1.

1 Introduction to Systems of Linear Equations

Lecture 1 a linear equation in n variables:

a1 x1+a2 x2 + a3 x3 + ! + an xn = b
Systems of Linear Equations a1,a2,a3,…,an, b: real number

- Matrices a1: leading coefficient – pivot if non-zero

x1: leading variable or unknown
n Notes:
1.1 Introduction to Systems of Linear Equations (1) Linear equations have no products or roots of variables and
1.2 Gaussian Elimination and Gauss-Jordan Elimination
no variables involved in trigonometric, exponential, or
logarithmic functions.

(2) Variables appear only to the first power.

Ex 1: (Linear or Nonlinear) n a solution of a linear equation in n variables:

1 a1 x1 + a2 x2 + a3 x3 + ! + an xn = b
Linear (a) 3x + 2 y = 7 ( b) x + y − π z = 2 Linear
2
x1 = s1 , x2 = s2 , x3 = s3 , !, xn = sn
π
Linear (c) x1 − 2 x2 + 10 x3 + x4 = 0 (d ) (sin ) x1 − 4 x2 = e
2
Linear such a1s1 + a2 s2 + a3s3 + ! + an sn = b
2
that
Exponential
( f ) ex − 2 y = 4 Solution set:
Nonlinear (e) xy + z = 2 Nonlinear
the set of all solutions of a linear equation
not the first power
1 1
Nonlinear ( g ) sinx1 + 2 x2 − 3x3 = 0 (h) + = 4 Nonlinear
x y
trigonometric functions not the first power

1
Ex 2 (Parametric representation of a solution set) a system of m linear equations in n variables:
x1 + 2 x2 = 4 a11 x1 + a12 x2 + a13 x3 + ! + a1n xn = b1
a21 x1 + a22 x2 + a23 x3 + ! + a 2 n xn = b2
a solution: (2, 1), i.e. x1 = 2, x2 = 1
a31 x1 + a32 x2 + a33 x3 + ! + a3n xn = b3
If you solve for x1 in terms of x2, you obtain "
x1 = 4 − 2x2 , am1 x1 + am 2 x2 + am3 x3 + ! + amn xn = bm

By letting x2 = t you can represent the solution set as n Consistent:

x1 = 4 − 2t A system of linear equations has at least one solution.

And the solutions are {( 4 − 2t , t ) | t ∈ R} or {( s, 2 − 12 s) | s ∈ R} n Inconsistent:

A system of linear equations has no solution.

Notes: Ex 4: (Solution of a system of linear equations)

(1) x + y = 3
Every system of linear equations has either
x − y = −1 exactly one solution
(1) exactly one solution, two intersecting lines
(2) infinitely many solutions, or (2) x + y = 3
(3) no solution. 2x + 2 y = 6 inifinite number
two coincident lines
(3) x + y = 3
x + y = 1 no solution
two parallel lines

2
Ex 5: (Using back substitution to solve a system in row echelon form) Ex 6: (Using back substitution to solve a system in row echelon form)
x − 2 y + 3z = 9 (1)
x − 2y = 5 (1) y + 3z = 5 (2)
y = −2 (2) z = 2 (3)
Sol: By substituting y = −2 into (1), you obtain Sol: Substitute z = 2 into (2)
y + 3(2) = 5
x − 2( −2) = 5
y = −1
x = 1
and substitute y = −1 and z = 2 into (1)
The system has exactly one solution: x = 1, y = −2
x − 2( −1) + 3(2) = 9
x = 1
The system has exactly one solution:
x = 1, y = −1, z = 2

Equivalent: Ex 7: Solve a system of linear equations (consistent system)

Two systems of linear equations are called equivalent
x − 2 y + 3z = 9 (1)
if they have precisely the same solution set. − x + 3y = −4 (2)
2 x − 5 y + 5z = 17 (3)
n Notes:
Sol: (1) + (2) → (2)
Eliminations: use following operations on a system of
linear equations to produce an equivalent system. x − 2 y + 3z = 9
y + 3z = 5 (4)
(1) Interchange two equations. 2 x − 5 y + 5z = 17
(2) Multiply an equation by a nonzero constant. (1) × ( −2) + (3) → (3)
x − 2 y + 3z = 9
(3) Add a multiple of an equation to another equation.
y + 3z = 5
− y − z = −1 (5)

