Adj A) = (Adj A) • A = A J . I.
Sol. In Ex. 2 Page 45 Ch. V. we have proved that if
A= II 1, then Adj. A = 3 -4 :5
1 2-3 -9 1 4
[2-1 -5 3 1
3J
A.(fldjA)= 11 1 x 3 -4 -5
1 2 -3 -9 I 4
2 -1 3 -5 3 I
= 3-9-5 -4+1+3 -5+4+1
3-18+15 -4+2-9 -5+8-3
-8-1+9 -10-4+3
L6915
=-1] 00
0 -11 0
0 0 -Ii
Also(4JjA).A= 3 -4 -5 x I I il
-9 I 4 1 2 -3
-5 3 1 2 -I 3
=3-4-10 3-8+5 3+12-IS
-9+1±8 -9+2-4 -9-3+12
-5+3+2 -5+6-I -5-9+3


Matrices 9,)9
¼
=-Il C)0 ..(ii)
o -ii o
o o -ji
Also IAI= II I 1 0 o replacing C2, C1

by
(iii)
I 2 -3 I I -4 C2-C1,C--C1
2 -I 3 2 -3
=- 1 -4 =1-12=-lI
3 1
From (i), (ii) and (iii) we get
.
A.(AdjA)=(Adj.A) 'A=-Il 1 0 0
010
0 0 I (Note)
(-lI)l 1 jA . I. Hence proved.
Ex. 1 (b). Find the adjoint of the matrix A =i 2] and verify the
13 -Si
theorem A • (Adj A) = (Adj A) • A A II. (Rundelkhand 93)
SoL For the given matrix A, we have
C11=-5,C12=-3,C21=-2,C22=1
C=[-5 -3

And soAdjA=C'=[-5 -2]

[-3 1]
A.(AdjA)=[l 2]4-5 -2
L3 L-3 _5]
1
=[-s-6 - + 2 21 = 1- Il 0
+ L- 15 15
-6-51 L o -ii
Also (AdjA).A=[-5 -
[-3
l 2 2] 1
JI
=1- 5 - 6 -io+iol=r-ii ...(ii)
[-3+3 -6-5 j [ o -ii
AIsoIAl= I 2=-5-6=-11
3
.. From (i) and (ii), we get
A.(AdjA)=(AdjA) A=I- II . 0
[ 0 -11
= - - 1I 11 0]=IAI12=AI1
[0 1]
Hence proved.


Inverse of Matrix 53
Ex. I W. Verify the theorem A • (Adj. A) = (Adj. A) • A
=IAI.IwhenA= 2 -1 3
5 31
3 23
Sol. Do as Ex, I (a) above.
Ex. 2 (a). Find the inverse of A =. 1 2 3
245
3 5 6 (Agra 91)
So!. For the given matrix A, we have
C= 4 5 -1;C 12 - 2 5 =3;C 1 2 4 =-2;
59
36 35
- 2 3 =3:C 22 = I 3 =-3;C23 = - 1 2 =1;
C21
56 36 135
C31 = 2 3 =-2;C32 =- 1 3 =1;C33 = 1 2 =0
4 5 25 24

C= - l 3-2
3 -3 1
-2 1 3

Adj. A=C= -1 3 -2
3-3 1
-2 1 0

Also A = i 2 3 = i 0 0 replacing C2, C3 , by

2 4 5 2 0 -1 C2-2C1,C3-3C1
3 5 61 -1
= 0 -1 =-1
-1 -3

IA 1

L
1
-
3 -2

0
=-I 2 -1 0 Ans.
Ex. Z).z1mndthe inverse ot the matrix A= 3 -2 -1
-4 1 -1
2 0 1 (Agra 96)
Sol. Here I A J 3 - 2 - 1 = 0 0 - 1
-4 1 -1 -7 3 -1
2 0 1 5 -2 1
replacing C 1 , C2 by C 1 + 3C3 , C-, - 2C 1 respectively.


54 Matrices
= - 1-7 3 , expanding w. r. to R1
5 -2

--[14--151=i
Also for the matrix A, we have
C 11 1 -!
C12 1-4 -J = 2; C 13 = -4 1 =-2;
U I 2 1 20
C21 - - 2 1 =2; C22 3 - =5; C 3 = 3 -2 -4;
0 1 2 1 2 0
C3 = - 2 -1 =3:C32 =- 3 -1 7;C 13 = 3 -2 =-5
1 -1 -4 -1 -4
C=l 2-2
2 5 -4
3 7 -5
Adj. A=C'= 1 2 3
2 5 7
-2 -4 -5

A_ 1 _ dj A _ 1 2 3
- jAJ 2 5 7

Ans.
Ex. IldtIlnvcrseofA

(Avadh 98, 91; Purvanchal 96)

Sol. Here A = I 2 3 = 1 2 3
I J

134 011
143 020

replacing R2 , R3 by !?2 - R 1 , R 3 - R1
or IAI= 1 1=-2
20 (•)
Also for the niatrix A, we have
C 11 = 3 4 =-7; G 12 =- 11 4 =1; C 13
= 1 3 1;
43 13 14
C21 =- 2 3 = 6; C22 = I 0; C 23 - I 2 -2;
43 13 14
C3 = 2 3 =-1; C32 =- 1 3 = -1;C33 = 1 2 =1
34 14 13


If) verse of Matrix 55

I 6 (1 -2
L- -
Adj.A=C 7 6 - H
I I
2
-t AJ.A
Al
-7 6 -i -3
- 2 L0 l 1
r)
-2 1 2
-
Ans
Ex. 3 (b). Find the inverse of the matrix A = 1 2 3
2 4 5 (A'c4h 92.
356

Hint: Do as Es. 3 (a) above. Ans. 1 -3 2

-3 3 -1
2 -1 0

Ex. 3 (c). Find the adjoint of the matrix 4 2 31 and henec

ii 3 5
Li 5

evaluate A . ( Kwnaun 94,
Hint. Do as Ex 3 (a). above. Ans. - 3 1

Ex. 3 (d). Find the inverse of A = 1 2 3

H
050
203
Flint. Do as Es. 3 (a) ahose. - -6 - IS
o 3 o:
-tO 4 51

Ex. 4 (a). Find the adjoint of the matrix A and evaluate A, where
A=T2 2 2
2 5 5
5 11
L2


56 M a tri ces

Sot. Here for the Inatri'c A. we have

C 1 5 5' =30;C12=- 1 2 5 1 = 12C 13 2 5 =0;
5 11 12 iii 2 5
C21 - 2 2 = - 12C22= 12 21 I8C23 =- 12 2 =-;
5 li 2 11 25
C31 = 2 2 =0;C 2 =-
12- 2 H -6;C 33 = 1 2 2 =6
55 2 5t . 25

C= 30 -12. 0
-12 18 -6
0 -6 6
Adj. A=C'= 30 -12 0
-12 18 -6
0 -6 6
A ns.
Also l A J = 2 2 21 2 0 oI applying C2 -C 1 , C3 --C1
2 5 5 2 3 3
2 5 II 2 3 c
=2 3 3 =2127-91=36
39
Adj. A - 30 -12 0
As-
- IAI 361- 12 18 -6
0 -. 6
[
5 -2 0=[ 5/6 -1/3 0
36 1 -2 3 -1 -1/3 1/2 -1/6
-i i 0 -1/6 1/6
[ L A ns.
Ex. 4 (b). Find the inverse of matrix A, where
A =1 4 3 31

- i
L - (Agra 94)
Sol. Here J A = 4 3 3 = 1 3 3 replacing C 1 by
-I • 0-1 0 0-1 c1-c3
--4 -' -3 , -I -4 --3
= - I 3 expanding w.r. to R2

=-4+3 -
Also for the matrix A, we have


Inverse of Matrix 57

C 11 =' 0 -I=-4:C1,=- - -1 =1:C='-1 0 =4;

--3 -.4 -3 -4 -4

C 21 =- 3 3 =-3:C2 7 14 3 =O;C 2 -[ 4 3 -4-

[-4 _

C 3 = 3 3 C32 4 31 =1:C 33 = 4 3 =3
-I -I -1 -1 0
C=-4 14
-3 0 4
-3 1 3
Adj. A=C=-4 -3 -3
I 0 I ..(ii)

L4
1 -4 -3 -3fm(i)and(ii)
A -- - 0 1
Al I 1
4 4 3
=1 4 3 3
--1 0-I
[-4
Ex. 5 (a). Find the inverse of A =11 - 1 1
Sot. Here I A I
1 - 1 i = lo -. 1 replacing C 1 , C1
4 1 0 5 1 1 by C 1 + C1, C 3 C2
$ 1 1 j 9 1 2
or Al= 5 1 =I0-9I#0
9
Also for the matrix A, we have
C 1 I 011:C 22 - 4 =-4:C 13 = 4 01
1 i 8 1 8 ij

C 2 =- -1 1 =2:C 22 1 1 =-7;C23 =- I -1 =-9:

Ii 81 3 1

C 31 -1 1 I.C32 1 1 =4:C 33 = 1 -1 =5
10 40 4 1

c=[ 1 -4
2 -7 -9
-1 . 4 5
-18


58 Matrices

Adj. A=C'= I 2 -I
-4 -7 4
-4 -9 5

2 -I ,from(i).
A t - - -4 -7 4
-4 --9 5
A ns.
E. S (h). Find the inverse A = 1 2 4
578
9 10 12

Hint Do as Ex. 5 (a) above.Ans. - 2

[
Ex. 5 (c). Find the inverse of the matrix A = 1 3 - 3
143
134
Hint : Do as Ex, 5 (a) Page 57 Ans. 7 - 3 -3
-I 1 0.
1 0
Ex. 6 (a). Find adj A and K when A = 1 3 3
143
134
(BundeliJiand 94; Kanpur 93)
I
Sol. Here A = 1 3 3 = '1 3 3 . replacing R 2 , R3 by
1 4 3 0 1 0 R 2 - R 1 , R- - k 1 respectively.
134 001
=10=1
0 1
Also for the matrix A, we have
C 11 = 4 3 =7;C 12 =_ 1 3 =-I;C 13 = 1 4 -l;
34 14 13
C 21 - 3 3 1 3;C22= I 3 =l:C23 = 1 3 =0;
34 14 13
C31 = 3 3 =-3:C12 =- I 3 1 =O,C= I 3 =1.
43 13 1 14


Inverse of Matrix 59

= 7 -I -1
-3 1 0
-3 0 1

AdjA=C= 7 -3 -3
-1 1 0 ...(ii)
-1 0 I

7 -3 -3from(i)and(ii)
1A11 0 Ans.
J
-1 0 1]

Ex. 6 (b). Find the inverse of the matrix A = 1 1 1

011
o 0 1 (Lucknow 91)

Sol. Here IA = 1 1 1 = 1 I , expanding with respect to C1

01 I 0 1
001
or AI=l-0=1.
Also for the matrix A, we have
C1 = 1 I I; C 12 =- 0 1 0;C 13 = 0 iI=o
0 I 0 I 0

C21 =- 1 1 =-1;C22 = I J 1 =I;C23 =- 1 =0;

01 01 00

C 11 - i = 0;C1 2 = - 1 =-1;C11= I 1 =1
1 I 01 01

C=1 0 0
-1 10
0 -1 1
Adj. A=C'= 1 -1 0
0 1 -1
0 0 1 ...(ii)
- I - I 0 . from (i) nd (ii)
A1
- Al 0 1-1
0 0 1 Aas.
*Ex 7 (a). If A = 3 - 1 11,, rind aij. A and A1.
-15 6 -51
5 -2 2] (Garhwal 95, 91; Meerut 95)
Sol. Here A I


60 Matrices

3 - i I I0 -1 o replacing C 1 , C3 by
=
-15 6 -5 I3 6 1 C1+3C1,C2+C1
5 -2 21 1 -1 -2 0
= 1_31=1
1 0
Also for the matrix A. we have
C = 6 -5 •2;C12- -15 -5 C13 -15 61 0
-2 2 5 2 5-21
C21 - - i I =0;C 2 = 3 1 =I;C23 =- 3 -1 =1;
1-2 2 5 2 5 -2
C31 -1I =-I;C32=- 3 1 =0;C3 = 3 -11=3
6 -5 -15-5 -15 6]

C=1250
Oil
0 3

Adj. A=C'= 2 0 --1

51 0
• 01 3

- 2 0 -l],fmm(i)
A--- -
- IAI 5 1 0
0 I Ans.
Ex. 7 (b). Find the inverse of the matrix
A= 1 2 -2
-1 3 0
0 -2 1 (Gorak/ipur 97)

Sol. Here I A = 1 2 -2 = 1 2 -2 ,replacing R2

-1 3 0 0 5 -2 byR2+R1
0 -2 1 0 -2 I

= 1_5 -2 =5-4=1. -
2 1
Also for the matrix A, we have
C 3 0 =3:C 12 =- -1o]=1;C 13 = -I 3 =2;
1-2 l 0 1] 0 -2

C21 - 2 -2 =2;C 2 = I -2 = 1;C'23 = - 12 =2: I

-2 1 0 1 0 -2


Inverse of Matrix 61

C 31 = 2 -26; C 32- 1 -2 2:C 33 = 1 2 =5

3 -1 0 -13
01
C3 I 2
212
625
Adj. A=C'=3 2 6
1 1 2
22 5 ...(ii)
- AdjA - 3 2 61, from (i) and (ii)
Al I I 21
2 2 5 Aiis.
Ex. 7 (C). if A = 1 4 01, find A1•
-1 2 2
L002
Hint Do as Ex. 7 (u) Page 59. Ans. (1/6) 2 -4 -4
I 1 -1
0 0 3
*Ex 7(d). 1111 0 - i1, find adj. A and A'
13 4 5

(fli'a(ih 94)
Hint Do as Ex. 7 (a). Page (. Ans. (1/20) 2 6 4
21 -7 8
-18 6 4
Ex. 8 (a). Find the reciprocal or inverse of the matrix
A= 2 1 2
2 2 .1
1 2 2 (Kunaun 92)
Sot. Here I A = 2 1 2 = 0 - 3 - 2 , applying
2 2 I 0 -2 -3 R 1 -2R3,R2-2R1
122 I 2 2

I-3 -2 =9-4=5
-2 -3
=
Also we have
C1 2 I 1 2:C2=- 2 1 =-3:C 13 = 2 2 =2;
22 12 12

c21=-11 2 =2;C22 = 12 2 =2:C21 =- 2 1 =-3;

2 2 2 12


62 Matrices
C 31 = 1 2 =-3:C 32 =- 2 2 =2;C33 = 2 1 1 =2
21
1.2 2
21

.C= 2 -3 2
2 2 -3
- .3 2 2

Adj. A=C'= 2 2 -3
-3 2 2
2-3 2
Reciprocal of A = A'
Adj.A i 2 2 -3
- Al 5 -3 2 2
2 -3 2
A ns.
Ex. 8 (b). Find the adjoint and inverse of A = 1 23
2 32
3 3 4)
Hint : Do as Ex. 8 (a) above. Ans.1_ I 5
'7 7 7
5 .4
7 7 7
3 3
7 7 7_
Ex. 9 (a). Find the inverse of A = 1 2 1
323
112

(Avadh 90; Bundelkhand 96, 95:

Gari1wal 96, 94; Gorakhpur 96; Purvanchal 97)
Sol. For the given matrix A, we have
C 11 = 2 3 =I:C 17 =- 3 3 = -3;C 13 = 3 2 =1;
12 12 21
C21 =- 2 I =-3:C, 2 = I I 1:C23 =-I 21 = 1;
I 2 1 2 I I
C31 = 2 1 =4:C32.=-.II I =0;c 3 = 1 =-4
2 3 3 3 3 2
V C= 1-3 I
-3 1 I
4 0 -4

Adj. A=C= I -34

-3 1 0
1 -4


Inverse of Matrix 63

A = I 2 1 = I 2 0 , replacing C- by C 3 - C1
and
323 320
112 III

= 1 2 =-4
32

The inverse of A = Adj

=- 1 -3 4=_I 1_j
-3 1 0
I -4 - 0
I
1 I
' Ans.

Ex. 9 (b). Find the inverse or the matrix A = 1 2-1

-1 1 2
2 - 1 1 (Meerut 9

i 2 - replacing R 2 , R 3 by
Sol. Here I A = i 2 -
-1 I 2 0 3 1 R2+R1,R3-2R1
2 - I I 0 - 5 3 respectively.

H-53 11 =9+5=I4

Also fur the matrix A, we have

C 11 I 2 =3:C 12 -1 2 =5;C 1 -1 1 =-I;
--I 1 2 1 2 -1

C, 1 =- 2 -1 1, C22 1 -1 =3;.C2 = - 1 2 =5;

-1 I 2 I 2-I

C3 = 2 -1 =5;C 1 ,- 1 -1 =-1C 3 I 2 =3
I 2 -I 2 -I 1

-1
I-I 3 5
[s_i 3

Adj. A=C'= 3 -1 5
5 3 -J
-1 5 3...(ii)

- 3 - 1 5 , from ( I ). ()
A =
Al 14 5 3 -1
-1 S I Ans.


64 Matrices 182/11/4
Ex. 9 (c). Find the inverse of the matrix A I 2 3
2 3 1
3 1 2J
(Purvw,c/,a( 95)
Hint : Do as Ex. 9 (a) above.
Ans. -5 1 7
I 7 -.5
- 7 -5 I•
Ex. 10. If A = ri - I I find
, A 2, and show that ,2 = A1.
2 -1 0
1 00
Sol. 2I._i IxI —11
2 -I () 2 -I 0
1 0 0 I 0 0
1-2±1 -1+1+0 1+0+0]=[0 0 I
2-2+0 -2+1+0 2+ 0 +01 0 12
1+0+0 -1+0+0 I+0+(lj
Also A i - i 1 I n 0 . replacing C2, C3 by
2 - I 0 2 I - 2 C, C 1 . C - C 1 respectively,
I 0 0 1 1 -I
= ji -2 =l +2-I.
i -i
Also for the matrix A, we have
C11r -1 0 0;C 12 =- 2 0 0;C 13 = 2
00 10 1 0
C21=-- l!=0:C22=tI 1-I;C,3=-[t
00 1 10 10
C31 = -1 1 1:C,=- I I =2:C33=
-1 0 2 0 2 -1
C=0 0 1
0 -1 -1
1 2 I

AdjA=C= 0 0 I
0 -1 2 (iii)
1 -1 I

- - 0 0 ii. from (ii) and (iii)

- IAI 0 -I 2
--I
182111/5 Inverse of Matrix 65

= A 2 .from (I) Hence proved.

Ex. 11. Find the adjoint of matrix A and hence find A'.
cosO — sin O0
A= sinG Cos O, 0
0 0 (Meeru( 96)
1
Sot. Here J A
cos B - sin 0 0 = 1 cos O - sin 0 expanding w.r. to C3
sin 0 cos 0 0 sin 0 cos 0
0 01
= cos2 e - (— sin 9) 1
Also we have
C 1 1 cosO 0 =cos0;C 12 =— sinG 0 =—sin0;
0 1 0 1 1

C 13 sine cosO rr OC2 1 =— —sinG 0 = sin 0;

0 0 0 1
C22 cosO 0 =cos0;C=— cosO -'sinG =0;
0 1 0 0

C31 = I —sinO 0 =0;C32 =— c 0 =0;

cosO 0 sin 4) 0

C33 cosO —sinG = I

sin cosO
^

C= cosO — sinG 0
sin cosO 0
0 0 1

Adj. AC' cosO sin 0

—sinG cosO 0
o 0 1 Ans.

And A' Adj. A cos 0 sin 0 01, from (i)

— A! — —sinG cos OOl
0 0 Ans.
*Ex. 12. How will you use the notion of determinant to compute the
inverse of a non-singular square matrix? Compute the inverse of the matrix
A= 1 2 3
45 6
7 8 10
Sot. For the first part See Theorem I. result (iv) Page 50 of this chapter
For the second part we have for the matrix A


S6 Matrices
C 11 5 6 = 2;C 12 = - 4 6 =2;C 13 = 4 5 =-3;
8 10 7 10 78
C21 =- 2 3 =4:C22 1 1 3 =-11;C23 =- 1 2 =6;
8 10 7 10 78
C31 = 2 3 = -3;C 32 = - 1 3 =6;C33 1 2 =-3
56 46 45

C= 2 2 -3
4 -II 6
-3 6 -3
Adj. A=C'= 2 4 -3
2 -11 6
-3 6 -3
Also A = 1 2 3 = 11 0 0 replacing C2, C3 by
-3 -6 C2-2C1,C3-3C1
4 5 6 4
7 8 10 7 6 -11 respectively.
-3 -6 =33-36=-3
-1 -6 -II
2 4 -3
AI - 3 2 -11
I 6
-3 6 -3
= 2
3 3
2 U
2
3 3
1 -2 1
Ans
*Ex. 13 If A' denotes the transpose of a matrix A and
A= 1 -2 3lr.nd(A5'
o -1 4
-2 21

Sd. A'= 1 0 -2 by definition of transpose of a matrix

-2-1 2
3 4 1
= B (say).
Now 18 1 0 -21 0 0
-2 -I 2 -2 -1 -2
3 4 1 3 4 7 replacing C3 by C3 + 2C1


Inverse of Matrix 67
= - I - 2 expanding with respect to R1
4 7
=(-I)(7)-(-2)(4)=--7+8=1#i
Also we have
C 11 = - I 2 =-9;C 12 =- -2 2 =8;C13 -2 -1 =-5:
41 31 3 4
C21 =- 0 -2 =-8C22 = I -2 =7;C23 = - 1 0 =-4;
4 1 3 1 34
C31 = 0 - 2 =-2;C32=- 1 -2 = 2;C33 = I 0 -I
-1 21 -2 2 -2-1
C=-9 8-5
-8 7 -4
-2 2 -1
Adj. B C' -9 -8 -2
8 7 2
-5 -4 -1

WiAdj.BJ9 -8 -2
IBI 1 8 7 2
[-5 -4 -I
-9 -8 -2
8 7 2
-5 -1 A.
*Ex. 14. Find the inverse of the matrix A, where
A= 1 0 -4
-2 2 5
3 -1 2
Hint: Do as Ex. 12 Page 65. Ans. (1/25) 9 4 8
19 14 3
-4 1 2
Ex. 15. Find the adjoint and inverse of the matrix
Icosa - sinai
Lsin a COS a]
(Bundelkhand 92)
Sot. Here I A = cos a - sin a
sin CO5 01
=Cos 2 a+ sin 2a=1*0.
C 11 = COS a,C 12 =- sin a,C21 =-(- sin a)= sin a and C22 = Cos a (Note)


58 Matrices 999

C=Icosa —sin a]
L sina cosaj
Adj. A=C'=I cosa sin
L- sina cos a
A-1 AdjAr
= A = cosa sin
[-sina cos a
substituting values from (1) and (ii). Ans.
EL 16. FInd the Inverse of A0 1 2
123
3 1 1 (Meerut 91S)
5.1. For the given nai,rix A, we have
C 1 2 3 =-1;C 12 =- 1 3 =8;C13= I 2 =-5;
11 - 31 31
C21 =- 1 2 1 . =I - C22 = 1 0 2 =-6;C23 =- 0 1 =3;
1.1 1 31 3 1
C31 = 1 2 = -1;C32 = - 0 2 =2;C33 = 0 1
23 13 12
-1 8-5
C= 1-6 3
-1 2 -1
Adj.A=C'=[-1 1 -1
8 -6 2
-1
and AI= 0 1 2 = 10 I 0 ,replacingC3byC3-QC2
1 2 3 1 2 -1
3 1 1 3 1 -1
or IAI=— 1 -1 =-2*0
3 -1

Inverse ofA= Adj. A t[- 1 -I -t

2 2
I 8-6 2 -8 6-2 Ans.
-5 3 -1 5 -3
Ex. 17 (a). Find the inverse of the matrix A = i -1 2i over the
2 02
-1 0
fkU of the complex numbers.


Inverse of Matrix 69

Sol. Hre A = - i 2j = - - i: 2z , interchanging C1

2 0 2 0 2 2 and C2
-1 01 0-Il

or I A I = 22 1, expanding with respect to C1

-1 I
=2-(-2)=40.
Also for this matrix A, we have
C 11 0 2 =0;C 12 =- 2 2 =-4;C 13 = 2 0 =0
0 1 -1 1 1-1 0

C2 1 =- -1 2i =1;C22 = i 2i =3i;C 23 = - j
o 1 -1 I -1 0

C 31 - - 121 =-2;C32 =-- i 2i =2LC33 I -I =2

02 22 2 0

C- 0 -4 0
1 3i 1
-2 2i 2
Adj. A=C'= 0 1 -2
-4 31 2i
01 2

0 1 -212 =- 0 -
2
- IAI 4 -4 31 2i
0 1 2 -1
A I I
U A.
*EL 17 (b). If A = 0 0 ii, then show that A - A
0 1 01
1 0 0] 91)
Sol. Here I A I = 0 0 I = 0 1 =-1
010 10
I 0 0 ...(i)

Also for the matrix A, we have

C 11 = 1 0 =0;C 12 - 0 0 0;C13 = 0 1 =-1;
00 10 10

c21 =- 0 1 1 =0;c22 = 0 i =-1;C23 = - 0 0 =0;

00 10 10

C3I = 0 1 1 =-1;C32=- 1 0 1 =0;C33 = 0 0 =0

10 00 01


70 Man-ices
C= 0 0-1
0 -I 0
-1 0 0
Adj. A=C'= 0 0 -1
0 -1 0
-1 0 0
A_1=AdjA=L 0 0 -I = [O 0 1 =A
AI 1)
0 -1 0 0 1 0
-1 0 0 100
Hence proved.
Lx. 18. Find the inverse of the matrix
a+Ib C+kIl,ifa2+b2+c2+d2=1
Al [-c+id a-lbJ
Sol. For this matrix, we have
C11 -a-ib; C12 =-(-c+ id) =c--id,
C21=-(c+iif);C22=a+ib
C1[ a-lb c - id
-c-id a+th
Adj. AC'=a-ib -c-id
Ic-id a+ib
Also J A l = a+ib c-i-id
-c+id a-lb
=(a+ib)(a-ib)-(c+ id) (-c-t-id)
a2_i2b2+c2_i2Ira2+b2+c2+d2=l^0.
InverseofA= AdjA Ia-ib -c-id=1
!AI L" a+:b
Ans.
**Ex. 19. If a + i3 = a 0 verify that

(a+i'=1 cX
[- aj
Sel. (a +i) = - (a - in)
a+: ((X+)((X-43)
multiplying num, and denoni. by a - i
= (a - i)/(tz2 + 2 ) (i)

Again let A=1 a

a


Inverse of Matrix 71
Then Az a 131 =a(a)-13(-13)=a2+132#O.
...
(ii)
1_0 a
Also for. matrix A. we have
C 11 =(X, C 12 = 13. C21 =-0; C22 = a
a 13landAdi A=C'=Ia -
[ _ 13 a j [13 a

ía -
i Adj.A, [13 a
A = = 2,from(ii)
IAI cz2+13

i.e. a 2 + 132)
=Fa -0]-(a
[-13 aj [13 a
Also we are given a+43=[ a 13
[-13 a
Replacing 13by _13we get (a_i13)=I a -13
ft a
From (i) we have (a + ( X - + 132)
[13 a]
Hence from (iii) and (iv), we have

( a + i13)-1 Hence puu'ed
= ] '
Ex. 20. If A ='-- 1 0 0
2 , find A 1 .
-9 1 1 0
1 0 2 -1
-4 1 -3 1

Sot. Here IAI= -1 0 0 2 = -10 0 0,

-9 1 0 1 -9 1 0 -17
1 0 2 -1 1 0 2
-4 1 -3 1 -4 1 -3 -7
I I -

replacing C4 by C4 + 2C1
1 0 - 17 , expanding with respect to R1
o 2 1
1 -3 -7
1 0 0 , replacing C3 by C3 + 17C1
0 21
1 -3 10


72 Matrices
2 I =-120+31=-23*0.
1_
3 10
Also for the matrixA, we have
C11 10 1 = I 0 0 2 -1 =-3;
0 2 -1 0 2 -I -3 0
I -3 1 I -3 0
C12 =- -9 0 1 - 0 0 1 =- -82 =- 14;
1 2 -1 -8 2 -1 5 -3
-4 -3 1 5 -3 1
C13 -91 1 0 1 0 - I -1 -'5;
1 0 -1 1 0 -I 5 0
-4 I 1 5 0 0
C14 = - 9 1 0 =- 0 10 = 1 2 =-13;
1 0 2 I 0 2 5 -3
-4 1 -3 5 1 -3
C21 - 00 2 =- 0 2 =4;
0 2 -1 2 -1
1 -3 1
Cfl = -I 0 2 = -1 0 0=- 2 I =11;
1 2 -1 1 2 I -3 -7
-4 -3 I -4 -3 -7
C23 -I 0 2 =- -1 0 0 = 0 I =-1;
I 0 -1 1 0 I 1 -7
- -4 1 1 -4 1 -7
C=-1 00=-0 2=2;
1 0 2 1 -3
-4 1 -3
C3 00 2 =2 1 0 =-6;
0 0 1 I -3
I -3 1
C32 = - C) 2
-I - -1 0 0 = 0 -17 =-51;
-9
0 1 -9 0 -17 -3 -7
-4 -3 1 -4 -3 -7
C33 -1 02=-IC) 0=-I 171 - 10-,
-9 1 1 -9 I -17 1 -7
-4 1-1 -4 I -7
Cy- 71 00=1 0=-3;
-9 I 0 1 -3
-41 -3.


Inverse of Matrix 73
C41 =-0 0 2=-2 I 0 =-4;
10 I 02
0 2 -I

C-12 = - I 0 2 =-2 -1 2 =-2(17)=-34;

-9 0 1 -9 1
1 2 -1

C=- -I 0 2 =- -1 0 0 = 1 -17 =
-9 I 1 -9 1 -17 0
I 0 -1 1 0 1
C= -1 0 01 1 0 =-2
-9 1 0 0 2
I 0 2
C= -3 -14 -5 -13
4 11 -1 2
-6 -51 -10 -3
-4 -34 I -2
Adj.A=C'= -3 4 -6 -4
-14 11 -51 -34
-5 -1 -10 I
-13 2 -3 -2
=- 3 -4 6 4
14 -11 51 34
5 1 10 -1
13 -2 3 2

.- t, 4L AI 3 -4 6 4
- A 23 14 -11 51 34
5 I 10 -1
13 -2 3 2 Ans.
*Ex. 21. Poe that Adi (Adj A) I = A - if JA 1;-, 0 and is any
n x n matrix. (Agra 90)
Sol. We know that

AdjA1=JAI.jfJAlO
(See Th. 11. Page 50 Ch. V)
Replacing A by Adj A in (I), we get
Adj AdjA) = IAJJA In_I
I

AdjA }' (Note)

In- } , from(i)
)2
= j A ,In- I)
= IA Hence proved,
S


74 Matrices
*Ex. 22. Prove that A dj (Adj A) = A 1" - 2 • A, where A is any n x n
matrix. (Agra 92, 90: Kanpur 90)
Sol. We know that
A .(AdjA)=JAJ.I (SeeTh. I Page 49 Ch. V)
or Adj{A.(AdJA)}=Adj(I AI.J)
or Adj (Ad) A) • (Adj A) = A .1 (See Th. UI. Page 50 Ch. V)
or Adj (AdjA) (AdjA)A =jA 1.1. A

or Adj(AdjA),IAI.l=JA"' Al, See Th. IP. 49 Ch. V

or Ad] (Ad] A) , I A J = I A I" .A (Note)
j n_ 2
or Ad) (Ad] A) =IA • A. Hence proved..
Exercises on § 5.09
Find the inverse of the following matrices
Ex. 1. 10 -1 Ans. -1- 9 1 2
1 2 3 1033_4
0-1 3 -11 2
Ex. 2.1 0 0 Ans. 1 0 0
1 1 0 -1 1 0
1 0 1 --1 0 1
Ex. 3. 3 2 -1 1 -3 7 -8
Ans.8 6 -6
-I 2 3 8
-3 1 3 -5 9 -8

Ex. 4. 2 -4 -2 Ans.- -1] -9 1

4 6 2 4-2-3
0 10 -4 10 -5

Ex. 5.0 1 Ans. - 1 1 1

1 0 1 1 -1 1
1 1 0
11 (Gorakhpur 91, Kanpur 94)
1 1-1

Ex. 6. 2 3 1 Ans..1 1 -5 7
123 7 J 5
3 1 2 -5 7 -1
Ex. 7. 2 5 3 Ans. I -3 1 7
3] 2 4 -1 -i 5
1 2 1 5 1 -13
*Ex 8. 1 1 1 Ans. i - 6 5 -
2 2 3 3 15 -8 1
149 -6 3 0


Inverse of Matrix 75
Ex. 9. 1 —2 — I 11 9 —11
2 3 1 —4 2 31
0 5 —2 —10 5 _7j
Ans.[
Ans.L! . -1
Ex. 10. 1 4 0 3 3 3!
J - II
—1 2 2 !
002
6 6
[0 0 -I
2J
Ex. 11. I 2 3
345 Ans. Not possible as j A I = 0.
678

Ex. 12. 0 0 2 Ans. 0 0 1

010 010
100
00
Ix. 13. 1 1 2 0 Ans. 2 - 1 1-1
0 1 1 —1 —5 —3 1 1
2 12 1 2 3-1 0
3 —2 1 6 —3 —1 0 1
Ex. 14, 0 2 1 3
1 1 —1 2
12 01
- J 1 2 6
(Kwnaun 90)
Ex. 15. 1 1
1 3 —3 1
1 2 3 Ans. - 3
(Kumaun 93) 5 -2
1 3 6 1 —2 1
Ex. 16. Verify that A • (Adj. A) = (Adjj. A) • A = A
I 1 1 3 , where 1 3 is the
identity matrix of order 3, and A = cos 0 - sin 0 0
sin cosO 0
0 0 1
Ex. 17.lfA= 1 2 —1],verify that
23 1
0 5 —2
A • (Adj. A) = (Adj. A) • A A • I (Meerut 96P)
*Ex. 18. Verify that A • (Adj. A) = (Adj. A) • A = A
J 11 2 , where
A=[l 2 ] andl 2 = 1 1 0
[3 -5j [0 1
76 Matrices
**15 . 10. Existence or Inverse.
An Important Theorem.
The necessary and sufficient condition that a square matrix ma y possess an
inverse is that it be non-singular. (Bundclklwnd 96. 92; Kumww 96;
Gorakhpur 99; Meerut 92: I'uriancha1 98)
Proof. The condition is necessary.
If A is an n x n matrix and B is its inverse then by definition of the inverse
we have AB = In (See Chapter II)
Taking the determinants of both sides we get
IAB=JII.
But ABI=IAI.IBI .... See Chapter lV
and J l, I = 1. where I,, is the n X n iaentiry matrix
From (i) we get IA . 1 B 1 = 1,
which implies that I A 0.
The matrix A is non-singular. See Chapter IV
The condition is sufficient.
If A is an n x n non-singular matrix and them be another matrix B defined
by Br1--1(Adj.4)

Then AB=A 1 (AdJ.A)j(A S Adj. A)

A I 1, ... See § 509 Th. I Page 49 Ch. V

= In

Similarly BA= j---j (Adj A) • A = --j [(Adj. A) • A]

AII ... See 5.09 Th. I Page 49 Ch. V

AB=BA=I -
B is the inverse of A and it exists.
§ 511. Some Important Theorems.
*Theorem I. If A is a non-singular matrix of order n such that AX = AY,
then X=Y.
Proof. If A is non-singular mat rix, then A 1 exists. See § 510 above
Given AX=AY
or A (AX) = A (AY)
or (AA)X=(A'A)Y
or IX=LY AA=I


Some Important Theorems 77

Hence proved.
or X = Y, by left cancellation law.
Theorem 11. The inverse of transpose of a matrix is she transpose of the
inverse.
Proof. Let A be the given matrix. Then its inverse is A'.
Also we have AA' = I = A A. by definition.
(AA')' = 1' (A A)', taking transpose.
' -1
or (A-I,
) A, = I A (A ) (AB)' = B'A' and F= I.
Hence A' is invertible i.e. A' possesses inverse
and (A) = (A
transpose of the inverse.
i.e. the inverse of a transpose of a matrix is the
Hence proved.
Theorem 111. If A, B are any two n x n matrices such that BA = 0, where
is .ngular.
o is the null matrix, then at least one of them
Proof. Since A, B are two it x n matrices
so AB = 0, where 0 is the null matrix
(Note)
IAHBN0
either A = 0, which means A is singular
or B 0, which means B is singular
or both I A and I B I are zero which means both A and B are singular.
Hence at least one of A and B is singular.
Theorem IV. The inverse of the inverse of matrix is the matrix itself i.e.
-1-I -t
(A ) = A, where A is the inverse of A.
Proof. Let A be the given matrix. Then its inverse is A
Also by definition AA = I = A 1A.
I
A is invertible and we have (A 'Y A.
i.e. the inverse of the inverse of A is A itself. Hence proved.

