Linear system: A finite set of linear equations in the variables x1,x2,...
xnis called a system of
linear equations or linear system.
Introduction to systems of linear equations:
An arbitrary system of m linear equations in n unknowns of form
a 11 x 1 +a12 x 2 +…+ a1 n x n=d 1
a 21 x1 + a22 x 2+ …+a2 n x n=d 2
a m 1 x 1+ am2 x2 +…+ amn x n =d m
Where x1,x2,...xn are unknowns and a’s and d’s denote constants, is known as non-homogeneous
system provided d ’ s not all zero (at least one of them is non-zero).
This system can be expressed as AX=b
where
a11 a12 … a1 n x1 d1
a
⋮
[
A= 21
a22 … a2 n
⋮ ⋮ ⋮
a m 1 am 2 … a mn
, X=
] [] [] x2
⋮
xn
, b=
d2
⋮
dm
A is called matrix of coefficients.
The double subscripted parameter symbol a ij represents the coefficients appearing in the ith
equation and attached to the jth variable. For example, a 21 is the coefficient appearing in the
second equation, attached to the variable x 1 .The parameter d i which is unattached to any
variable, on the other hand, represents the constant term in the ith equation. For instance, d 1 is
the constant term in the first equation. All subscripts are therefore keyed to the specific locations
of the variables and parameters.
Examples of Linear Models from Economics
Example 1 (National-income model)
Y =C + I 0+ G0 ,
C=a+ bY ( a>0 , 0<b< 1 )
Where Y and C stand for the endogenous (dependent) variables national income and (planned)
consumption expenditure, respectively, and I 0 and G 0 represent the exogenously (independent)
determined investment and government expenditures.
Example 2 (National-income model)
Y =C + I 0+ G0 ,
C=a+ b(Y −T ) ( a>0 , 0<b< 1 ) [T: taxes]
T =d +tY , ( d > 0 ,0< t<1 ) [t: income tax rate]
whereY , C and T are endogenous variables, and I 0∧G0 represent the exogenous variables.
Example 3 (National-income model)
Y =C + I 0+ G ,
C=a+ b ( Y −T 0 ) ( a>0 , 0<b <1 )
G=gY , ( 0< g<1 ) [G: Government expenditure]
whereY , C and G are endogenous variables, and I 0 ,∧T 0 represent the exogenous variables.
Linear dependence of two vectors:
Two vectors v1 and v 2 are linearly dependent if one vector is multiple of other; otherwise they
are linearly independent. For example
1. The two row vectors v ' 1=[5 12] and v ' 2=[10 24] are linearly dependent because
2 v ' 1=2 [ 5 12 ]= [ 10 24 ]=v ' 2 .
2. The two row vectors v ' 1=[2 6 ] and v ' 2=[−4 12] are linearly independent because
one vector is not a multiple of the other vector.
3. The two column vectors v1 = [−24] and v =[−84 ] are linearly dependent because
2
−2 v 1=−2 −2 = 4 =v 2 .
4 [ ][ ]
−8
4. The two column vectors v1 = [ 37] and v =[ 149 ] are linearly independent because one
2
column vector is not a multiple of the other column vector.
Example 1: Are the rows linearly independent in following coefficient matrix
A= 3 4 ? Hence determine it singular or nonsingular?
[ ]
6 8
Solution: We can write
A= [ 36 48 ]=[ vv '' ]
1
2
v ' 1=[ 3 4 ] and v ' 2=[ 6 8 ].
The two row vectors are linearly dependent because
2 v ' 1=2 [ 3 4 ] =[ 6 8 ] =v ' 2 .
Hence matrix A is singular.
Example 2: Are the rows linearly independent in following coefficient matrix
A= [ 03 42 ]? Hence determine if it is singular or nonsingular?
Solution: We can write
0 4 v'
A= [ ][ ]
3 2
= 1
v '2
v ' 1=[ 0 4 ]and v ' 2=[ 3 2 ] .
The two row vectors are linearly independent hence matrix A is nonsingular.
Determinantal Criterion for linear independence:
Ann × n matrix coefficient A (where AX=d, a linear system) can be considered as an
ordered set of row vectors, i.e., as a column vector whose elements are themselves row
vetors:
a11 a 12 ⋯ a1 n v'1
a
⋮
[
A= 21
a 22
⋮
an 1 an 2
⋯ a2 n
⋮ ⋮
⋯ ann ][ ]
v'
= 2
⋮
v'n
Where
v ' 1=[ a11 a12 ⋯ a 1 n ]
v ' 2=[ a21 a 22 ⋯ a2 n ]
⋯ ⋯⋯⋯ ⋯
v ' n=[ an 1 an 2 ⋯ a nn ].
1. | A|≠ 0if and only if row vectors v ' 1 , v ' 2 , ⋯ v ' n are linearly independent.
2. | A|=0 if and only if row vectors v ' 1 , v ' 2 , ⋯ v ' n are linearly dependent
3. For a square matrix there is row independence if and only if there is column
independence.
Example 2: Are the rows linearly independent in following coefficient matrix
A= [ 03 42 ]? Hence determine if singular or nonsingular?
Solution: We can write
A= [ 03 42 ]=[ vv '' ]
1
v ' 1=[ 0 4 ]and v ' 2=[ 3 2 ] .
The two row vectors are linearly independent hence matrix A is nonsingular.