3
 Ex 8: Solve a system of linear equations (inconsistent system)
(4) + (5) → (5)
x − 2 y + 3z = 9 x1 − 3x2 + x3 = 1 (1)
y + 3z = 5 2 x1 − x2 − 2 x3 = 2 (2)
2z = 4 (6) x1 + 2 x2 − 3x3 = − 1 (3)
(6) × → (6)
1
2 Sol: (1) × ( −2) + (2) → (2)
x − 2 y + 3z = 9 (1) × ( −1) + (3) → (3)
y + 3z = 5
x1 − 3x2 + x3 = 1
z = 2
5 x 2 − 4 x3 = 0 ( 4)
So the solution is x = 1, y = −1, z = 2 (only one solution) 5 x 2 − 4 x3 = − 2 (5)

Ex 9: Solve a system of linear equations (infinitely many solutions)

( 4) × ( −1) + (5) → (5)
x 2 − x3 = 0 (1)
x1 − 3x2 + x3 = 1 x1 − 3 x3 = − 1 (2)
5 x 2 − 4 x3 = 0 − x1 + 3 x2 = 1 (3)
0 = −2 (a false statement)
Sol: (1) ↔ ( 2)
So the system has no solution (an inconsistent system).
x1 − 3 x3 = − 1 (1)
x 2 − x3 = 0 (2)
− x1 + 3 x2 = 1 (3)

(1) + (3) → (3)

x1 − 3 x3 = − 1
x 2 − x3 = 0
3 x 2 − 3 x3 = 0 (4)

4
 1.2 Gaussian Elimination and Gauss-Jordan Elimination
x1 − 3x3 = −1 mxn matrix:
x2 − x3 = 0 ⎡ a11 a12 a13 ! a1n ⎤
⎢ a21 a22 a23 ! a2 n ⎥
⇒ x2 = x3 , x1 = −1 + 3x3 ⎢ a31 a32 a33 ! a3n ⎥ m rows
⎢ " ⎥
let x3 = t ⎢ ⎥
⎣am1 am 2 am 3 ! amn ⎦
then x1 = 3t − 1, n columns
Notes:
x2 = t , t∈R
n

(1) Every entry aij in a matrix is a number.

x3 = t ,
(2) A matrix with m rows and n columns is said to be of size mxn .
So this system has infinitely many solutions.
(3) If m = n , then the matrix is called square of order n.
(4) For a square matrix, the entries a11, a22, …, ann are called
the main diagonal entries.

Ex 1: Matrix Size a system of m equations in n variables:

[2] 1× 1 a11 x1 + a12 x2 + a13 x3 + … + a1n xn = b1
a21 x1 + a22 x2 + a23 x3 + … + a 2 n xn = b2
⎡0 0⎤ 2×2
⎢⎣0 0⎥⎦ a31 x1 + a32 x2 + a33 x3 + … + a3n xn = b3
!
⎡1 − 3 0 1 ⎤
⎢⎣ 1× 4 am1 x1 + am 2 x2 + am3 x3 + … + amn xn = bm
2 ⎥⎦
⎡ e π ⎤ Matrix form: Ax = b
⎢2 2⎥ 3× 2
⎣⎢ − 7 4 ⎦⎥ ⎡ a11 a12 a13 ! a1n ⎤
⎡ x1 ⎤ ⎡ b1 ⎤
⎢ a21 a22 a23 ! a2 n ⎥
Note: ⎢ ⎥ ⎢ ⎥
A = ⎢ a31 a33 ! a3n ⎥ x = ⎢ x2 ⎥ b = ⎢ b2 ⎥
n
a32
One very common use of matrices is to represent a system ⎢ ⎥ ! !
" ⎢⎣ xn ⎥⎦ ⎢⎣bm ⎥⎦
⎢ ⎥
of linear equations. ⎣ a m1 am 2 am 3 ! amn ⎦