Theorem V. if a non singular matrx A is .svmmetric, tlu'ti A_t is also

symmetric.
Proof. If A is symmetric, then A A'
Also by definition if A is non-singular. then
A=I
= I', since V = I
= (AA )' since 1 , = A A=AA -
= (A 5' A', since (AB') = B' A'
i.e A A = (A 5' A, since A = A', from (i).
or A = (A - 5', by right cancellation law.
is symmetric by definition. Hence proved.
Hence A


78 - Matrices
Theorem Vi. The inverse of the transposed conjugate of a non-singular
matrix A is the transposed conjugate of the inverse of A
i.e. (A)- = ( A 1)8
Proof. If A is a non-singular matrix, then A is invertible and we have
AA I AA
or (AA 58 = 18 (AA)8
or (A 5 A = I = A (A 58, since (AS) 8 = B0A0, J = 1.
A E) invertible and we have (A 8)- = (A 1)8 Hence proved.
***5 . 12 Theorem, if r be the rank of a matrix A of order in x n,
Ar be the
normal form of A, R be the product of elementary matrices of order in S be
the product of elementary matrices of order n, then A,. = RAS.
Proof. Since R and S are non-singular (i.e. their inverses exist), therefore
R 1 A r S A, where R and S are the inverses of R and S
respectively.
Of ABA,.C, where Bk.CrS
or Ar=;B'AC',
(Note)
Now if A is a non-singular matrix of order n, then r = n and

Hence A=8IC,
which is of the form A B, since B and C are the product of elementary
matrices.
Cor. If two matrices A and B are of the same order in x n and same rank,
then there exists non-singular square matrices P, Q such that B = PAQ.
Proof:. From above theorem we find that
A = CA,,D, B C 1 Ar
where C, C1 are product of elementary matrices of order in and D, D 1
of order
n.
From A = CA,. D, we get Ar = C' AD
Substituting this in B = C 1 A,.. D 1 , we get
B = C 1 (CAJY')D(C 1 C- A (D- ' D1)
which is of the form B = PAQ.
Solved Examples on § 5•12
Ex. 1 (a). Find the non-singular matrices R and 5, such that RAS is the
normal form, where A =
1-i
2 2 - 6
2 2
Sol. Here we find that A is a 2 x 3 matrix


Important Solved Examples 79

[A2i=1?A1]

or
1=[ ?.0? ]
we are to bring L.H.S. to the normal form by applying elementary row
NOW
and column operations.
[ 1 1 -31 1 0 •A . I 0 0'.byRi()
2
i 2 2j 0 I 0
01 001

1 0 0byR2+R
or F1 I 3 1[ ! ol.A.
0 1 0
[0 3 —1]
L2 J 0 0 I

or [i 0 0 1 = [! o1 —1 i.byC2_Ci
[0 3 —1] 0 1 OI and C3+3C1
L2 J 0 0 1]
• - 3 replacing C2 by C2
or 0 01 . 0 • A
[0 I —1] [ 0 0
001

i 3 byC1+C2
Fl 0 o l = rj o1 . A.
or [0 1
0 01

Since L. H. S. is in the normal form, so

IS
^
R=[1 olandS
I 3 3

013
'] 0 0 1 Ans.
Ex. 1 (b). Determine two non-singular matrices P and Q such that PAQ
is in the normal form, where
A=3 2-1 5
1 4 - 2 (Garhwa! 93
5
1 —4 11 —19
Sol. Here we find that A is a 3 x 4 matrix
[A]34=I3AI4
or 3 2 —1 5 =1 0 0 •A.1 0 0 0
5 I 4-2 010
1 —4 11 —19 0 0 1


80
Matrices
182111/5
or [ 0 0 -0 0l.A 1 II 0 0 01
117 9 4 18] I lo
I 0 01 ol Jo
[34 18 11 36 [0 0 13 2 1 iJ
[0 0 0 iJ
applying C 1 + 3C3 . C2 + 2C3,
C4 - 5C-
o 0 -1 o11j r 0.A.1 0 0 01. applying
17 9 4 i8 Jo i 1° 2i Io oJ R3-2R2
0 3 Oj [0 -2 1]
10 0 0 Ii
or [0 0 -10 1 =[i 0 01.A.r1 0 0
117 9 0 1s 4 ol I 1
0 0] 3 -2 I] 132 i sJ
[o 0 0 1]
applying R2 + 4R 1 , R 3 + 3R1
or i 0 0 -101=r1 1 0 0 1
0 01.A
I_ I 9 0 ol 14 I oJ H2 1 0
0 0 0 oJ [3 -2 ij 1 1 2 1
L [ 0 0 0 1]
applying C1 - 2C2, ('4 - 2C2
or [0 0 -1ir1 0 0 01
0 O1.4.[
Li 0 0 0 J4 I ol H2 -17 0
[
0 0 0 O [3 -2 iJ I' - 7 1 ii
[0 00 iJ.
applying C2 + 9C1
or0 1 01=[_ 1 0 01.A.[ 1 0 0 o]
0 0 0j -4 -i oj -2 -17 0 21'
1101, 0 0 0] [ 3 -2 1]
1
°
-7 1
0 0 L i l
1]
applying - R 1 and - R,
or0 0 o1=[i
0 0.A.0 -2 1 -170 1
i o oJ j-4 -1 00
100, 0 0 0] [ 3 -2 1]
-I -7
f 0 0 1]
i n terchaijn p C(IIlr1nc
L.H.S. is in the normal form, so we have
P =[- 1 0 OJandQ=fg

-21
0 0 0 1
Ans.


182/11/6 Important Solved Examples 81

Ex. 2. Find two non-singular matrices P and Q such that PAQ is In the
normal Iórm, where
A=1 1 1
1 -1 -1
3 1 1
(Garhwal 96; Meerul 91)
SoL Here we find that A isa3x3metrix
(A]33=I3A13
or 1 1 1=1 00A.l 00
1 -1 -1 0 1 0 0 1 0
3 1 1 001 001

or 1 1 1 = 1 0 01- A - 1 0 0 , applying R2+R1,R3-R1

200 110 010
2 0 0 -1 0 1 0 0 1

or I I I = 10 01-A- 1 0 0applying R 3 - R2
2 0 0 1 1 0 0 1 0
0 0 0 -2 -1 1 0 0 1

1 0 01
or 111= l.A. [1 0 01. applying R2 (-i)
100 ol 10 1
2 I I
I 000 0 0
-2

or o 1 1 = 2 - 2 • A • 1 0 0 R 1 - R2

100 -1 -0 010
2 2
000 001
-2 -1 1

0
or j Q 0 = 2 2 •A. 1 0 0. interchanging R 1 and R2
0 1 1 1 0 0 1 0
0 0 0 -2 -1 1 0 0 1
• L
or 1 0 0 = 1/2 1/2 01- A e 1 0 0, applying C3-C2
0 1 0 1/2 -1/2 0 0 1 -1
0 0 0 -2 - 1 1 0 0 Ij (Note)
Since L. H. S. is in the normal form, so we have
}= 0
2 2 0 0
=
- - 0 0 I -1
-, -1 I 0 0 I
Ans.


82 Matrices
Ex. 3 (a). Using the matrix A = r s 3 14 41., find two non-singular
0 1 3 i
1 L'
2 0]
matrices P and Q such that PAQ is in the normal form. (Agra 95)
SoL Here we find that A is a 3x4 matrix
[A J34 = 13 • A • 14
or [ 5 3 14 41=r1 0 OlsASI Ii 0 0 01
01 3 if fo i 01 OO.i I
—11 2OJ[OO 1] l o 0 1
000
or [ 0 8 24 41=[01 0 51.A.'' 0 0 0applyingR1+5R3
0 1 3 ii O1 of 1 0 lo
o 0 i 01
200 1J L0001]

O 8 24 4 1=r l 0 51.A.10 Ii 1 2 01 applying

•
0 ' I 1 1 o f
o
1 0 0 C2+C1.C3+2c1
0 i 0
- 1 0 0 0 ] O 0 0 Ij 11
C [o 0 0 1] (Now)
o 0 0 - 4 1=1 1 — 8
o I 3 511A'1 applying R1-8R21
i i jO 1 01 lo 0 1 01
- 1 0 0 Oj [0 0 lj
{ [0 0 0 ij
1 1 2
I 0 0 0] = [00 01. inteihanging
0 1 0 OjR1andR3
-I_ o I 3 '1
0 0 0 —4] I 5j 10 01
o 0 o ij
_Ii 0 0 01 1 0 0 — ii 1 1 2 0' applying
1
01 ii
0 1 0 01_ R and I
0 0 ij= I ° l
L 114 2 —4j
S/
10 0 1 Oi—(1/4)R3
[o o 0 ij

_li 01=1 0 0 1 1 —1
—1•.
0]
0 1 —3
1 0 1 0 0 0 i A 10

0 0 1] [-1/4 2 5/4 1
L° 0 0 [0 0 0 ij
applying C 3 - 3C2 , C4 - C2
-Ii o 0 01=[ 0 0 1 1 —1 —11
—11.A. 0 1 —1
Jo 1 0 O I 0 1 of
[0 0 1 0] [- 1/4 2 -5/4 1 0 ii
[o 0 1 0]
interchanging C3 and C4


Important Solved b.xamples 83

L.H.S. is in the normal form, so we have
P = 0 0 -' 1wdQ=
01 0
-1/4 2 -5/4 00 1 1 0 Ans.
Ex. 3. (b). Find non-singular matrices R and S such that RAS Is in
normal form, where A = 1 2 3
321
132
213
Sol. Here we find that A is a 4 x 3 matrix
[A143=14.A.I3
or 1 2 3- = - 1 000.A'1 00
321 0100 010
132 0010 001
2 1 3 0001
or I 2 3 100 0.A.1 00
0 -4 -8 -3 1 0 0 0 I 0
o t -1 -1010 001
0 -3 -3 -2 0 0 0
applying R2-3R1.R3-R1,R4-2R1
or 1 2 3= 1 00 0 . A . i 00
01 2 I 0
-1 0 0
o i - i •001
o 0 -6 0
-1 0 10
-5 0 3 1
applying R2 (- ) and R4 + 3R3
1 2 3= 1 0 0 0At 00
or
00 3 0 1 0
0
01-1 4 4 001
0 0 -1 -1 0 1 0
0 ! _!
6 2 6
applying R2 -R3 , R4( )

or i 2 0 = 0:fl.A. 1 0 0
0 0 1 0 1 0
0 1 0 i - 3 " o0i
001 _! o ii
6 2 6
0 - 2 - plying R 1 - 3R4 , R2 (-i ), R3 + R4


84 Matrices 999
[i 0 O= [_ 1 0 1 5'. A . r1 0 01.
JOl °iI jo 1 01
0 12 - O
10 1 [0 1 0]
L0 0 0j o !2
61

'J applying R - 2R 3 , R 4 - R2
or 1 0 0 15 1 . A . J 0
[0 2 6 01.
i-1
o
iJ I
12 12 2
100
Oo 1 0]
L
0 , 0] I_ i 0 _!
2 6
I 3 6 1 -1
12
[ 2 interchanpinc' ('P. nA ('
L.H.S. is in the normal form, so we have
R=_? 0 1 5 S=I 00
62 6
7 I
12 o
12 2 0 1 0
1
6
o
2 6
1 I I I
4 12 6 6
Ans.
Exercise on § 512
Ex. 1. Reduce A= 1 1 2 to normal form N and compute the
1 23
0
matrices P and Q, such that PAQ N.
Ex. 2. Determine two non-singular matrics P and Q such that PAQ is in
the normal form., where
A=1 1 2]
1 2 3
1
- j
(Garhwal 94)
MISCELLANEOUS SOLVED EXAMPLES
Ex. 1. Find the reciprocal (or inverse) of the matrix
S = 1 1 and show that the transform of the matrix
Ii o -i
1 1 0
A= 4 b + c c - a b - a by S i.e. SAS 1 is a diagonal matrix.
c—b c+a a—b
b—c a—c a+b
Sol. In the usual way we can show that


Miscellaneous Solved Examples 85

inverse of S = I - I I I (To be proved in the examination)

1 -1 1
I 1 -1

SA= 0 1 ' 1 I x b + c c-a b-a

1 0 1 c-b c+a a-b
1 1 0 h - c a-c a+b

= 0 2a 2a, multiplying the matrices in

2b 0 2b the usual way.
2c 2c 0]

Or SA 0 a a
bOb
cc0

SAS' = 0 a a - i • 1 i
b 0 b 1 -1 1
ccO I 1-1

= -2a 0 0 . multiplying the two matrices in

o 2b 0 the usual way.
o 0 2c]
a 0 0 j. which is a diagonal matrix,
o b 0 [See Chapter II)
o o cj
Ex. 2.. If A Is Invertible show that A is invertible.
Sol. If A is invenbie, then we know that
AA'=I=AA
or ()=I=(A1A) (No(e)

or A5IK5. .
Hence A is invertible and we have =( A)
Hence proved.
Ex. 3.(a). If A= a 1 0 OwherenoneOfa'SiSlerO,tbenthoWth
o 22 0
o o 83
A Is invertible. Also evaluate
Sot. jA = a 1 0 0 =a 102a3, on evaluating
oa2 0
0 0 a3

i.e. I A 0. Hence A is invertible. ...'sec Cb. 1'\

Also


86
Matrices
C 1 = a2 0 =a2a3;C12=-
0a3 0a3
0 01 0;C 13 = 0 a2 =0;
00
C21 = - o 0 = 0;C22 = l a j 0 a 1 a 3 ;C- ja i 0 =0
o a3 0 a3
00
C31 = 0 0 0;C32 =- j a l 0=0;C 3 = a1 0 =a f a2
a 2 0
0 0 0 a2
C=a2a3 0 0
0 a-as 0
o o a1a2

AdjA=C'= a 2a 3 0 0
0 a3a 1 0
o 0 a1a2

A_IA4jAJ [ a 2a3 0 0
JAJ a1a2a3 0 a 3 a 1 0
0 0 a1a2

or A1= 1/a 1 0()3

0
• 0 0 1/a3
Ans.
Ex.3 (b). Show that the matrix A 0 1 01 Is its own
inverse.
100
001
Sol A J = 0 1 0 = - I () expanding wr. to
100 01
001
=-l*0
AlsoC11= 0 0 0;C 12 =- I
01 -I;C13 = 1 00;
01 01] 00
C21 =- 1 0 = -1;C22
= 0 0 0;C23 =- 0 1 =0;
01 01 00
C31 = 1 0 z O;c32 = 0 0 0;C33 =
0 J =-i
00 10 10
C=f 0 -1 0 and adjAC', 0 -I 0
J-i 0 01 -1 0 0
[00-I]
00-1
 Miscellaneous Solved Examples 87

0 -1 0
- jAj (-1) -I 0 0
0 0 -1

0-1 0=01 0=A

-1 0 0 1 0 0
0 0 -1 0 0 1 Hence proved.

Ex. 3 (c). Compute the inverse of the matrix A, if

A= 3 -2 0 -1
o 2 2 1
1 -2 -3 -2
o i 2 1

S0 flAI = 3 -2 0 -1
0 2 2 1
1 -2 -3 -2
0 1 2 1

, replacing R 1 byR1-3R2

4 9 5 , expanding wr. to C1
221
121
= 0 1 1 , applying R 1 - 4R 3 and R2 -
100
121
= 0 0 1 , applying C2 - C3
100
111
1 0 . expanding w.r. to R1
11
=1 0
AlsoC 11 2 2 1 = 1 0 0 -3 -2
-2 -3 -2 -2 -3 -2 2 1
1 2 1 1 2 1

=-3+4=1
C 12 - 0 21 =0;C 13 0 2 1 =-1;
1 -3 -2 1 -2 -2
0 2 1 0 1 1


88 Matrices
C 14 =- 02 2 2 2 =2.
1 -2 -3 I 2
0 1 2
C21 =- -2 0 -1 =- 0 4 1 =1;
-2 -3 -2 0 1 0
1 2 1 12!
C22 - 30 -I = 0
0 -1 =- -5 -3 =1;
1 -3 -2 -5 -3 -2 3 2
o 2 1 3 2 1
C23 = - 3 -2 -1 =- 0 0 -'= -5 2 -I;
1 -2 -2 -5 2 -2 3 -1
o 1 1 3 -1 1
C24 3 -2 0=0 4 9-49=1;
1 -2 -3 1 -2 -3 12
0 1 20 1 2
C31 . -20 -1 = 0 0 -1 -2;
22 1 02 1
1 2 1 -1 2 1
C32 =- 3 0 -1 =0
.02 I
02 1
C33 = 3 -2 -1 =3 2 1 =3;
0 2 1 11
0 1 1
C34 3 -2 0 =-3 2 2 -6;
0 22 12
0 12.
C4 - -2 0 -J =- 0 2 0 =2 2 0 -4;
2 2 1 2 2 1 -1 -1
-2 -3 -2 0 -1 -1
C42 = 30 -1 = 0 9 5 = 9 5 =-1;
0 2 1 0 2 1 21
1 -3 -2 1 -3 -2
C43 =- 3 -2 -1 10
=- 4 5 =- 4 5 6;
0 2 1 0 2 1 21
1 -? -2 -, 1 -2 -2


Miscellaneous Solved Examples

3 —2 00 4 9
14
C= 0 2 2 = 0 2 2
= 2 2
1 —2 —3 I —2 —3 1^

C= I 0—I 2
1 1—I
—2 0 3 —6
—4 —1 6 —10

• AdjA=C'= 1 1 —2 —4
0 1 0 —1
—1 —1 3 6
2 1 —6 —10

Adj. • A - 1 1 -2 4-1 from (i) and (ii).

• A-
Al 0 1 0 1
• —1 —1 3 —6
—2 1 —6 —10 Ans.
**Ex 4. Find A', If A= 1 1 ii,, where w is the cube root of
1

unity. (Agra 9,3)

S0LIAI Zz 1 1 1

= 1 0 0 replacing C2, C3 by
1 (0-1 ()2_j
c2-c1,c3-c1
1 (02 - respectively.
-

= - 1 W - 1 ; expanding w.r. to R1
(021 i
- 1) 2 1 U) + 1 , taking out common factors
=
(0+1 1
)2]
=(w— 1)211 —(w+
I),
0 or co2+2wol
1)^0.
Also C 1 I (0 (02 = 2 - 4 = 2-

•.
=

-20


90 MaLrices

C 12 =- Ii 0)21

o:
=-(CO -(02)=0)2-W
I i
c13 = Ii ( 0) 2 0) .C_I I

t w
c22 = Ii ii 0)-1;C23 = - i ii 2 1);

I' (0)2
= c32 - Ii 1
C31
I 2j
C33 r11 1=w—I
I
°l
c =[ w 2_ w 0)20) 0)20)
1
w-1 (0)2_I)j
0)2 0) (0)2 (j)—I j

Adj. A=C 1)
0 2._ W

2_ 2 1
w—i _(0)2_1)
(0)2_i) (fl-I
L(0 2_ W
=(w-1)Ico (1) 1
Ico 1
[U) — ( 0 + 1 ) I
j
w c1
(o W 2I'
[0) W
2
j

or Adj. A=(U)-1)WrI 1 1 where 1 =-

0)3
0)
I i/w 0)
1 1/0)1

=w(w— l)[1 1 11
k0)2
2j

At, t Adj. AcO(W — [ i I I l from (j) and (jj)

1)3 0)2 Wr
Al —(w—
[1 ()

1i 1 i
(co - j)2 1 0)2 (0 •.(iii)
-
(I) 0)2]
 Miscellaneous Solved Examples 91

Now - (w- 1) 2 ((0 2 + 1-2w)

r-[(-w)-(2(U)], :w2+w+J=Oorw2+1_w
=3w
From (iii), we gct

A=--1
3w'
1 1 1= 1 i 1
W 2 W 3 1 w2
1 w2 I
Aiis.
Ex. 5 (a). Find the rank of the matrix
A= 1 2 22 32 42

22 32 42 52
32 42 52 62
42 52 62 .2
(Agra 96; Bundclkhand 96)
Sol. Given A = 1 9 4 16
4 9 16 25
9 16 25 36
16 25 36 49

-1 o 0-, replacing C 2 , C3 , C4, by

0
4 -7 -20 - 39 C 2 - 4C, C 3 - 9C 1 and

9 - 20 -56 - 108 C4 - 16G 1 respectively

16 -39 -108 -207
0 replacing R 2 , R 3 , R4
0 0
0 -7 -20 -39 byR2-4R1,R3-2R2
-6 -16 30 and R4 - ' 2 respectively
0 -11 -28 -51

-1 0 0 0 replacing R 3 , R4 by
0 -7 -20 -39 R 3 -R 1 and R4-R2respectively
0 -6 -16 -30
0 -4 -8 -12

-I 0 0 0 replacing R 2 , R 3 , R 4 by - ( R 2 -
0 1 4 9 - (R 3 - R 4 ) and - R 4 respectively
0 2 8 18
4 8 12
Lo
0 } replacing R 3 . R4 by R 3
- I 0 0 - 2R2,
0 1 4 9 R - 4R2 respectively
0 0 0 0
0 0 -8


92 Matrices

- I T o o I replacing C 3 , C4 by C 1 - 4C2,
o 0 0 C4 - 9C2 respectively
0 I o -0
0 -8 -24

0 replacing R4 by -
- 1
0 A o00
0 00
0 13

0 0 interchanging R3 and R4
0 00
0 13
0 00

-1 0 0 replacing C4byC4-3C3
0 00
0 , 10
0 00

0
-10 [0 0

The rank of martixA is 3. A ns.

Ex. S (b). Find the rank of the matrix
A= 1 2 3 0
2432
3213
L J (Agra 94; Bundelkhand 93)

Sol. GivenA -. 1 2 R 2, R 3 . R4
3 0 replacing
2 o 2 byR2 - R I , R3-R2,R4-R3
-2 - 2 1 respectively
3 6 62

0 0 o]. replacing C2 , C 3 by
0 -3 2 C2 - 2C 1 , C3 - 3C 1 respectively

-4 -5 1
3 0 -3

0 0 o' replacing R2, R3, R4 by

0 0 - 3 2 It?2 - R 1 , R 3 - R 1 , R4 - 3R 1 respectively
0 -4 -5 1
0 0 -3 2


Miscellaneous Solved Examples 93

- I o o o replacing R 1, R4 by
o 0 -3 2 R 3 - R 2 , R4 - R2 respectively
o -4 -2 -1
o 0 0 0
- 1 0 0 oreplacing C, C3 by - (1/4) C2
o a 3 2 and - C3 respectively
o 1 2 -I
000 0
- 1 0 0 o replacing C3, C4 by
0 0 3 2 C3 - 2C2 and C4 + C 1 respectiely
0100
0000
- 1 0 C) 0 replacing C 3 by C3 - C4
0012
0100
0000
1 0 0 0 replacing C4byC4-2C3
0010
0100
0000
- I 0 0 ' 0 interchanging R2 and R3
0100
.0 0 1 0
0 0 0 0
130
00
The rank of the given martix A is 3. Ans.
*Ex 6. If A = 3 - 3 41., show that A 3 = A1.
2 -3 4
0 -1 1
So!. I A I = 3 -3 4
2 -3 4
0 -1 1
or A = 1 0 0 , replacing R 1 by R - R2
2 -3 4
0 -1 1
= -3 4 =-3+4=1 0.
-1 1


94 Matrices
Also for the matrix A. we have
C1 1 = -3 4 =1;C 12 =- 2 4 =-2;C 13 = 2 -3 =-2;
-11 01 0-1
'2I =- -3 4 =-1;C22 = 3 4 = 3;C23 = - 3 -3 =3;
-11 01 0-I
C31 -3 4 =0;C32 =- 3 4 =-4;C33 = 3 -3 =-3;
-3 4 2 4 2 -3
C= 1 -2 -2
-1 3 3
0 -4 -3
Adj.A=C'= 1 -1 0
-2 3 -4 ...(ii)
-2 3 -3
A_ 1 _ Adj A _1 1 —1Oljrocn(i),Oi)
Al [-2 3 -4 ...(iii)
3 -3j
Also A2 = [03 -3 4 x 3 -3 4
2 -3 4 2 -3 4
-1 1 0 -1 1
9 -6+0 -9+9-4 12-12+4 - 3 -4 4
6-6+0 -6+9-4 8-12+4 - 0 -1 0
=[0 -2+0 0+3-1 0-4+1 -2 2 -3
A3 =A2 .A= 3 -4 4 x 3 -3 4
0 -1 0 2 -3 4
-2 2 -i 0 -I I
= 9-8+0 -9+12-4 12-16+4
0-2+0 0+3+0 0-4+0
-6+4+0 6-6+3 -8+8-3
= 1 -1 0
-2 3 -4.
-2 3 -3
i.e. A3 = A 1 . from (iii). Hence proved.
EL 7. If two non-singular symmetric matrices A and B be such that
AB = BA (i.e. commute under mlt1p1lcatloo), then prove that
A 1 B and A 1 B 1 are symmetric.
' SoL Hem we are given that AB=BA.
We have A 1 AD = A 1 BA, premultiplying by A1


Miscellaneous Solved Examples 95

I
or IBA'BA. A- A = I

or B=A1BA, . IB=B
-I=A- 1 BAA , -I . . -1
or BA post multiplying by A
=A'BI=A#B,
since AA- I and B! = B.
-1 B)' = B' (A)', •.. (AR)' = WA'
Again (A
= B' (A') 1 (A-')'= (A')
...See Th. II Page 77 Chapter V

=BA', . A=A,B'=B as A and Bare symmetric

I
i.e. (A- B)' = A 4 B, from (i)
Hence A'B is symmetric.
I 1
Similarly (A- B- )' = ( B 4)' (A t )', as (CD)' = D' C'
(A1 B 1 )' = ( B'' (A'Y'. See Th. H Page 77 Ch. V
or
I A'= A, B' = B
= B -1A -
= (AR)'.
(AB)-' = B 1 A1
=(BA)' AB=BA(given)
or (A1 B)' = A 1 B1.
Hence A 1 is symmetric.
Ex.1LFtnd the rankOIA= [ i 2 3 1
12 4 6 2
Li 2 3 2
Sol. A, replacing R2, R3 by
R2 - 2R 1 and R3 - R 1 respectively
o

1 2 3 ol. replacing R 1 by R 1 - R3
0000
0001

- 1 0 0 01. replacing C2, C3 by

0 0 0 0 C2 - 2C 1 and C3 - 3C1 respectively
0 0 0 1]
0000l.
1 0 0 interchanging
o C2 and C4

0100
96 Matrices 182/11/6
- 1 0 0 01. interchanging R2 and R3
0100
0000
41 2 0
L0o
..
The rank of matrix Ais2. Ans.
**Ex. 9. J1nd the rank of an m xi, matrix, every element of which is
unity.
Sol. Let an ,n x n matrix be A I I I ,...
1 1 1... 1

.1 1 1 1
Then we find that every square submatrix of A higher than I x I will be a
matrix each element of which is unity and therefore the value of the determinant
will be always zero, since its rows and columns are identical. But the square
sub-matrices of order 1 x I are [1] and the determinants of these are
IA I = 1 # 0.
Hence the rank of A it 1. Ans.
Ex. 10. bow that the matrix A = I a a aa Is of rank 3 provided
• • I Ii 0 bfi
-
. Icycy
no two of a, b, c are equal and no tvo of a, 13, y are equal.
Sol. A - I a a aa ]. replacing R2. R3 by R2 - R1,
0 b—a 13—a b13—aaR3_Rjrespectively
0 c—a y—a cY_aczj

- 1 0 0 0 replacing C2 , C3, C4 by
0 b—a 13-a 143—act C2—aC1,C3—cxC1ajtd
0 c - a - a cy - na C4 - aaC1 respectively
-. I 0 0 o ],replacing c4byC4_ac2
0 b—a 13 —a b13—ba
0 c—a y-a cf - ca
- 1 0 0 0 replacing C4byc4_bc3
o b — a 13 —a 0
0 c — a y — a cy — ca— by+ba

1 0 0 0 =B(say);
d b—a 13 —a 0
-[0 c—a y—a (c—b)(y—ct)
Now a minor of order 3 of B


182111/7 Miscellaneous Solved Examples 97

0 0 =b - a 0
(c-b)(y-a)
=1 o b-a 0 c-a
10 c-a (c-b)('j-CL)
expanding with respect to R1
= (b - a) (c h) (y - a) #- 0, as no two of a, b, c and no two of
a, P, 'y are equal (given)
p(B)>3
Also the matrix B does not possess any minor of order 4 i.e. of order 3 + 1,.
so p (B) < 3.
From (i) and (ii) we get p (B) = 3
and therefore p (A) = 3, as A B. Hence proved.
Ex. 11(a). Find A 1 if A = I - 1 2
1 0.-i
2 -1 1
Sot. Here A I = 1 - 1 2 = 1 - 1 2 , replacing R3 by R3 - R1
1 0 -1 1 0 -1
2 -1 1 1 0 -1
= 0, since two rows are identical.
Hence the matrix A is not non-singular (i.e. is singular) and so A does not
exist. (See § 5 . 10 Page 76 Ch. V)
Ex. 11(h). Find adjoint and inverse of the matrix
3 -1 1
-15 6 -5
5-2 2
Sot. Do yourself.
Ans. Adj. A=C'= 2 0 -11; A - ' 2 0 -1
51 0 51 0
01 3 01 3
-Ex. 12. IfA= [ 1 1 -1 and B= -1 -2 -1
2-3 4 .612 6
3 -2 3 5 10 5
then show that p (AB) * p (BA), where p denotes its rank.
(Rohilkhajzd 93)
Sol. AB=[1 1 _ix[_i -2 -1

1][
=-1 * 6-5 -2+12-10 -1+6-5 00 0
-2- l8+2( -4-36+40 -2-18+20 0 0. 0
3-12+ 15 -6-24+30 -3-12+ 15
- 0 0 0


98 Matrices
=0
By definition p (AB) i.e. rak of AB is 0.
Sec § 5.02 Note 2 Page 2.
Again BA= -1 -2 -lxi1-1
6 12 6 2 -3 4
5 10 5 3 -2 3
-1-4-3 -1+6+2 1-8-3
= 6+24+18 6-36- 12, -6+48+18
5+20+15 5-30-10 -5+40+ 15
- 8 7 -10 =C.say
4 -42 ()
40 -35 50
=[
Now C- -8 7 - 10j.replacing R2byR2-R3'
8 -7 10
40 -35 50
- 8 7 - io] replacing R 2 , R3 byR 2 + R1
0 0 Oland R3+SR1 respectively
0 0 0]
-[
- 1 i 1k replacing C1 , C2 , C3 by (- 1/8) C1
0 0 O(1/7)C,(...1/I0)C3reSpcctjvcly
.000]
1 o o1 replacing C2, C3 by C'2 - C1
0 0 0 Jand C3 - C 1 respectively
0 0 0]

0
[o 0.
p(C)=1 or p(BA)=1, from (ij) (iii)
From (i) and (iii), p (AB) * p (BA) Hence proved.
Ex. 13. If A = I I
123
11
and B = 31 then show that
2 1 2
1.4 9 2 1

(AB) = B' A1
Sol. HereA= 1 1 1 = 1 0 0= 1 2 =2
123 11 2 38
149 13 8
ri
Also for the matrix A. we have


Miscellaneous Solved Examples 99

C 11 = 2 3 =6;C 12 - 1 3 =-6;C 13 I 2 =2;

19 14
49
1 1 =-5;C22 I I =8;C23 - 1 1 =-3;
C21 -
19 14
49

- 1 1 =-2;C33 I I = I;
C31 = 1 1 =1;C32
13 12
23

C = 6-6 2
-5 8 -3
1 -2 1

Adj. AC' 6 -5 1
-6 8 -2
2 -3 1

Adj. A_. 6 -5 1
A_
IAI 2 -6 8 -2
2 -3 1

1 1 - 1 1 =4
Again I BI= 2 5 3 = 0
312 O_5_151
121 1 2 1

Also for the matrix B, we have

C 11 = 1 2 =-3;C 12 - 1321
=-1;C 3 3 1 =5;
11 12
21

2 3 =-1;C 23 - 2 5=1,
C21 - 5 3 =1;C22
21 11 12

- 2 3 =5;C33 2 5 =-13.
C31 = 5 3 =7;C32
12 32 31

C=-3 -1 5
1 -1 1
7 5 -13

Adj. BC' -3 1
-1 -1
7
5
5 I -13

1 7
• B- -1 -1 5
5 1 -13


..100 • Matrices 999

Again AB 1 1 x 2 5 3
1 23. 3 12
I 4 9 I 2 1'
2+3+1 5+1+2 3+2+1 = 6 8 6 =D(say)
2+6+3 5+2+6 3+4+3 11 13 10
=[2+12+9 5+4+18 3+8+9 23 27 20
.NOWDJ 6 8 6 = 6 8 6 = 0 8 6
II 13 10 11 13 10 1 13 10
23 27 20 1 1 0 I 1 0
0 8 6 = 8 6 =80-72=8^o.
0 12 10 12 10
• 1 10
For this matrix D, we have
C11 =— 13 10
-10;C12=-111 10 10;C3= 111 13 =-2;
27 23 20 23 27
C21 =- 8 6 =2;C22 = 6 6 -18;C=-
6 8 22;
27 20
. 123 20 . 123 27
C31 = 8 6 =2;C32 =- 6 6 =6;C33 =
6 8 =-10
• 113 10 11 10 111 13
10 -2
12-18 22
6 -10
[ 2
Adj. D=C'=f' -io 2 2
10 -18
I 6
-2 22 -10
1J. D 1-io
• (AB)' -
- D1 • 2 21 ... (iii)
- JD'j 81 10 -18 6
[-2 22 -10
From (1) and (ii) we get
B'A'-! -3 1 7 6 -5 1
4 -1-1 5 2-6 8-2
5 1 -13 2 -3 1
i 7 18-6+14 15+8-21 -3-2+7 _i -10 2 2
8 -6+6+10 5-8-15 -1±2+5 8 10 -18 6
• 30-6-26 -5+8+39 5-2-13 -2 22 -10]
= (AB)', from (iii).
Hence proved.
*Ex. 14. Prove that (Adj. A) 1 (Adj. A 1 ), where
A Is any n x n matrix.
(Agra 95; Avadh 99; Bundelkhand 92; Pun'ancha( %)
Sol. We know from Ex. 22 Page 74 Ch. V (To be proved in the examination)
that


Miscellaneous Solved Examples 101

Adj.(Adj. A)=IA2sA
Also from Theorems I and II Pages 49-50, Ch. V we know that
In- '
I Adj. AI=IA
Al = Adj. A
and
Al
Now Adj. A'=Ad). {'l from (iii)
IAI

1=Adj.—.Adj.(Adj.A) ...
11A1 J See Th. III Page iO Ch. V
1
• l AI2.A, from (i) and (ii)
'
or ,..(iv)

= Adj. (Adj. (Note)

'
Also (Adj. A)from
I Adj. A
IA A
or (Adj. A1 - —, from 0) and (ii)
A In
or (Adj. A)-'=

Hence from (iv) and (v), we get (Ad). A) = (Ad). A). Hence proved.
Ex. 15. If A Is of order m x n,' R is a non-singular matrix of order m,
show that Rank of RA = Rank of A.
Sot. Let A=EA r FandR=E j See 5.12 Page 78
Then RA = E i (E A r F) = E l EArF
i.e. RA has been expressed as the result of elementary operations on A,.
Thus Rank of (RA) = Rank A,. = Rank A.
**Ex. 16. Prove that the rank of a matrix remains unaltered by the
application of elementary row and column operations.
or Prove that two equivalent matrices have the same rank. (Avadh 99)
Sot. Let an ni X n matrix A be given by
A= a 11 0 12 ...
cj 21 a22

0 m1 am2 ...
Let M be any minor of order r belonging in the first r rows of A I

Now firstly if we inerchange any two rows or columns of A, then the minor
M either remains unaltered or changes sign.
Secondly if we multiply one row or column of A by a number A, then either
the minor M remains unaltered or changes into AM.


102 Matrices
• Thirdly if we replace any row R, (for column C) by R + AR, (or C, + ),C),
then either the minor M remains unaltered or changes into a sum or difference of
two of the original minors.
Let B be the matrix obtained from A by the application of any one of the
above three elementary row or column operations.
Thus if all the minor of order r in B are zero, then all the minors of order
I

r in I H I are also zero.

rank of B 15 rank of A.
Similarly if all the minors of order r in B J are zero, then all the minors of
order r in A I are also zero.
I

rank ofA:5 rank ofB ...(ii)

From (i) and (ii) we get
rank of A = rank of B. Hence proved.
Exercises on Chapter V
EL 1. Are the following pairs of matrices equivalent?
4 —1 2 and 2 1 4 7
3 40 3621
1 0 0 0 0 1 5
Ans.No.
EL 2. Show that the rank of a matrix is not altered if a column of it is
multiplied by a non-zero scalar.
EL 3. Show that the inverse of
1 2 3 1 is 1 —2 1 0
1 3 32 1 —2 2 —3
2 4 3 3 0 1 —1 —1
1 1 1 1 —2 3 —2 3
EL 4. Compute the adjoint and inverse of
Ans. 1 1 1and ! !j
1--I 0
—1 I I _!
0 1' —1 2 2 2
—1 -' 1
1 0 1 I II
2 2

Ex. 5. Show that the adjoint and inverse of the matrix

1 2 3 are j 3 —2 and—! 1 3 —2
345 : 3_9 4 3-9 4
679 —3 5-2 —3.5-2
EL 6. If A a l 0 0 ... 0 where a * 0 for all 1 :5 i :5 n , then
Q a 2 0 ...
0 show that A is invertible. Also
o o a3
0 evaluate
::
.


Exercises on Chapter V 103

1/a t 0 0 0
Ans.A1
o 1/a 2 0 0
o 0 /03 0

.'. .. ..
I /a,

Ex. 7. Show that the reciprocal (or inverse) of

-2 — 3
s
Fi i
2 3 5— -
Ii 49
L —2 I 0

Ex. 8. Show that the inverse of

1 1 2
2 -1 3 is
I 1 1 0 -
I —1 I 1 —.!. _2
2 2

Ex. 9. If A= a 1 h1 Cl prove that AA' 13

'2 h2 c2
a3 h3 c3

Ex. 10. Prove that Adj. (A') = (At/i. A)'.

Ex. ii. Let A be a non-singular matrix. Will the adjoint of A also be
non-singular ?
Ex. 12. Show that A = 3 7 1 is non-singular.
5 9 —1
7 13 —5

[J-flnt Prove that I A I 01.