Determinantal Criterion for linear independence:
Ann × n matrix coefficient A (where AX=d, a linear system) can be considered as an
ordered set of row vectors, i.e., as a column vector whose elements are themselves row
vetors:
a11 a 12 ⋯ a1 n v'1
a
⋮
[
A= 21
a 22
⋮
an 1 an 2
⋯ a2 n
⋮ ⋮
⋯ ann ][ ]
v'
= 2
⋮
v'n
Where
v ' 1=[ a11 a12 ⋯ a 1 n ]
v ' 2=[ a21 a 22 ⋯ a2 n ]
⋯ ⋯⋯⋯ ⋯
v ' n=[ an 1 an 2 ⋯ a nn ].
1. | A|≠ 0if and only if row vectors v ' 1 , v ' 2 , ⋯ v ' n are linearly independent.
2. | A|=0 if and only if row vectors v ' 1 , v ' 2 , ⋯ v ' n are linearly dependent
3. For a square matrix there is row independence if and only if there is column
independence.
Example 3: Check row vectors are linearly independent or not for the following 3×3 matrix
1 3 2 v'1
[ ][ ]
A= 4 1 3 = v ' 2
2 5 2 v'3
Solution: The determinant of A is
| A|=17 ≠ 0
Hence row vectors are linearly independent.
Determinantal Criterion for nonsingularity:
A square matrix A is said to be non-singular if and only if its determinant value is non-zero. A
square matrix A is said to be singular if and only if |A| = 0.
TEST OF NONSINGULARITY USING DETERMINANTS
Given a linear equation system
AX=d
where A is an n x n coefficient ¿
| A|≠ 0 ↔ t h ere is row ( column ) independence
¿¿
↔ A is nonsingular
↔ A−1 ∃
↔ a unique solution ∃
Example: Does the equation system given below have a unique solution?
7 x−3 y−3 z=7
2 x+ 4 y + z=0
−2 y−z=2
Solution:
7 −3 −3
|
| A|= 2 4
|
1 =7 (−2 )+ 3 (−2 )−3 (−4 )=−8≠ 0
0 −2 −1
Hence a unique solution does exist.
Rank of a Matrix
If determinant of an n x n matrix A is nonzero, then its rank = n. IF determinant of A is
zero then its rank will be less than n!!!
Example: Test whether the following matrix is nonsingular. What can you say about its
rank?
−4 9 5
[
B= 3 0 1
10 8 6 ]
Solution: Here determinant of B is
−4 9 5
| |
|B|= 3 0 1 =−4 (−8 )−9 ( 18−10 )+5 ( 24 )=32−72+120 ≠ 0
10 8 6
Hence B is nonsingular.
Also Rank(B)=3
12 1 3
(
Find the rank of A = −4 0 −1
4 5 1 )
Since det(A) = 0 hence rank(A)≠3. It could be 1 or 2.
1. Write the following systems of linear equations in matrix form AX = b.
Find the determinant of A. Is A singular or nonsingular. Does a unique solution exist?
i)2 x+ y =5 ; 3 x −4 y =2
AX =b where A= (23 1
−4
5
) () ()
; b= ; X =
2
x
y
Det ( A )=−8−3=−11 ≠ 0 Hence A is non−singular .
A unique solution ∃
ii) 3 x+ 2 y =5; 6 x + 4 y=10
3 2 5 x
AX =b where A= ( ) ( ) ()
6 4
; b=
10
;X=
y
Det ( A )=12−12=0. Hence A is singular .
A unique solution does not exist .
Given the national income model
Y =C + I 0+ G0 ,
C=a+ bY ( a>0 , 0<b< 1 )
whereY and C stand for endogenous variables are national income and
consumption expenditure, respectively, and I 0∧G 0 represent the
exogenously determined investment and government expenditure.
(a)Rewrite national-income model in the AX=d (with Y as the first
variable in vector X).
(b) Are the rows linearly independent in the coefficient matrix A.
Hence determine if A is singular or nonsingular?
(c)Test whether the coefficient matrix A is non-singular using
determinant criterion.
Solution: (a) System can be rewritten as
Y −C=I 0 +G 0
−bY +C=a
[−b1 −11 ][ YC ]=[ I +Ga ]
0 0
where
I +G
A= 1 −1 , X= Y , d = 0 0
−b 1[ C ] [] [a ]
1 −1 v'
(b) A=[
−b 1 ] =
[v ' ] 1
where v ' 1=[ 1 −1 ] and v ' 2=[ −b 1 ]. The two row vectors are linearly
independent because one vector is not multiple of other vector.
Hence A is a non-singular matrix.
(c)| A|= −b|1 −1
1 |
=1−b≠ 0 . Hence A is a non-singular matrix.
Given the following model
Y =C + I 0+ G0 ,
C=a+ b(Y −T ) ( a>0 , 0<b< 1 ) [T: taxes]
T =d +tY , ( d > 0 ,0< t<1 ) [t: income tax rate]
whereY ,C and T are endogenous variables, and I 0∧G 0 represent the
exogenous variables.
(a)Rewrite national-income model in the AX=d (with Y as the first
variable in vector X).
(b) Test whether the coefficient matrix A is non-singular.
Solution: (a) System can be rewritten as
Y −C=I 0 +G 0 ,
−bY +C +bT =a
−ty+T =d
1 −1 0 Y I 0+ G 0
[ −b 1 b C
−t 0 1 T ][ ] [ ]
= a
d
where
1 −1 0 Y I 0 +G 0
[
−t 0 1 T ] [] [ ]
A= −b 1 b , X= C , d= a
d
1 −1 0
(b) | |
| A|= −b 1 b
−t 0 1
Expanding along first row yields
| A|=1 1 b −(−1 ) −b b +0=1−b +bt
|0 1| |−t 1|
| A|=1−b+bt ≠ 0. Hence A is non-singular matrix.