5
Augmented matrix: Elementary row operation:

⎡ a11 a12 a13 ! a1n b1 ⎤ (1) Interchange two rows. rij : Ri ↔ R j

⎢ a21 a22 a23 ! a2 n b2 ⎥ ri( k ) : (k ) Ri → Ri
⎢ a31 (2) Multiply a row by a nonzero constant.
a32 a33 ! a3n b3 ⎥ = [ A b]
⎢ " ⎥ (3) Add a multiple of a row to another row. rij( k ) : (k ) Ri + R j → R j
⎢ ⎥
⎣am1 am 2 am 3 ! amn bm ⎦
n Row equivalent:
n Coefficient matrix:
Two matrices are said to be row equivalent if one can be obtained
⎡ a11 a12 a13 ! a1n ⎤ from the other by a finite sequence of elementary row operation.
⎢ a21 a22 a23 ! a2 n ⎥
⎢ a31 a32 a33 ! a3n ⎥ = A
⎢ " ⎥
⎢ ⎥
⎣ a m1 am 2 am 3 ! amn ⎦

Ex 2: (Elementary row operation) Ex 3: Using elementary row operations to solve a system

Linear System Associated Elementary
⎡ 0 1 3 4⎤ r12 ⎡ − 1 2 0 3⎤ Augemented Matrix Row Operation
⎢ − 1 2 0 3⎥ ⎢ 0 1 3 4⎥ x − 2 y + 3z = 9
⎢⎣ 2 − 3 4 1⎥⎦ ⎢⎣ 2 − 3 4 1⎥⎦ ⎡ 1 −2 3 9⎤
− x + 3y = −4 ⎢− 1 3 0 − 4⎥
⎢ ⎥
⎡2 − 4 6 − 2⎤ (1)
⎡ 1 − 2 3 − 1⎤ 2 x − 5 y + 5z = 17 ⎢⎣ 2 − 5 5 17 ⎥⎦
r1 2
⎢ 1 3 − 3 0⎥ ⎢ 1 3 − 3 0⎥
⎢⎣ 5 − 2 1 2⎥⎦ ⎢⎣5 − 2 1 2⎥⎦ x − 2 y + 3z = 9 ⎡ 1 − 2 3 9⎤
y + 3z = 5 ⎢0 1 3 5⎥ r12(1) : (1) R1 + R2 → R2
⎢ ⎥
⎡ 1 2 − 4 3⎤ r13( −2 ) ⎡ 1 2 − 4 3⎤ 2 x − 5 y + 5 z = 17
⎢ 0 3 − 2 − 1⎥ ⎢0 3 − 2 − 1⎥ ⎣⎢2 − 5 5 17 ⎦⎥
⎣⎢2 1 5 − 2⎦⎥ ⎣⎢0 − 3 13 − 8⎦⎥ x − 2 y + 3z = 9 ⎡ 1 − 2 3 9⎤
⎢0 1 3 5⎥⎥ r13( −2 ) : (−2) R1 + R3 → R3
y + 3z = 5 ⎢
− y − z = −1 ⎣⎢0 − 1 − 1 − 1⎦⎥

6
 Associated Elementary
Row Operation
Row-echelon form: (1, 2, 3)
Linear System Augemented Matrix
n Reduced row-echelon form: (1, 2, 3, 4)
x − 2 y + 3z = 9 ⎡ 1 − 2 3 9⎤ r23(1) : (1) R2 + R3 → R3
⎢0 (1) All row consisting entirely of zeros occur at the bottom
y + 3z = 5 1 3 5⎥
⎢ ⎥
2z = 4 ⎢⎣0 0 2 4⎥⎦ of the matrix.
(2) For each row that does not consist entirely of zeros,
x − 2 y + 3z = 9 ⎡ 1 − 2 3 9⎤ 1
1
⎢0
( ) the first nonzero entry is 1 (called a leading 1).
1 3 5⎥ r : ( ) R3 → R3
2
y + 3z = 5 ⎢ ⎥
3
2
(3) For two successive (nonzero) rows, the leading 1 in the higher
z = 2 ⎣⎢0 0 1 2⎥⎦
row is farther to the left than the leading 1 in the lower row.
x = 1 (4) Every column that has a leading 1 has zeros in every position
y = −1
above and below its leading 1.
z = 2