Ex. 13. Show that if A is a square matrix of order n then
Adj. A (Adj. (Adj. A)) =(det. A)' 'I.
Ex. 14. What is the rank of a non-singular matrix of order n 7
Ex. 15 (a). Show that the inverse of
2 —1 31 is 7 —9 10
—5 3 1 —12 —IS 17
—3 2 3 1 1
Ex. 15 (b). Find the inverse of the matrix
A=l 00 2 Ans. 24 10 —2 —6
0 —1 1 —1 —5 —3
2 1 2 1 —16 —6 2 4

3 — 2 1 6 (Agra 95) 11 — 1 3


Matrices
Ex. 16. Compute rank of the matrix 1 '2 3
456
789
Ans. 2
Ex. 17. Prove that each non-singular matrix has a unique inverse matrix.
*Ex. 18. Define Rank of a matrix. Determine rank of the matrix
A= 1 a b 0
Ocdl
labO
Ocdl

[Hint: See § 5.02 Pages 1-2 Ex. 13. Page 10 Ch VI

Ex. 19. Find the rank of the matrix A, given by
A=3 2 71
4 1 32
1 -1 -4 1
Ans.2
EL 20. Find an invertible matrix P such that PAP_ 1 is a diagonal matrix,
where A= 1 -3 3
3 -5 3
6 -6 4
Ex. 21. Prove that every matrix of rank r can be reduced by means of
elementary transformation to the form Ir 0
1o 0
Ex. 22. Show that for any matrix A. rank (AA) rank (A).
Hence or otherwise show.that if n be the rank of an m X matrix A, then
A'A is a non-singular matrix.
Ex. 23. Find inverse of
(1) 1 - 1 0 (ii) 1 0 1
0 11 3 44
1 11 0-4-7

EL 24. If adj. B=A and I P 1 =1 =IQI. then prove that adj. (Q- I BP_I
PAQ. (Kanpur 95, 93)
Ex.25. Prove that the inverse of a matrix is unique.
Ex. 26. Prove that for every matrix A there exist two non-singular matrices
P and Q such that PAQ is in normal form. (Rohilkhand 90)
EL 27. Show that every elementary matrix is non-singular i.e. it is
invertible and its inverse is also an elementary matrix of the same type.
• (Purva,ichal 93)
EL 28. Find the rank of the matrix A, where
A= 1 3 4 3
3 9 12 3
1 3 4 1
(Rohilkharzd 99) Ans. 1
Chapter VI

Solution of Linear Equations

§ 6.01. Matrix of coefficients of a system of equations.
Definiton. Let the system of m simultaneous equations in n unknowns
X1, X2, xbe

a21x1 + a 22x 2 + a 2 x 3 + ... + a2trxn= k,,

a 3 1 x 1 +a 32x 2 +a 33 + ... + a.,rxn=k3,

a m 1XI + am2 + a m3x 3 + + = km

or written in a compact form
'1
ax=k.wherei=1,2.
1=I
Then the matrix A = [ a 1 ) = a 11 a1
a 2l a22 02,,

0 m1 am2 mn,nJ

of order m x n is known as the matrix of coefficients of the system of equa-

tions given by (I).
The determinant of the matrix A, [if there be ,t equations in (1)] viz.
IAI= a ll a12

°21 0 22 ... 111

amj 0 m2 (1mn

is called the determinant of coefficients of the system of equations given by (1).

Note. If all the k's are zeio, then the system of equations given by (i) is said
to be homogeneous and if at least one of k's is not zero, then the above system
of equations is said to be non-homogeneous.
§ 6.02. System of equations in the Matrix Form.
The system of equations given by (i) in § 6.01 above can be written in the
matrix form as
a ll °12 a1 x 1 1 = [kjl - (Note)
a, 1 a 22 a21 a2,, ..
X 2 k1

a, 1 0 12 ... Q jj ...

oJ,[ 0 m2 mj1rn Xi; km

106 Matrices
or in more compact form it may be written as
AX = K.
'where A = [a] i.e. the matrix of coefficients of the system ot'equations given by
(i) in § 6.01 on Page 105;
X = the transposed matrix of 1x 1 , x2. x3 , ... x,j and
K = the transposed matrix of [k 1 , k2 , k3, ..., km].
Here students should note that the product AX is a matrix of order ni x 1, as
A is a matrix of order m x n and X is a matrix of order n x 1. And X is also a
matrix of order m x 1.
* 6.03. Consistent and Inconsistent Equations.
Consider the system of equations given by (i) of § 6.01 Page 105;
If the above system has a solution (i.e. a set of values of x 1 , x2 , x3, .., x,
satisfy simultaneously these m equations), then the equations are said to be
consistent otherwise the equations are said to be Inconsistent.
A consistent system of equations has either one solution or infinitely many
solutions.
§ 6.04. Solulion of non-homogeneous Simultaneous equations.
Solution of equation given by (i) of* 6.01 Page 105 Ch. VI when rn = n and
'he matrix A is non-singular
We know from, § 6.02 Page 105 Ch. VI that the matrix form of the given
quations is AX.= K
Also we, know that if A is non-singular, its inverse matrix i.e., A' exists
uch that A'A=I. ...(ii)
where I is the identity matrix.
Hence by multiplying both sides of (i) by A 1 , we have
A- I AX = A-
or IX=A1K, from (ii)
or X = A- which is the required solution
of the given equations and is unique.
Solved Examples on § 6.02—* 6.04.
Ex. 1. Express in matrix form the system of equations
9x-f7y+3z=6;5x+y+4z1;6x+8y+2z4. (Gorakhpur97,94)
Sot. The given equations are
9x+7y+3z6
5x+y+4z1
6x + 8y + 2z = 4
The required matrix form of these equations is
AX = K,


Solution of Non-homogeneous Linear Equations 107

'where A=9 7 3 X= x andK= 6
514 y I
2 z 4
L6 8
Ex. 2 (a). Find the matrix X.from the equations AX =B,
wlrre A=1 —1 OandB=2
0 1-1 1
1 1 1 7
Sot. Let X =then from AX=B,wehave

Yzj
1 —10 [x=2
0 1 —1 y 1
1 1 lz 7
= I - 1 0 x = [21. by the elementary row
• 0 1 —1 y Jll operation R3—R3+R2
1 2 Ozt8j
0x = 21, by the elementary row
0 1 —1 y 1I operations Ri—R3+2Rj
3 0 Oz 12j
=. x—y=2,y—z=1,3x12 (Note)
= y=x-2.z=y-1,x=4
y=4-2-2,z=2--1=I,x=4
= x=4,y=2,z=1.
X=[x]=[4]
Ans.

Ex. 2 (b). Solve by matrix method

x-2y+3z=2,2x-3z=O,x+y+z=O.
Sot. Do as Ex, 2 (a) above. Ans. x = (6/19), y - (10/19), z = (4/19)
**EX . 3 (a). Solve by matrix method:
x +y +z = 6,x —y +z = 2, lv +y —z = I
(Gorakhpur 96: Kanpur 95, Rohilkhand 95)
Sot. Given equations are

x - : +z=2
2x + y - z 1
Let A = 1 1 11. K = 6 and assume that there exists a matrix
I —I II 2
2 1 —1] 1


108 Matrix
X= x such that AX =K
y
z
Then 1I I I x=6
I --1 1 y 2
2 I -I z 1
= it i] = 6 . by the elementary row
o - 2 0 y -4 operations R 2 -+ - R1
o -1 -3j z -11 •R3-*R3-2R,
1 0-2lx -5 , by the elementary row
I -
operationsR1-R1+R3
0 3] '
.r-2z=-5.-2y=-4,-y-3z=-1l
• x-2z=-5,y=2,2+3z= 11
= x=2z-5,)=2,Z=3
.x=6- 5,y=2,z=3=x= 1.y=2,z 3. Ans.
*Ex. 3 (b). Solve by matrix method
x+2y+3z=14,3x+y+.2z=11,2x+3y+z=1].
•
(Bundelkhand 92; Purwrnchal 93; Rohiikhand 98)
Sol. The given equations are
x+2y+3z=14
3x + y + 2z = ii
2+3y+z=11
Let A = 1 2 3 and K = 14 and assume that there exists a matrix
3 1 2 11
2 3 111
X= xYsuch that AX = K.

z
Then r1 2 3]rxl=1141
3 1 2 1 Y 111
[2 3 1j[z j [U
1 2 3 x = 141, by the elementary row
4 3 5 y 25 operation R2 -3 R 1 +
2 3 1 z 11]
2 4 6 x j = 281, by the elementary row
2 0 4 14 operation R1-2R1
2 3 1 z II] R-,--*R,-R1


Solution of Non-homogeneous Linear Equations 109

2 x = 141. by the elementary row

0 4
2 0 4 y 14 operation R 1 —* R 1 - R2,
0 3 -3 -3] R3-+R3-R2
71' by the elementary row
2 1
j operation P 1 -4 R1,
0 2
1 -1 z -_ij R2-4-R2,R1--*R3
ro
[0 3 0 X =6 by the elementary row
-- j operations R 1 -4 R 1 + R3

(Note)
3y6,x+27,YZ=
. y=2,z=y+I+I3.X72Z76=1
Ans.
x=1,y2,z3.
Ex. 3 (ci. Solve the following equations by matrix method
(Agra 96, 93)
x+2y+3z=4,2-r+3Y+8Z=7,x-Y+9Z
Sot. The given equations are

2x + 3y + 8z 7
X - I, + 9z 1

Let A =[i 2 3 K 4 and assume that there eistS a matrix X = x

7 y
2 38
-19
z

such that AX = K.
Then 1 23 x4
2 38y 7
1 -1 9 r 1

= i 2 3 [ by elementary row
41.
i operations R2 -4 R2 - 2R1
0 - 1 2 v I -
0-3 6 z L-3 and R3-RyRi
12 3 x = 4 1. by elementary row
o - 1 2 y - 1 operations R 3 —* R3 - 3R2
0 0 0 z 0]
x+2y+3z4,-Y+2Z=i
x+7z2.--Y+2Zl
Ans.
= x = 2 - 7z, y = 2z + I and z can take any finite value.
**Ex. 4 (a). Using matrix method, solve the following equations-
2x -y +3z = 9,x +y +z = 6 and - y +z = 2.
(Avadh 98. Agra 95; Garhwa.1 95; Gorakhpur 99: Kanpur 90;
Meenil 92P. 91; Rohilkhand 97)
 110 Matrices
Sol. The given equations arc
2x—y+3zr9
x+y+z=6
x—y+z=2
Let A=2 —1 3 K=[691andassumetha[thereexistsamatnxX_IYZI
x
1 Ii
1-11 2
such that AX = K.
Then 1 2 -1 3 x=9
1 1 1 y 6 (Kanpur 90)
—1 1 z 2
3 0 41[Yx =[is3 by the cicmentaiy row
I61 operations R 1 -p R 1 + R2.
2 0 2z 8] R3 -4 R3 + R2
L

3x+4z=15,x+y+z6,+2z8
3x+4z=15,x-I.y+z=6,x+•z=4
y=6—(x+)=6--4 :x+z=4
or y=2
Also 3x+4z=1s gives 3x+4(4....x)=5 . x+z=4
or 3x+16-4x=15 or x=1
z4—x=4-1=3
=.
EL 4 (b). Solve by matrix method only the equations:
X+y+z=6;x+2,+314,x+4y+9Z=36
- (Gorakhpur 98, 91; Kanpur 94, Rohilkhand 99)
Hint: Do as Ex. 4 (a) above.Ans.x=I,y=2,z3,
**EX S (a).
Solve the following equations by matrix method:
x+2y+z=2,3+sy+=+4y+..3 (Bundelkhand9J)
Sol. Let A = 1 2 J] K = 2 and asume that there exists a matrix
3 5 51 4
2 43 J 3
X= xl,suchthatAx=K,
y
z
Then 1 2 11 X]:= 2
5 5y
2 4 3
31zJ
 Linear Equations Ill
Solution of Non.h O mOgCfleOUS
the elementary row
1 2 tlIxl = 21. by
-21 operationsR2-4R23R1
o -1 2 11 y 1
o 0 ljLzj -1] and R3-*R32Ri

x+2y+Z2,-Y+2Z2. I

x+2y3,Y=0,Z=1
x+O3,yO,Z
Ans
x=3,y0,Z=l'
**Ex. S (b). Solve lbt equations:
Z 0. (Gorakhpur 92)
x +y +z9,2x +5y +7z=52,2x +y -
Sol. Let A = I 11 1 K = 952and assume that there exists a matrix
2 5 7 1
2 1 -1] 0

X= xY such that AX=K

z-
Then I I 1 9
2 5 152
2 1.-itij 0
JL
i x =
9]. by the elementary row
ii
34 I operations R2 -4 R2
- 2R1
o 3 5 y
o -1 -3 z _l8jandR3-4R32RI

1 1 x = 9]. by the elementary row

= 1
o 0 - 4 y - 20 operations R2 -4 R2 + 3R3
o -1 -3 z _18 j

x+y+z9;4z20Y3Z8
z=5;y 18-3z=18-3(S)3
and x9-y-z=935 1
Ans.
x=1,y3,Z=5.
Ex. 6. Solve the following equations by matrix method-
2x-y+3=8,-x+2y+z43Y4 (Rohilkhand9i)
Sot. The given equations are
2x - y + 3z = 8
-x+2y+z4
3x+y-4z=O
Sol. Let A = 2 - 1 31 K = 8 n assume that there exists a matrix
-1 2 1 4!
3 1-4 0
 112 Matrices 18211117
Y
X
i such that AX = K.

X= [ z
Then 2 - I 3[x= 8
-1 2 ly 4
3 I --4 zj
_ 10-j
0 3 5 x = so], by the elementary row
- 1 2 I y 4 operations R' 1 -. R 1 + 2R2
0 7 - z 12J R3-R3+3R2

[ 0 3 .S][Zx]=
. 16]. by the elementary row
- 1 2 1 J 4 operation R 3 R 3 - 2R1
0 i -lI J z j -20]
0 0
- 1 2
38 x] f
76]. by the elementary row
1 y j 4 operation R 1 - R 1 - 3R3
0 1 -II z] L-20i

' 38z=76,-x+2y+z4,y_lIz..20 '-

z=2.-x+2y=4_z2,y_20+ llz=-20+22=2
-x=2-2y=2_42y2z2
= X2,y=2,=2
Am;.
Ex. 7. Solve the equations—
XI +2X2+x34,x_xi+x35,1+32_X.i
Sol. Let A =[1 2 Iland K = [4 ].ind assume that there exists a
Il -1 Il Is
2 3-1 . 1

matrix X= [X j such that AXK -

[x3

Then1 2 •
Xj
J4
-2 15
2 3 -1 x3

o 3 0x1x.1
= [— i1 by the elementary row
I - 1 i x2 J operations R 1 -* R 1 - R2;
- [ 9] R 3 -4 R, - 2R2
01 L-
3x2=_I,xJ_x2+x35,5x2_3x3..9
- x2=— I13,x 1 +(I/3)+x3r5,5(_ 1/3)-3x3=-9
X21/3,x 1 +x3 = l4/3,x3='22/9.


182/11/8 Solution of Non-homogeneous Linear Equations 113

= x2 - 1/3, x 1 + ( 22/9) = 14/3, x 3 = 22/9.
= 20/9, x 2 = - 1/3, x 3 = 22/9. Ans.
Ex. 8. Solve the following equations by matrix method:
2x+3y+z=9;x+2y+3z=6,3x-t-y+2z=8. (Gorak.hpur95)
Sol. Let A = 2 3 1 K = 9 and assume that there exists a matrix
123 6
• 312 8
X= .4
Y
such that AX=K

z
- Then 231 x=9
123y 6
3 1 2 z, 8

o - 1 -5 . = - 31. by the elcmcntaty row

1 2 3 6 operatonsR 1 7-+ R1 -
o -5 -7 z -. 10] and R3-*R3-3R2
0 - 1 - 5 .r = - 31. by the elementaty row
1 2 3 y 6 operation R3 - R3 - 5R1
0 0 18 z 5]
-y-5z=--3,x+2y+3z=6,18z=5
I
25 5 5
-y--1j=-3.x+2y+6.z-j1

29 (29. 31 5
Y=ii,x+2JJ=-.-,z=T8_
• 35 29 5
Xjy j ,Z= jj Ans.
Exercises on 6.02—* 6.04
Solve the following equations by the matrix method -
Ex. 1. 3x+y-z=2,x+2y+z=3,-x+y+4z=9 (Puh'ancha.198)
Ans. x=24y=-1,z=3
Ex. 2. x+2y-z= 1,x+y+2z=9.2x+y-z=2 Ans.xJ2,y=1,z=3
Ex.3. x+2y+3z= 14,2x-y+5z= 15,3x-2y+4z=- 13
ho
Ans.x=- l7_§_,y=-1-,z= iO-j-
1

*Ex. 4. x±v+z=3,x.+2y+3=4,x+4y -9z=6. Ans. x=2.y=1,z=0

Ex. 5. 4x+3y+6z=25,x+5y+7z=13,2.x+9y+z1
I
Ans. x4Jy=- 1,z2
Ex. 6.x+y=5,2x-y=1. A/s.x=2,y=3
Ex.7. x-v+2z=3, 2x+z= 1, 3x+2 y +z=4. Ans. x=j-1,y=2,z=3


114 Matrices
Ex. 8. x-2y+3z=I1,3x+y--z=2,5x+3y+2z=3.
Ans.x =2,y=-3,z=1
Ex.9. x+y+z=9,2x+5y+7z=50,2x+y-z=2.Ans.x=l,y4.z=4
Ex. 10. x+y+zr4 , 2x_y+2z=5 , x_2y_z_3. (Agra 92)
Ans. x = 1, y = 1, z = 2
Ex. II. x-y+2z=4,3xy+4z6,x+y+z I. (Meerut 94)
Ans. x = 0, y = - 2/3, z = 5/3
Ex. 12. x+2y-z=3.3x-y+2z1,2x-2y+3z=2. (GrakJipur93)
An.x =- 1,v=4,z=4
Ex. 13. x + 2y + 3z = 6, 2x + 4y + z = 7, 3x +2v+ 91- 14. (Meerut 97)
Ans. x= l, y = l,z = I
Ex. 14. x - 2y + 3z = 6. 3x + y - = -. 7, .5x - 3y + 2z = 5. (Meerut 95)
Ans. .t = - 8/7, v = - 25/7, z = 0
Ex. 15. 5x-6y+4z= 15, 7x+4y-3z= 19: 2+y+6z=46. (Meerut 93)
Ans. x3,v=4.z=6.
* 6.05. To compute inverse of a square matrix with the help of the
linear equations.
Let asystem of n linear equations in n unknowns x 1 , x2, x3, Xn be
a11x1+a12x2--..+a1,t=k1
a21x1 +a 22x2 + ,. +a2flxfl_2
a 3 1x1+ a32-X2 + ... + a 3 x= k3

an I x I + a 2x2 + .. + anirxn = k.
Then ths system can be written in the matr i x from as
AX=K,
whereA= Pu a 12 X1 and K=
aufll.X_
a22 •-- a 1 x2 Ic,
• 131 a 32 a3
.,. ... x n Ic fl -J
I a 2a
nn
La ...See § 6.02 Page 105 Ch. VI
If A # b, the matrix A is non-singular and the inverse of A exists,
..,(Sce Ch. V and Ch. IV)
Hence premu tiplying (i) by A 1 . we have
• AAX=A'K or (A'A)X=A'K
or IX=AK. A'A.-1
or X = A- which gives the value of
 Inverse with the help of Linear Equations 115
Solved Examples on § 6.05.
Ex.1. If A 2 - 1 31, then find A and hence solve the equations
1 II
1 -1 1
2x - y + 3z = 9,x +y + z + 6 and -y -1- z = 2. (Kanpur 97)
Sot. The matrix equation AX = K is here equivilent to the equations
2x-y+3z=k1
x+y+z=k 2 ...(ii)
x-y+z=k 3 ...(iii)
Adding (i) and (ii) we get 3x + 4z = k 1 + k2 (iv)
Adding (ii) and (iii) we get 2x + 2z = k + k3
Multiplying (v) by 2, we get 4x + 4: = 2k 2 + 2k3 .,.(vi)
Subtracting (iv) from (vi), we get x - k 1 + k 2 + 2k 3 ...(vii)
From (vi), (vii), we get 4z = 2k2 ± 2k3 -4 (- *1 + k 2 + 2k3)
or 4z = 4k 1 - 2k2 - 6k3

or z=k1 -k2-k3
...(viii)
From (iii), y=x+z-k3
I
= (- k + k 2 + 2* 3) + ( k 1 - k 2 k 3) k3

or y=Ok 1 +k2 -k 3 ...(ix)

From (vii), (viii) and (ix) we get

X = - 1 I 2 i i.e.X=A1K
,-1

z - 3 k3

1 2

0 1 - I . ... ( x)Ans.
I _i -
.2 2
[
Also given equations can be written in the matrix from as AX --' K, ...(xi)
where A= 2 -1 3,X= x andK= 9
I I I v 6
I -1 I z 2

Now from (xi), we also have X = A 1 K

or x=-1 1 2 9
y 0 1/2 -1/2 6
1 -1/2 -3/2 2


116 Matrices 999

x=(- 1).9+ 1.6+2.2=1
y=(0).9+(1/2). 6+(— 1/2).2=2
z =(1) . 9 + (— 1/2) .6 +(— 3/2). 2 = 3
xl,y=2,z=3. A ns.
Ex. 2. Find the invesre of the matrix A = 1 2 3
050
243
Sol. The matrix equation AX = K is here eqaivalent to the equations
X I + 2 + 3x 3 = k (i)
O.Xt + 5x 2 + 0.x 3 = k2 ...(ii)
2x 1 + 4x2 + 3x 3 = k3
From (ii)we get X=!k2=O..k1+!k2+o.k3
...(iv)
Subtracting (i) from (iii), we get
x1+2x2=k3—k,

or x1 =k 3 —k 1 —2x2 =k 3 —k 1 —2.k 2 . from (iv)
or xj=—k1—(2/5)k2+k3 .(v)
Also from (I), 3x3 = k 1 -x1 -
=k 1 +k 1 +k 2 — k3 — k2, from (iv) and (v)
or _2,.
r k2
X3-A.1+u.--i2 I,
...(vi)
From (iv), (v) and (vi) we have
XI = k1
i.e.X=A1K
x1 0 0
_! k3

A'=

3 Ans.
Ex. 3. Solve the equations by finding the inverse of the coefficient
matrix:
x-6y+4z = l5;7x+4y-3z=19,2x+y+6z=4. (Gorakhpur9O)
Sol. The-coefficient matrix 5 - 6 4 A (say)
7 4 —3
2 1 6
The matrix equation AX = K is here equivalent tothe equations
5x-6y+4z=k1


Inverse with the help of Linear Equations 117

7x + 4y - 3z k 2 ...(ii)
2x+y+6z-k3 ...(iii)

Multiplying (iii) y 6 and adding to (i) we get

17x+40zk 1 +6k 3 ...(iv)

Multiplying (iii) by - 4 and adding to (ii) we get

x+27z=4k3-k2
Multiplying (v) by 17 and subtracting from (iv) we get
...(vi)
z=_(kj+17k262k3)
Substituting this values of z in (v) and simplifying we get
...(vii)
x=--(27k,+40k2+2k3)
Substituting values of x and z in (iii) we get on simplifying,
...(Viii)
Y = (- 48kg + 22k 2 + 43k3)

From (vi), (vii) and (viii) we get

x_j_ 27 40 2 k1
Z 419 -48
y 22 43 k2
-1 -17 62 1(3

i.e.

A'=(l/419) 27 40 2 -
-48 22 43 Ans.
-1 -17 62
Also the given equations can be written in the matrix form as AX = K or

X A- where
YX x;A 1 27 40 2;K 15
-48 22 43 19
z •-1 -17 62 46

or Z = j 27 40 2 15
419 -48 22 43 19
-1 -17 62 46
x = (I/ 4l9 )[ 27 ( 15 ) +40 ( l9) +2(46)I3 -
y(1/419)[-48(15)+22(19)+43(46)14
z(l/4l9) [-1(15)- 17(19)+62(46)16
xz3,yz4,z6. Ans.
Exercises on § 6.05
Find the inverses of the following matrices
Ex. 1. 1 2 3 Ans.! 11 -9 1
13 5 3-7 9-2
1 5 12 2 -3 1


118 Matrices
Lx. 2. 5 -2 4 Ans. 1 . I -4 6
-2 1 1 37_4 16 13
4 10 6 13-I
Lx. 3. 0 2 1 3 -17 -3 15
Ans. I9 7
1 1 - 1 2 1 - 5 -4
1 2 0 1 -10 -5 10 5
-11 26 -t I 0 1
Ex. 4. Find the inverse of the coefficient matrix of the following system of
e4ution s—
X +Z = I,X+ 2y+ I =0
and hence solve them. Ans. 2 -1 0;x=1,y=lz=-I
-1 2 -1
0-1 1
Ex. 5. Solve the following equations by finding the inverse of coefficient
matrix
+ y + z = 9, 2x + 5y + 7 = 52, 2 + Y - z 0 Ans. x = 1, y = 3 z = 5
Ex. 6. Find the inverse of the matrix A 2 5 3 and apply the results
31 2
1 2 -1
to solve the equations.
+ Sy + 3z = 9, 3x + y + 2z = 3; x z = 6.
+ - Ans.- r = 1, y = 2, z = - 1
§ 6.06. Augmented Matrix.
Definition : The matrix A. a ll 012
0 2l a 22 a

0 m1 0m2

augmented by the matrix K k1

is called augmented matrix of A and is
k2

written as A* or [A, K]
a ll°12 a1, k1
A'=
0 21 a 22 ... a 2, k

0m1 0 m2 -.. a,,,, km

Also it is evident that the order of the matrix A or [A, K] is m x (a + 1).

 lundamental i tieorem 119

* 6.07. Fundamental Theorem.

A svstcnz oJm linear equations in n unknowns given by AX = K is consistent
(i.e. has a SOtl(tU)n) i (Old ozzlv tj ilie matrix of coefficients A arid the augmented
,naIrzx A* of the s ystem have the came rank. (Agra 94, 92)
If the above corn/non rank is r then r of the unknowns can be expressed as
lmn('ar conibuzalion.' 'f t h e ,cnzalnlnz,' n - r unknowns. When these n - r
Unknowns are assigneil arbitrary values, the system has an infinite number of
solutions out of which (n - r -m- ] ) are linearly independent whereas the rest are
linear co,zibi/1a1iVis of iheni. I
Proof. Consider in non-homogeneous linear equations in n unknowns given
by AX = K, where
A a 11 '12 .,. 0 1n ,X= x ll and K= ki
a21 0 22 .. 12n
X)] I k

aflil a, •. a,,.4 Xp .

Let r he the rank of the matrix A and C 1 , C-,, C2..., C, be the column
vectors of the matrix A, then A = (C1, C2 , ... C r2 ] and so AX = K reduces to
[C 1 , C 2 .....C] =K

xn

or C1-Vi + C 2x1 + - -. + C,7 = K.

Necessary Condition. I..ct the given system of equations possess a solution
(i.e. be consistent), then there must exist n scalars h 1 , b2.....b,.1 which satisfy (i)
i.e. C1b1 + C 2b2 + ... + C,,h = K.
Since tank of A is r, so each it - r columns viz. C + 1' C + 2', C n is a
linear combination of C 1 , C2..... C.
From (ii) we find that K is a linear combination of C 1 . C 2 , .., Cr. SH1C
C r+ Cr+ 2' C,1 . in (ii) can be expressed in terms of C 1 , C 2 . ........ C
The maximum number of linearly independent columns of the augmented
A* is r.
matrix [A, K] or A' is also r. Hence the rank of
Thus A and A* are of the same rank r.
Sufficient Condition. Let the matrices A and A* be of the same rank r.
Theii the number of linearly independent columns of the matrix A is r. But the
column vectors C 1 . C 2 , ... Cr of th matrix A already form a linearly
independent set and thus the matrix K can he expressed as a linear combination
of the columns C 1 . C2.....Cr.
 120 Matrices
There exist r scalars b 1 , b 2 ,.., hr such that
biCi+b2C2t --- +b r C r +OC +i + ...+OC=K
(iii)
From (i) and (ii) on comparing, we get
x 1 =b 1 ,x2 =b2 , ...,Xr=brxr+ 1=0.....x=0 and these are the solutions
of the system of equations given by AX K.
* 6.08; Theorem.
If A be an n X n matrix, X and K be n x 1 matrices, then the system of
equations AX = K possess a unique solution if matrix A is nc5n-singular
Proof. Let A [a 1 ) and I A I ;t 0.
Then rank of A and augmented matrix [A, K] or A* are both n. Thus from
§6.07 Page 120 Ch. VI we conclude that the system of the equations AX = K is
consistent.
From AX = K, we have
A' (AX) = AK, premultiplying both sides by A'
or (A1 A) X = A 1 K or IX = A'K, A- I A = I
or X = A'K is the solution of the given system of equations.
Now let X 1 and X2 be two sets of solutions of AX = K.
then AX1=K;AX2=K
= AX 1 = AX2, as each is equal to K
A 1 (AX 1 ) = A' (AX2 ), premultiplying both sides by A'
=t (A' A) X 1 = (A' A) X2

IX l =1X2, A'A=I
= X1=X2
- the solution is unique.
* 6.09. Reduced Echelon Form of a Matrix.
Definition. If in an Echelon Form matrix (See § 5.04 Page 36 Ch. V) the
first non-zero element in the ith row ties injth column and all other elements in
the fin column are zero, then the matrix is said to be in reduced Echelon form,
For example: In1 0 2 5J the first non-zero element in the second row
.0132
0000
iies in the second column and all other elements in the second column are zero.
Solved Examples on 6.07 6.09.
Ex. 1. Solve the system of equations:
x + 2y - 3z —4w = 6
x+3y+z-2w=4
 Solution of Non-homogeneous Linear Equations 121
..x+Sy—Zz-5w10
Sot. The given equations In the matrix form AX = K is
12-3-4 x=6
1 3 1 1 2 y 4
2 5 -2 —5 z 10
W

The augmented matrix A t = 1 2 -3 —4 6

1 3 1 —2 4
2 5 —2 —5 10

or At I 2 -3 -4 6], replacing R2 and R 3 by

o I 4 2 —2 R2—R1andR3-2R1
o i 3 - 2] respectively
- 1 0 - 11 -8 io] replacing R 1 , R 2 by R I - 2R2,
o i 4 2 - 2R 3 - R 2 respectively
o 0 0 1
This is a matrix in the reduced Echelon form having three non-zero row
and hence the rank of A is 3.
Simultaneously we get the reduced Echelon from of A viz.
I 0 -11 - 8 having three non—zero rows and hence the
o 1 4 2 rank of A is also 3
00 0 1

Thus we observe that A and A have the same rank and as such the gi.
equations have solutions which can be obtained as follows
The matrix equation is
1 0 —Il —81 - x -= 10
01 4 2 y —2
00 0 1 Z, 0
W

or x+Oy—llz-8w=l0;
y+41-+2w=-2andw=O
or y-4z-2:w Oz; x= llz+ 10. (Notes
Thus we rind (See § 6.07 Page 119 Ch. VI) that as the rank of A and A t is
3, so three of the unknowns viz, x, y and w are expressed as a linear function of
the remaining 4 - 3 i.e. one unknown viz. z.
By assigning arbitrary values to Z. an infinite number of corresponding
values of x, 3, and w can be obtained. Hence the system of equations has infinite
number of solutions.
Now we can show that the system has only n - r+ I i.e. 4 - 3 + I i.e. 2
linearly independent solutions. (See § 6.07 Page 119 Ch. VI)
-22


122 Matrices
Assigning two arbitrary values 0. 1 to z, we have two sets of solutions of
the given equations as

ti±Ii
x 0 21
-6
y_L2

Let any other solution of the given equations be

x= - 1,v=2,z=-l.w=O,
corresponding to the value - 1 of z.
If this third solution is a linear combination of the first two solutions then
a, b can be found as follows
IOa+21b- 1
• -2a-6b2
(Note)
O.a+ l.b- I
0.a + 0,b = 0
if lOa + 21b -1 (i)
-2a--6b2 (ii)
b = - 1.
Solving (I) and (iii) we get a = 2, b
These values of a and b satisfy (ii) also. Hence the third solution is a linear
:ombination of the first two solutions.
Ex. 2 (a). Examine if the following equations are consistent ? If yes,
solve it
r+y + 4z = 6,3x + 2y - 2z = 9, Sx +y + 2z 13. (Meerut 96)
Sol. The given equations in the matrix from AX = K can be written as
1 1 4 x= 6
3 2 -2 y 9
5 1 2 z 13

The augmented matrix

A * = l 1 4 6
3 2 -2 9
51 2 13

or 1 4 61 replacing R2. R 1 by
A* - 1
1) -1 -14 -9 R 2 - 3R 1 , R 3 -
4 0 -2 7] respectively
6 replacing R by R - 4R1
-Ii 1 4
-9
-1 -14
-4 13
-
 Solution of Non-homogeneous Linear Equations 123

- I 1 4 replacing R 3 by R 3 - 4K,
61.
o —i —14 —9
o 0 38 19

1 1 46j replacing R 3 by (1/38) !?3

o —i 14 —9
o () 1 1/2

- i i o 4]. replacing R 1 . R2 by R 1 - 4R3

o - 1 0 - 2 R 2 + 14R 3 respectively
O 0 1 1/2

- I 0 0 replacing R 1 by R 1 + R2
21.
o —1 0 —2
0 ' 0 1 1/2

I 0 0 2 1 replacing R2 by - R2
010 2
o 0 1 1/2

This is a matrix in the reduced Echelon form having three non-zero rows
and hence the rank of A is. 3.
Simultaneously we get the reduced Echelon form of A viz 1 0 01 which
010
001
is equal to 1 3 and so the rank of Ais also 3.

Thus we observe that A and A* have the same rank and as such the given
equations are consistent i.e. have solutions which can be obtained as follows
The matrix equation is
100 x= 2
0 1 0 y 2
. 0 0 1 z 1/2]

= x=2,v=2,z = 1/2. Arts,

Ex. 2 (b). Examine, if the system of following equations is consistent. If
consistent find the solution
x 4-y ±z = 6, Zr + 3 y - 2z = 2, 5x +y + 2z = 11
Sol. Do as Ex. 2 a) above.
Ans. Given equations are consistent. x 1, y = 2, z = 3
Ex. 2 (c). Apply rank test to examine if the following system of
equations is consistent and if consistent then find the complete solution
Zr -Y + 3z = 8, - x + 2y -i z = 4,31 +Y - 4z = 0. (Garhwal 92)
Sol. The given equations in the matrix form can be written as
124 Matrices
2—! 3 x=8
—1 2 1 y 4
3 1 —4 0
The augmented matrix A 2 - I 38
—I 2 1 4
3 I —4 0

or A*- 0 3 5 16], replacing R L, R3 by R, + 2R2

- 1 2 1 4 and R3 + 3R2 respectively
o 7 —1 12]
- 0 3 5 16 1' replacing R1 by R 3 - 2R1
—12 1 4
o 1 —1! —20
- 0 0 38 76 ]. replacing R, by R 1 - 3R2
—1 2 1 4
0 1 —11 —20
- 0 0 12 ], replacing R by (1/38)R1
—1 2 1 4
0 1 —11 —20
- 0 0 1 21, reolacing R2 , R 1 by R2 - R1
- 1 2 0 2 R3 + I 1R 1 respectively
0 1 0 2]
- 0 0 12],, replacing R2 by - ( R 2 - R3)
1 —1 0 0
o 102
I - i 0 011 rearranging rows
0 102
0 012
-[
- 10 d 21 replacing R1 by R 1 + R2
0102
0.0 1 2
This is a matrix in the rçduced Echelon form having three non-zero rows
and hence the rank of At is 3.
Simultaneously we get the reduced Echelon from of A viz.
I 0 Oj i.e!3
010
001
and so the rank of A is also 3.


Solution of Non-homogeneous Linear Equations 125

Thus we find that the ranks of A and A' are the same and so the given
equations are consistent i.e. have solutions which can be obtained as follows—
The matrix equation is
1 0 0 x=2
OlOy 2
.001 z 2
x=2,y=2,2. . Ans.
or
*Ex. 3 (a). Apply rank test to examine if the following system of
equations is consistent and if consistent, find the complete solution:
x+2y-z6,3x-y-2Z3,4X+3Y+Z=9. (Meerut 98)
Sol. Given e uations can be written in the matrix form as
1 2-1 x6
3 -1 -2 y 3
4 3 lz 9

The augmented matrix A' 1 2 - 1 6

3 -1 -2 3
4 3 19

or A'- I 2 -1 6 1, replacing , '2, R 3 by R 2 - 2R1

-5 0 -9 and l? 3 + R 1 respectively
5 5 0 15]

7 15]. replacing R 1 , R3 by R 1 -R2

-5 0 -9 (1/5) R 3 respectively
0 3]
7 is]. replacing R2 by R2 - R3
-6 0 -12
0 3
-1/7 15/7]. replacing R 1 , R2 by (1/7) R1
0 .2 (- 1/6) R2 respectively
0 3]
- 1/7 15/71. replacing R1 by R3 - R1
0 2
1/7 6/7
- 1/7 1/71 replacing R 1 by R 1 - R2
0
1/7 6/7

Iii o
1
1/7

0 -1/7
0
6/71 rearranging rows
2
1/7


126 Matrices

- i 0 0 i , replacing R 1 by R 1 + R3 and
o i 0 2 R1by-7R3
o o I -1
This is a matrix in the reduced Echelon form having three non-zero rows
and hence the rank of A is 3.
Simultaneoausly we get the reduced Echelon form of A viz. 1 0 0
010
001
i.e. 1 3 and so the rank of A is also 3.

Thus we find that the ranks of A and A* are the same and so the given
system of equations is consistent i.e. have solutions which can be obtained as
follows—
The matrix equation is
lOOx= I
010)' 2
001 z -1
x= 1,3 , = 2,z=- 1. Ans.
.*Ex. 3 (b). Apply rank test to examine if the following system of
equations Is consistent, solve them
Zr + 4y - z = 9, 3x - y + Sz = 5, 8x + 2y + 9z = 12.
Sot. The given equations in the matrix form AX = K can be written as
2. 4-1 x= 9
3 -1 5 v 5
82 9 z 19

The augmented matrix A = 2 4 - 1 9

3-1 55
8 2 9 19
or At =2 4 - 1 9 . replacing R by R 3 - 2R2
-1 3 55
2 4 -1 9

-2 4 -1 9, replacing R2 , R 3 by R2 - R1

-5 6 -4 and R- R 1 respectively
0 00
-2 4 -I 91, replacing R2 by R,- R1
0 -7 13/2 - 17/2
0 0 0 0

-2 0 19/7 29 . replacing R 2 by - (1/7) R2

0 I -13/14 17/14 land then R 1 by R 1 -- 4R-,
0 0 0
0j

Solution of on-homogeneous Linear Equal IOflS 127

- 1 0 L9 Lfl,
4 replacing R 1 by R1
14
3
o 1 14 14

o 0 0 0
and
This is a matrix in the reduced Echelon form having two non-zero rows

hence the rank of A* is 2.

Ion form of' A viz
Simultaneously we get the redv d
sand as such the rank of
I 0 (19/14) 1 which also has two non-zero
0 I --(13/14) A is also 2.
00 0
Thus we observe that the rank of A and A are the same and as such ih
have solutions which can he obtained a
given equations are consistent i.e.
follows -
The matrix equation is
1 0 (19/14) x =1(29/14)
0 1 —(13/14) 1(17/14)
() () o z 0
[
(29/14): y - (13/14) (17/14)
or x + (19/1 71

13li 29 19
y=_±jjZ
or
19 29 13 17
or 14 14 14 14'
Thus we find that as the rank of A and A is 2, so two of the unknowns vu
one
x andy are expressed as a linear function of the remaining 3-2 i.e.
unknown viz.
By assining arbitrary values to z, an infinite number of corresponding values
of x and car) be obtained. Hence the given system of equations has an Infinite
number of solutions.
i.e. 2
Now we can sho that the system has onl y n - r + 1 i.e. 3 - 2 + I
linearly independent solution (See § 6.07 Page 120 Ch Vi).
Assigning to arbitrary values 0, I to z, we have two sets of solution s of the
'en equations as
x 29 101]
14 14

r 17
14

11111


128 Matrices 182/11/8

Let any other Solution of the given equation be x = (24/7), y=(2/7),
Z = - 1, corresponding to the value - I of z.

If this third solution is a linear combination of the first two solutions then
a and b can be found as follows
(29/14)a+(10/14)br-(24/7)
(17/14)a + (30/14)b =(2/7) (Note)
0.a + 1.b = -
or 29at lOb =48
17a+30b=4
...(ii)
b=- 1. ..(iii)
Solving (i) and (iii) we get a = 2. b = - 1, which satisfy (ii) also. Hence the
third solution is a linear combination of the first two solutions.
Ex. 4. Apply rank test to examine if the following system of equations
is consistent and if consistent, find the complete solution.
I +y+z = 6,x+2y +3Z r! O,X+2y+4Zz 1.
Sol. The given equations in the matrix form AX = K can be written as
1 1 1 x= 6
1 2 3 y 10
I 24z I

The augmented matrix ,A' = I I I 6

1 2 3 10
124 1
or A -. [1 1 I 6]. replacing R3 by R3 - R2
jI 2 3 10
0 0 I -9
1 1 1 6]. replacing R 2 by R2 -- R1
012 4
o 0 1 -9
- 1 0 - I 2], replacing R 1 by R 1 -
01 2' 4
0 0 1 -9
or A' -[l 0 0 -7 , replacing R 1 , R 2 by R 1 + R3
0 1 0 22 and R2 - 2R3 respectively
0 0 A -9

This is a matrix in the reduced Echelon form having three non-zero rows
and hence the rank of A is 3.
Simultaneously we get the reduced Echelon form of the matrix A viz.


192/11/9 Solution of Non-homogeneous Linear Equations 129

I 0 01 i.e. l and hence the rank of A is 3.
o 1
o o
Thus we find that the ranks of A and A are the same and so the given
equations are consistent.
The matrix equation is F 0 ol x T-
1 OH 22
[0 0 I] z
or v=-7,y=22,z=-9. Arts.
Ex. 5. Are the following equations consistent?
x +y 4-2Z + w = 5
Zr + - z - 2w = 2
4r+5y+3z=7. (Agra 91)
Sot. The given equations in the matrix form AX = K can be written as
1 1 2 1 x=5
2 3 —1 —2 y 2
45 3 0 z 7
W (Note)
The augmented matrix A = 1 1 2 1 5
2 3 —1 —2 2
45 3 07

or A— I 1 2 I 5 1 , replacing R-, and R3 by

01 —5 —4 —8 .R2 -2R 1 and R3-4R1
01 - 5 4 - 13] respectively

—10 7 5 131. replacing R I , R 3 byR 1 —R2

01 —5 —4 - 8 R3 - R2 respectively
00 0 0 - 5]

This is a matrix in the reduced Echelon form having three non-zero rows,
hence its rank is 3. -
Simultaneously we get the reduced Echelon form of A viz.
I 0 7 5 which has two non-zero rows hence its rank is 2. -
0 I —5 —4
00 0 0

Thus we find that the ranks of A and A are not the same, hence the given
equations are not consistent i.e. they cannot have any solutions.
Ex. 6. Discuss the consistency and find the solution set of the following
equations
x + 2y + 2z = 1, Zr +y +z =2,3x+ 2y + 2z = 3,y +z =0.
130 Matrices
Sol. The given equations in the fnatrix form AX = K can be written as

322z
1
1 2 2 x1
21
3
2

011 0
J
, 1221
The augmented matrix A*
2 1 1 2
3223
0110

A' - 1 2 2 1 replacing R2 , R3 by R2 - R4 and

or
2 0 0 2 R3 - R 1 respectively
2002
0110
1 0 0 1 replacing P 1 . R2, R 3 by R 1 - 2R4
1 0 0 1 1 R2 , 2R respectively
1001
0110
- 1 0 0 11 replacing R,. R 3 by R., - R1
0 0 0 0 R 3 - R 1 respectively
0000
0 11 0
- 1 0 0 1 interchanging R2 and R4
0110
0Qi0
0. (.1 0 0
This is a matrix in the reduced Echelon form having two non-zero rws,
hence its rank is 2.
Simultaneously we get the reduced Echelon form of A viz.
1 0 o1 which has two non-zero rows and hence its rank is also 2.
011
000
00 0]
Thus we find that the ranks of A and A' are the same and so the given
equations are Consistent.
:.The matrix equation reduces to
1 0 0 1 x =
011Oy 0
0000z 0
0000 0


Solution of Non-homogeneous Linear Equations 131

x=1,y+zO. ...(i)
or
Also here we find that the rank of A andA' is each 2 i.e. less than the
number of unknowns viz. x, y and z. So the number of solutions of given
equations will be infinite given by (1) above, which gives x = 1 and + z = 0 can
be satisfied by.an infinite number of values e.g. 0,0; 1, - 1; 2, —2; etc.
= 9,
Ex. 7. Show that the equations Sx + 3y + 7z = 4, 3x + 26y + 2z
(Bursdelkhand 96)
7x + 2y ± lOz = 5 are consistent and solve them.
Sol. The given equations in the matrix form AX = K can be written as
5 3 71x4
3 26 2y 9
7 2 l0z 5

The augmented matrix A * = 5 3 7 4

3 26 2 9
7 2 10 5

or A - —4 75 i - 231, replacing R 1 . R3 by
3 26 2 9 - 3R2 and R3 - 5R2
—8 —128 0 _40] respectively.

4 75 - i 231, replacing R I , R3 by - R 1 and

3 26 2 9 R3 - 2R 1 respectively
0 22 —2 6]
- 4 75 - I 231 replacing R3 by 1 R2
3 26 2 9
• 0 11 —1 3

- 4 64 0 20]. replacing R 1 , R2 by R 1 - R3
3 48 0 15 R2 + 2R3 respectively.
0 11 —1 3]
1, replacing R I , R2 and R3bY
1 16 5
I 16 I R 1 ,-R2 and (1/l1)R3
o 0
1 —(1/11) (3/11)] respectively.
I 0 (16/1I) (7/11) ]. replacing R2byR2Rl
o o 0 0 and then R 1 by R 1 16R3
LO 1 —(1/11) (3/11)]

I 1) (16/1I) (7/11)1, interchanging R 2 andR3

o I --(1/11) (3/11)1
JJ
This k a matrix in the reduced Echelon form having two non-zero rows,
hence its rank is 2.
132 Matrices 999
Simultaneously we get the reduced Echelon form of A
viz. 1 0 (16/11) which also has two non zero rows
o 1 - (1/11)and so its rank is also 2.
00 0

Thus we find that the rank of A and A' are the same and as such the given
equations are consistent. And so the matrix equation reduces to
1 0 (16/11) x = (7/Il)
0 1 —(1/11) y (3/11)
00 0 z 0
or x+(l6/ll)z=(7/ll);y—(l/11)z=(3/ll)
or llx=7-16z,lly=z+3.
Thus we find that as the ranks of A and A' is 2, so two of the unknowns viz.
x and are expressed as a linear function of remaining 3 - 2 i.e., one known viz.
Z.
By assigning arbitrary values to z, an infinite number of corresponding
values ofxand'y can be obtained. Thus given system of equations has an infinite
number of solitions. Now we can show that the system has only (n - r + I) i.e.
(3 - 2 + 1) i.e. 2 linearly independent solutions (See § 6.07 Page 119 Ch. VI).
Assigning two arbitrary values 0, 1 to z, we have two sets of solutions of the
given equations as
X 7 .9
11 - 11
Y 3 4
________ 11 11
Z 0 1
Let any other solution of the given equation be x=(23/ll),y=(2/lI),
z = - 1 corresponding to the value - 1 of z.
If this third solution is a linear combination of the first two solutions, then
a and b can be found as follows
(7/11)a—(9/11)b=(23/11)
(3/11)a+(4/11)b=(2/1I) ...(ii)
0.a+l.b=—l. ...(iii)
From (i) and (iii) we get a=2,b=— 1, which satisfy (ii) also. Hence the
third solution is a linear combination of the first two solutions.
Ex. 8. Apply test of rank to examine if the equations x + y + z = 6,
x + 2y + 3z = 14, x + 4y + 7z = 30 are consistent and if consistent find the
complete solution. (Kumaun 91; Meerut 96P)
Sol. In the matrix form AX K. the given equations can be written as
 Solution of Non-homogeneous Linear Equations 133

1 1 I x= 6
1 2 3 y 14
1 4 7 z 30

The augmented matrix A' = 1 1 1 6

1 2 3 14
1 4 7 30

or 6 replacing R 2 and R1 by R 2 - R1
0 1 2 8 and R 3 - R t respectively.
I1 1
0 3 6 24

- 1 0 - 1 - 2. replacing R 1 and R 3 by R 1 -R2

0 1 2 8 and R 3 - 3R2 respectively.
0 0 0 0]
This is a matrix in the reduced Echelon form having two non-zero rows,
hence its rank is 2.
Simultaneously we get the reduced Echelon form A viz-
1 0 1 which has two non-zero rows and hence its rank is 2.
01 2
00 0

Thus we find that the ranks of A and A are the same and as such the given
equations are consistent.
Now the matrix form of the given equations reduce to
10-1 [rI=-2
01 2y 8
00 Oz 0

which is equivalent to
I .x+0.y_1.z_2;0.x+1.y+2.Z=8;0.X+O.Y+OZ°
or x—z-2,y+2z=8

As the rank of A and A' is 2. so two of the unknowns viz. x and y are
expressed as a linear function of the remaining unknown z viz.
x=-2+z,v8-2z, where ziSarbitrary.
By assigning arbitrary values to z an infinite number of corresponding
values of x and y can be obtained. Hence the system of equations has infinite
number of solutions.
And x=_2+k,yr8-2k,z=k forms the general solution of the given
equations.
In the matrix form the solution can be written as
[Yx I =-2+k I ]
8 —2
z 0 1


134 Matrices
*EX. 9. Find the values of A so that the equations
ax + by + g = 0, hx + by + f= 0, gx + fy + c = A
are cólsistent.
Sot. The given equations in the matrix form AX = K can be written as
a h[xl= -g
h b[yj -f
g f A-c

The augmented matrix A =[a h - g

h b -f
g f X-C

or A*-[] h/a - g/a 1 replacing R 1 , R2 and R3 by

I b/h -f/h R11a,R21h and R319
I f/g {(), - c)/g}j respectively.

-1 h/a - gi l replacing R2 and

0 (b/h) - (a/a) - (f/h) + (g/a) R 3 by R . - R 1 and
2
0 (f/g) - ( h/a) {(A - c)/g} + ( g/a)j R 3 - R1

- 1 h/a -g/a
0 (ba - h2 )/ha (gh - aj)/ah
0 (af- gh)/ga ((Aa + g - ac)/ag)

- I h/a - g/a

o (ba - h2)/ha (gh-aj)/ah

(ia-h2) Aa + 2 - ac) (ba - h2)
o
ha ah(af-gh)

ha - h 2) R
(
replacing R2 by -
(af- gh)
-1 h/a - g/a replacing R3
0 (ha - h2)/ha (gh - af)/ah
0 0 {X(ab-h2)-(abc+2fgh
bg2 _ch2))/(h(af_gh)j

1 h/a - g/a, replacing R2 by

0 1 (gh-aJ)/(ba-h 2j ah R21(ba -h2)
0 0
 Solution of Non-homogeneous Linear Equations 135

Here h(af — gli)

Simultaneously we get the reduced Echelon form of A viz.
I h/a which has two non-zero rows hence its rank is 2
o I

o ()

From (i) and (iii) we conclude that if the given equations have solution then
the ranks of A and Amust be the same viz. 2 and from (i) if the rank of A' is 2,
then it must have two non-zero rows i.e. p. = 0.
- ch2)1(ab - 112) from (I). Ans.
i.e. X = (abe + 2fg1 - aJ2 -
**Ex. 10. For what values of X, the equations x +y +z = 1,
solution and solve completely in
x + 2y -i- 4z = x + 4y + JOz = A 2 have a
each case.
(Garhwal 90; Kanpur 97, 93, 91; Rohilkhand 92)
Sol. The given equation in the matrix form AX = K can be written as
I I I 1
1 2 4 y A.
1 4 iü z

The augmented matrix A 1 1 1

124 A
1 4 10 A2

r I 1 1 1 replacing R2 and R 3 by
or A
0 1 3 A - - R 1 and R 3 - R 1 respectively.
U 3 9

1 0 -2 2 - A replacing R 1 by R 1 -
0 3 9 3X - 3 and then R2 by 3R2

0 3 9 x2i
2—A 1rep1acingR3byR3_R2
1 0 —2
o 1 3 A—I and R2by--R2

0 0 0 X23A+2
Simultaneously we get the reduced Echelon form of A viz.
1 0 -2 has two non-zero rows hence its rank is 2.
01 3
00 0 . ...(ii)
From (i), (ii) we conclude that if the given equations have solution tlle n the
ranks of A and A' must be the same viz. 2 and from (i) if the rank A' is 2, then•
it must have two non-zero rows


M.arncci

i.e. X2 -3X+2=0 or X=1,2. Ans.

The\ matrix form of the given equations reduces to

0 11
0
I;
0 z X2-3A+2
which is equivalent to
I .x+0.y-2 .z=2-X,0.x+ 1. y + 3 .z=A - t
and 0.x+0.y+0.z=A2-3A-f2
IfA=I, then these arex-2z=J,y+3z=O.
As the rank of A and A is 2, so two of the unknowns viz, x and y are
expressed as a linear function of the remaining unknown z viz
x=2z+l,y= -3z.
By assigning arbitrary values to z, an infinite number of corresponding
values of .x and y can be obtained. Hence the system of equations has infinite
number of solutions.
Assigning two arbitrary values 0, 1 to z. we have two sets of solutions of the
given equations as

t 1"]
Let any other solution of the given equations be x=- 1,y=3.z=- 1
corresponding to the value - 1 of z.
If this third solution is a linear combination of the first two solutions then
a, b can be found as follows :
La + 3.b =- 1
0.a - 3.b = 3
0.a+l.b=-1 ...(iv)
or a+3b=-1,3b=-3 or b=-1
i.e. b = - 1, a = 2. These values of a and b satisfy all the three equations given
by (iv). Hence the third solution is a linear combination of the first two solutions.
We can similarly solve for A = 2 also.
*Ex. 11. Express the following system of equations into the matrix
equations AX=K
4x-y+6z=16,x-4y-3z=-16.
2x+7y+ 12z=48,5x-5y+3z=0
Determine If the system of equations is consistent and if so find its
SolutiolL
SoL In the given system of equations, we observe that the number of
unknQwns are not equal to the number of equations.


Solution of Non-homogeneous Linear Equations 131

The single matrix equation of these is
4 - I 6 x = 16 i.e. AX = K
1 -4 -3 y -16
2 7 12 z 48
5-5 3 0
The augmented matrix A' = 4 - I 6 16
1 -4 -3 -16
2 7 12 48
5 -5 3 0

A- o 15 18 80 re-placing R I , R3 arid R4hy

I -4 ,- 3 - 16 R 1 - 4R2 , R3 - 2R2 and
(7 15 18 80 R4 - 5R2 respectively.
o 15 18 80
or -4 -3 - 16 interchanging R 1 and R2
A' -
0 15 18 80 and then replacing R3.R4
o 0 0 0 byR-R2,R4-R2
1) 0 0 0
- I 0 (9/5) (16/3) replacing R2hy(1/15)R2
0 I (6/5) (16/3) and then R 1 by R 1 +4R2
00 0 0
00 0 0
This is a matrix in the reduced Echelon form having two non-zero rows
hence its rank is 2.
Simultaneously we get the reduced Echelon form of A viz.
1 0 (9/5) which has two non-zero rows and hence its rank is also 2.
0 1 (6/5)
00 0
00 0
Thus we find that the ranks of A and A' are the same and as such the given
equations are consistent.
Now the matrix form of the given equations reduce to
[i 0 ( 9 / 5 )1 =r(1o/3)
L0 1 (6/5)] y Loo/3
which is equivalent to
1 + 0. +(9/5) z =(16/3); Ox+ l.v+(6/5)z =(16/3)
or 15x+27z=80, lSy+ 18z=80.
As the rank of A and A' is 2, so two of the unknowns viz. x and y are
expressed as a t rneas function of the remaining unknown z viz.

-23


138 Matrices
15x = 80 — 27z and 15y=80- 18z
or x—(9/5)z+(16/) and y —(6/5)z+(16/3),
where z is arbitrary.
By assigning arbitrary values to z, an infinite number of corresponding
values of x andy can be obtained. Hence the given equations have infinite
number of solutions.
Also x = (16/3) - (9/5) k, y = ( 16/3) - (6/5) k, z = k forms the general
solutions of the given equations.
In the matrix form the solutions can be written as
. i6/3)1+k —(9/5)
16/3] -(6/5)
YZ =I 1 I Ans.
*Ex 12. Examine if the system of equations x + y + 4z = 6, 3x + 2y - 2z
= 9, Sx + y + 2z = 13 is consistent? Find also the solution if it is consistent.
So!. The given equations are

3x + 2y - 2z 9
5x+y+2z=l3
In the matrix form AX = K, these can be written as
1 1 4x= 6
32-2)' 9
5 1 2 z 13

The arugmented matrix A = 1 I 4 6

3 2 —2 9
5 1 2 13

or 1 1 61 replacing R 2 , R 3 by R 2 - 3R1
4
0 - I - 14-9 and R 3 - 5R 1 respectively.
0-4 —18 --17]

0 - 10 - 3], replacing R 1 , R 2 by R 1
-f R2,
- i
0 - 1 - 14 - 9 - 4R2 respectively.
0 0 38 19]

- 10 - 10 - 31 replacing R2, R3 by - R2,

0 1 14 9 (1/38) R 3 respectively.
0 0 I 1/2]

- 1 0 0 2], replacing R1byR1+lOR3

0 1 14 9
0 0 1 1/2
This is a matrix having three non-zero rows and in the rcduced Echelon
form, hence its rank is 3.


Solution of Non-homogeneous Linear Equations . 139

Simultaneously we get the reduced Echelon form of A viz.
1 0 0 which also has three non-zero rows and hence its rank is also 3.
o 1 14
00 1

Thus we find that the ranks of A and A are the same and as such the given
equations are consistent.
Now the matrix form of the given equations reduces to
I 0 0 xl= 2
0 1 14
00 1 Zj

which is equivalent to
Lz+O.y+O.Zr2O.X+1y+I4Zr9.OX+Oy+lZ=I

or xzz2;y+14z=9;z=i or x=2;y=9-14zz=1
or X2,yr9_7-=2,=.
Ans.
**Ex. 13. Show the equations - 2x + y + a = a, x - + z = b, x + y - 2z
= c have no solution unless a + b + c = 0 in which case they have infinitely
many solutions. Find the solution when a = 1, b = 1, c - 2. (Lucknow 90)
Sol. The given equations are
- 2i + y + z a
x-2y+z=b
x+y - =c
In the matrix form AX = K. these can be written as
—2 1 1 x=a
1 —2 1 y b
1 I —2 z c

The augmented matrix A * = - 2 1 1 a

1-2 lb
1 I —2 c
or A 0 0 0 a + b + ci, replacing R 1 by
1 —2 1 b 1R1+R2+R3
1 1 —2 c
j
1 -2 1 b ] interchanging R 1 and
0 0 a + b + c R2 and then replacing
0 1 —1 1(cb)] R3by(R3—RI)


14U Matrices

1 0 - 1 - (2c + b) interchanging R 2 and R 1 and then

o i - i 1 (c b) replacing R 1 by R 1 .+ 2R2
o 0 0 a -+-b+c

This is 'a matrix in the reduced Echelon form and has three non-zero rows,
hence its rank is 3.
Simultaneously we get the reduced form of A viz.
[1 0 -1 i. which has two non-zero rows and hence its rank is 2.
lo 1 —1
[0 o o

Thus we find that the tank of A and A are not the same and as such the
given equations are inconsistent i.e. have no solution.
But in case a -f h + c = 0, the augmented matrix A has two non-zero rows
i.e. the rank of A is also 2 and thus A and A* have the same rank 2.
Consequently the given equations have solutions if a + b + c = 0 and in this case
from (i) we have
A= 1 0 —1 (2c+b)

o i —i
00 0 0'

The matrix form of given equations reduce to,

1 0 - I x = ! (2c + F,) which is equivalent to
0 1 —1 y
- (c — h)
o o o
0

1.x+O.y— 1.z = (2c +b);0+ l.y — l.z=(c—b)

or x_z(2c+b),y—z(cb).

As a, b, c can take different values we shall get different solutions rather

infinitely many solutions (as will be evident below also) of the given equations.
If a = 1, b = 1, c = - 2, then from (ii) we get
x — z — 1,yz=— I
or x=z-1.yZ—1
By assigning arbitrary values to z, an infinite number of corresponding
values of xand y can be obtained. Hence in this case the given system of
equations has infinite number of solutions.
Assigning two arhitrary values 0, 1 to z, we have two sets of solutions of the
equations as
 Solution of Non-homogeneous Line;ir Equations 141

Let any other solution of the given equations be x = - 2, y = - 2, z = - 1,

corresponding to the value - 1 of z.
If this third solution is a linear combination of the first two solutions, then
two constants A, p. can be found as follows -
- I .A+O.t=-2
- 1,A+0.p.=-2
O.X+1.jt=—I

These values of A, M satisfy all the equations of (iii) and as such third
solution corresponding to z = - 1 is a linear solution' of the first two solutions.
*Ex. 14. Solve the equations with the help of matrices considering
specially the case when A = 2 :-
Ax+2y-2z=1,4x+2Xy—z=2.,6x+6y+Xz=3. (Kumaun9O)
So!. The given equations in the matrix form AX = K can be written as
A 2-2 x= 1
4 2A —1 y 2
6 6 A z 3.

The augmented matrix A A 2 -2 1

4 2A —1 2
6 6 A3

or A .. oA 12 -12 6 replacing R 1 . R3 by
-
12A 6A2 - 3X 6?, 6R 1 , 3XR2 , AR3 respectively.

6?. 6A A2 3Xj

or A— 6A 12 —12 6 replacing R 2 , R 3 by
o 6X 2 —24 24-3?. 6X-12 R 2 - 2R 1 . R3 - R 1 respectively.
o 6X-12 A2+12 3X-6

This is a matrix having th ree non-zero rows and in the reduced Echelon
form. Hence the rank of A is I
Simultaneously we get reduced Echelon form of A viz.
6A 12 - 12 which also has three non-zero rows and so its rank is
o 6A2.24 24-3?.
o 6X-12 A2-i-12
also 3.
142 Matrices

Thus the ranks of A and A are the same and as such the given equations
have solutions.
The matrix form of the given equations then reduce to
6X 12 l21= 6
- 24 24 - 3X y 6X - 12
o 6A2
3X-6
o 6X-12 X2 +12 zl[
This gives 6A- + 12y - 12z = 6
(6A 2 - 24 ) y + (24— 3X) z = 6X - 12
and (6X— 12)y+(),2 + 12)z=3X-6
If A = 2, then these equatons reduce to
12x+ 12y— 12z = ó; ISz=O
which gives z = 0 and 2x + 2y = 1
i.e. x=—v+fz0.y (Note)
By assigning arbitrary values to y, an infinite number of corresponding
values of x and z can be obtained. Hence the given system of equations has an
infinite number of solutions. (This can be ascertained from the ranks of A and
Aa1so in the case when A = 2).
Assigning two arbitrary values 0. 1 to y we have two sets of solutions of the
given equations as

2 2

Y 0 1

z 0 0
Let any other solution of the given equation be x = 3/2, y = - I, z = 0
corresponding to the value - 1 of y.
If this third solution is a linear combination of the first two solutions. then a
and can be found as follows
b
a—.h=3/2;O.a+ l.b— 1;0.a+0.b0.

Solving the first two of these we get a = 1, b = - I which satisfy the third
also.
Hence this solution is a linear combination of the first two solutions. In this
way we can get two linearly independent solutions of the given set of equations
for A = 2.
**EX . 15. Solve the system of linear equations:
• 2x-3y+4z=3,X-3Z=2.
Sol. The given equations in the matrix form AX = K can be written as


Solution of Non-Homogeneous Linear Equations 143

[2 -3 4 1 x =[ 3
[i 0 _3]y L-2
z
The augmented matrix A = 12 - 3 4 3
[I 0 -3 -2
01 A -3 10 7, replacing R j by -2R2
[1 0 -3 -2]
- 21. interchanging R 1 and R2 and
- [i 0 -3
[o I -(10/3) (- 7/3)] replacing R2 by i R3
This is a matrix having two non-zero rows, hence its rank is 2.
Simultaneously we get A -1' 0 - 3 1 which is also in the reduced
L 0 1 -(10/3)]
Echelon form having two non-zero rows. Hence its rank is 2.
Thus we find that A and A* have the same rank 2 and then two of t1.
unknowns can be expressed as a linear function of the remaining unknown.
Now the matrix form of the given equations reduce to
[I 0--31 x =[ -2
-(10/3)] y
L°
10 7
which is equivalent to l+ O.y_3.z = -2; 0.x+ l.y -

i.e. x-3z = -2, --3y+ 10z=7i.e.x=3z-2,y=z, for all

i.e. x = 3k -2, v = (1013)k - (7/3), z k, for all k. Ans.
Hence the complete solution of tLe given system of equations is x = 3k - 2,
y = (10/3) k - (7/3), z = k, for all k. Ans.
Exercises on § 6.07 - § 6.09
Ex. 1. Show that the following equations are consistatit
3x-4v=2,5v-2y12,-x+3y=1.
• 2. Show that the equations
2i 6' + ii 0, 6x+ 20v - 61- + 3 = 0, 6'- 13: + 1 = 0
a' uot consistent.
Show that ii in the following problems the given equations are consistent,
then solve them.
Ex. 3, 5k4 3v + 7.: =4, 3x+ 20v+2z=), 7x+ 2v+ lOz= 5.
(Kanpur 84; 1Jeerui 86) Ans. x = 5, v = 0,: = - 3
2.=2,21+4i*3=3.3xt±6.v2+5x3=4.
Ex.5. 1 1-x,+x=2,3.v1-x,+2.=-6,3xj+x2+x3-l8.
I'X. j . x -. 3x 2 + 2, 2r 1 +x 2 + 3x1 =3,x 1 -t-5x, + 5x 3 = 2.
Ans. x = I, x2 = - ( 1/5), . 2/5
144 Matrices 182/11/9
Ex. 7. Can the following equations have solutions ?
(1) x-y+3z+2w=3,3x+2y+z+w= 1,4s+v+2z-i- 2w=3.
(il)x-4y+7z=8.3x+8y- 2z=6,7x-8y+ 16z=31.
Ex. 8. Prove, without actually solving that the following system of
equations have a unique solution-
5x + 3y + 14 = 4, y +2z = 1,x-y+2z0.
.Ex. 9. 'Can the following equations have solutions?
(I)x+2y+3z=4,2x+3y+8z=7,x-y+9z=1. Ans. Yes
(il)x+2y+3z=2,2x+3y+4z5,3x+4y+5z=9. (Agra 92) Ans. No.
Ex. 10. Show that the following equations
xi-2y--z=3.3x-y+2z=1.2x-2y+3z=2,x-y+z=-- I
are consistent and solve them by the use of matrices.
(Rohilkhand 94; Garhwal 93)
Ex. 11. Show that the following equations are consistent and find their
solutions by matrix method.
x1+x2+x3=2,4x1-x2+2x3=-6,3x1+x1+x3=-18.
Ans.x1 10- X 2 10/3,x3=46/3
Ex. 12. Solve 3x-4y=2,5x+2y=12,-x+3y=l. Ans. x=2,y=1
Solution of Homogeneous Linear Equations
§ 6.10. Definition : A linear equation of the type
a1x1+a.x2+a3x3+ax4+...+a,x=O
is called a homogeneous linear equation.
§ 6.11. Definition : A system of homogeneous linear equations is given in
the matrx form by AX = 0, where A and X are the same notations as used in
§6.02 Page 105 Chapter VI.
Note I : Here K is zero matrix.
Note 2 The matrix of coefficients A and the augmented matrix A being
the same have equal ranks and thus the system is always consistent.
Note 3 : x 1 = 0 = x 2 = ... = x, is always a solution and is called the trivial
solution.
An Important Theorem (Without Proof).
A homogeneous system of 'n linear equations in n unknowns, whose
determinants of coefficients does not vanish, has only the trivial solution.
§ 6.12. Theorem (Without Proof)
A system m of in homogeneous equations in n unknowns x 1 , x 2 ..... x hasa
solution other than the trivial solution viz. x 1 = 0 = X2 = ... x, if and only ifihe
rank r of the matrix of coefficients A is less than n, the number of unknowns.
If r = n, then n of the equations can be solved by Cramer's Rule for the
unique solution x 1 = 0 = x2 = ... = x and the given system has non-trival
solutions,


145
12I11/10 Solution of Homogeneous Linear Equations
Xl,X2.... . xM can be,
expressed
Tfristhe rank of A. then rof the unknowns
as linear combination of the remaining n - r unknowns to which arbitrary values
ay be assigned. Hence the system will have an infinite number of solutions out
M
are lineatly independent and the remaining can be expressed a
of which n - r
linear combination of these n - r.
The above theorem is illustrated in the following examples
x+y+Z+WO
Ex. 1.Solve
+ 3y + 2z +4w 0
2x + 2- w =0
Sol. The given equations in the matrix form AX =0 is given by
liz =fO
1 3 I' i
2 4 1 y 10
2 01 —ljz [0
w

The matrix A of coefficients [ l 1 1 1

11324
2 0 I —1

A 1 1 1 I' replacing R 2 arid R3 by

or
R2_Ri and R3_2RlreSPti'eLy
o 2 1 I
0-2—I -3j
1 1, replacing R3bYR34R2
-11 1
10 2 1 3
0000

1 1 replacing R2bY2
jO 1 1/2 (3/2)
1.00 • 0 0
or A 1 0 1 i]. replacing C2 by C2 - C1
0 1 1/2 (3/2)
00 0 0
This is a matrix in the reduced Echelon form having two non-zero rows,
i.e. 4-
hence the rank of A is 2 and is less than the number of unknowns x, y, z and w
The matrix form of the given equations reduces to

1 0 1 1
1=0
Y 0
0 1 - (3/2) z 0

000 0

1.x+O.y+ 1.z+ l.w=O:

or O+ l.+ 1 .Z+ .w=0;

0..r+0.y+ 0,z+0.wO,


146 Matrices
From the first two equations we have

y-z-(3/2)
i.e. two of the unknowns viz. x
4
and have been expressed as linear combinations
Of the remaining two unknowns viz. z and w.
An infinite number of solutions of the given equations can be obtained by
assigning arbitrary values to z and W.
Also according to * 6.11 Page 144 Ch. VI we know that the system has
a.- ri. e. 4-2 i.e. 2 linearly independent solutions.
Take any two solutions of the system by assigning the following arbitrary
values to z and w
z=2,4
w = 0,2
Then the solutions are given in the tabular farm as

L_x -2
- 1
-6 J(Nole : The corresponding values
-5
J
of
and
are calculated from the equations
Z .2 4__j G) above)
L o 2J
Let any other sokion be
X=-2y=-3,z=(j,w2 ...(ii)
ob(aincd by assigning z and w the value 0 and 2 in equations (I).
11 this solution is a linear combination of the first two solutions (givqn in the
above table) then we can always rind Iwo constants A and z such
IJ'a*
-2A-6g-2 ...(iii)
-X-5u=--3 ...(iv)
2A+4t=0 No(e
0.A+2t=2 (vi)
From (vi) we get 1.
From (v) We get A -2.
These values of A and z satisfy (iii) and (iv) also.
Hence the third solution [given by (ii) abovel is a linear combination of the
First two solutions (given in the tabular fprm above).
Ex. 2. Find the solution of the following equations by the matrix
method:
2X1_X2+X30,3xi+2x2+x3.0x1_3x2+Sx_O
SoL The given equations in the matrix form AX =0 is given by
1 2 -1 i] x1

1 3 2 ijx2 o
Li -3 sJ [o


141
Solution of Homogeneous Linear Equations
The matrix A of coefficients
= 2-1 1
321
1 -3 5
1 replacing Ci,C2bYC2C3
4 0 0
C2 + C3 peY
I I 3 I
25]
L9
Iep1acingC2,C3byC23C1
10 0
11.0 ojc3-C1itY
29 14]
L9
0 0 (rcp1acingR3byR39R2
100
r
o 29 14
by
0 replac ing R3 R 3 141R I and C2
0 Ol by ( l/29 ) C2
1 0.1
r
.-1 i 0
Ia 0 I
010
- I 0 o]imterdiangingR2aMR3
010
001
=
is equal to the number of unknowns viz. x 1 . x, A
The rank of A 1s3 and
Howe XIOX23 (See 6.I2 pap 144Cb
mdz,
Es. 3. Show by consèdering rank of an appropriate the irMal
following stem of equations, possseS no ,udon other "a
otuttonsx=O = YZ :-
3x -y ±zO, - 15x 4-6y - 5z=0,5x -2y +2z=O
Sol. The give equations in the matrix form AX = 0 is given by
3-1il[x =10
-15 6 -5)y Ia
5 -2 2 jLz L°
The matrix A of coefficients
= 3-I 1
-15 6 5
5-2 2


148 Matrices 99°

-. 0 0 1 replacing C,, C2 by C,t. 3C 3, C2 -f C3
0 1 - 5 J respectively
—10 2j
- 0 0 i] replacing R2 , R3 by R2 + 5R,, - R3
0 1 0 respectively
'
1 0 —2j
0 0( replacing R3byR3+2R,
010
100
-Ii 0 0]. interchanging Rt and R3
10 •1.o
001

.. The rank of is 3 and is equal to the number of unknowns viz. x,

y, Z.
Hence x = 0 = y = z and the given system has no non-trivial solutions. (See
* 6.12 Page 144 Chapter VI).
Ex. 4.Sce:
4x - y +z + w=0
3X-2y+3z+4w0
x - 3y + 7z + 6w 0 (Kumaupz 94)
The given equations in the matrix form AX = 0 is given by
2 —2 5 3 x 0
4—I 1 1 y_O
3 —2 3 4 z
1 —3 7 6 w 0
The matrix A of coefficients
= 2' —2 53
• 4 -i 11
3 —2 34
1 —3 76
-.0 4 _9]. replacing R I , R2 and R4by
9
o ii —27 —23 R, — 2 R4, R2 — 4R4 and
0 , 7 -18 -14 R3 - 3R4 respectively
1 —3 7 6
- 1 —3 7 6]. interchanging R 1 and R4
0 11 —27 —23
0 7 —18 —14
0 4 —9 —9


149
Solution of Homogeneous Linear Equations
7 6replacing R2 by R2 - R3
- I -3
o 4 -9 -9
o 7 -18 -14
o 4 -9 -9

-3 7 6 replacing R4 by R4 - R2
o 4 -9 -9
o 7 -18 -14
o 0 0 -o

7 6replacingC2byC2+3C1
- 1 0
o 4 -9 -9 l
o 7 -18 -14
o 0 0 -0

- 1 0 7 6replacing R 3 by R3 - 2 R2
o 4 -9 -9
o 0 -9/4 7/4
00 0 0
V

6 replacing R2 by R2 and
1 0 7 -
V

0 1 -9/4 -9/4 R3by-R3

0 0 1 -7/9
00 0 0
non-Zero rows,
This is a matrix in the reduced Echelon form having three
hence the rank of A is 3 and is less than the number of unknowns viz. 4.
The matrix form of the given equations reduces to
1 0 7 6 x_O
o 1 -9/4 -9/4 y - 0
1 -7/9 z 0
o 0
00 0 Ow 0
V

- V -
f'•
or
...(ii)
Y ?z_w_0
z-W=O
and
From (iii) we get z (7/9) w
= - (103/9)w
From (I) x = - 7z - 6w = [- (49/9) - 61 w
w)[(7/4) + (9/4) w = 4w.
and y = (9/4) z + (9/4
expressed in
Thus we find that the three of the unknowns viz x,y and z are
terms of the 4th unknown viz. w.
be obtained by
An infinite number of solutions of the given equations can
assigning arbitrary values to w.