Ex 4: (Row-echelon form or reduced row-echelon form) Gaussian elimination:

⎡ 1 2 − 1 4⎤ (row - echelon ⎡0 1 0 5⎤ (reduced row - The procedure for reducing a matrix to a row-echelon form.
⎢0 1 0 3⎥ ⎢0 0 1 3⎥ echelon form) Gauss-Jordan elimination:
⎣⎢0 0 1 − 2⎦⎥ form) ⎣⎢0 0 0 0⎦⎥
n

The procedure for reducing a matrix to a reduced row-echelon

⎡ 1 − 5 2 − 1 3⎤ ⎡1 0 0 − 1⎤
⎢0 1 0 2⎥ (reduced row - form.
⎢0 0 1 3 − 2⎥ (row - echelon
⎢0 0 0 1 4⎥ form) ⎢0 0 1 3⎥ echelon form) n Notes:
⎣⎢0 0 0 0 1⎦⎥ ⎣⎢0 0 0 0⎦⎥
(1) Every matrix has an unique reduced row echelon
⎡ 1 2 − 3 4⎤ ⎡ 1 2 − 1 2⎤ form.
⎢0 2 1 − 1⎥ ⎢ 0 0 0 0⎥ (2) A row-echelon form of a given matrix is not unique.
⎣⎢0 0 1 − 3⎦⎥ ⎣⎢0 1 2 − 4⎦⎥
(Different sequences of row operations can produce
different row-echelon forms.)

7
 leading 1
Ex: (Procedure of Gaussian elimination and Gauss-Jordan elimination) ⎡1 4 − 3 2 14⎤ 1 4 −3 2
r23( −5) ⎡⎢
( − 12 ) 6 6 14⎤
Produce leading 1 r2 ⎢0 0 1 0 − 4 − 6⎥ 0 0 1 0 − 4 − 6⎥
⎡0 0 − 2 0 8 12 ⎤ ⎡2 8 − 6 4 12 28⎤ ⎢ ⎥ ⎢ ⎥
⎢2 8 − 6 4 12 28⎥
r12 ⎢0 0 − 2 0 ⎣⎢0 0 5 0 − 17 − 24⎦⎥ ⎣⎢0 0 0 0 3 6⎦⎥
⎢ ⎥ ⎢ 8 12 ⎥⎥ Submatrix
Zeros elements below leading 1
⎢⎣2 4 − 5 6 − 5 4⎥⎦ ⎢⎣2 4 − 5 6 − 5 4⎥⎦ Produce leading 1
Zeros elsewhere
The first nonzero column
1
r3
(
3
)
⎡1 4 − 3 2 6 14⎤ r31( −6 ) ⎡1 4 − 3 2 0 2⎤
⎢0 0 ⎢0 0 1 0 − 4 − 6⎥
leading 1 Produce leading 1 1 0 − 4 − 6⎥ ⎢ ⎥
⎢ ⎥
(1)
r1 2 ⎡1 4 − 3 2 6 14⎤ r13( −2 ) ⎡1 4 − 3 2 6 14⎤ ⎢⎣0 0 0 0 1 2⎥⎦
⎢0 0 − 2 0 ⎣⎢0 0 0 0 1 2⎦⎥
⎢0 0 − 2 0 8 12⎥ 8 12⎥ leading 1
⎢ ⎥ ⎢ ⎥ (row - echelon form) (row - echelon form)
⎣⎢2 4 − 5 6 − 5 4⎦⎥ ⎣⎢ 0 0 5 0 − 17 − 24⎦⎥
Zeros elements below leading 1 The first nonzero Submatrix
column
r32( 4 ) ⎡⎢ 1 4 − 3 2 0 2⎤⎥ r21( 3) ⎡ 1 4 0 2 0 8⎤
0 0 1 0 0 2 ⎢0 0 1 0 0 2 ⎥
⎢ ⎥ ⎢ ⎥
⎢⎣0 0 0 0 1 2⎥⎦
⎣⎢0 0 0 0 1 2⎦⎥
(row - echelon form) (reduced row - echelon form)