ISO Matrices
Also we know that the system has n - Ti. C. 4-31. e. I linearly independent
solution.
- Assigning w one arbitrary value 9, we have a set of solution as
- - lU, ) = .iO. Z I, w = 9.
lothersolution (by assigning IX to w) be
x 206. y 72, z 14 and w= 18.
It Is evident that this second set of values are
nothing but double of the first
set Of values, Hen ce the theorem of 6.12 Page 144
Chapter VI is fully verified.
EL 5 SoJve cosepletety the system of equatho;
* - 2y + z - w =0
S + y - 2* +3w =
4x + y - 5* + Sw = I
S*-7y+2*—wO (Kumuun93)
SoL The given equation in the matrix form AX =0 is given by
I — 2 1 —1 X
1 1-2 3y 0
4 1 —5 8 z 0
5 —7 2 —J w 0
The matrix A of coefficients

=1 —2 1 —1
1 1 —2 3
4 I —5 8
5 —7 2 —1

—1 —2 i - 1]. replacing R2 . R3 and R4 by

0 3 —3
0
41R2—R!,R3_4RIanI,JR4_5R,
9 . 9
12 respectively
3 '-3 0
4J
o —I 5/3].replacingR3 and R4byR1_32
0 I - 1 4/3 and R4 - R2
respectively, then
0
0 R2bYR2 and fiillYRIbyRI+2R2
gj
This is a matrix in the reduced Echelon form having two non_
hence the rank of A is 2 zerorows,
and is less than the number of unknowns viz. 4.
The matrix form of the given equations reduces to
1 0 —1 (5/3) x =10
0 I —1 (4/3) y 10
00 0 0 z 0
00 0 ' w 0
which is equivalent to


Miscellaneous solved examples 15'

i.e. yZ-(4/3)W,XZ(S/3)W.
X X = z-(5/3)w

Y z-(4/3)w
1 Z
W w
By multiplication we rind that
AX= I 0 - I ( 5 / 3 )1 z - ( 5/3)
o i - i ( 4/3)1 z-(4/3)w
o 0 0 0
0 0]
LO o
= z-(5/3)w-z+(5/3)w =0 =o•
0+z-(4/3)w-l+(4/3)W 0
0+0+0+() 0
0+0+0+() Q

whatever the values of z and w may be.

take any values, as the complete solution of the given system of equations.
**Ex. 6. Find the general solution of she matrix:
2 3-1-I 1=0
I-I -2-4 y
3 I 3-2 v
6 2 9-7 t
Sol. The given equation in the matrix form AX = 0 is given by
3 - I - I x 0
1-I -2-4 y _0
3 I 3-2 v
6 2 9-7 t 0
The matrix A of coefficients
=2 3-1-1
- I -2 -4
3 I 3-2
6 2 9-7
o 53 7 replacing R 1 . R 2 and R4 by
I - I -2-4
R 1 -2R2 .R3 -3R2 and R4-6R2
o 4 9 10
0 8 21 17 respectively.

() 1-6-3 replacing R 1 . R4 by R 1 - R2
I - I -2 -4 and R4 - 2 R2 respectively.
0 4 9 10
0 0 3-3


152 Matrices

-01 -6 -3 replacing R1 by R - 4R 1 and R2

'C -8-7 byR2+R1.
of 33 22
oc 3 -3

-o -9 -_0].replacingR 1 andR3 byR 1 -R4

0 -8 -7 and R3-l1R4repectively.
0 0 0 55
0 0 3 -3
-.0 0 R 1 , R2, R 3 and R4 by
-9 replacing

0 g —'j R,+3R4.R2+(8/3)R4,(V55)R3
0 - and (1/3) R4 respectively.

-0
1 0
o 0 replacing Rt, R 2 , and R4 by
0 0 1IRI+9R3,R2+l5R3 and R4+R3
0 0 I oj respectively.

0 0 01, rearranging the rows

0 00
0 0 10
0 0 01
[14).
:.The rank of the matrix A is 4 and is equal to the number or unknowns viz.
X. Y. V, t.
Hence x = 0, y = 0, v = 0,1 = 0 [See 6.12 Page 144 Ch. VII,
Exerdaeon*6jO-6.12
Ex. 1. Solve the following equations :-
x1 -x2 +x3 =0,x 1 +2.x2 - x3 =0,2x 1 +x2+3x3=0.
(Lucknow 92)
Ex. 2. Find the rank of the coefficient matrix for the following system of
hcnnogeneous equations over the field of real numbers and compute all the
solutions :-
x1+2x23x3+4x4=0,XI+3x2_x3O.&I4X2+3=O
Ex. 3. Solve completely the following equations using matrices :-
x+3y-2z0,2x-y+4z0,x_ lly+ 14z=0. (Lucknow 90)
Ex. 4. Solve completely the following equations with the help of matrices:
(i)x-y-z+t=0,x_y+_,o,+y+,o.
(ü)2w+x-y-z=0,4w_6x_2y+2zo,_6w+12X+3_4zo
MISCELLANEOUS SOLVED EXAMPLES
Ex. 1. Show that the only real value of A for which the following
eqIs have non-zero solution is 6:


Miscellaneous Solved Examples 153

(Kanpur 95)
x + 2y + 3z = Ax, 3x + y + 2z Xy, 2 + 2y + i = Az.
Sot. The given equations can be rewritten as
(I -X)x+2y+3z=O;
3x+(l -A)y+2zO;
and 2x+3y+(l-A)z=O.
The equations in the matrix form AX = K can be rewritten as
i -A 2 3 x10 0
3 I - A 2 yI
2 3 1-A zj 0

If the given system of equations has a non-zero solution then the matrix A
must have a rank < the number of unknown quantities x, y, z i.e. 3 and I A = 0
i.e. i - A 2 3 1 =0
3 i-A 2
2 3 1-A

or 6-A 6-A 6-A =Ø , replacing R 1 by R, +R2+R3

3 1-A 2
2 3 1-A

or (6 - A) 1 i 1 = 0, taking out (6- A) common from R1

31-A 2
2 3 I-A

or (6 - A) 1 0 0 = 0, applying C2 - C 1 and C3 - C1
3-2-A -1
2 1 -1-A

or (6-A) -2-A -J =0
I -I-A

or (6_A)[(2+X)(1+A)+110
or (6-A)[A2+3A+3100r A6,[-3±(9 12)]
A for which
or A = 6 (the other roots being imaginary) is the only real value of
the given system of equations has a non-zero solution.
Ex. 2. Prove that if the system of equations
ay + a, y = z + ax, z =x+y
is consistent (having non-zero solutions) then a + I = 0.
Sot. The given equations can be rewritten as
x - ay - zO
ax - y+ZO
X +y-z=0
These equations in the matrix form AX = K can be rewritten as

-24


154 Matrices
1
(I l I y 0
1 I - I
X1 = (
z 0
If the given system of equations has a no -zero solution then the matrix A
:iave a rank < the number o unknown quantities A, y, z i.e. 3 and I A = 0. Here
we have
I -a -I
a -I I

r A - 0 - - - ]. replacing C 1 , C by ( + C3,
a + I (1 + C1 respectively
0 0 -1]

() - (a + 1) j replacing R 1 , I?2 by k 1 - R3,

0].
a + 1 0 01 R2 + R 3 respectively (i)
() 0 _l

- 0 I o]. replacing C l . C2 .C3 by

I 0 () C1 /(a + l). - ('2 /(a+ I),
C) I] respectively 0

I 0 o]. interchanging R 1 and R2

0 I 0
001
= 13
i.e. the rank of A is 3 i.e. equal to the number of unknowns viz. x, y, z.
But ire, + 1=0. then from (i) we get
A- 0 00], which has one non- zero row and so the rank of A is I
00 0
0 0 -I
i.e. <3, the number of unknowns vi,., x, y, z.
Also AI = I
a -1 I
1 I -i
= a - a-I .adding C3 to C 1 and C2
a+I 0
0 0 -1
= 0 0 -1 ,ifa+t0
00 I
0 (1 -1

= 0, two rows being identical

 Miscellaneous Solved Examples 155

Hence the given equa ti ons are consistent if a + I = 0.

*Ex. 3. investigate for what values of A, ji the simultaneous equations:
x+2y+z=8,2s+y+3=13,3x+4y-Xz=p.t
have (i) no solution (0) a unique solution and (10) infinitely many solutions.
Sol. The given equations in the matrix form AX = K can be written as
I 2 1 x= 8
2 1 3 y 13
3 4 -A z t

The augmented matrix A' I 2 I 8

2 I 3 13
3 4 -A i

or ' i 2 18]. replacing R2 . R 1 . by

0 -3 1 -3 R2-2R,,R-3R1
() -2 A - 3 ii - 24j respectively,
- o ii]. replacing R 1 , R 2 , R1
o -6 2' _6IbyRu_R2,2R2and
o
I -6
1 - 3A -9 3 1i - 72j 3-R 3 respectively;
- I 5 0 II replacing R and R2
() I - I by R 3 - R2 , and -
o 0 -3X- II 3t 66 respectively

Case I If 3A+ II 0.3M-66() i.e. A;P-(lI/3),p.#22, then from (i)

matrix A' has three non-zero rows and is in the reduced Echelon form. Thus the
matrix A' is of rank 3. Also then the matrix 4 = I 5 0 is also in
o i - ii
o 0 -3k- II
reduced Echelon form having three non-zero rows and thus its rank is also 3.
Also there are three unknown quantities x, y, z so, in the case
A ^ - (11/3),;1 * 22, the given equations have a unique solution.
Case H. If 3A + II 0, 3j.t - 66 ;t 0 i.e. A = - (11/3), i * 22, then the rank
of the matrix A' is 3 but that of A is 2, since in this case both A'and A are in the
reduced Echelon form but A' has three rsrn-cro rows, wiicarcs A has two
non-zero rows. Thus the ranks of A' and A are not the same and so there is no
solution of the given equations.
Case HI. If 3A + II = 0, 3p. - 66 0i.e.A= - (11/3). t= 22, then the ranks
of A as well as A' arc the same and each is 2 i.e. less than the number of
unknowns viz. x. y and z. Hence in this case two unknowns will be expressed in
terms of the third and thus we shall have an Infinite number of solutions.


156 Matrices
Ex. 4. Investigate for what values of A and j.t, the simultaneous
equations : x+y+z6,x+2y+3z=10 and x+2y+Xz=j have (i) no
solution, (Ii) unique solution (iii) infinite solutions.
(Agra 96. 93, 91; Garhwal 91, 90; Kanpur 96, 94;
Meerut 91 S 90; Rohilkhand 96, 90)
SoL The given equations in the matrix form AX K can be written as
I I I xl=6
12 3yjlO
1 2 A zj L

The augmented matrix A' = I 1 1 6

1 2 3 10
12A ja

or A' = 1 0 - 12 replacing R 2 , R3 by R 2 - R, R 3 -
0 1 2 4 respectively and then R 1 by R 1 - R2
o 0'A-3 ja-10

Now following cases arise :-

Case I. If A —3 = 0, t — 10 # 0 i.e. A= 3, 10, then from (i), A ' is in the
reduced Echelon form having three non-zero rows and A is in the reduced
Echelon form having two non-zero rows.
The ranks of A' and A are 3 and 2 respectively which being different, the
given equations have no solution.
Case fl.IfA-3#0,ja-100i.e. A#3,ja#10, then from (i)we find that
both A' añdA are in the reduced Echelon form having three non-zero rows and
hence the ranks of A' and A are each 3 and these being the same the given
equations are consistent. Also there are three unknowns viz. x, v, z so the solution
is unique in this case.
Case Ill. IfA —.3 = 0, j.t - 10 = 0 i.e. A 3, j = 10, then from (i) we find that
the matrices A' and A are in the reduced Echelon form having two non-zero
rows each. Hence the ranks of A' and A are the same each being 2, which is less
than the number of unknowns x, y, z. Therefore in this case two unknowns will
be expressed in terms of the third and thus we shall have an infinite number of
solutions.
EL 5. Show that the following equations are consistent
x+y+z=-3,x+y-2z=-2,2x+4y+7z=7. (Kumaun 96)
Sol. Given equations can be written in the matrix form as
I II x=-3
1' 1 —2 y —2
24 7 z 7


Miscellaneous Solved Examples 157

The augmented matrix A s = I I I -3

1 1 —2 —2

24 7 7

or replacing R 2 , R 3 by R 2 -
- - .
3- R1 respectively
I 3 6 ioj R

- 11 I - 31 replacing R 3 by R 3 + 2R2
00-31
1 3 0 12

3, replacing R2.R3bY(I/3)R2.
-i
R3 - R 1 respectively
o 2 - I is]

- 1 i o _g /3 ].replacing R I , R3bYRI+R2
0 0 - 1/3 R3 - R 2 respectively

o 2 0 44/3]

- 1 1 0 —8/3]. replacing R 2 , R 3 by - R2.

0 0 1 - 1/3 j (1/2) R 3 respectively
o i 0 22/3]

1 0 0 _1o] . replacing RIbyRLR3

0 0 1 1/3
o i 0 22/3

- 1 0 0 - io]. interchanging R 2 and R3

0 1 0 22/3
0 0 1 —1/3
This is a matrix in the reduced Echelon form having three non-zero rows, so
its rank is 3.
ioo]
Simultaneously we get the reduced form of A viz. 0 I 0 ti.e. 13 and so the
0 0 I]

rank of A is also 3.
Thus we find that the ranks of A and A am the same and so the given
equations are consistent i.e. have solutions given by 1 0 0 ] - 10
o I 0 y 22/3
0 0 11 zj —1/3

which gives
Ans.
x=—l0.y22/3.zl/3.
 158 Matrices
EXERCISKS ON CHAPTER VI
Ex. I. Fxaminc whether the following linear equations are consistent, and if
C(InSiStCflt solve them :-
x 1 +x2 +x-4 +x 4 =ftZr 1 --x2+3x+44=4
and 3x1+4x1+5x4=l.
Ex. 2. Solve the equations by matrix method
x+2y+3zl4:x+j+z=6.x+3y+(z=25.
Ex. 3. Solve by matrix method :-
x-2y+3z=2,2x--3z=3;x+y+z=O.
Ex. 4. Solve by matrix method :-

Ex. S. Solve X+2y+3z=14.2x+v+2zzI1,2x+3y+z=11 with the

help of matrices. Ass. x 18/1I,y = Il/Il. z = 34/I1
EL 6. Investigate k such that the following system of linear equations is
consistent and ,btain its solution :-
2x+y—z= 12.x-y-2z=-3;3y+3z=k.
Ex. 7. Investigate for which values of A and p the following sysicmn of
equations will have
(I) No solutiqu and (ii) a unique soluliitn
x+2v+3'-S.3-'v+2z=l2.3x"v+Az=p.
EL 11. Investigate br what values tiC A and p. the simultaneous equa*ions
x + y + z = 16. x + 2 y + Sz = p. x + 2y + Az = 10 have unique solution. (Agra 90)
EL 9. Does the fi4kwing system of linear equations pus=s a unique
solution 7 if so, then solve them. if not, why?
x+2yi-3z=6,2x+4y+ z =7,3x,-2y+9z= 14.
EL 10. Find the inverse of the matrix I 2 31 and use it to soh'e the
2 3 2
3 3 4j
equations X+2y+3Z3.2x+3y+2ZrO.3x+3y+4z=5.
Ex. 11. Show that the following ystem of equations have unique solution:
x+2z=O,y+2z=l.Sx+3y+I4=4.
Ex. 12: Solve : Zr1 '2 + + = 0: 1 +x2 - x4=0.
- 3r2 + = 0: 4x, - +6X 3 -r4 =0-
(Hint. See Ex. I Page 145 Chapter VI).
F;x. 13. Stmtc the conditions under which a system of non-homogeneous
equations will have (i) no solution, (ii) a unique solution, (iii) infinity of
solutions.
Ex. 14. For what values of a, 1 do the system of equations
x+2 y + 3z=6.x+3y+5z=9;2x+5y+az=b have (i) no solution: (ii) a
unique solution: (iii) more than one solutions I


lxcrciscs On Chapter VI 159

Ex. 15. Solve x 2v-:--3;31-y+2z.= I; 2x- 2 y + 3z= 2; x - y + z = - 1.

Ex. 16. Find the inverse of the coefficient matrix and hence solve the
lollowing equations :-
4- 2y - 3z = I, 3 + y + z = 8, x - 2 0.
Ex. 17. Apply rank-test to examine lithe lollowing system of equations is
Consistent and 1 . consistent, then only find the complete SOIuIR)fl

Ex. 18. Apply rank test to find whether

x + 2 y r 3: 4 4w = 0, 8x + 5y + z + 4w = 0. 5- + ôy + Xz + 14' = 0,
8 + 3y + 7z + 2w =
have any solution other than x = y z = w = 0.
Ex. 19. Solve the following equations by the use of matrices
(a) x-y-2z-41=O.2x 4 3y-z--t=0.6x+3y-710.
3x+y + 3z -21 = 0.
(b) x+y+z0.2x-y-3z0,3.r-5Y+4Z0.X+17Y+4Z=O
(c) x+2y+3zrO.3x+4y+4'().7x+lOy+l2Z=O.
Ex. 20. Solve the equlions by matrix inversion
x+y+z=.'-y+3zI.3x+2y-z=l.
Ex. 21. Show by rn method that the following systcm of equations is
Consistent and solve it
4 3x 2 + 3x 1 + X4 =(). x 1 + 2 + x + x4 = 2;

3i1+4x2+2x3-x48;2x1+3x2+4X3+5X4=5
- =
X I - -2
Ex. 22. Show that the equations x - 3y - g + 10 = 0. 3x + y - 4z =
2x + Sy + 6, - 13 = 0 arc consistent and solve them. (Men,I 92)
Ex. 23. Solve by matrix method :-
x+y+z=4,x-y+5.2X+3y-ZI.
Ex. 24. Show that the system of equations is consistent
2x+6'=- lI,6x+20y-6z-3,6y- 18z=-l.
Ex. 25. Show by matrix method, the following equations are consistent and
have infinite number of solutions
4- + x = 0, 2x + Sx, + 6x 1 = 0.

Ex. 26. Solve by matrix method

x + v + z + w = I, 2 - ' + z - 2w 2, 3x + 2y - z - w 3.
Ex. 27. Slove the following equations :-
2x 1 -- 2x 3x - 3, x 1 + 2 2 - - 5 4 = 4, 2x, 2x1 + 7x4 = 5
(Rohilkhand 93)
Ex. 28. II x, v, are not all zero and i O = 0,
hi + y + az = 0, cx ay
4- -4 b - 0, prove that,
= I : I : I or I : () : ( 12 or I : ()2
where IL) i ,,aconipics cubc riot of Li ty. (I'urvanchal 98)
Chapter VII

Characteristic Equation of a
Matrix
* 7.01. Zero Divisors.
We have previously read in § 5.10 Page 76 of Chapter V that the necessary
and sufficient condition for a square matrix to possess an inverse is that it must
be non-singular. From this result, we get a very important result which does not
hold for ordinary multiplication of numbers viz. 1f A and B are two singular
matrices, it is possible to obtain the result AR = 0, where neither A = 0 nor
B = 0, 0 being the null matrix'. In such a case A and B are called proper
divisors of zero.
If however, A and B be two square matrices of order it such that
AB=O,
then if .A is non-singular, A * 0, A exists and A 0.
Pre-multiplying (I) with A - get
A 1 AB=A 1 0 orIB=0,: A.O=0andA1A=I
or B=0.
Hence we conclude that
If A*O, then AB=O=B=O
Similarly if B 7e 0. then AB = 0 = A 0
Hence non-singular matrices are not proper divisiors of zero.
Note. If AB = 0 and B# 0 then A is called a left zero divisor and if
AR = 0 and A 0, then B is called a right zero divisor.
§ 7.02. Characteristic equation and roots of a matrix.
(Agra 94; Rohilkhand 92)
Let A = [a11] be an n x n matrix.
(I) Characteristic Matrix of A The matrix A - XI is called the
characteristic matrix of A, where I is the identity matrix.
(ii) Characteristic polynomial of A :- The determinant JA - X11 is called
the characteristic polynomial of A
(iii) Characteristic equation of A :- The equation JA - 0 is known as
the characterisctic equation of A and its roots are called the characteristic roots
or latent roots or cigenvalues or characteristic values or latent values or
proper values of A: (Avadh 99)
(iv) Spectrum of A :- The set of all eigen values of the matrix A is called
the spectrum of A.
(v) Eigen value problems :- The problem of finding the eigcn valucs of a
matrix is known as an cigen-value problem. 182/11/10
 Characteristic Equation of a Matrix 161
182111/I i
Solved Examples on * 7.02.
*Ex. 1. Find the characteristic roots of the matrix
A=[ cosO - sine
[-sinO - cosO

Sol. Here A= cosO — sinoland i[i 0

[—sinG _ Cos 6] [0 1

A-A1 cos6-A -sinG-0

I-sinO-O -cos0-A
(sin
= (cos 0 - A) (- cos e - A) - 2 6)

(Cos 2 0_A 2 sin 2 O=X 2_ 1.

The characteristic equalton of the matrix A is A 2 - 1= 0 and its roots
i.e.
Ans.
charactcrisctiC roots are ± I.
Ex. 2. (a) Find the characteristic roots of the matrix
123
0 2 3 (Kanpur 96)
002

Sot. Here A = I 2 31 and I = 1 0 0

023 010
0 0 20 0 1
A-All = i - A 2-0 3-0 = i-A 2 3
0-0 2-A 3-0 0 2-A 3
0-0 0-0 2-A 0 0 2-A

= (1 - A) 2- A 3 ,expanding with respect to C1

0 2-A

=(1 -A)(2-A)2.
The characteristic equation of the matrix A is

(1 - A) (2- A) 2 =0 .... See § 7.02 (iii) Page 160 Ch. VII and its roots
Ans.
i.e. required characteristic roots are I, 2. 2.
Ex. 2 (b). Find the characteristic roots of the matrix
A= 1 -3 3
3 -5 3
6 -6 4

Hint : Do as Ex. 2 (a) above. Ans. - 2,4, 2

Ex. 3. Find the characteristic roots of the matrix
Ar2 2 1
131
122


162 Matrices
SoL Here A = 2 2 I and I = I 0 0
131 010
122 001

IA-AI!= 2-A 2-01-0 = 2-A 2

1-0 3-A 1-0 1 3-A
1-0 2-02-A 1 2 2-A

= - 0 1,
1 applying C 1 - 2C3 and
-1 1-A 1 C2-2C3
-3+24 - 2+2A 2 -XI
= 0 0 1 applying C 1 - AC3
A-i I-A 1
-3+4A-A 2 -2+2A 2-A

= A-i i-X(A1)2 I -i
1) ( 3 3-A 2
=(A_1)2(2+3_A)=(X_1)2(5_A).
The characteristic equation of the matrix A is (A - 1)2 - A) 0
and its roots (or characteristic roots of A) are 1, 5. Ans.
Ex. 4. Obtain the characteristic roots of the matrix
A= b c a
cab
a b c
SoL Here
IA-AIf= b-A c-U a-U = b - A c a
c-0 a - A h-0 c a-A b
a-0 b - U c-A a b c - A

= b-A+c±a ca ,applying C1+C2+c3

-c+a-X+b a-A b
a -s-b+c - A b c-A

=(a+h+c-X) 1 c a
I a - A b
I h c-A

= (a + b + c - A) I C a ,applying
0 a-A-c b - a R2 - R I , R3-R1
0 b - c c -- A- a


Characteristic Equation of a Matrix 16

= (a * b + c - A) [(a c) (c - a) - A ((c -a) + (a - c))
+ A 2 - ( b - a) (b - c)]

= (a + /, i c - A) (A + 2a<' - a 2 -e 2 - b 2 + he + rib - ac]

The c)I:Lr,cIcr1tic NILIM1011 of A is Alp =

i.e. a
(+1+c-A)(A-a-b2 -c 4 411) +bccu)0
and the characteristic roots of A are
) 2
a - I' + c. ± (a b + C, be ca) Ans.

S. Find the cigenvalues (or latent roots) of

As 8 6 2
1-6
[ 2-4 3
(Avadh 99; Kumaun 91, Lacknow 90)

Sol. Here JA - A I = 8 - k - 6 2
-6 7-A --4
2 -4 3-A

=(8 - A) ((7 -A) (3-A) - 16) 6 (- 6(3 - ),) + 8) +2 (24 -2(7 -A))
= (8-A)(5- 10x 6(6A- 10) f2(l0-2A)
2
=-45X+ 18X X3
The Characteristic equation of A is - 45 A + IS A2 - A 3 0
which gives
A (A 2 - 18A + 45) = 0 or A (A - 3) (A - 15)= 0 or A = 0, 3, 15.
The required latent roots or eigeiivalues of A are 0. 3 and 15.
Ex. 6. If a t, a ,. a 3 ...... a are the characteristic roots of the n-square

matrix A and .i. is a scalar, then show that the characteristic roots of A -
are it - - p.

Sol. Since a 1 , (12......a, are the characteristic roots of the matrix A, so form
§ 7.02 Page 162 we have
A-XI)=(A--a1)(A-a2) ... (A-an).
Now the characteristic function of A - It!

=( (Ii +A)-a1) ((ft+A)-a} ((t+A)-a 3 ) ... ((t+A)-a}, from )

= (A - (a - )) (A - ( a - t )) ... ((A -- (a, i) 1'

rearranging the terms in each bracket.

The characteristic roots of A - tI are (a 1 - I'), (a 2 - M). etc.


164 Matrices 999

Exercises on § 7.02

0b
Lx. 1. Fina the characteristic roots of A = a 1

O
a 2 a1

0
0
C31
Ans. a l , h i, 1
Lx. 2. Find the eigenvalues of the matrix 0 I 2
I O -1
2 - 1 01 Ans. 2, - 1 ± I3.
• Lx. 3. Find the eigen values of the matrix - 2 2 -3
2 I -6
-1 -2 0 Ans. 5,-3,-3.
Ex. 4. Find the eigenvalues of the matrix 1 - I 2
0 10
1 2 I
Ans.I,-1±2
• Lx. 5. Find the characteristic roots of the matrix
130
3 -2-1
0-I 1 Ans. I,3,-4
Ex. 6. Determine the characteristic equation and roots of the matrix
1 -14
0 37
0 05 Ans. (X-1)(A-3)(X--5)=0;1,3,5.
Lx. 7. Find the eigenvalues of the matrix
-3 2 2
-6 5 2
Ans.l,2,3.
Lx. 8. Find the eigenvalues of the matrix
314
026
0 0 5
(Agra 92, L.ucknow 92) Ans. 2,3,5
Lx. 9. Find the, latent roots of the matrix
0 sin cc cos ct sin l3
-sinex 0 Cos aCos
-cos a sin 1i cos a cos, (3
Ans. O,i,-i.
Lx 10. Show that the matrices
o cb l. 0 a C- 0 b a
c 0- a ! a 0 b b 0 c
b a Oj c b 0 a c 0
have the sane characteristic equation. (Kumaun 90)
 Cayely Hamilton Theorem 165

§ 7.03. tCayely Hamilton Theorem (Agra 95, 93, 91,Avadh 99;

Gathwal 96; Kanpur 97. ; Kumaun 92; Lcknow 92;
Meerut 98, 92, 91; Rohilkhand 94, 93, 92, 90)
Statement : Every square matrix satisfies its characteristic equation or if
A - XI = ( 1)" [X" + 0 1 x'1 + a 2 X" -2 -i- ... + a 1 ] be the characteristic
polynomial of n x n matrix A = [ a 11 ], then the matrix equation
X,,+a1X''+...+anI=OisSat1SfleYX".4

i.e. A"+aLA"'+ ... +aI=O.

Proof. the elements (A - XI) are at the most of first degree in A.

The elements of Adj (A - XI) are at the most of degree (it - 1) in A and
the coefficients of various powers of A being polymomials in the a.
Adj (A - XI) can be written as
B = B 0 Xn -2 + B 1 A" - + - + B -
.
where B 0, B 1 .... B - i are it x it matrices, their elements being polynomials
in a11.
Also from § 5.09 Page 49 Ch. V we know that if A = [a, 1J be an n X n
matrix, then A.(Adj A)=(Adj A). where I is an n X n identity

matrix.
Therefore (A— XI) •Adj(A XI) 1AX111

or (A - XI) • B = A - Xfl • I, B = Adj (A . — XI)

or (A_ XI) (B O X"'+B 1 X"+ ... +B1_i)

n n + ... a1]I.
(—I) [A +a 1 X + n - a2X
1
Comparing coefficients of like powers of A on .both sides, we get
—1B0 =(— 1)"!;
- AB0—IB1(-1)°a1I;

Ali 1 —1B 2 =(— 1)"a21;

AB - (- 1)" a
Now pre-multiplying these equations by A", An A, I respectively and
adding the results so obtained we get
2
A" (- 1B 0 )+ A" (AB 0 — 1B 1 ) + A" (AB 1 - 1B2 ) + ... + I (AB,_ )
=(_1)f( [An +at IAfh +a 2 [A n 2+0n111

tThis theorem was first established by Hamilton in 1883 for a particular type of
matricei and was later on stated by Cayley in 1885.
166 Matrices
or O=(l)n (An C1An+Q2A2+.. -4u,,1].
where 0 is the null matrix.
Hence A° + a 1 A n- + a 2 A' -2 + ... + a I = 0 Hence the theorem.
Cor. I. Multiplying the result of § 7.03 above by A"' -" , where
m n and m is a positive integer, we get
Am
+a, Am I + a2 A m 2 + ... + Q,,Amn=O

i.e. any positive integral power A' of A can he linearly expressed in terms of
I, A_, Ar'.
Cor. H. In § 7.03 above we have proved that
A+a1A'+a2A2++aJ=() (i)
or - a I = A [A - l + a 1 A -- 2 + (12A" + ... + a - 1 fl (Note)
or --aAI=A'+a1A2+...+a1I ,..(ii)
or (1)AB_tA'=_A'-aiA"2+..._a1I
AB, - = (- 1)" an I ( 7.03 above)
or ... _a
Bn_1=([_A'.aiA2_ ni flAdjA ...(iii)
Cor. III. From result (ii) of cor. H above we have
-I
A =--[A 1 [ An-' +a 1 A n-2 +...+a1I]

which show that A- i can be expressed linearly in terms of A , A" 2 I.

* 7.04. Characteristic vectors (or Eigenvcctoi's). (Agra 94)
Let us consider the linear transformation
KzAX
which transforms a column vector X by means of a square matrix A into
another column vector K.
If X be a vector which transforms to its multiple p. X by the above
transformation (i), then we have p. X = AX ...(ii)
or AX - p.LX = 0 or (A -- p.1) X 0. (iii)
This equation (iii) when written in full gives n homogeneous-equations in
x1,x2.....x,, which are n unknowns. The .' equations will have a non-zero
solution only if I A - p. I j = 0 i.e. the coefficients matrix is singular. (Note)
This equation is called the characteristic equation of transformation and is
the same as the characteristic equation of the matrix A. (See § 7.02 Page 160 Ch,
VII). This equation has n roots and corresponding to each root, the equation (iii)
has a non-zero solution


lheoiems on characteristic roots 167

x xi
x-'

Xn
ij
which is defined as characteristic vector or Elgenvector or Intent vector or
invariant vector.
7.05. Theorems on latent roots (or characteristics roots).
Theorem 1. If square matrix A of order n has latent roots A 1 , A2..... A then
A' has also the same latent roots. (Kw,un 95)
Proof. IAMA=- a, a 1 3 ..........
a 21 a 22 a ..........

t nI 1n2 a, 3 ..........

The characteristic equation of A is

011 —X a12 ... 171

=0
"21 a22 - A . . - a
. (1)

a, 2 ...

Also A' all '21 (43 j ..........

a 1 -, a 22 a 32 ..........

2n (13 ..........

The characteristic equation of A' is

A' A1 'I x I a,1 =

(1 12 a,) —A a2

o a ........ a,,,,—A

Also we know that the value of a determinant remains unaltered if rows are
changed i nto columns and thus we find that the determinants given by (i) and (ii)
are the same, the diagonal elements being the same.
Hence 1mm (i) and (ii) we conclude that the characteristic equations of A
and A' are the same. Consequently the latent roots of A and A' are the same.
Theorem 11. If A is an n x r triangular matrii, then the elements of the
principal dtaooah are the c/ia racterislic roots of A.


168 Matrices
a 11 1 12 a 13 . a1,
Proof. Let A'=
o a 22 a23 ..........
o 0 a33 ..........

o 0 0
(Here we have taken upper triangular matrix).
The characteristic equation of A is
a ll a12 a3 ... 11In _
IA — AII=
0 a 2 —A 023 ... 02n
o 0 a33 —X ...

o 0
a 22 —)- a23 ... a 2, 0. expanding
or (all
o .033 - A . . with respect to C1

o 0 ...

... a3, = 0, expanding

or (a ll - A) (a22 - A) 033 - A . a with respect -
U ... a4n
to C2
0. 0 ... a,,,,—A
or (a1 - A) (a22 - A) (a 3 - A) ... (a,.,, - A) = 0, proceeding in this way
or X=a11,a22,a33.....
i.e. the element roots (or characteristic roots) of A are the elements of the prin-
cipal diagonal of A. . Hence proved.
Theorem Ill. The characteristic roots of a hermizian matrix are all real.
• (Agra 95, Kbnpur 95, 94)
Proof. Let A be the hermitian matrix. Then from § 7.04 (ii) Page 166 Ch.
VIII we know that AX = XX,
where A is a characteristic root of A and X the corresponding characteristic
vector.
From (i) we get X8AX = xe AX, premultiplying both sides by xe
or . X8 AX=AX0 X, ...(ii)
where Xe is the transponsed conjugate of X (See chapter 11)
Also if A is a hermitian matrix, then by definition we have
= A (See Chapter II).
Now taking transposed conjugate of both sides of (ii) we get


Theorem on Characteristic Roots 169

= X X X. using (iii) also

From (ii) and (iv) we get AXe X = Axe x
or (A-J)X6X=O or X—X=O, :XeX*O
or A = A or A is real. Hence proved.
**Theorem IV. The characteristic roots of a real symmetric matrix are all
real.
Proof. Do as Theorem Ill above. Here all the elements of A are real and as
such it is particular case of Theorem Ill above.
**Theorem V. The characteristic roots of a skew-herrnitian matrix are
either purely imaginary or zero.
Proof. Let A be a skew-hermitian matrix, then (see Chapter II) we know
that iA is hermitian.
If I be a charcteristic root of A, then IA - XI] = 0
or i J A - A I] = 0 or] iA - (iX) I] = 0, where iA is hermitian.
or (iA) is real, since the characteristic roots of a hemiitian matrix are all real.
(See Theorem Ill Page 168 Ch. VII)
or I is either purely imaginary or zero.
i.e. the characteristic roots of a skew her-mitian matrix A are either purely
imaginary or zero. Hence proved.
Theorem VI. The characteristic roots of real skew-symmetric matrix are
purely imaginary or zero.
Proof. Do as Theorem V above. Here all the elements of A are real and as
such it is a particular case of Theorem V above.
VII. The characteristic roots of a unitary matrix are of unit
modulus.
Proof. Let A be a unitary matrix (Sec Chapter 11). Let I- be a characteristic
root of A and X the corresponding characteristic vector.
Then AX AX ...(i) (See § 7.04 (ii) Page 166 Ch. VII)
Taking transposed conjugate of both sides of (i), we get
(AX)0 = (AX) () or X0 A0 = A X°
From (I) and (ii)we get X 0 A° AX = A X°AX (Note)
or X°(A°A)X=XXX°X
or X0 (1) X = X X° X, . A is unitary (See Chapter II)
or X°X(l—IA)=O
or l—XX=0.
o
or r
AX=l jA]2 =XX=I.
i.e the characteritic roots of A are of unit modulus. Hence proved.

—25
170 Matrices
"Theorem VIII. The characterise roots of an orthogonal matrix are of unit
modulus.
Proof. Do as Theorem VII above remembering that if all the elements of the
unitary matrix A are real then A is an orthogonal matrix.
** 7.06. An Important Theorem.
The scalar A is a characteristic root of the matrix A if only if the nuliriA
A - XI is singular
Proof. Let A = [a,J , X = 1x 1 , x2 ..... .v,J.
then AX = ...See § 7.04 Page 166
reduces to a11 012 ... a 1 , x1 =
02 1 a22 2n X2

an a,12 . . a,,, x, À

=A 1 0 0 ... 0 xi
0 1 0 ... 0 x2

0 0 ... 1

= A 0 ... 0 Xi
0 A ... 0 x2

00...

or all°l2 =0
- i
{ a21 (122 02n }
00... A
an a2

or a11—A a 12 ...
"21 a22-X ... a2, x2

a., ar12 1.1 a

or (A— II) X=0

i.e. AX=XX(A—AI)X=0,
which is a homogeneous system of linear equations whose coefficient matrix is
A — A!.
Now as we require a vector X * 0, so we must have
IA-X11=0


Solved Examples on Cayley's Theorem 171

i.e. the matrix A - Xl must be singular.
Solved Examples on § 7.03 to § 7.06
Ex. I. (a). Verify Cayley Hamilton's Theorem for the matrix
A= 101
010
0 0 1
(Agra 93)
Sot. Here 1= 1 0 and A 01 1) rfi
o
1 0 10 I 0
0 0 I [o 0 I

A -XiI= 1-X 0
0 i - X 0
1 1
0 0 1-A

rr(I -A) I -X 0 ,expandingw.r. toC1

0 1-A

Characteristic equation of A is (1 - A) 3 = 0
or X33X2+3A10
Now A2=1 0 IxI 01=102
0 1 0 0 1 0 0 •I 0
0 0 I 0 0 1 0 0 1

A 3 =A 2 .A= 1 0 2 x 1 0 1 = 1 0 3
0 1 0 0 1 0 0 1 0
0 0 I 0 0 1 0 0 1

A3-3A2+3A-1
= ri 03 -3 1 0 2 +3 1 0 1 - 1 0 0
010 010 010 010
00] 001 001 001

103-306+3 03-I 00
0 1 0 0 3 0 0 3 0 0 1 0
001 003 003 001

1-343-I 0-0+0-0 3-6+3-0= 0 0 0 =0,

0-0+0-() 1-3+3-1 0-0+0-0 0 0 0
0-0+0-0 0-0+0-0 1-3+3-1 0 0 0
where 0 is the null matrix.
Hence the given matrix A satisfies its characteristics equation given by (i)
Hence proved.


172 Matrices
Ex. 1 (b) Show that the matrix A = 2 2 1 satisfies ('ayley
-1 03
2-11
Hamilton Theorem. (Meerut 92 1')
Sol. Here I = 1 0 0 and A = I 21
C) 1 0 -1 0 3
O C) 1 2 -I I
JA - AlI= 1-A 2 I
-1 -X 3 (Note)
2 -1 1-A
= I - A 2 I adding 2R 2 to R
-I -A 3
o -l-2A7-A
=(1-X) -X 3 + 2 1
-1-2X7-A -1-2)7-X
=(1 -A)(- 7 X+A2 +3+6XJ+114-2X+] +2X1
=(1 A)(A 2 A+3)+ 15=-X 3 +2A 2 -4A+ IS
Characteristic equation of A is
A3-2X2+4A-18=0
Now A2= 1 21 I 21=1 18
-1 0 3 -1 0 3 5 -5 2
2 -I 1 2 -I I 5 3 0
Now A'118 1 21 16-6 12
5-52 --1 03 14 8 -8
5 30 2-11 2 10 14

A 3 - 2A 2 + 4A - 181
= [16 -6 12 -2 1 I 8 +4 I 2 I
14 8 -8 5 -5 2 -1 0 3
210 14 5 30 2-11
-18 I 0 ()
0 1 0
0 0
[16 -6 12 + -2 -2 -16 + 4'8 4 + - 18 0 0
14 8-8 -10 10 -4 -4 012 0-18 0
2 10 14 -10 -6 0 8 -4 4 0 0 -18
=
= 16-2+4-18 -6-2+8+0 12-16+4+0
14-10-4+0 8+10+0-18 -8-4+12+0
2-10+8+0 10-6-4+0 14+0+4-18
 Solved Examples on Cayleys Theorem 173

= 0 0 0 = 0 where 0 is the null matrix

000
000
Hence the given matrix A satisfies its characteristic equaiton given by (i).
Ex. 2 (a). Use Cayley Hamilton Theorem to find the inverse of the
matrix 12 3
1 3 5 (Ro/iilkhand 95)
1 5 12

Sol. here 1= 1 0 0 and A = I 23

010 135
0 0 I 1 5 12

A_XI = iA 2 3
1 3—A 5
1 5 12—A
=i - A 2 3 applying R 3 -
I 3—A 5
0 2+X 7—A -

= 0 X)f(3_A)(7_A)_5(2+X)I—[2(7—))-3(2+.
expanding w.r. to C1

=(1—X)121 —IOA+A 2 — 10-5X1—[14-2A-6-3A]

=(l —A)(1l - ISX+ A2 )— (8-5 A)
= 3-21 A + ioA2 - A3 , on simplifying.
The characteristic equation of A is
A3-1oA2+21X-3=0.
Now as A muSt satisfy is characteristic equation (I), so we have
3
A —]6A+21 A-31=0
or 31=A3-16A2+21A
Multiplying both sides by A we get

3A '=A 2 — 16A+211, :AA-

or A '= (1/3) A 2 - (16/3) A + 71. (ii)

Now A2=A.Ar 1 23 1 2 3
13 5x1 3 5
I 5 12 1 5 12
• = 1+243 2+6+ 15 3+10+36 = 6 23 49
1+3+5 2+9+25 3+V+60 9 36 78
1+5+12 2+15+6() 3+25+144 1877 172

Froii (ii), we get



174 Matrices
A - 1 49 6 23
16 1 23 + 7 1 0 0
3 936
7831 3 5 010
172 18 77
I 5 12 001
- (16/3) + 7 (23/6) - (32/3) + 0 (49/3) - (48/3) • 01
!3-(16/3)+O 12-16+7 26--(80/3)+O
L 6 - 16 / 3 +0 (77/3)-(80/3) +0 (172/3) - (192/3)+ 7

= 11/3-3 1/3 =(1/3) II -9 1

-7/3 3 -2/3 -7 9 -2 Ans.
2/3-1 1/3 2-3 1
Ex. 2 (b). Verify Cayley Hamilton's Theorem for the matrix
A 0 0 ii. Hence compute A -
310
- 2 1 4j .
(Agra 94; Kanpur 96: Meerut 96)
Sol. Here
I = [01 0 0 and A = 0 0 1
010 310
0 1 -2 1 4
A-AI= 0-A
J 0 1
3 1-A 0
-2 1 4-A
=-A{(l-X)(4-X)}+1 (3+2(1-A)}

=5-6),+5A2-X3
The characteristic equation of A is
X3-5X2+6A5---0
Also we have
• A2= 00 100I=-21 4
• 3 I 0 3 1 0 3 1 3 ...(ii)
-2 1 4 -2 1 4 -5 5 14
and A3=A2.A= -2 14 0 0 1 = -5 5 14
3 1 3 3 1 0 -3 4 15
-5 5 14 -2 1 4 -13 19 51
• A3-5A2+6A-5I
= 1 -5 5 14 -5 -2 1 4 +6 0 0 1 -5 1 0 0
1-3 415 31 3 310 010
L- 13 19 51 -5 5 14 -2 1 4 001
-5 5 141- -10 5 20+0 0 6-5 0 0
-3 4 15 15 5 15 18 6 0 0 5 0
-13 19 ij -25 25 70 -12 6 24 005


Solved Examples on Characteristic Equations 175

= -5+10+0-5 5-5+0-0 14 --20+6-0

-3-l5+1-0 4-5+6-5 15-15+0-0
-13+25-12-0 19-25+6-0 51-70+24-5
= 0 0 0 = 0, where 0 is the null matrix.
000
000
Hence the matrix A satisfies its characterstic equation given by (I). Hence
Caylcy-Flamilton'S Theorem is satisfied by the matrix A.

Again A3_5A2+6A_51=0

or 51=A3-5A2+6A
Multiplying both sides by A , we get
5A=A 2 -5A+6I, AA' =1
-2 1 41-5 0 0 1 +61 0 Q, from (ii)
3 1 3 3 10 01 0'
-5 5 i4j -2 1 4 0 0 1

=-2 14+ 0 0 -5+600

3 1 3 -15 -5 0 0 6 0
-5 5 14 10 -5 -20 0 0 6

= -2+0+6 1+0+0 4-5+0= 4 1 -1

3-15+0 1-5+6 3+0+0 -12 2 3
-5+10+0 5-5+0 14-20+6 5 0 0

or A1_ 4 1 -1.
5 -12 2 3
500 Ans.

Ex. 3 (a). Verify Cayley-Hamilton Theorem for the matrix

A = 1 20 and hence find A
2-1 0
0 0-1 (Kanpur 97)
Sol. Here we have
(A-XI1 = 12 0 1 -X I 0 0 = 1-A 2 0
2 -1 0 0 1 0 2 -i-A 0
0 0 -1 0 0 1 0 0 -I-A

A-X1l= I -X 2 0
2 -i-A 0
0 0 -1-A
(- I -A)[(1 - A)(- I -A)-4}, expanding wr. to C3


176
Matrices I 7jjjj J]
+A 2 -41=(I +A)(sA2)
= 5 + S A - A2 -
The characteristic equation of A is A 3 + A 2 —5 A - 5
00)
Also we have

A 2 =AA= 1 2Ox 1 2 0 =[0, 0 0

2 —1 0 2 —1 0 0 5 0
O 0 —I 0 0 —1 0 I

A 3 =A 2 A 5 0 0 x 1 2 0 = 5 10 0
O 5 0 2 —1 0 10 —5 0
001 0 0-1 0 0-1
A3+A2—SA-51
5 10 0 + 5 0 0 1-5 1 2 0}-5[01 0 0
10 —5 0 0 5 01 2 —I oj 0 I 0
O 0-1 OOij 0 0-1] o.i
=[ 5+5-5-5 10 +0-10-0 0+0-0-0
I10+0-10-0 — 5+5+5-5 O+O—O—O
[ 0+0-0-0 0+0-0-0 —1+1+5-5
= [0 0 01 = 0, where 0 is the null matrix.
Jo o 0
[0 o o

Hence the matrix A satisfies its characteristic equation given by (i). Hence
Cayley Hamilton theorem is verified by the matrix A.
Again A 3 + A 2 - SA —51 = 0 gives

SI A 3 + A 2 - 5A
& 5 A = A2 + A —51, multiplying both sides by A 'and AA = i
= [s 0 0 1 4- [i 2 0 —5[i 0 0 =[i 2 0
10 5 oJ J2 —i o Jo 1 0 2 —1 0
L o 0 ij [o o —i [0 0 1 [o 0 -s
2 o].
5[0
22 —i of
o _5j
Ans,
Ex. 3 W. Show that the matrix A= 1 2 Ij satisfies Cayley-
0 1-1
3-1 1
Hamilton Theorem and hence complile A


92111h12 Solved Examples on Characteristic Equation 177

Hint Do as Ex. 3(a) above. Ans. (1/9) 03 3
3 2-.1
3 -7-1
Ex. 4 (a). Determine the characteristic equation of the matrix
1 3 71
and verify that A satisfies its characteristic equation.
423
1 2 1 (Garhwai9ó)
Sol. Here we have A - Al j = 1 - A 3 7
4 2-A 3
1 2 I-A

3 -3 4 3 +74 2-A
- 1 x 2-A
2 I-A 1 1-A 1 2
expanding w.r. to R1
= (1 - A) [(2- A) (1 - A) -6] -3[4- 4A -3] + 7 [8-2 + A]
= ( l X)[ 3X+X2 413 [1 -4A]+7 (6+X}
),
= 35 + 20 A + 4A2 - 3, on simplifying.
Characteristic equation of matrix A is
A3 - 4A2 - 20A - 35 = 0.
Also we have
A 2 =1 3 7x1 37=202323
4 2 3 4 2 3 15 22 37
1 2 1 1 2 1 10 9 14

A 3 =A2 .A= 20 23 231 x 1 3 7

15 22 37 4 2 3
10 9 14 1 2 1

= 20+92+23 60+46+46 140+69+23

15+88+37 45+44+74 105+66+37
10+36+14 30+l8+28 70+27+14
135 152 232
140 163 208
60 76 III

A3-4A2-20A-351
= 135 152 232 -4 20 23 23 -20 1 3 7 -35 1 0 0
140 163 208 15 22 37 4 2 3 C) 1 0
60 76 111 10 9 14 1 2 1 0 0 1
178 Matrices
o 0 = 0,
0( where is the null matrix
000
000
Hence the matrix A satisfies its characteristic equation (i). Hence proved
EL 4 W. Find the characteristic equaiton of the matru
A = 1 13 and show that A satisfies this equation.
526
--2 —1 —3
Hint :Doas Ex. 4(a)above.
Ans. X'
Ex. 4 (c). Verify that matrix A = 2 2 1 satisfies its characteristic
131
122
equalton. (Rohilkhand 99)
Hint: Do as Ex. 4 (a) above.
EL 4 (d). Determine the characteristic equation of the matrix
A = 0 1 2 and verify that A satisfies Its characteristic equation;
i 0 —1
[2 - 1 0 (Garhwal 94)

HIa(:Doas Ex. 4(a) above.

Ans. A3-6A+4=O
'Ex. S. Find the characteristic vectors of the matrix A given in Ex. 4 (c)
above.
SoL As in Ex. 4 (a) above we can find that the characteristic equation of A
is (A - 1) 2 (5—A) = 0 and so the characteristic roots of A are 1, 1, 5.
The equation (A - XI) X = 0 of the matrix A is
2—A 2 1 Xi=Ø]
I 3—A I 12 01
1 2 2 - X3 °I..See § 7.04 Page 166 Chapter VII
Putting A = 1 in the above equation, we get
• 121 Xf
1 2 1 x2 to
1 2 1 x 3 [0
The corresponding characteristic vector is given by the equation
xI + 2x2+ = 0.
The characteristic vector corresponding to A = I may be taken as
—1,1). Ans.
Putting A = 5 in (i), we get


Solved Examples on Characteristic Equation 179

2-5 2 2 ,1=0
Oj
1 3-5 1 x 2 0
1 2 2-5 X3

or -3 2 1 X10
1 -2 1 X2 0
1 2-3 0
The corresponding characteristic vector is given by the equations
-3x 1 +2x2 +x1 =O,x 1 -2x2 +x3 =O and x 1 +2x2-3x3=0.
x 2 x3
Solving these we get
-j-- = x
-j- -j-.
The characteristic vector corresponding to A = 5 may be taken as (1, 1, 1).
Ex. 6 (a). Verify Cayley Hamilton's Theorem for the matrix
A = 2 - 1 1 and hence find A -
1 2-1
1-1 2
(Agra 96: Garhwoi 93; Kanpur 95, 93, Kumaun 95; Lucknow 91;
Meerut 98, 97; Rohilkhand 97) -
or find the characteristic equation of the matrix A = 2 -1 1 and
-1 2-1
1-1 2
verify that it Is satisfied by A. (Kanpur 91)
Sol. Uere IA-XII= 2-A -1
-I 2-A -1
1 -1 2-A

= 2-A 0 I ,replacing C2 by C2+C3

1 I-A 2-A
=(l-A) 2-A 0 1
1 -1
1 2-A
=(1-A)f(2-X) 1 -1 + -1
1 2-A I
= ( 1 -A) ((2-A) (3-A)+(- 1- 1)}
=(1 -X) (4- 5X + A2 ) = 4 _ SA ; A2 - 4A + 5A2 - A3
+6),2
= - A3 - 9X + 4.
The characteristic equation of the matrix A is
A3 - 6A2 + 9A - 4 = 0


ISO - Matrices 999

Now A 2 = 2 —! 1 2 —1 1 = 6 —5 5
—1 2 - —1 2 —I —5 6 —5
1 —1 2 1 —1 2 5 —5 6
and A3=A2.A= 6 —5 5 2 —1 1
—5 6 —5 —I 2 —1
5 —5 6 1 —1 2
= 22 —21 21
—21 -22 —21
21 —21 22
A3-6A2+9A-41
= 22 —21 21 —6 6 —5 5 +9 2 —1 I
—21 22 —21 —5 6 —5 —1 2 —1
21-21 22 5-5 6 1-1 2
—4100
010
001
=1 22 —21 21 + —36 30 —30 + 18 —9 9
1-21 22 —21 30 —36 30 —9 18 —9
[ 21 —21 22 —30 30 —36 9 —9 18
+-4 0 0
0-4 0
0 0-4
= 1 22 - 36+18 - 4 —21+30-9+0 21-30+9+0
1-21+30-9+0 22-36+18-4 —21 +30-9+0
• [21-30+9+0 — 21+30-9+0 22-36+18-4
= 00 0 =0, where 0 is the null matrix
000
000
Hence A satisfies its characteristic equation given by (i).
Hence Cayley Hamilton theorem is satisfied by the matrix A.
Again A3 —6A 2 + 9A-41= 0 gives 41=A3-6A2+9A
Multiplying both sides by A, we get
4A=A2 -6A+9I :AA=j
=1 6 —5 5 —6 2 —1 1 +9 1 0 0
j-5 6 —5 —1 2 —1 0 1 0
[5-5 6 1-1 2 001


Solved Examples on Characteristic Equation 181

= 6-12+9 -5+6+0 5-6+0 = 3 1 -1
-5+6+0 6-12+9 -5+6+0 1 3 1
5-6+0 -5+6+0 6-12+9 -1 1 3.

3 1 -1
or 4 13 1
-1 1 3 A.
Ex. 6 (b). Find the characteristic equation of the matrix
A= 00 2 and verify that itis satisfied byA and hence obtain
2-3 0
1 1-1
A-
Hint; Do as Ex. 6 (a) above.
Ans. X3+4A2+A-l0=O,

Ex. 6 (c). Verify Cayley Hamilton Theorem for the matrix

A = 0 1 21. Find out the inverse If possible.
23 O
1 1 -2] (Rohilk.hand94)
Hint: Do as Ex. 6 (a) above. -
Ans.!6 4-6
2 4-2 4
-1 1-2

Ex. 7. Find the characteristic roots of the matrix A [1 4] verify

[2 3]
Cayley Hamilton theorem for the matrix.
Sol. Do as Ex. 6 (a) above. AnL 5,- 1
Ex. 8. Find the characteristic root and inverse of the matrix A:= 15 0
[12
Sol. Here IA-?,1I= 5-A 6
I 2-A

=(5 -),)(2 - A) -6=4 - 7A + A2

The characteristic equation of A is A2 - 7A + 4 = 0
Now as A must satisfy Cayley Hamilton's Theorem, so we get
A 2 - 7A + 41 = 0, where 0 is the null matrix
I=--A2+A.

Multiplying both sides by A we get



182 Matrices

__±r5 6]+ . {I 0-) = [_ 14-[l 01

4 [i 2 0 1] 4 2j 4 I
I I iJ
[4 _j [0
=[
0 1=(1/4)F 2 -61
14 4 2 I 2 2 Ans.

Aiso the characteristic roots of A are the roots of (i).

A = [7 ± (49 -16)] (7 ± (33)). Ans.

Ex. 9. Using Cayley Hamilton's Theorem find A 2 where

A= 1 2 0
2 - 1 0 (Agra 95)
0 0-1
Sol. Here I A - AIJ= . I-A 20
2 -1-A 0
0 0 -I-A
= 0-X-1-A 0 -2 2 0
0 - I --A 0-1-A
(1-A)(1 +X) 2 +4(1 +A)=(l ±A)[(I X2)+4]
=(1 +A)(5A2)5A2+5XA3
The characteristic equation of A is
A3+A2-sA-5=o
Now A2= 1 2 0 1 2 0=500
2 -1 0 2 -1 0 0 5 0
o 0 -1 0 0 -I 0 0 I
And A3=A2.A= 5 0 0 1 2 0 = 5 10 0
0 5 0 2 -I 0 10 -5 0
.00l0 0-1 0 0-1
By Cayley Hamilton's Theorem from (i) we have
A 3 + A 2 - 5A - 51 = 0
or 51=A3+A2-5A. ...(ii)
Multiplying both sides by A ,we get
5A=A 2 +A-5I, AA- I= I


Solved Examples on Characteristic Equation 183

= 50 0+1 20-5100
0 5 0 2 —I 0 0 I 0
0 0 1 0 0 --1 0 0 I
= 5+1-5 0-+2-*() 0+0+0 = 1 2 0
0+2+() 5—I-5 0+0+0 2 —1 0
0+0-i0 0+0+0 1-1-5 0 0 —5
nr 2 0
A --
5 1 2 -1 0
[0 0
Again multiplying both sides of (ii) by A 2 we get
5A_2=A+I_5A_L. (Note)
Dr 2=
5A 0 + 1 0 0 - 1 2 0 from (iii)

0 2? - g
10
= I fl — I 2*0-2 0+0+0 = 1 0 0
2+0-2 —1+1+1 0+0-0 0 1 0
0+0+0 0+0+0 —1+1+5 0 0 5
)r 2_ I 0 0
A
5 0 1 0
0 0 5 Axis.
Ex. 10. Verify Cayley Hamilton's Theorem for the matrix A = 1 0 2
021
203
fence of otherwise compute A . (Garhwal 92; Kwnaun 94, 92;
Luc-know 92; Meerut 96 P; RohiLkhand 98)
Sol. Here IA-A1I= 1—A 0 2
0 2—A i
2 0 3—A
=(l -A) {(2—A)(3—A)-0) +2 (0-2(2—A)}
= (1 - A) (2— A) (3— A) —4(2—A)
= (2—A) (1 - A) (3—A) —4) = (2 —A) (3 —4A + A2 4;

= (2 —?.) (A2 -
4X- 1) = - A3 + X2 - 7A —2
The characteristic equation of A is
A3-6A2^-7A+2=o.
Now A2=l 0 2 1 0 2 = 5 08
021021 24 5
2 0 3 2 0 3 8 0 13
184 Matrices
and A3=A2.A= 5 08 1 0 2 = 21 0 34
2 4 5 0 2 1 I? 8 23
8 0 13 2 0 3 34 0 55
A3-6A2+7A+21
[21 0 34 -6 5 08 +7 I 0 2 +2 1 0 0
12 8 23 2 4 5 0 2 1 0 I 0
34 0 55 8 0 13 2 0 3 0 0 1
=
=21034+ -30 0-48+7 014+200
12 8 23 -12 -24 -30 0 14 7 0 2 0
34 0 55 -48 0 -78 14 0 21 0 0 2
= 21-30+7+2 0+0+0+0 34-48+14+0
12-12+0+0 8-24+14+2 23-30+7+0
34-48+14+0 0+0+0+0 55-78+21+2
0
= 0 0 0 = 0, where 0 is the null matrix.
000
00 0
Hence A satisfies its characteristic equation given by (i)
Hence Cayley Hamilton's Theorem is satisfied by A i.e.
A 3 -6A2 -i-7A+21=O or 21=-A3+6A2--7A,
Multiplying both sides by A we get
2A=-A 2 +6A-7I, AA'=I, IA - =A
508+6102-7100
245 021 010
8 0 13 2 0 3 0 0 1
= -5 0 -8 + 6 0 12 + -7 0 0
-2 -4 -5 0 12 6 0 -7 0
-8 0 -13 12 0 18 0 0 -7
= -5+6-7 0+0+0 . -8+12-0 = -6 0_
-4+12-7 -5+6-0 -2 1 1
-8+12+0 0+0+0 -13+18-7 4 0 -2
or A'=(1/2) -6 0.4 = -3 0 2
-2 1 1 -1 1/2 1/2
4 0 -2 2 0 -1 Am
Ex. 11(a). Find the characteristic equation of the matrix A 1 3 7
423
1 2 ij


Solved Examples on Characteristic Equation 185

1.
and hence find A Also verify Cayley Hamilton's Theorem for A.
(Kanpur 90)
Sol. Here A - Al!
= I - X 3 7= - A 1 6 + A applying
4 2-A 3 O-6-X-1+4A R,-R3,
1 2 I - A 1 2 1-). R-4R3
=-X -A + 1 6+).
2 I-A -6-X-1+4A
=-A[-(6+A)(l -A)-2(4X- l)]+ [(4).- 1)+(6+A)j
=-A(-f+6X--X+A2-8A+2)+2+ 16X+35)
A3+42+20X+35.
The characteristic equation of the matrix A is
A34X2_20A_35=0
By Cayley Hamilton's Theorem, we have
A 3 - 4A 2 - 20A - 351 = 0, where 0 is the null matrix.
or 3
, 51 A 3 - 4A2 - 20A.
Multiplying both sides by A ', we get
35A '= A 2 - 4A - 201, AA I (ii)
Now A 2 = 1 3 7f1 37
4 2 3 j4 2 3
1 2 1 [i 2 1
= 1+12+7 3+6+14 7+9+7 = 20 23 23
4+8+3 12+4+6 28+6+3 15 22 37
1+8+1 3+4+2 7+6+1 10 9 14

From (ii) we get

35A -I = 20 23 23 -4 1 3 7 -20 1 0 0
15 22 37 4 2 3 0 1 0
10 9 14 1 2 1 0 0 1
= 20-4-20 23-12-0 23-28-0 = -4 11 -si
15-16-0 22-8-20 37-12-0 -1 -6 251 Ans.
10-4-0 9-8-0 14-4-20 6 1 -10 1
Verify Cayley Hamilton's Theorem for yourself.
Ex. 11 (b). Find the characteristic equation of the matrix
A= 1 1 -2 and hence flnd A1
-2-1 2
341 (Meerut 91 S)
Sol. Do as Ex. 11(a) above.

-26


186 Matrices
Ex. 11(c). Using the characteristic equation of the matrix A = 1 0 1
010
001
find A
Sot. Do as Ex. 11(a) above. Ans. f 1 0 —1
01 0
00 1
Ex. 12. If A 8 -6 2 find the characteristic roots of A.
—6 7-8
2-4 3

Verify Cayley Hamilton's Theorem and hence find A'.

Sol. Here IA-XII= 8—A —6 21
—6 7—A —8
2 —4 3—A
=(8—A) {(7—A)(3—X)-32 J+6(-6(3—X)+16)
+2{24-2(7—A)}
=(8—A)(— 11— IOX+X 2 } +6 (6)L—2J +2 {10+2XJ
=_88_69A+18A2_A3+36A_12+20+4A=_Al+18A2_29A_80
The characteristic equation of the matrix A is
A3 — 18A2+29X+80=0.
Its roots are the required characteristic roots of A (students can calculate it
I they have read solution of cubic equations, so left as an exercise for the
,tudents).
Now A 2 = 8 —6 2 8 —6 2 = 104 —98 70
—6 7 —8 —6 7 —8 —106 117 —92
2 —4 3 2 —4 3 46 —52 45
And A3 =A 2 .A= 104 —9870 8 —6 2
—106 117 —92 —6 7 —8
46 —52 45 2 —4 3
= 1560159012021,
- on evaluating
- 1734 1823 - 1424
770 -820 643
A3-18A2±29A+801
= 1560 - 1590 1202 - 18 104 98 70
—1734 1823 —1424 —106 117 —92
770 —820 643 46 —52 45
+29 8-6 2+80100
—6 7 -8 0 1 0
2 —4 3 0 0 1


Solved Examples on Characteristic Equation 187

= 1560 - 1590 1202 + - 1872 1764 -1260

-1734 1823 14241 1908 -2160 1656
760 -820 643 -828 936 -810

+ 232 -174 58+80 0 0

-174 203 -232 0 80 0
58 116 87 0.080

= 0 0 0 = 0, where 0 is the null matrix.

000
000

Hence A satisfies the characteristic equation given by (I).

Hence Cayley Hamilton's Theorem is satisfied by the given matrix A i.e.
from (i) we get A 3 - 18A 2 + 29A + 801 = 0
or 801 = - A 3 + 18A 2 - 29A
or 80 A = - A 2 + 18A - 291, multiplying both sides by A -
104 -9870 ^18[ 8 -6 2 -29 1 0 0
-106 117 -92 1-6 7 -8 0 1 0
46 -52 45 , 2 --4 3 0 0 1

= -104+144-29 98-108+0 -70+36+0

106-108+0 -117+126-29 92-144+0
-46+36+0 52-72+0 -45+54-29

or 80A'= 11 -10 -34

-2 -20 -52
-10 -20 -20

or A-
1 = (1/80) 11 -10 -34
-2 -20 -52
-10 -20 -20 Ans.
Ex. 13 (a) Verify that A = 12 0 satisfies its own characteristic
2-1 0
0 0-1
equation. (b) Is it true of every square matrix ? (c) State the theorem that
applies here. (d) find A 1 . (Meerux 93, 90r Rohilkhand 9])
Sol. (a). Here as in Ex. 9 Page 182 Ch. VII we can prove that the
characteristic equation of the matrix A is
X3+A2-5X-5=o
Now A2= 1 2 0 1 2 0 = 5 0 0
2-1 0 2 -1 0 0 5 0
0 0 -1 0 0 -1 0 0 1


188 Matrices
And A 3=A 2 .A= 5 0 0 1 2 0 = 5 10 0
0502-1 0 10 —5 0
0010 0-1 0 0—I
A3+A2-5A--51
= 5 100+50 0-51 2 0-5100
10-5 0 050 2-1 0 010
0 0-1 001 0 0-1 001
= 5+5-5-5 10+0-10+0 0+0+0+0
10+0-10+0 —5+5+5-5 0+0+0+0
0. +0+0+0 0+0+0-s-0 —1+1+5-5
=000
000
000
0, where 0 is the null matrix.
Hence the matrix A satisfies its characteristic equation given by (i).
(b) Every square matrix satisfies its characteristic equation.
(c) Cayley Hamilton's Theorem. (d) Do yourself
Ex. U. Find the characteristc equation of the matrix
A = 1 0 2 and hence compute its cube.
012
120
Sot. Here we have
JA — X1l = 1—A 0 2
0 1—A 2
1 2 0—A
(1—A){—A(1—A)-4)+1 (-2(1—A))
2-6(1—A)=—A3+2A2+5X-6
The characteristic equation of the matrix A is
X3-2A2-5A+6=0.
By Cayle'-Hamilton theorem we have
• A 3 -2A2 -5A-i-61=O . ...(ii)
Now A 2= 1 02 1 x 1 0 2 = 3 4 2
01 2'. 012 252
1 2 0] 1 2 0 1 2 6

.. From(ii)we have A3=2A2+5A-61

2 3 4 2 +5 1 0 2 _6[1 0 0
252 0 12 010.
1 2 6 1 2 0 10 0 1


Solved Examples on Characteristic Equation 189

=[6 8 4 + 5 0 10 + -6 0 0
14104 0510 0-6 0
412 5100 0 0-6
12
=5814
4 9 14
7 14 6 Ans.

Ex. 15 (a). Using Cdy!ey Hamilton Theorem calculate

2A5 + 3A4 + A2 - iii where A = 1 3 ii. (Rohilkhand 93)
H 1 2]
Sol. Here IA-All = 3-A ii
-1 2-Al
=(3-X)(2-A)-(--1)(1)=A2-5X+7.
The characteristic equation of the matrix A is X - 5X +7=0.
Now A 2 =A.A=[ 3 1]x 3 1 1=1 8 5
[-1 2] [- 2] [-5 3
• A2-5A+71= [ 8 51- 5 1 3 1 1 +7 1 1 0
[-5 3] [-1 2] Lo 1
8-15+7 5 - 5+0 1 = [ 0 01=0,
=t 5+5+0 3-10+7] L° 0]
where 0 is the null matrix.
Hence A satisfies the characteristic equation of A given by (I). Thus we have
A 2 -5A+71=O. ..,(ii)
Now 2A5+3A4-A2-III
=2A3(A2-5A+71)+13A4-14A3-A2-l1I
= 2A 3 (0) + 13A 2 (A2 - 5A +71)-t-51A 3 -92A 2 - 11!, from (i)
= 13A2 (0)+ 51A (A 2 - 5A +7!) + 163A2 - 355A -111
= 5 I (0) + 163(A2 -5A + 71)+460A - 11521
163 (0)+460A -11521 =460A - 11521
=4601 3 1 1- 1152 1 1 0 1 = 1 1380 - 1152 460-0
[-1 2] [o 1] [ -460-0 920- 1152
= 1 2284601=41 57 1151.
[-460 -232] [-115 -58]

Ex. 15 (b). Verify Cayley Hamilton's Theorem for the following and
compute 2A 8 - 3A5 + A4 + A2 - 41.


190 Matrices
A= 102
0 -1 1 (Kumaun 90)
0 10
So!. Here IA-All
= 1-A 0 2
=(I-A)-1-A 1
o i -- x I
I -x
o 1 0-A

=(I-A)(A+A2-l)=-A3+2A-1
The characteristic equation of the matrix A is
A3 - 2X+ 1 =0. (i)
Now A2=A.A
[01 021 02=1 2 2
01 1 0-1 1 0 2 -1
10010 0-1 1
A3 = A2 'A
=1 2 21 02=1 0 4
0 2 -1 0 -1 I 0 -3 2
0 -1 1 0 1 0 0 2 -1
A3-2A+I
=1 0 4-21 02+100
0 -3 2 0 -.1 1 0 1 0
0 2-1 0 10 001
I -2+ 1 0+0+0 4-4+0 = 0 0 0
0+0+0 -3+2+1 2-2+0 0 0 0
0+0+0 2-2+0 -1+0+1 0 0 0
=[
= 0, where 0 is the ilull matrix.
Hence A satisfies the characteristic equation of A given by (i) and so Cayley
Hamilton's Theorem for the matrix A is verified.
Thus we have A3 - 2A + I = 0
Now 2A8_3AS+A4+A2_41
=2A5(A3_2A+I)+4A6_5A5+A4+A2_41
=2A$(0)+4A3(A3_2A+1)_5A5+9A4_4A3+A2_41,
from (ii)
=4A 3 (0)-5A 2 (A3 -2A+I)49A- 14A3+6A2-41
=-5A2(0)-+-9A(A3---2A-i-1)- 14A34-24A2-9A-41
= 9A(0)-14(A 3 -2A+J)+24A 2 -37A+ 101


Solved Example on Characteristic Equation 191

=-14(0)+24 1 2 2-37 1 0 2 +10 100

0 2-1 0-1 1 010
0-1 1 0 1 0 001

= 24-37+ 10 48+0+0 48-74+0

0+0+0 48+37+10 -24-37+0
0+0+0 -24-37+0 24+0+10

=-3 48 -26
85 -61
Ans.
o
0 -61 34

Ex. 16 Evaluate the matrix 2A 4 - 7 A3 - 4A2, where A = 1 1 0 2

01 2
120
I
Sol. As in Ex. 10 Page 185 Ch. VII it can be calculated that the matrix
satisfies A3 - 2A 2 - 5A + 61 = 0
+1
Now 2A4 - 7A 3 _4A 2 (A 3 - 2A2 - 5A -i- 61) (2A - 31) - 27A
= -27 A + 181, from (I). Now calculate.
Ex. 17. If A 1 0 0 , show that for every integer n 2: 4,
I ) 1
010
n n-2 3 20
A = A + A - A. Hence evaluate A
Sol. Here we have
A-XI = i-A 0 0
I 0-A I
0 1 0-A

=0 A) Y- X) (- X ) -1) 0 X)(A 2 1)
= - + X2 + A - 1
I =0.
The characteristic equation of A is A 3 - A2 - A +
And so by Cayley-Hamilton's theorem we have
A3 - A2 - A+ I =0
A(A2-1)=A2-I ...(i) (Note)
or
Premultiplying both sides of by A 3 , we have
A
2 (A 2 - 1)=A r- 3 (A 2 - I). ...(ii)
Putting r=n,n-l.n-2.... 4 in (ii) we get
- 2 (A - I) A" (A 2 - I)
A 3 ( A 2 - I)= A 4 (A 2 - I)


192 Matrices 182/11112

A2(A2_I),=A(A2_I)
Multiplying these n - 3 identities we have
A 2 (A 2 -I) = A ( A 2 - I). (Note)
or AA'2A3_A or A'=A2+A3_Aforn>4
Now we have An-A'72 = A(A2 - I)
• Putting a 20, 18, 16,..... 4 in (iii) we get (iii)
(Note)
• A20_A18=A(A2_J)
A 18 - A 16 = A ( A 2 - 1)

A-AA(A2_J)
Adding these nine identities, we get
A20 - A2 = 9A (A2 - 1) (Note)
=9A3_ 9A = 9(A 2 +A_I)_9A from (j)
or A20= 10A2-91. ...(iv)
Now A2$1 0 O l x I I 0 0 1=1 1 0 0
Ii 0 l I'! i 0 Ij 1 1 0
[o 1 0] [o 1
0J 1' 0 1
From (iv) we have
A20 .= 10 1 0 01- 9 1 0 01
II 1 Oj 0 1 0J
[1 0 1] 0 0 1]
.110 0 0 1+ -9 0 0
o 10 00-9 0
=[ 10 0 10] 0 0 -9
=Ii 0 0
10 1 0 Ans.
[10 0 1 ]
Ex. 18. Compute A2. In Example 17 above.
SoLin Ex. 17 above we have already proved that
A3-A2-A+10
• Also in Cor. III Page 168 Ch. VII we have proved that

...(ii)
provided A'T +a 1 A'7 f a2 A'72 +... +a_ j A+a,1=O
...(iii)


182/11/13 Solved Example on Characteristic Equation 193

Comparing (1) and (iii) we have

1, a ,, 2 etc. ora1-1,a2=letC.

From (ii) we have

=—-{A' '+(- 1)A+ +(l)I}, where n3

=A2—A+1}
= 1 0 0 - 1 0 01 + r 1 0 0
1 1 0 1 0 1 0 1 0
1 0 1 0 I 0 01

From Ex. 17 Pages 191-192

or A= 10 0
0 2 -1 ...(iv)
1 —1 2
2
Also from (i) pre-multiplying each term by A we have
A—I—A+A 2 =O. (Note)
or A2=—A+1+A1
=-100+l00+10 0
101 010 0 2-1
010 001 1-1 2

100
—1 3-2
1-2 3 •A.
Ex. 19 (a). Determine the eigen values and the corresponding elgen
vectors of the matrix A = 2 2 1
131
1 2 2
(Garhwal 95)
Sol. Here I A - Xfl
= 2 -A 2 I= 2 - A 0 I applying-
I 3—A 1 1 1—A 1 C2-2C3
1 2 2-. X I —2-2A2—A
=(2—A) I—A 1 + 1 i —A
2X-2 2-A 1 2A-2

(2 - A) ((2 - 3A + A 2) - (2A - 2)1 + [(2A —2) - (1 - A)]

=A 3 +7A2 l]A+5=—(A—l)2(A-5)
The characteristic equation of the matrix A is (A - 1)2 (A - 5) = 0
Its roots i.e. required eigen values of A are 1. 5. Ans.


194 1

Matnces
Now the equation (A -Al) X = 0, for the matrix A is
[2-A. 2 1 X1=0

1
1 3-A 1 x2 0 G)
1 2 2-A x 3 0
[
...See §7.04 P. 166 Ch. VII
Puting A 1 in (i), we get
1 2 1 X1 =[o
1 2 1 2 0
1 2 l X3 [0
The corresponding eigen-vector are given by the equations
X I + 2x2 + x3 = 0, which docs not give any non-zero solution.

The eigen vector corresponding to A = I cannot be evaluated.

Putting A 5 in (i), we get
-3 2 I i=0
1-2 1x2 0
I 2 -3 3 0

f-2 0 2 X1 = [o applying
1 2 I X2 R1 - + 82,
0 4 -4 x3 [0 R 3 -4 R - R2
[
= -2x1+2x3.=0,x1-2x2+x3=0,4x2-4x3=0
= X1 = x3 , x 1 - 2x2 + X3 = 0. x2•
=x1=x2=x3,x1-2x1+x1=O
X = x3 and A I can take any value.
Corresponding eigen vector is (x 1 . x 1 , xt), where x can take any non-zero
value.
Ex. 19 (b). Find the eigen-values and eigen vectors of the matrix
A=311
151
1 I 3
SoLHerejA - )JI
3-A 1 '1- 3-A 0 ,applying C2 C3
1 5--A 1 I 4-A I
I I 3-A 1 A-2 3-A
=(3-A) 4-A 1 +1 14-A
A-2 3-A 1 X-2
=(3 -A) ((4 -A) (3- A)- (A -2)1 + [(X-2)- (4-A)]
(3-A)[A2-8),+ 141+(2X-6)
 Solved Examples on Characteristic Equation 195

=(3-X)[A2-8A+ 14-21=(3-A)(A2-8A+12)

The characteristic equation of A is (A - 2) (), - 3) (A - 6) = 0.