Ex 7: Solve a system by Gauss-Jordan elimination method Ex 8 Solve a system by Gauss-Jordan elimination method
(only one solution) (infinitely many solutions)
x − 2 y + 3z = 9 2 x1 + 4 x1 − 2 x3 = 0
− x + 3y = −4 3x1 + 5 x2 = 1
2 x − 5 y + 5z = 17
Sol: Sol: augmented matrix
( 12 ) ( −3) ( −1) ( −2 )
augmented matrix ⎡2 4 − 2 0⎤ r1 , r12 , r2 , r21 ⎡ 1 0 5 2⎤ (reduced row -
⎡ 1 − 2 3 9⎤ r12(1) , r13( −2 ) ⎡ 1 − 2 3 9⎤ r23(1) ⎡ 1 − 2 3 9⎤
⎢0
⎢⎣ 3 5 0 1⎥⎦ ⎢⎣0 1 − 3 − 1⎥⎦ echelon form)
⎢ − 1 3 0 − 4⎥ ⎢0 1 3 5⎥⎥ ⎢ 1 3 5⎥⎥
⎢
⎣⎢ 2 − 5 5 17⎦⎥ ⎢⎣0 − 1 − 1 − 1⎥⎦ ⎢⎣0 0 2 4⎥⎦ the corresponding system of equations is
1
x1 + 5 x3 = 2
( )
r 2 ( 2) ( −3) ( −9 )
⎡ 1 − 2 3 9⎤ r21 , r32 , r31 ⎡ 1 0 0 1⎤ x = 1 x2 − 3 x3 = − 1
3
⎢0 1 3 5⎥ ⎢0 1 0 − 1⎥ y = −1 leading variable：x1 , x2
⎢⎣0 0 1 2⎥⎦ ⎢⎣0 0 1 2⎥⎦ z = 2
(row - echelon form) (reduced row - echelon form) free variable： x3

8
 x1 = 2 − 5 x3 Homogeneous systems of linear equations:
x 2 = − 1 + 3 x3 A system of linear equations is said to be homogeneous
Let x3 = t if all the constant terms are zero.
x1 = 2 − 5t , a11 x1 + a12 x2 + a13 x3 + ! + a1n xn = 0
x2 = −1 + 3t , t∈R a21 x1 + a22 x2 + a23 x3 + ! + a2 n xn = 0
x3 = t , a31 x1 + a32 x2 + a33 x3 + ! + a3n xn = 0
"
So this system has infinitely many solutions. am1 x1 + am 2 x2 + am 3 x3 + ! + amn xn = 0

Trivial solution: Ex 9: Solve the following homogeneous system

x1 = x2 = x3 = ! = xn = 0 x1 − x2 + 3x3 = 0
n Nontrivial solution: 2 x1 + x2 + 3x3 = 0
other solutions Sol: augmented matrix
( −2) (1)
(1)
⎡ 1 − 1 3 0⎤ r12 , r2 , r21
3
n Notes: ⎡ 1 0 2 0⎤ (reduced row -
(1) Every homogeneous system of linear equations is consistent. ⎢⎣2 1 3 0⎥⎦ ⎢⎣0 1 − 1 0⎥⎦ echelon form)
(2) If the homogenous system has fewer equations than leading variable：x1 , x2
variables, then it must have an infinite number of solutions. free variable： x3
(3) For a homogeneous system, exactly one of the following is true. Let x3 = t
(a) The system has only the trivial solution. x1 = −2t, x2 = t, x3 = t, t ∈ R
(b) The system has infinitely many nontrivial solutions in
When t = 0, x1 = x2 = x3 = 0 (trivial solution)
addition to the trivial solution.