Its roots i.e. required eigen values of A are 2, 3, 6. - Ans.
Now the equation (A - Xl) X= 0, for the matrix A is
3-A 1
I 5-A 1 x2

1 I 3 A X3 (See § 7.04 Page 166 Ch. VII)

Putting A = 2 in (i), we get
Xj =0=: 0
.
1 3 1 12 0
1 1 1 0
The corresponding cigen-vector is given by the equations
x1+x2+x3=0,x1+3x2+x10
1 1 12 X3 X1 X2 13
These give --= or
The characteristic vector corresponding toA = 2 may be taken as (1, 0, 11.
Ans.
Similarly calculate for A = 3 and 6 also.
Ex. 20. Find the eigen-values and elgen vectors of the matrix
A= 3 -5 -4
-5 -6 -5
-4 -5 3
SoI.HerelA - XIt= 3-A -5 -4
-5-6-A -5
-4 • -5 3-A
=(3-),)-6-A -5 1+5 1-5 -5 1-4 -5-6-A
1-5 3-A -43-A -4 -5

= (3- A) [- (6+A) (3-A) -251+51- 15 + 5X-201

-4[25+4(-6-X)
_30:=_(A_4)(A2+4A_77)
=-(A--4)(A-7)(A+ 11).
- Characteristic equation of A is (A -4) (A - 7) (A + II) = 0
Its roots i.e. required eigen values of A are 4, 7, - ii.
Now the equation (A - XI) X = 0, for the matrix A is
r3_A - 41[xi=O
--5 -6-A -5 Iix2j ...(ii)
-4 -5 3 Aj[3j ...See § 7.04 Page 166 Ch. VII
 196 Matrices 999
Putting A 4 in (i), we get
-1 5 4 Xj =0
-5 -10 -5 x2
-4 -5 -1 x3
The corresponding cigen-vectors is given by the equations
x j +5x 2 +4x 3 =05x 1 + lOx2+5x3=O
and 4x1+5x2+x3 =0, which giveex 1 =O=x2 =x3.
These being all zero, the eigen vector corresponding to X = 4 cannot be
e'alutated.
Putting X=7in(i), we get
[4 5 4 X =0
-5 -13 -5 x2
-5 -4

The Corresponding eigen vectors is given by the equations

4x 1 -P5x 2 + 4x 3 = 0 and 5x 1 + 13x 2 + 5x3 = 0
which give x2 = O and x3 = -x1
X1 X3
These give -j-= X2
--= ---.

The characteristic vector corresponding to A = 7 may be taken as (1,0,- 1).

Ans
Similarly calculate for A - 11 also.
Ex. 21. Find the eigen-vectors of A = - 2 2 -3
2 1-6

Sol. Here we can'calculate that

JA -AIJ= -2-A 2 -3
2 1-X-6
-1 -2 -A
and the eigen values are given by A = 5, - 3, - 3.
(Students are to find these in exam.)
Now the equation (A -;U) X = 0, for the matrix A is
-2-A 2 _3 x1 =0
2 1-A-6 x2
-1 -2 - A X3
(See § 7.04 Page 166 Ch. VII)
Putting A = 5, in (i) we get
-7 2 _3 X1
=0
2 -4 -6 x2
-1 -2 -5


Solved Examples on Characteristic Equation 197

The corresponding eigen-vector is given by the equations

-7.rl+2x2-3x3=O.-4x2-6x3=O,-Xl-2X2SX3=O
Xj -2 X3
These give = = — j

The cigen-vector corresponding to A = 5 may be taken as (1, 2, - 1).

Ans.
Putting - 3, in (i) we get 1 2 -3 =0
2 4 -6 -2
-1 -2 3 x3

The corresponding eigen-vector is given by the equations

X1 +2x23x3=O,2x1+4x26x30x1-Zr2+3X3=O.

These reduce to x 1 + 3x3 = 0 only and so non-zero solution of this

2-r 2 -
cannot be found, hence no eigen-vector car( be derived for A = - 3.
Ex. 22 (a). Determine the characteristic roots and characteristic vector
of the matrix A = '1
[1 2
j (Garhwal 93. 92)
Sol. Here I A - XI = 5 -A 4
1 2-A
=(5-X)(2-X)-4=6-7X+X2
The characteristic equation of the matrix A is
A2 -7X+6=O or (A-1)(A-6)=0 or A=1,6.
The characteristic mots of the matrix A are 1, 6. Ans.
Now the equation (A - XI) X = 0, for the matrix A is
[5-x 4 1[xtl=o=[ol
.
[
I 2-A][x 2 ] [ o]

Putting A I in (i), we get.

[441[x1=[o1xl+x2=0=2
[I 1j[x2j [0]
X1 -2
I -'-1
The eigen-vector corresponding to A = 1 may be taken as (I, - 1). Ans.
Putting A = 6 in (I), we get
r-i 41[xI1[014x20
[1 -4j[x2 j [0]
X x2

The eigen-vector corresponding to A = 6 may be taken as (-4. 1). Ans.



198 Matrices
Ex. 22 (b). Find the characteristic vector of A = 1 2 3
023
0 0 2 (Aç'r 901
Sol. Here we get ( A - Al I = 1 - A 2 3
o 2—A 3
o 0 2—A
and the cigen-values are given by A = 1, 2, 2.
(Students are to find these in the exam.)
Now the equaiton (A - Al) X =0, for the matrix A is
I - X 2 3 X1' 0, see 7.04 Page 166 Ch, VII.
0 2—A 3 x2
0 0 —1 [X3
Putting X= 1 in (i) we get
0 2 3 X1
0 1 3 x2
o 0 1

The corresponding characteristic vector is given by the equations

22 + x3 = 0, x2 + x3 0, x3 = 0 which do not give non-zero solution of these
equations and hence no characteristic vector can be derived for A
Putting X=2in(i)weget
1
—1 2 3 [xx3i = 0
0 0 3 x2
0 0 0
The corresponding characteristic vector is given by the equations
- x 1 + 2x2 - 3x3 = 0, 3x3 =

The gives =0.

.. The characteristic vector corresponding to A = 2 may be taken as

(21,0).
Ans.
EL 23. Find the elgen-vector of the matrix A = [03 1 4
026
0 5 (Agra 92)
SoL Here we have JAiJ= 3—A 1 4
0 2—A 6
0 0 5—A
and the characteristic roots (or eigen values) are given by A = 2, 3, 5.
(Students are to find these in the exam.)
Now the equation (A -XI) X = 0, for the matrix A is


Solved Examples on Charactcistic Equation 199

3—A I 4
0 2—A 6 x2
0 05—A
(See § 7.04 Page 166 Ch. Vil)
Putting A 5 in (i) we get
—2 14 X1=Ø
0 -36 x2
0 00 x3

The corresponding eigen-vector is given h the equations

- + x 2 + 4x 3 0, -2 6x 3 0,
X2 X3
which give 3x2 = 6x3 or --
j.
=
X2 x3
Now '-- = k (say), then from - 2x 1 +x + 4x 3 = 0 we get

2 -2 + 4x3 = 2k + 4k 6k or x = 3k
X xl x2 x1
or -=k. Sowc get --=-=.

The characteristic vector corresponding to A = 5 may be taken as

(3, 2, 1).
For A 2, 3 we find that I A - Al I = 0 and so non-zero solutions of (I)
cannot be evaluated in these cases i.e. cigen-vectors Cannot be calculated.
*Ex. 24. If a ± li * c = 0, find the characteristic roots of the matrix
A a c b
c b a (Garhwal 96)
b a c
SoL Here we have
IA - AlI= a—A c-0 b-0 a—A c b
c-0 b--A a-0 c b — A a
h-0 a-0 c—A h a c—A

= a + h + - A c b replacing C 1 by
c+h+a — X b—A a C1+C2+C3
b+a+c — X a c--A'

= —A c b ,since a+h+c=0(given)

—A a c—A

= - A c b . replacing R , R 1 by
0 b-A--c a—h R2 —R 1 and R3—R1
0 a — c c— A — b
 200 Matrices
- A b - c - A a - b ,expanding with respect
a-c c - b - A toC1

=_A[bc_b2_bA_c2+cb+cx_Ac+bA+j2_Q2+ac+_bc]
or AyAIJ,=rA[(a2+b2+c2_ab_bc_ Ca) _2J

Also a+b+c=0(a+b+c)2=O

=a2+b2+c2+2ah+2hc+2ca=O
=2 (ab + bc + Ca) = -(a 2 + b2 + c 2) ( Note)
_(ab+bc+ ca) =(a2+b2+c2)

... From (i)weget

JA-AlJ=A[ [(a2+h2+c2)+(a2+b2+c2))_A2]
(Note)

= A [(3/2) (a 2 + b2 + c 2) - A21
The characteristic equation of A is A [(3/2) (2 -1- b2 + c2 ) - A 21 = 0 which
gives A = 0 or A 2 = ( 3/2) (a 2 + b2 + c2).

The required roots are 0. ± [ ( 3/2) (a 2 + b2 + c 2 )} 2. Ans.

Ex. ' (a). Find latent roots and latent vectors of the matrix
A= a h g
o b 0 (Kanpur 94)
OOc
SoI.Here 1A-XII = a — A h g
o b-A 0
0 0 c - A
=(a-A)(b-A)(c-A)
The characteristic equation of the matrix A is
(A - a) (),-b) (),-c)=0
and the characteristic or latent roots of A are a, h, c. Ans.
Again the equation (A - Al) X = 0 for the matrix A is
-A h g X10
0 -A 0 x 2 0
0 0 c-A x 3 0
r
Putting A = a in the above equation we get
0 h g XIr0
0 b - a 0 x 2 0
0 0 c - a x 3 0


Solved Examples on Characteristic Equation 201

The corresponding characteristic vector is given by the equations
Ox 1 + 12x 2 + 9x 3 = 0, (b - a) x-) = 0. (c - a) x 3 0.
The characteristic vector corresponding A a may he taken as ((X, 0, 0).
Ans.
Similarly we can find the characteristic vector corresponding to A = b and
A ( as (- h, a - b, 0) and (- h, 0, a - c). Ans.
Ex. 25 (h). Find the latent roots and latent vector of the matrix
A h g
Oh f
OOc
So!. Do as Ex. 25 (a), above.
Ex. 25 (c). Find the latent roots and latent vectors of the matrix
A= 2 5 1
O 3 0 (Agra 95)
004
Hint : Do as Ex. 25(a) above.
Ex. 26. If B = r 2 '121 then find the characteristic equation of B and
['12 1]
verify that the matrix B satisfies the equation. Also find the characteristic
roots and the corresponding characteristic vectors of B. (Garhwal 94)
Sol. Here B = [^2 '121 and I= [10
L
2 ij [oi
'12
'12 1—A
=j(2—A)(1 —A)—'12'121
= 2 - 24 - A + - 2
= - 3X = A (A - 3).
The characteristic equation of the matrix B is A (A - 3) = 0 Ans.
and the characteristic roots of B are 0 and 3. Ans.
Also B'= [ 2 V2 1 x [ 2 12
['12 ij ['12 1

2.2
r2.2+'12.'12 2.'12+12.l1={ 6 3'12
[ 12.2+ l.'12 42. '12+ 1.1j [3'12 3

B 2 -3B= [ 6 3'121 — 3[2 12

[3'12 3j ['12 I

= [ 6 3'121+[ _— 6
3'12 3 j3'12 --3

- 27


202 Matrices

=[ 6-6 342-3I21=[00
L3 2 - 3\2 3-3 ] L° o

= 0, where 0 is the null matrix.

Hence the matrix B satisfies its characteristic equation given by
or X23A0
The equation (B - XI) = 0 for the matrix B is
12XI2 i[xil=o
2 I _X][x2j
L
(see § 7.04 Page 166 chapter VII)
Putting X=0 in (I), we get 2 '211XI1=0
I 2 1][x2]

The corresponding characteristic vector is given by the equations

(Note)
III+42X2=0 and 'I2i + x2 0.
Taking any one of them we get -
X1 .r 2 X1 X2
01

.'. The characteristic vector corresponding to A 0 may be taken as

Putting A 3 in (i), we get

[-i 421['il=0
•i 42 -2j[x2j
The corresponding characteristic vector is given by the equations
_x 1 +2x2 =0 and I2-2x2=0.
X I x2 x1 X-
Taking any one of them we get -or or - =

The characteristic vector corresponding to A = 3 can be taken as (12, 1). Ans.

**Ex 27. Show that the matrix A= 0 c - b satisfies its
-c 0 a
b -a 0

characteristic equation. Also find A . (Agra 91. Kwnaun 91 Rohi/khand 92)

Sol. Here (A - AIH 0-A c-0 -h-0
-c-0 0-A a-O
h-0 -a0 0—A

a1_bIc -A
=4

-a -Ab -A]
I
11-
[b -a
C

= - A (A2 + a ) - C (c), - ab) - h (ca + bX)



Solveds Examples on Characteristic Equation 203

(e,2 2 2
+ 1, + c )
_X3 - X
The characteiistic equation of A is
i+A(a2+b2+c2)-O.
Also we have A 2 [ 0 c - h 0 c - b
C 0 a -c 0 a
b -a 0 b -a 0
2 2
= --(c +b) ab ca
2 2
ab -(c +a ) he
2
ac he -(a +b 2)
43
= A2
2 4 c2 ) ab Ca 0 c -b
= [- (h 2)
bc ab - (c 2 + a - C 0 a
I22 b-a 0
[ ac be -(a +b )
' 2 2 ' 2
- c(b+c +a ) b(b+c+a )
02
= 2 2
C. + a +b) 0 -a (b +c +a)
' ' 2 - 2 2 2
-h(c+a+h) a(c +a +b) 0

0 C -b
-c 0 a
b -a 0

taking out - (a 2 + b 2 + c2) common. (Note)

or A3 = - (2 + b 2 + c 2 ) A .(ii)
or A 3 + (a 2 + b 2 + c2 ) A = 0, which shows that the matrix A satisfies its char-
acteristic equation given by (i).
Again multiplying both sides of (ii) by A 2 we get
( 2
A - + h 2 + c 2 ) A (Note)
which gives A = - _ 12 A
(a+b +c')
- 1
0 c-b
c 0 2a Ans.
- (a 2 + b + c2 ) -
b -a 0

Exercises on § 7.03-7.06
EL I. Show that the characteristic equation of the matrix
A= 1 1 0 2 is(? - 1)(X+2)(X-3)=0
012
1 2 0


204 Matrices
Ex. 2. Show that the matrix A= 1 21 satisfies Cayley-HaniiRon
L' Ii
Theorem,
Ex. 3. Let A = 2 20 Find the characteristic equation of A and
21 1
-i 2 —3
verify that matrix A satisfies this equation. Also find the characteristic roots and
the corresponding vectors of A.
Ex. 4. Using Cayley-Hamilton Theorem or otherwise determine the inverse
of the matrix 4= 2 2 1
- 12
1-22
Ex. 5. Verify Cayley-Hamilton Theorem in the case of the matrix
A = 01 2 . Hence find A
2-3 0
1 11

*Ex. 6. If matrix A= 2 - I
—1 2 —1
i1 find the characteristic roots of A.

1 —1 • 2
Verify Cayley-Hamilton's Theorem and hence compute A
- Ans. 1, 1,4, !f 3 1 -
1 3 1
L- 1 1 3
Ex. 7. Verif the Cayley-Hamilton Theorem and find the characteristic
roots. where A= 1 2 2
212
211
(Garhwal 91)
Ex. 8. Show that the matrix A = 3 0 - 1 satisfies Caylcy Hamilton
20 1
00 4
Theorem.
Ex. 9. Using Cayley-Hamilton Theorem find the inverse of
A=2 4 3
0-1 1
2 2 - I (Rohilkhand 96. 90)

Ex. 10. Verify Cayley-Hamilton Theorem and verify it for the matrix A and
1,
hence find A where


Exercises on Characteristic Equation 205

A= 1 'ho
'J2 -1 0
0 0 0 (Lucknow 92, 90)

**Ex. 11. Using Cayley-Hamilton's Theorem, compute the inverse of the

matrix A 1 2 1 1 Ans. 1 3 -2 1
032 420-2
I 0 1 -3 2 3

Ex. 12. Show that A = 1 2 2 satisfies the matrix equation

• 212
22!

A - 4A - 51 = 0, where I is the unit matirx, Deduce A

Ans.i -3 2 2
5 2-3 2
2 2-3

Ex. 13. Obtain the characteristic equation of the matrix A = 1 0 2 and

012
120
hence evaluate A (Kumaun 93)

Ans. X3_2X2_5X+6=O;A1= 4 -4 2
-2 2 2
1.2-1

Ex. 14 If A =[i 21 calculate A with the help of Cayley-Hamilton's

[4 3j
Ans. ir-3 2
Theorem.4 - J
I with the help of
Ex. 15. If A 0 0 0 , calculate A-
310
-2 1 4

Cayley-Hamilton's Theorem. S Ans. i 5 0 0

4-3 40
2-I I

Ex. 16. Find the cigen-values and eigen-vectors of the matrix

211
-11 4 5
- 1 1 0 Ans. - 1 1 1 [7 ± 'i(- 39)J, no eigen vectors.

Ex. 17. Find the eigen-vectors of the following matrices :-



206 Matrices
(a) -- 3 2 2 (h) I 30 (c) 3 1 4
—6 5 2 3 —2 —I 0 2 6
—7 4 4 0 —1 1 0 0 5 (Agra 92)
Ex. 18. Find the characteristic vectors of the matrix
A= 8 —6 2
- 6 7 -4 (Kurnaun 91)
2-4 3
Ex. 19. Find the eigen values and eigen-vectors for the matrix
A= 3 —5 —4
—5 —6 —5
—4-5 3
Ex. 20. Find the characteristic roots of
—2-8-12
14 4
0 0 1
(Kumaun 96)
Ex, 21. Find the invariant vector of the matrix
A=[2 2 1
131
I I I
Ex. 22. Find the characteristic roots and vectors of the matrix
A=[-2 -11
[5
] Ans. —J,3;(1,—l)jl-5)
Ex. 23. Determine the characteristic roots and associated invariant vectors.
given .A= 2 3 1
131
I 3 2 (Lucnow 91)

MISCELLANEOUS SOLVED EXAMPLES

*Ex I Prove that matrices A and 13 AB have the same latent rouIs.
Sol. We know that two matrices have the same latent roo c (or characteristic
roots) if their characteristic equations are the same. (See definition of latent roots
in § 7.02 Page 162).
Let B 1 AB C. then C - XI = AB - XI. (i)
Also B,JBB'AB=XB 'B=XI
From (i)we get C-A1=B AB — BXIB
=1 (A — XI)B
or . jC — XII=BI!A - AIJItfl
=jA - AIIIBIIBI


Miscellaneous Solves Examples 207

= I A - A1 HB'BI
A_AlII1j=IA-ALl
tC_-X1I = 0 = I A -Al 1=0
Hence the characteristic equation of C and A are the same i.e. C and A or

AB and A have the same latent roots. Hence proved.

Ex. 2. Prove that the cigen values of a diagonal matrix are given by its
diagonal elements.
0 ... 0.
Sot. Let A= all
0 a22 ... C)

0 0 .

be a diagonal matrix with a11 a2 2 ,.... as diagonal elements.

Then the characteristic equation is

all- 0 ... 0 =0
IA)1l 0 or I
0 '22 - A .. 0

0 0 ...

or (a - A) (a2 2 - A)... (a,,,, - A) 0

or A = a, a22......1trn are the eigen values of the matrix A and are given
by the diagonal elements of the diagonal matrix A. Hence proved

Ex. 3. Find the spectrum of the matrix

A= --2 2
2 1 —6
--1 —2 0
Sol. We can iind that the eigen values of the matrix A are 5. - 3. - 3.
Also we know spectrum of A is the set of elgen values of A.
[Sec § 7.02 (iv) Page 1601
Reuuired spectrum of A = (5. -- 3. - 3) 15. - 3 Ans.
Ir:. 4. 1 he equation AK = A X has non-trivial solution X 1ff
A is a
characteristic value of A.
Sol. Let A 1 be a haracteristic value of A and X 1 hc the corresponding
characteiistic vector of A. then
AXX 1 X 1 =A 1 1X1.=(X11)X1
AX - 'A 1 l't X 1 = 0. where 0 is the null matrix
or

or (A - A 1 1) X 1 = 0 or (A - X 1 1) =0. X1 ;4 0

or -- =0



20.8 Matrices 182111/13

Hence every characteristic value A of A is a root of its characteristic
equation.
Conversely if 1 1 be any root of the characteristic equation j A - A
1 = 0,
then the equation (A - X 1 1) X =r o must possess a non-zero vector X1,
such that AX1 =A 1 1X 1 =X 1 X1.
Hence every root A of the characteristic equation of A is a characteristic
value of A.
EXERCISES ON CHAPTER VII
EL 1. Evaluate the matrix A 5 - 27A 3 + 65A 2 , where
A= 001 Ans. 40 2 48
3 1 0 128 —3 0
—2 1 4 86 —43 —132'
Ex. 2. Evaluate A 50 in Ex. 17 Page 193 Ch. VII.
Ex. 3. If the matrix A = 2 - I ii. find the characteristic roots of A.
—I 2-1
1 —1 2
Verify Cayley-Hamilton Theorem and hence compute A
**Ex. 4. When do you say that two matrices A and B are smilar ? Prove that
the similar matrices have the same characteristic roots,
Ex. 5. Find the characteristic polynomials of the matrix A and hence
compute 2A 6 - 3A 5 .+ A 4 +A 2 - 41, where A is the matrixI 02
1o
1 1
Also determine one of the characteristic roots corresponding the
characteristic vectors.
Ex. 6. Verify the Cayley Hamilton's Tlieorern and find the latent roots where
A= 1 2 2
212
221
Ex. 7. Determine the characteristic roots of the matrix
1 0 0:—I
—1 0 0 0
0 —1 .0 0
o 0 1 0
Ex. 8. Let A and B be two square matrices over thc field of real numbers,
and let B be non singular, obtain the characteristic roots of A. if
- BAB=—
- 300
010
- 007


Exercises on Chapter VU 205

182/11/114

Ex. 9. Use Cayley-Hamilton Theorem to find A*f

A=2!2
023
005
Ex. 10. Find the characteristic roots and characteristic vectors of the
following matrix A 1 0 - 2
0 0 2
—2 0 4

Ex. 11. Find cigen-vectOrs of the matrix

A= 8 —6 2
—6 7 —8
2 —4 1

**Ex. 12. If A 1 . 2. •-.. X, are the characteristic roots of a square matrix A of

1/As, 1/A 2 ,..., l/A are the characteristic roots of the
order n, then show that
(Agra 92; Kwnaun 94)
matrix A .
Ex. 13. Does the matrix A = I 2 3}atisfy Cayley Hamilton Theorem?
023
003
(Agra 90)
Find eigen-values and eigen vectors of A.
A is 2, and find the
Ex. 14. Prove that one characteristic root of
corresponding characteristic vectors where A = 2 0 1
02 0
102
the matrix
Ex. 15. Find the characteristic roots and characteristic vectors of
2-2 3
I 1 1
I 3_i]
tools of a matrix A, then
Ex. 16. Show that if X 1 . A2 , A3 ,... A are the latent
A'1 has the latent roots
X, X, 4 ... A. (Agra 96)

Ex. 17. Find the characteristic roots and characteristic vectors of the matrix
A= 1 —1 -ii
1 —1 oj (Agra 96)
0 —1]

- Ex. 18. Show that the characteristic roots of an idempotent matix are either
(Bundelkhafld 91)
zero or utity.
0


Chapter VIII

Linear Dependence of Vectors

• 8.01. Two dimensional vector.
We know that the ordered pair of
real numbers (x 1 , x 2 ) is used to denote
a point P in a plane where Ox and
Ox2 are th.. coordinate-axes.
A two dimensional vector or
2-vector OP is denoted by the same
pair of numbers written as {x 1 , x21.
If A1=[x11,x12] and
A2 = [x21 , x27 1 are two distinct (Fig. 1)
two-dimensional vectors, then
their sum by parallelogram law of
addition is given by
A 3 A 1 + A2
Ex 11 +x21,x12+x22]
[Here OM=OL+LM
0L+ON=x 11 +x21 etc.]
If we treat A 1 and A 2 as I x 2
($1112)
matrices, we find that the above is
the rule for adding matrices as 1T
given in chapter 1. L Al
Also we observe that
k. Al = [kx 11 , kx j2 ], where Ic is any scalar.
2)
§ 8.02. n-dimensional vector or n-vector.
Definition. An ordered set of n elements x, of a field F, w1itten as
A[x 1 , x2 , x3......
is called an n-dimensional vector or n-vector A over F and the elements
x 1 , x2 . ... . x,1 are called, the first, second, ..., nth components of A.
We find it more convenient to write the components of a vector in a column
as A'[x1x2..... x,j'= x1

xn
0i)


211
Linear Dependence of Vectors
(i) is called a row-vector and (ii) is called a column-vector.
Thus we consider the p x q matrix as defining q column vectors or p row
(Note)
vectors.
Note 1. The sum or difference of two rwo (or column) vectors is formed by
the rule governing matrices as given in chapter I.
Note 2. The product of a scalar and a vector is formed by the rule governing
matrices as given in chapter I.
null
Note 3. The vector whose all the components are zero is known as the
vector or zero vector and is written as 0.
Solved Examples on § 8.01 - § 8.02
Ex. 1. Given the 3-vectors
evaluate 2A 1 + A 2, 5A 1 - 2A3
A 1 = [1. 2, 11, A 2 = [2, 1, 4 ], A 3 = [2,3,6],
Solution.
= [2, 4, 21 + [2, 1, 41
Ans.
=[2+2.4+1,2+4]=[4.56].
5A 1 -2A3 =5 1 1 , 2 , l]-2[2,3,6]
= [5 , 10,5] — [4 , 6 . 12]
=[5-4, 10-6,5-12][L4,71.
Ex. 2. Given the four-dimensional column-vectors
, evaluate 3A 1 + 2A2.
A 1 I and A2 =
0 5
2 7
9
3J
Solution. 3A 1 + 2A 2 = 3 1 +2 2 = 3 + 4 = 3 + 4 = 7
0 5 0 10 0+10 10
2 7 6 14 6+14 20
3 9 9 18 9+18 27 Ans.

Ex. 3. Given the three-dimensional row vectors

evaluate 2A 1 - A2 - 3A3.
A 1 = [ 3, 1, 4], A 2 = [0, —4, 11; A 3 = [2, 2, — 31,
1 [2, 2,— 31
Solution. 2A 1 - A 2 - 3A 3 =2 [3, 1, —4]— [0, —4, 1- 3
=[6,2,_8]_[0,_4,l1—[6.6-91
Aris.
=[0,0,0]0.
8.03. Linear dependence and independence of vectors.
(Agra 94; Pur.'anchal 97)

The nm-vectors over the field F,

A 1 = ( x 11 . x12.....x1,J, A 2 = [x 21 , .r22,..x2,l.....
A n = [Xnl, Xn2.....X,,,,,]
n elements
are called linearly dependent over F if there exists a set of
). X2. ... . X, of F, X's being not all zero, such that
 212 Matrices 999

Otherwise the n-vectors are called linearly independent over F.

For example the 3-vectors given in Ex. 3 above are linearly dependent.
Note 'A vector A + can be expressed as a linear combination of the
vectors A 1 , A 2, ..., A n if there exist elements 4 1
, A2 , ..., A of F such that
An I = A 1 A 1 + A2A2 + -I- A,,A
Solved Examples on § 8.03.
Ex. 1. Examine whether the set of vector V 1 = [1, 2, 31, V 2
and V3 = (0, 1, 0) are linearly dependent or not (1, 0, 1)
(Pu,','anc1 94)
Solution. Let the given set of vectors be linearly dependent,
so that
AIVI+A2y2^A3v3o
or A1 (1,2,3) +A2 (11 0, 1)+A3 ( 0,- 1 , =0= (0,0,0)
or (Note)
(A +A 2,2A 1 +)L 3,3X 1 +A2)={O,O,O)
)1 1 + ?12 = 0,2? I +A3=O,3X 1 +A2=o
Al 9 .. 2 =O,A 3 0j.e.As are all .zero.
Hence the given set of vectors are not linearly dependent
i.e. these are
linearly independent.
Ex. 2. Examine the following set of vectors of
the real field for linear
dependence or independence
AI=[2,_1,3,2);A2=[1,342];A3..[35Z2]
Also express A1 as a linear combination of A
1 , A2.
Solutjo. Suppose the given set of
vectors is linearly dependent, so that
A 1 Ai+A2A2+A3A3.O
or A1 [2,— 1,3,21 + A2 [1,3,4,2] + A3 [3, — 5,2.2) -
= 0 [0,0,0,0]

or (Note)

41 + A2 + 3X3 = 0 =10,0,0,01
...(i); - A 1 + 3A - 5A = 0 ...(ii)
3A 1 +4A2 +2A3 =o ...(iij); and 2 (X I +) 2 +A3 )0 ...(iv)
From (i) and (iv)we get, 1 + 3 0 or A1=-2A3
From (iv) we get A2 = — A 1 — A3 2 A3 — A 3 = A3
From (ii). (v) and (vi) we get ...(vi)
- A 1 + 3A2 - 5A 3 = - (— 2A) + 3A3 - 5A3 =0. Hence (ii) is
satisfied.
Again from (iii) we get 3X1 + 4A2 +'2X3
= 3 (— 2A 3 ) + 4 01 3) + 2X3 , from (v), (vi)
= 0. Hence- (iii) is also satisfied.


Linear Dependence of Vectors 213

Thus for X - 2X 3 and X = A all the equations (1), (ii), (iii) and (iv) are
satisfied and therefore the given Set of vectors are linearly dependent.
- A2 , which
From (i) we get - 2X 3A 1 + A 3A 2 + A.3A 3 = 0 or A 3 = 2A 1
expresses A 1 as a linear combination of A 1 and A2.
Ex. 3. Show, using a matrix, that the set of vectors
is linearly
= [2, 3, 1, - 1 J, X 2 = [ 2,3, 1, - 2 ], X 3 = [4, 6, 2, -3]
(Agra 96)
dependent.
Solution. Let the given set of vectors be linearly dependent, so that
XX1+X2X2+A3X30
or X1(2,3,l,_1l+A2[2,3,12I+A3[4,6,2,31O°').°*
, - A 1 2A2 - 3X31
or [2X1 + 2X 2 + 4A 3 , 3X 1 + 3X2 + 6A3 , A 1 + A2 + 2X 3
(0.0,0,01

2)9 + 2A 2 + 4A3 = 0, 3)9 + 3A 2 + 6A 3 = 0

X 1 +X2 +2A3 -0; _A1_2A2-3A3=O

which reduce toA 1 +X2 +2XO. A1+2X2+3X30

whence solving we get A 1 +X3=0, A 2 + X3=0

I 2 '
which give
-i--=--i--=_i_
From (i) we get X 1 + X 2 - = 0, where A 1 , A2 , A 3 are not all zero.

Hence given set of vectors is linearly dependent.

Ex. 4. Examine the following set of vectors over the real field for linear
dependence or independence
A1=[1,2,1];A2=[2,1,41;A3[4,5,6]A4=[1S,'3I
Solution. Suppose that the given set of vectors is linearly dependent, so that

or A 1 [1,2, 1]+ A 2[ 2 , 1,41+A3[4,5,6]+A4E1.8,_3]0=[00.01

or
...(ii); 2)9 + A2 + 5A 3 + 8A4 = 0 ...(iii)
+ 2X2 + 4A 3 + A4 = 0
A1 + 4A2 + 6A3 - 3X4 = 0 ...(iv)
and
From (ii) and (iv) we have 2A 2 + 2A3 - 4? = 0

or A2+A3-2X4=0
From (iii) and (v) we have 2)9 i-A 2 + 9 3 + 4 012 + A 3) = 0
2A1+5A2+9A30 ...(vi)
or
From (v) and (vi) we have 2)9 + 5A 2 + 9 (2A - A 2) = 0
or 18X4 4A 2 + 2)9 = 0 or 9A4 2A2 - A 1 ...(vii)
(v), (vi) and (vii) are satisfied by A 1 = 0 = A = A 3 = A4


214 . Matrices
Hence the given set of vectors are linearly independent.
8.04. Basic Theorems on Linear dependence of vectors.
Theorem 1. If there be n linearly dependent vectors, then some one of them
can always be expressed as a linear combination of the remaining ones.
Proof. Let Ai=[Xji,x12,,.,x1JA2(x21x22.....
... Ar1XnI,X2,
be n rn-vectors over the field F, such that
A 1 A 1 +X2 A 2 +A.3 A3 + +XA"O.
where A 1 , A2.......are elements of F and not all zero.
Let Ar # 0 then solving (i) we get

Ar_[A l A I +A2A2 + A A + +AA]

or
...(ii)
Hence proved.
Theorem U. If there be n linearly independent vectors
A 1 , A 2 ,..., A,
whereas the set obtained .5y adding another vector A
+ I is linearly dependent,
then A + can be expressed as linear combination of A 1 , A 2 , ... A.
Proof. Given A,A 1 +A2A2 +. +XA;40
where A, X2, ..., X, arc elements of the field F.
.
And (XI A 1 + )L2A2 + ... + + A + 1 0
(Note)
or [A1A1 +A2A2 + ... +A,A]
Hence provea.
Example. Consider three 3-vectors
A 1 =[4,5,6],A2=f2, I,4],A3=[I,2, 11
Let A 1 and A3 be linearly dependent then we musthave
A 1 A 1 + A2A3 = 0, where A 1 and A 2 are to be determined.
or

Solving first and third of these we get A 1 = 0= A2

which satisfies the second
also. But as all X's are zero, so A2 and A3 are not linearly dependent [See § 8.03
Page 211 of this chapter]
But we find that if we take A l , A2, A3 to be linearly dependent
then A1A1 + A2A 2 + A3A3 0.
(i)
or Al[45.6}+X2[2.1,4]+A3[l2lJO_[000
4X 1 + A3 = 0 ...(ii), 5X1 + X2 + 2X3 = 0
and (iii)
6A1+4)12+A30
.(iv)
From (ii) and (iv) on subtracting we get A, +,k2 = 0
or . . X2 X1
 Linear Dependence of Veêtors 215

From (ii) and (v) we get

4A 1 4 2(— A 1 ) + A 3 = 0 or A 3 = - 2 X i ...(vi)
Substituting values 1mm (v) and (vi) in (I) we get
- 2A 3 = 0 ...(vii)
X 1 A 1 - X 1 A, - 2XA 3 = 0 or A 1 - A 2
or A-,= A t - 2A3-
are not linearly dependent yet A1
Thus we find that though A 1 and A 3
can be expressed
satisfy the relation (vii) and from (viii) we find that A 2
A 2. A 3
as a linear combination of,A 1 and A3.
*Theorem M. if tiure be a subset of r linearly dependent vectors among
such that r < n, then the vectors of the whole set are
the n vectors A 1 , A7. ....A
linearly dependent.
subsets A 1 . A,.....A of the given n vectors
Proof. Let th
A 1 , A2.... . A ,1 be linearly dependent, then we have

X 1 A 1 +X,A,+ ... X=O, where all X's are not zero

We can rewrite this as
XiAi+X2A2+..+r+0Arl+OA2+OO (Note)

where all A's are not zero.

definition are
Hence the set of vectors A 1 , A 2 ..... A,., Ar+ i,,... A, by
Hence proved.
linearly dependent.
of n rn-vector,
"Theo re m IV. If the rank of the matrix associated with a set
is r where r < n, then there are exactly r vectors which are linearly independent
of
while each of remaining n - r vectos can be expressed as a linear combination
these r vectors.
**Theorem V ,4 necessary and sufficient condition that the vectors
..., A, = [ XnI,Xn2.... . X,.,1) i1
A 1 = [x 11 . x12,...,.u1m], A 2 = 1x 21 , x 22 .....
linearly dependent is that the matrix A = -x l 1 X12 x1m of the
X21 X12 ... ... .X2rn

X,1 Xn2 ....... . Xnm

vectors is of rank r < n. if the rank is n, the vectors are linearly independent.
Proofs of Theorem IV and V above are beyond the scope of this book.
§ 8.05. Linear Form.
Definition. A linear form over F in m variables AI,X2, ....X,.,, is a
polynomial of the type

= a 1x1 + a 2 + ... +

where the coefficients are in F



216 . Matrices
Consider a system of n linear forms in m variables
A =a 11x 1 +ax2 + ... +a1,,,x,,
f2 = a 21x 1 + a22x2 + -. + a2,nxm

f = an + a n2X 2 + . - +
and the associated matrix formed by their coefficients is
A= a ll 0 12 aim-
a21 a22 ... a2m

an i afl 2 anm

If them exist elements A 1 , A2 .....Xin F, A's being not all ero such
that 6
X + A 2 + ... + X,,j = 0,
then the forms (i) are said to be linearly dependent, otherwise they are said to
be linearly Independent.
Thus, 'the linear dependence or independence of the forms (i) is equivalent
to the linear dependence or independence of the row vectors of the matrix A.
More Solved Examples:
Ex. 1. Show that the set of vectors A 1 = (1, 1, 1), A 2 (1,2,3),
A3 = (2,3,8) is linearly independent.
Solution. Suppose that the given set of vectors is linearly dependent, so
that 41A1+A2A2+A3A3=0,
where X's are to be determined.
or
or (A1 + A2 + 2A3, A 1 + 2X 2 + 3A 3 , A 1 + 3X2 + 8A 3 ) = ( 0, 0, 0)
A 1 +X2 +2A 3 =0 ...(ii), A+2X2+3A3=o
and X1+3X+8A3=0. ...(iv)
From (ii) and (iii) we get A 2 + A 3 0.
From (ii) and (iv)we get 2A 2 +6A3 0 or A2+3A3=0
From (v) and (vi) we get A 3 = 0.
From (v) we get A 2 = 0 and from (ii) we get X = 0. when A2 = 0 A3.
Thus all the A's are zero and hence the given set of vectors is linearly
independent. -
Ex. 2 (a). Show that set of vectors A 1 [1, 2, 3 ], A, = [3,2, 11,
A3 = 11, 1, 11 is linearly dependent.
(Ara 93)
Solution : Suppose that the given set of vectors is linearly dependent, so
that AIAI+A2A2+A,A30
 Linear Dependence of 'vectors 217

or AL1,2,3]+12[3,2,11+1311.1.110=10,0,01
or + A 2 + X 31 10,
[11 + 31 2 + 29. 229 + 212 + 1 3 , 3A I = 0. Oj
29 + 312 -+ 1 3 = 0 (i): 2X + 21 2 + 29 = 0
and 329 + 1 2 + 13 = 0
From (i) and tii) we get 29 - X2 = 0 or 11 = 12 ...(iv)

From (iv) and (iii) we get 4X + 1 3 = 0 or 1 3 = - 4X

A 12 13
Then we have - - and from here we do not get all the values of
=-j- =
A as zero.
Hence the given set of vectors is linearly dependent. Hence proved.
Ex. 2 (b). Show that the set of vectors [1, 2, 3), [3, - 2, 11, [1,- 6, - 51 is
linearly dependent. (Agra 92)
Solution : Let A1=[l.2,3},A2=[3,-2,l}andA3=[],-6,-5}beaset
of linearly dependent vectors.
Then ?c 1 A 1 + 13 A 3 = 0, where all A's are not zero.
or A11L2.3]+12(3,-2.1J+)L3[l,-6,-51=O=[0,0,0J
or [29 + 317 + X . , 229 - 21 2 - 613, 3A + 12 - 51 3 ] = [ 0, 0, O]
29 + 3 1 2 + A 3 = 0 ...(i), 2A - 21 2 - 61 3 = 0
and 329 + 12 - 91 3 0
From (I) and (ii). we get 81 2 + 81 3 = 0 or 12 + 1 3 = 0
From (ii) and (iii), we get 8A 1 - 161 3 = 0 or 29 = 213

XI 12 13
From (iv) and (v) we getand
-1 and this does not give all the va
-j-
of A as zero.
Hence the given set of vectors is linearly dependent. Hence prov
Ex. 3. Find a linear relation, if any, between the linear forms of t;
following system f1 = x + v + z,f2 = y - 2z ,f3 = ir + 1y.
Sollution: Let Lf1 +A2h+AJ3=0.
Then 29 (.x+y+)+X,(y-2z)+A3(2x+3v)=0
or (A + 21 3 ) x + (A 1 4 X2 + 31 3 ) y + (X - 21 2) z = 0
=X 1 +229=0,29 +17+313=O,Aj 20
whence we get 1 3 = -X I, X2 = X1,which satisfy 29 + ? + 31 3 = 0

Hence from (i)we get AJ1+.A1f-A1f3=0

or -f = 0. which is the required relation.
2.f + f2

Ex. 4. Find a linear relation, if any, between the polynomials

3
= 2x 3 - 3x2 + 4x - 2;f2 3x 2x2 - 2r + 5;f3 = 5x 3 -x2+ Zr + 1.

-28
218 Matrices
Solution. Let X 1/1 + A 2/ + Af1 = 0
• Then A1(23-3x2+4x-2)+X2(3x3+22-21+5)
+X3(5x3-x2+2xfl)=0
or (2X3 + 3X 2 + 9. 3 )x3 + (- 3A 1 + 24 A) 2 + (4A 1 - 2A 2 + 2A3)x
+(-2X +5A,+X3)=0
2X 1 + 3X 2 + 5X3 = 0 •••(i'); 3X1 - 2A 2 + A3 = 0 (iii)
2X 1 - X 2 ±X3 =0 (iv); and 2X1-5A2-A3=0
Solving (ii) and (iv) we get A2 - A 3 = 0 (vi)
From (iii) and (iv) we get 3X1 + 321 3 = 0 or A 1 + A 3 = 0 (vii)
From (v), (vi) and (vii) we get
2(-X3)-5(-)3)-A3=0 or 2A3 =0 or A3=0
vhich gives Al = 0 = A2.
Hence from (I) no linear relation exists betwecqf 1 , f2 and f3.
Ex. 5. If the vectors (0. 1, a), (1,a, 1), (a, 1,0) are linearly dependent,
hen find the value of a. (Agra 94)
Solution Jet A t = (0, 1, a), A 2 = (1, a, 1), A 3 = (a, 1,0) be a set of
nearly dependent vectors.
Then X 1 A 1 + X 2 A 2 + A 3 A 3 = 0, where X's are not all zero.
A 1 (0, La) + A (I, a, 1) + A 3 (a, 1,0) =0= (0,0,0)
(A +aX 3 ,A 1 + X 3 , IX I + A 2) = (0.0,0)
X 2 +aA 3 =0 ...(i), A1+aX2+X3=0 ...(ii)
and aA1+A2=0 .,.(iii)
From (i), A 3 = - (1/a) 42
From (iii), A 1 = - ( 1/a) A2
From (ii), -(1/a) A 2 + aX 2 - (1/a) A 2 0

or [a-(2/a)]A2=0 or (a2-2)X2=0

Either a2-2=0 or A2=0.

But A2 = 0 gives A 1 = 0, A3 = 0, from (i). (iii)
Then A 1 , A 2, A 3 are not linearly dependent.
Hence a 2 - 2 = 0 or a = ± '[i. Ans.
Exercises on Chapter VIII
Ex. 1. Show that the vectors (1, 0, 0), (0, 1,0) and (0, 0, 1) are linearly
ndependent.
Ex. 2. Show that the vectors [1, 2, 01, [8, 13, 01 and [2, 3, 01 are linearly
ependent.

Linear Dependence of Vectors 219
Ex. 3. Prove that the set of three vectors
[1,2,— 1, 31, [0, —2,1,—I] and 12.2,— 1,5]
is linearly dependent and obtain a relation connecting these vectors.
Ex. 4. Find a linear relation, if any, between the linear forms of the
system :-
f1=2x1-22-3+x4;f2=x1—x2+x3+x4:f3=5x2+3x3+x4.
Ex. S. Prove that any non-empty subset of a linearly independent set is
linearly independent. (Agra 94)
00
Chapter IX

Quadratic Forms
• § 9.01. Quadratic Form.
Deliniton. A homogeneous polynomial of the type

q=X'AX=Z Lajjx1xj
i=I j=1
whose coefficients aii are elements of the held F is known as a quadratic form
over F in the variables , x 2 . ... . x,.
For Example. X I + 2-x2 +5x3+ 8xlx3-6x2x3 is a quadratic form in the
variables x 1 , x2, x3 . Here the matrix of the form can be written in many ways
according as the cross product terms 8x 1 x3 and - 6x2-x3 are seperted to form the
terms a 13x 1 x 3 , a 3 1 x, 1 and a23 x2 x3, a32 x 3 x2.
Here we shall agree that the matrix A of quadratic form be symmetric and
shall always separate the cross-product terms so that a, =

q=4+24+54+8.r1.t3-6x2x3
[x 1 x2 x 3J 10 4 X1 ...see Ex. 1(b) Page 223 of this chapter
0 2 —3 x2
4 —3 5 x3

=X'1 0 41 X=X'AX
0 2 —3
4 —3 5
The symmetric matrix A [a] is known as the matrix of the quadratic form
and the rank of A is called the rank of the quadratic form.
If the rank of the form is r < n, then the quadratic form is called singular
otherwise non-singular.
Transformations.
The linear transformation X = BY over F carries the quadratic form (i)
above with symmetric matrix A over F into the quadratic form
(BY)'A (BY) = (Y'B') A (BY) = Y' (B' AB) V ..(ii)
with symmetric matrix B' AB.
Equivalent Quadratic Forms.
Definition. Two quadratic forms in the same variables x, x 2,..., x, are
called equivalent if and only if there exists a non-singular linear transformation
X = BY which together with V = IX. where I is the identity matrix, carries one
of the forms into the other.
 Quadratic Forms 221

As B'A 11 is congruent to A, we have

1. The rank of a quadratic form is invariant under a non-singular
transformation of the variables.
2. Two quadratic forms over F are equivalent over F iff their matrices are
congruent over F.
A quadratic turin of rank -r can be reduced to the form
b1y + b 2y + by + ... +
h1 ;1 0 ..(iii)

in which only tcnns in the squares of the variables occur by a non-singular linear
transformation X = BY.
§ 9.02. Lagrange's Reduction.
The reduction of a quadratic form to the form (iii) of § 9.01 above can be
carried out by a method or procedure called Lagrange's Reduction, which
consists of repeated completing of the square.
From example q = 4 + 2x + 5x + 8x 1 x3 - 6xx1 can be reduced to form
(iii) of § 9.01 above as follows :-
2 2 2
q=x 1 +2.x+5x3+8x1x3-6x2x3
2 2 2 2
= (x 1 + 8XL X 3 + 16x3 ) + 2x2 — 1 1x - 61C2x3

12x
= (x + 4x 3) 2 + (44 - 2x3) - I i4
2
=(x 1 +4x 3 +(4x - l2 2x1 +-9x)—xi- 11x3

21 2 3t2
= (xt + 4x 1) + ( 2x - 3x3 ) - -- x

=4 + ( 1/2)4 —(31/2)4.
where Yi = x 1 +4x3, Y2 - 3x3 , Y.3 x3.
§ 9.03. Real Quadratic Forms.
Let a real quadratic, form q=X'AX be reduced by a real non-singular
2
transformation to the form b 1 y + b2y + ... + by . h t 0. If one or more of the
bi are negative, then there exists a non-singular transformation X = CZ, where
C is obtained from B (see § 9.01 Page 220 of this chapter) by a sequence of row
and coluffn transformations which carries q into
2 2 2 2 2
h 1 z + + ... + hz - h + t Zp+ ... /ig
in which the terms with positive coefficients precede those with negative coeffi-
cients.
Now the non-singular transformation
1,2.....r

carries (i) into the canonical form



222 Matrices
2 2 2 2 2 2
S2+S2+S3+ .,.

Thus, as the product of non-singular transfdnnations is a non-singular

transformation, we have every real quadratic form can be reduced by a
non-singular transformation to the canonical form (ii) above, where p, the
number of positive terms is called the index and r is the rank of the given
quadratic form.
Example. In Ex. 3 (c) Page, 225 the quadratic form q = + 2x - 7x 4
- 4x 1 x 2 + 8x 1 x3 was reduced to q, = 4-
2y + 9y. The non-singular
transformation Yt = Zi. Y = z. z2 carries q 1 into q2 = + 94 - 24 and the
y =
non-singular transformation z1 = s, z2 = s2/3, z 3 s/I2 reduces q 2 to
4
= +S
2 - S1.
Also in Ex. 3 (c) Page 225 we have
Yi = xi Z2+4x1,y2=x2-4x3,y3=x3
or x 1 +2y24 4'3,x2=y2+4y3,x3=y3
or X 1 =ZI + 2z3 + 4 :2, -2 = Z3 + 4z2, x3 =
or = s, + 2 (s/12) + 4 (s 2 13), x 2 = (5 3 /'12) + 4 (s13),
x3 =
or x1 =.c + (4/3) s 2 4 'I2s, x 2 = (4/3) s2 + (1/2) '12s 3 , x 3 (1/3) S2
or X= I 4/3 ''2 S
O 4/3 (1/2)42
o 1/3 0

is the non-singular linear transformation that reduces . q to

The quadratic form is of rank 3 and index 2.
=4- -4.
§ 9.04. Complex Quadratic Forms
Let the complex quadratic form VAX be reduced by a non-singular
transformation to the form b 1 y + b,y + .. + by,, b, ^ 0.
Evidently the non-singular transformation
zi = V(h 1 )y, i I, 2,...,r
j=r+1,r+2,., n
2
carries by + b2v,2 + ... b
2. 2 2
into Z + Z2 + 2
+Z r , (r)
Solved Examples on § 9.01 to § 9.04
*Ex . 1. Write the following quadratic forms in matrix notation
(a) - f 4. _x., +
(h) 4- 2.16 -- 3x + 4x 1, + 6.v 13 -- 8x1x3.
(Ga,1it'iI 95)
Sol. Let
4 + 4.v 1 v 3. X'AX = 1x 1 x, a-)
n2lHl
blj[x2J
V'2-


Quadratic Forms 223

since A is a symmetric matrix

IC.
4 + + 34 = ( a jxj + a 2 +/) I x,1 I
[x2

= ((a l x i + a2x2) 11 + (( 1 -s 1 + h v) x21

+ 2av 1 x b1x
= 4 + 4xx 2 + 34 = a 1 4 -4

h sides, we get
Comparing coefficients of 4, x 1 x 2 and

a1 = I,a2=2,h=3.

From (i) 4 + 4.111x2 + 34 = x'[i 2] X Ans

(h) Let x i - 2x 2 - 3x + 4x 1 x 2 + 6x 1 x 3 -

X'AX = [A1 .t2 x i Ih1. c 1 .r

b 1 b 2
" c2 -2 -.
('I C2 C3 -

A is a symmetric matrix

= [a 1 x l-2 -F c 1 '. Al -4 h 2 x2 + " 2-3 CI Jr + C 2 1Q) + c 3,x 31x

X2

X3

= ((a 11 + c 1 3) x 1 + (L x t + C2A3) x2 F (('iAl + C2-v2 + cv 3) xii

4- - 3s + 4xx7 + 6x 1x 3 -

-f b-v + CV2b 1 x 1x2 + 2c' i x j x i + 2C2x-x3

a1=1b2=_2.c=-3,b1=2,ci=3.C2-4-
, .) ,,
Fro m (ii), we have x - 2x2 3x 3 4V142 4 6 Ix . ,-
=X I 2 3X
Ans.
2 -2 -4
3 --4 -3
2
Ex. 2 (a). 1'n1 the matrix of the quadratic form x+ 2 - 5x - x1x
4x1-X3 - 3xx 1 and verify that it can be written as a matrix product X'AX.
(Garliwa! 94, 9311

So!. Let 4 i 2.v -- 5. - + 4x 2x 3 -

= X'AX - [x 1 x 7 x i i CI b1 c [- -
h1 b2
C I 2 C- -

since A is a s y mmetric matrix

 224 Matrices 182111/14
= [a 1 x 1 + b 1 x 2 + c j x 3 bx 1 + b 2 X2 + c2x3 ('P'I + c,x 2 + c 3 x 3 1 X1

x3
= [(a 1 x 1 + b 1 x2 + c 1 x3) x 1 + (b1x1 + b2,x2 + c2x3 ) x2 + ( c 1 x 1 + c2x2 + c 3x 3 ) x31

4 + 2x - 5x - x 1 x 2 + 4x2x 3 - 3x3x1

= a1 4 + bx + + 2b 1 x 1 x2 + 2c 2 x2 x 3 + 2C1X3X1
Equating coefficients of like terms on both sides, we get
a 1 =l,b 2,c3=-5,2b1=-1,2=4,2c1=_3
From (i), we have the given quadratic form
= X1 X2 x3 1 1 - 1/2 -3/2 Xj
-1/2 2 2 -2
-3/2 2 -5
The required matrix of the given quadratic form
= 1 -1/2 -3/2
-1/2 2 2 Ans.
-3/2 2 -5

Also [x 1 x 2 x 3 1 I - 1/2 -3/2 X1

-1/2 2 2 x2
-3/2 2 -5

= [x 1-x2- 3 I+2+3 -x j +2x 2 -5x 1 J X1

1.3
= (x1 - X2 - x 3 ) x I + (-
12 X I +2X2 + 2X3) X2 + (- + 2x, - 5x3) x3

-
= 4 + 2x - 5x - x 1 x 2 + 4x 2 x 3 - 3x3. = Given quadratic from.

Hence proved.
Ex. 2 (b). Find the matrix of the quadratic form G +Y 2 + 3z2 ±
4xy + 5yz + 6zx and express it is the form G = VAX, where X'= (x,
y, z)
(Garhwal 96)
Sol. Let Grx2+y2+3z2+4xy+5yz+ozx
= VAX = [x y zj a b
1 Ct
b 1 b2 c 2 y ( i)
C t C2 C3 Z
since A is a symmetric matrix
= [ a 1 x±b 1 yc 1z b i x+b2y+ciz CIx+C2y+c3y}

182/11/15 Quadratic Forms 225

= [(a 1 x + 1) 1 y + c 1 z)x + (hx + b,v + C.-2Z) y

+ (c 1 x + c2y + c3z) zJ

x2+y2+3z2+4xy+5y+6u
2
= a 1 x2 + h 2y + c 3z' + 2b 1 xy + 2c 2yz + 2c1u
a 1 = 1,1)2 = 1, c- = 3,21)1 4, 2c 2 = 5, 2c 1 = 6
'i =1.b2=1,c3.h12.c25/2,C13.
From (i), we get G = VAX = [, y z] i 2 3
2 1 5/2 y...(ii)
3 5/2 3z
which gives the required matrix A 1 2 3
2 1 5/2
3 5/2 3 Ans.
and G can be expressed in the form X' A X by (ii).
Ex. 2 (c). Write out in full the quadratic form in x 1 , x 2, x 3 whose matrix
is 2 —2 5
—2 3 0 -
5 04]

So!. Here VAX = [ x1 X 2 '3] 2 -2 5 x1

—2 3 0 x2
5 04x3

[2x - 22 + 5X 3 - + 3x, + 0.x 3 5x 1 + Ox2 + 4x31 X1

X2

x3

= [(2.x 1 - + 5) x i + (- 2'i 3) x 2 + (5x 1 + 4x3 ) x31

= Ii - 2'I + - 12 + 34 + 5x1x3 + 44]

= [2 + 34 + 4x —4x 1 x2 + lOx jx3)
Required quadratic form is
2 2 2
2x I + 3x2 + 4x 3 - 4x 1 x 2 + 10x 1 x3 . Ans.
Ex. 3. Reduce the following by Lagrange's Reduction
(a) X' 12 4 X; (b) X' 0 0 1X
2 6-2 0 1-2
4 —2 18 1 —2 3
2 2
(c) q x1 + 2x 22 - 7x 3 - 4x 1 x 2 + 8x1x3

Sol. (a)X' 12 4 X=[x j x2 x3 ] 1 2 4 X1

2 6 —2 2 6-2 x2
4 —2 18 4 —2 18
 226 Matrices

= [ x 1+ 2x2+ 4 x3 2x,+&2 3 4x 1 -2x2 +18x3 1 X1

X2
= [(x 1 + 2x2 + 4X3) XI + ( ZXj + 6X2 - 2X3 ) x2 + (4x 1 - 2x +
2 18X3) X31

q=4+64+184+4x1x2_4x2x3+8x1x3

(4+4x1(x2+23))+64+184-423
= t4 +4x 1 (x2 + 2x3)+4(x2 + 2x)} + 64
+ 18 4— 4x2x3 —4(x2 + 2x3)2

-44-
= (x1 + 2 (x2 + 23)}2 + 64 + 184 - 4x3 164 - 1
6x3x2
2 2
=(x1+2x 2 +4x3) +2x2 +2x3-2Ox23

= (XI + 2x2 + 4x3) 2 + 2 (4 - 2x3 ) + 24

10x
= (x1 + 2 + 4x3)2 + 2 - 1Oxr3 + 254) -
42

= (x 1 + 2 + 4x3) 2 + 2 (x2 - 5x 3 ) 2 - 484

=y-1-2y-48y.
Ans.
where y 1 =x 1 +ZX2+4X3,y2rX2_5x3,y3X3.

(b)X' 0 0 1 X=[x 1 x x3 ) 0 0
0 1 —2 0 1 —2 x2
1 —2 3 1 —2 31
j X3
XI

ExI0 +x20+x31x10+x2123x1_2X2+3x31xi

X3
= [(x3 ) x 1 + (x2 - 2x3 ) x2 + (x 1 - 2x2 + 3x3 ) x31
2
= [x 2 + 3x3
2 + 2x 1 x3 - 4x2x31

= [4 + 3z + 2z3z 1 - 4 z1z2], Using X1 = Z, X2 = Z1, X3

2 2 2
= [( Z 1 —4z 1 z+4z} —(z2 —2z2z3 )J
2 2 2 2
= [(z1 - 2z2) - (z2 - 2Z2Z3 + Z3) + z31
2 22
= [(z1 - 2z2) - (z2 - Z3) + z31
lip q
22
= (z j -2z2) — (z2 —z3) +Z3
2

=y2_ +
where Yi
Z12Z2X22x3,y2=z2-z3=x3_x1,y3z3x1
(c) q = 4 + 247 - 4 - 4x 1x2 + 8x1x3
 Quadratic Forms 227

= (x- 4x 1 (x 2 - 2X 3 )) + - 7x
=(x_4x1(x2_3)+4(x2_3)2)+7x4(X23)

= {x 1 - 2 (x2 - Z 3)} 2 +2x 2 - 74 —4 (4 - 4x2x3 + 44)

2 2 2
= (x1 - -2 + 4x1 ) - 2 2 + 16x2x3 - 23x3

= (x 1 - 2x 2 + 4x 3 ) 2 —2(4- 8xx 3 + 164) + 94 (Note)

2 2 2'
(X I —Zz 2 +4x3 ) —2(x2 -4x3 ) +9x3
2 2 2
Y 1 2Y2+9Y3 Ans.
wherey 1 = X i- 2 + 4x3. Y2 - 4x3, y =
Exercise on § 9.01 - § 9.02
Ex. 1. Write 2x - 6x 1 x2 + 4 in matrix notation
Ans.X' 2 —3 0 X
• —3 0 0
0 01

Ex. 2. Write out in full the quadratic form in x 1 , x2, x3 whose matrix is
2 —3 1
—3 2 4
1 2 2 2
Ans. 2x fux 2 + 2x 1 x3+ 2x2 + 8xr3 5x3

Ex. 3. Reduce the following by Lagrange's reduction:

X' 01 2X 2 2
1 1 —1 Ans. >12 — Y2 +
2 —1 0
[Hint : Use x 1 = Z3. X2 = z, x3 = z21

§ 9.05. Definite and Semi-definite Forms.

Definition (i) A real non-singular quadratic form q = X'AX, A # 0, in n
variables is known as positive definite if its rank and index are equal. Thus, in
the real field a positive definite quadratic form can be reduced to the form
+ y + ... + y, and for any non-trivial set of values of the x's, q >0.
(ii) A real singluar quadratic form q = X'AX, I A I = 0 is as positive
semi-definite if its rank and index are equal i.e. r = p <n.
Thus in the real field a positive semi-definite qudratic form can be reduced
to the form y + + ... + y, r < n and for any non-trivial set of values of the
x's. q ^! 0.
(iii) A real non-singular quadratic form q = X'AX is known as negative
defnite if its index p = 0 i.e. r = n, p = 0.
Thus in the real field a negative definite form can be reduced to the form
- y - y - ... - y and for any non-trivial set of values of the x's. q <0.
 228 Matrices 999
(iv) A real singular quadratic form q VAX is known as negative
semi-definite if its index p = 0 i.e. r < n, p =0.
- Thus in the real field a negative semi-defnite form can be reduced to the
form - y - Y2 ... and for any non-trivial set of values of the k's, q S 0.
Note 1. If q is negative defnite (semi-definite), then - q is positive definine
(semi-defnite).
Note 2. For positive definite quadratic form, if q = X'AX is positive definite
then I A J>O.
§ 9. Definite and Semi-definite Matrices.
Definition. The matrix A of a real quadratic form q = X AX is known as
definite or semi-definite according as the quadratic form is definite or
semi-definite. Thus
(i) A real symmetric matrix A is positive definite iff there exists a
non-singular matrix C, such that A = C'C.
(ii) A real symmetric matrix of rank r is positive semi-definite iff there exists
a matrix C of rank r, such that A = C'C.
§ 9.07. Principal Minors.
Definition. A minor of matrix A is known as principal if it is obtained by
deleting certain rows and the same numbered columns of the matrix A.
Note 1. The diagonal elements of a principal minor of the matrix A are
diagonal elements of the matrix A.
Note 2. Every symmetric matrix of rank r has at least one principal minor
of order r different from zero.
Note 3. If the matrix A is positive definite, then every principal minor of
the matrix A is positive.
Note 4. If the mattjx A is positive semi-definite, then every principal minor
of the matrix A is non-negative.
Exercises on Chapter IX
• Ex. 1. Reduce the square matrix A into diagonal form and interpret the
result in terms of quadratic form:
A= 6 —2 2
-2 3 - I (Gar/nval 94).
2 —.1 3
Ex. 2. Reduce the quardative form 2x 2 2
1 +x 2
2 -3x 3-82x3_4xj + 12xjx-,
to normal form. (Garhwal 93)


Objective Type - Questions Ch. V to IX

(A) VERY SHORT AND SHORT ANSWER TYPE QUESTIONS
Ch. V Rank and Ad joint of a Matrix
1. Define rank of a matrix. (Purvwiclial 2000) (See § 5 .02 Pages 1-2]
2. Find the rank of the matrix A, where -
A= 1-12-3
4 102
o 304
o 1 0 2 (Meerui 2001)
[Hint: DoasEx. 11 (b)P.28]
Ans.4
3. Reduce the matrix 1 3 4 to normal form.
-2 1 -1 (Kanpur 2001)

Ans.[g]

4. When is a matrix said to be in Echelon farm? (Purvanchal 98)

[See §5-04 Page 361
5. Write down the four normal forms of a.matnx. [See §5 . 03 Page 15]
6. Define adjoint of a matrix. [See § 5-08 Page 43]

7. Show that adjoint of the matrix a 00 is

be 0 0
0 b0 0 ca 0
- OOc 00 ab

8. How will you use the notion of determinant to complete the inverse of a
non-singular square matrix ? [Sec 1k I result (iv) Page 501
9. Find the inverse of the matrix [cos a - sin al
• [sin a cos a] [See Ex. 15 Page 671

Ans. i
r
cos a sin
sin (X cos 1^11

10. If a2 + b2 + c2 + d2 = 1, then show that the inverse of the matrix

a+ib c+id].[a-ib —c--idI
a + ib [See Ex. 18 Page 701
^ -c+id a - ib][c -id
11. Find the rank of an x n matrix, every element of which is unity.
Ans.l.
[Hint: See Ex. 9 Page 96]
12. A, B, P and Q are matrices such that adj. B = A, P I
Q = l then
adj (Q 1 B P 1 ) = PAQ. (Kanpur 2001)
Ch.V1 Solution of Linear Equations
13. Express in matrix form the system of equations:
9x+7v+3z=6.5x+y+4z=l;6X+8Y+2Z 4 - [See Ex. iPage 106
182/11/00
 2 Matrices
14. Define a homogeneous linear equation. [See § 6 . 10 Page 1441
15. Solve the simultaneous equations given below:
x+y+2z=3;2x+2y+3z=7; 3x—y+2z= 1,2x—y--z=2
(Kanpur 2001) A ns. x = 2, y = 3, z = I
Ch. VII Characteristics Equations of a Matrix
16. What do you understand by the characteristic equatioii of the matrix A7
ESee7•02 (iii) Page 1601
17. What is eigen value problem 7 [See § 7•02 (v) Page 160]
18. Obtain the characteristic equation of the matrix I 0 2
021
20 3
(Meerut 2001) Ans. - 6X + 7X .+ 2 = 0
19. State Cayèly-Hamilton's Theorem. -
20. Find latent vectors of the matrix a h g
ObO
0 0 c [See Ex 25(a) Page 2001
Ch. VIII Llneat Dependence of Vectors
21. Define linearly dependent and linearly independent set of vectors.
(Kanpur 2001) [See § 8-03 Page 2111
22. Show that the set of vectors V 1 = (1,2,3), V 2 = ( 1 , 0, I) and V3 =
(0,A, 0) are linearly independent. [See Ex. 1 Page 2121
23. Find a linear relation, if any, between the linear forms of the following
systemf1 =x+y+z,f2=y-2z,f3=2x+3y.
Ans. 2f f2 j'3 [See Ex, 3 Page 217]
Ch. IX Quadratic Forms
24. Define a quadratic form. [See § 9-01 Page 220]
25. What do you understand by the rank of a quadratic form.
- [See § 9 .01 Page 2201
26. Write the quadratic form corresponding to the matrix

L -
(Kanpur2 01) Ans. - 4x] x3
(B) OBJECTIVE TYPE QUESTIONS
(I) MULTIPLE CHOICE TYPE:
Select (I). (ii), (iii) or (iv) whichever is correct
Ch. V. Rank and Adjoint of a Martlx
1. The rank of the matrix' 4 5 is
456
- 234
 Objective Type Questions Ch. V to IX 3

(ii) I (iii) 2 (iv) none of these

(i) 0
(Kanpur 2001)

2. The rank of the matrix 0 2is

046
069
(1) 0 (ii) I iii) 2 (iv) 3
3.lfA be matrix, which one of the following is number:
(iii) rankA (iv) none of these
(i) A 1 (ii) adj A
4. 11 A' be the transpose of the matrix A, then
(i) rank A' > rank A (ii) çank A' = rank A
(iii) rank A' < rank A (iv) none of these
5. If by a series of elementary transformations an n-rowed square matrix A

is reduced to the form [Ir the rank of A is

],
(i) n + r (ii) r (iii) n - r (iv) n
6. The rank of the matrix I 1 1 is
234
4 9 16
(1)0 (ii) I
[1.
(111)2
2 -I 3 15] is
7. The rank of the matnx [3 2 0 21
(iv) 3

(i) I (ii) 2 (uI) 3 (iv) none of these

8. The rank of the matrix 1 6 5 5 is
3 18 15 3
1651

(I) 4 (ii) 3 (iii) 2 (iv) -1

9. The value of a for which the matrix 2 3 1 is singular. is
3 2 -1
al 3

(I) - 2 (ii) - 1 (iii) 1 (iv) .2

10. The necessary and sufficient condition that a square matrix may
possess an inverse is that it be
(I) singular (ii) triangular
(iii) non-singular (iv) none of these
11. If A is non-singular matrix, then (A - is
(1) 1 (ii) A A (iv) AA''
(Kanpur 2001)

12. If a non-singular matrix A is c.'mmetric, then A 1 is

(i) skew-symmetric (ii) Hthnitian
(iii) diaeonal (iv) symmetric
 Matrices
[Hint : See 5. Il Th. V Page 77
Ch. VI Solution of Linear Euations
13. The system of equations
x+2y +z =2 ,3x+5y+5z4,+4 y +3z=3 has a
(i) unique solution (ii) infinite solution
(iii) trivial solution (iv) none of these
[See Ex 5(a) P. 1 101
14. The system of equations
3x_y +z0,_l5x+6y_5z0,5,_2y+2z=0 has a
(1) unique solution (ii) trivial solution
(iii) infinite solution (iv) none of these
15. The t beorem every square matrix satisfies its characteristic equation'
is named after -
(i) Cramer (ii) Hamilton (iii) Newton (iv) none of them
Ch. VII Characteristic Equations of a Matrix
16. If A bt any matrix and I the identity matrix, then A - XI is known as
(i) characteristic polynomial of A (ii) spectrum of A
(lii) characteristic matrix of A (iv) none of these
(Ii) TRUE OR FALSE TYPE:
- Write 'T' or "F" aécording as the following statements are true or false:
Ch. V Rank and Adjoirit of a Matrix
1. The rank of A = 1 00is2.
010 -
O 0 1] . (Agra 90)
• 2. All equivalent matrices have the same rank.
3. If every minor of order p of a matrix A is zero, then every minor of order
higher than p is not necessarily zero.
4. • If at least one minor of order r of the matrix A is not equal to zero, then
rank of A > r.
2 3
5. The rank of = 2 5 8is2.
4 10 18
(See Ex 1(a) P. 21
2 3 and - 6
2 are equivalent.
6. The matrices2 5
3 7 5 —2 —9 14
4 —2 —4 8

7. If the elements of a row of a matrix are multiplied by a non-zero

number, then the rank of the matrix remains unaffected.
8.The rank of a matrix is equal to the rank of the transposed matrix.
 Objective Type Quc'tion\ Ch. V to IX 5

9. The rank of the product mail ix All of two maricc' A and B is )ss than
the rank of either of the matrices A and B.
10. If A and B are two ri x n matrices, then
Adj (AB) (Adj B) • (Adj A)
0 0 ii
iI.lfA= () I 0thenA'=,A
iooj
12. The necessary and sufficient condition that a square matrix may
Pos s ess an inverse is that it be singular.
13. The inverse of transpose of a matrix is not the transpose of the inverse.
14. The inverse of the inverse of a matrix is the matrix itself.
15. If A is any square matrix, then (adj. A) t = adj (A1).
16. Inverse of a matrix A exists if A is singular.
17. If A is a matrix of order n x n, then A' is also of the same order.
Ch. VI Solution of Linear Equations
18. A consistent system of equations has no solution.
I9.A consistent system of equations has either one solution or infinitely
many solutions.
20. A system of m linear equations in n unknowns given by AX = K is
consistent if the matrix A and the augmented matrix A' of the system have the
same rank.
Ch. VU Characteistic Equation of a Matrix
21. The matrix A - XI is known as the characteristics matrix of A, when I
is the identity matrix.
22. Every square matrix satisfies its characteristic equation.
23. The characteristic roots of a Hermitian matrix are either purely
imaginary or zero.
24. The characteristic roots of real skew-symmetric matrix are purely
imaginary or zero.
25. The characteristic roots of a unitary matrix are of unit modulus.
Ch. VIII Linear Dependence of Vectors
26. The set of vectors V 1 = { I. 2, 3), V 2 = ( I, 0. I] and V 3 = 10. 1,0) are
lkiearly dependent.
27. If there be n linearly dependent vectors, then none of these can be
expressed as a linear combination of the remaining ones.
ISee Th. I § 804 Page 2141
Ch. IX Quadratic Forms

28.4+ 24 + - 8x 1 x2 + ox 1 .x 3 is a quadratic form in the variables

- T I' .'r2, x1.
 6 Mitrjces
(III) FILL IN THE BLANKS TYPE:
Fill in the blanks in the following -
Ch. V Rank and Adjoint of Matrix
1. Rank ottho•nuIl martix is
(Kanpur 2001)
2. Rank of the matrix I 01...................
1
• (Kanpur2001)
3. if a matrix A of order m x it can be expressed as{ ], then rank o A
is
(Meerut 2001)
4. All ..............matrices have the same rank.
.. If every minor of order p of a matrix A is zero, then every minor of order
..........p is definitely zero.
6. if a matrix A does not possess any minor of .
order(r+ 1) then rank of A

7.The rank of the matrix [ I jis..............

Ill
S. The rank of the matrix 2 2 2 is ..............
333
9. The rank of a matrix is .........to the rank of the transposed matrix.
10. The adjoint of the matrix [g ] ' ...............

11.Jf Abe a square matrix of order n, then
A (adj A) = (adj A) A = ( Meerut 2001)
12. If A and B are two n x n matrices, then
(Adj B). (Adj A) = [See § 509 Th. Ill P. 501
13. if A be an n x n matrix and I A 0, then
I Adi A I =.... [See 5•O9 Th. UP.5oJ
14. A and B are two matrices such that AB = I, then
adj B
(Kanpur 2001)
(Hint: We know B_ and here B' = A. AB =

.adjB=A.IBI]
15. If A is a non-singular matrix, then (A 1
) = ........ where A' is the
inverse of A.
16. The inverse of a matrix is ..............
17. A singular matrix has no ..............
IS. A matrx when multiplied by its inverse given the ...............matrix.


Objective Type Questions Ch. V to IX 7

Ch. VI Solution of Linear Equations

19. Inconsistent equations have ...............solution.
20. A consistant system of equations has either oe solution or .........
sol u lions.
21. A homogeneous system of n linear equations in a unknowns, whose
determinants of Coefficients does not vanish, has only the .........solution.
22. A set of simultaneous homogeneous equations expressed in matrix
form as AX = 0 has non-trivial solution if ............. (Kanpur 2001)

23. A system of m non-homogeneous linear equation AX = B in n

unknowns is called ...........iff ranks of A and [A, B) are equal. (Meerut 2001)
[See 607P. 119,606P. 1181
Ch. VII Characteristic Equation of a Matrix
24. The set of all cigcn values of the matrix A is called the .............of A.
[See § 7 . 02 (iv) Page 160]
25. The determinant I A - XI I is called the characteristic ............of the
matrix A. when I is the identity matrix.
26. Every square matrix .....................its characteristic equation.
(Meerut 2001)

27. Characteristic roots of skew-Hermitian matrix are either zero or ...........

(Kanpur 2001)
28. All the characteristic roots of a real symmetric matrix are .............
29. The characteristic roots of a Hermitian matrix are all ....................
30. The characteristic roots of an orthogonal matrix are of .........modulus.
31. Two .............matrices have the same characteristic roots. -
- (Kanpur 2001)
[See § 705Th. 1 Page 1671
Ch. VIII Linear Dependence of Vectors
32. The set of vectors V 1 = 1, 2, 3), V 2 (1,0, 1) and V 3 (0, 1,0) are
linearly [See Ex. 10. 2121
33: The set of vectors X1=[2,3,l,— 11, X 2 =[2, 3, 1, 7 21 and
X3 = [4,6.2,— 31 is linearly [See Ex. 1 Page 213]
Ch. IX Quadratic Forms
34. The quadratic form in X 1 , X2, x3 of the matrix 2 —2 5 is
—2 30
5 04

ANSWERS TO OBJECTIVE TYPE QUESTIONS

(I) Multiple Choice Type:

2. . (ii) 3. (iii); 4. (ii); 5. (ii); 6. (iv



S
UF .) 1X2111100
7. (t'
12. (i
13.

(H) True & False Type:

1. F 2. T, 3. F: 6. F 7. 1.
8. T. 9. T; 10. F, II. 1. 42. F: 13. I 44. r
15. T; 16. F: 17. T. 18. F, 49. 1. 20. T. 24. F
22. T; 23. F. 24. T. 25. J. 26. F. 27. F: 2. I
(Ill) Fill in the blanks Type:
1.0:2. 2; 3. r; 4. cquivalcnt: 5. liihcr Fh..in: 6. L. 7. I: 8. 1.9. ctiitI.
0
a
11.1 Al • 1,, 12. Adj (AR): 13.1 A I' :14. A • B :15. A; 16. unique:
17.inverse; 18. unit; 19. no; 20. inlinitely many. 24. lri'ial:
22. the rank of A < number of unknowns: 23. conslstcnr: 24. speciruili;
25. polynominaf 26. sati s fies, 27. purely imaginary: 28. real: 29. real:
30. unit; 31. mutually reciprocal; 32. independent; 33. dependent:
44-
34. 24 + 3x F' 4.p r, + lUx 1 .v