NUMERICAL METHODS 2.

If the digit to the immediate right of the last significant digit is

greater than 5, you round up the last significant figure.
Numerical methods are techniques to approximate mathematical
processes. E.g.

ACCURACY. Accuracy refers to how close a measurement is to the o 234.87 – Since the number to the right of 8 is 7, this will
true or accepted value. become 234.9

PRECISION. Precision refers to how close measurements are to each

other. 3. If the number to the immediate right of the last significant is a 5,
and that 5 is followed by a non zero digit, round up.

SIGNIFICANT FIGURES E.g.

1. Every non-zero numbers are ALWAYS significant figures. o 78.657 – The number to the right of 6 is 5. And the number
to its right is a non-zero, therefore, it will become 78.7
E.g.
o 846 has 3 sigfigs
4. If the number to the immediate right of the last significant is a 5,
and that 5 is followed by a zero, you look at the last
2. Zeroes in-between non-zero numbers are ALWAYS significant. significant digit and make it even.

E.g. E.g.

o 704 has 3 sigfigs o 2.5350 – The number to the right of 3 is 5. And the number
o 5006 has 4 sigfigs to its right is 0. The 3 will become 2.54.
o 2.5250 – If the number is already even, it is what it is. In
this case, it will be 2.52.
3. Trailing zeroes, or the zeroes to the right of non-zero numbers, are
siginificant only if there’s a decimal point.
E.g. SCIENTIFIC NOTATIONS
o 500 has 1 sigfig A useful way to represent very huge/very small numbers.
o 500. has 3 sigfigs E.g.
o 500.050 has 6 sigfigs
o 45000 = 4.5 × 104
o 3750 = 4.75 × 103
4. Leading zeroes, zeroes to the left, are NEVER significant. o 580000 = 5.8 × 105
o 0.0023 = 2.3 × 10−3
E.g.
o 0.00076 = 7.6 × 10−4
o 006 has 1 sigfig
o 0.0067 has 2 sigfigs
o 0.05087 has 4 sigfigs
o 0.0508700 has 6 sigfigs
ROOT FINDING
5. All counting numbers and constants have an infinite number of Solving for 𝑥 such that 𝑓(𝑥) = 0. Root is where the function
significant digits. becomes zero; it’s point intesecting the x-axis.

E.g. E.g. 𝑥 2 − 4𝑥 + 4 = 0
o 1 hour = 60 minutes (𝑥 − 2)2 = 0
o 12 inches = 1 foot
o 24 hours = 1 day √(𝑥 − 2)2 = √0

𝑥−2=0

Rounding Significant Digits 𝑥=2

1. If the digit to the immediate right of the last significant digit is less But in numerical methods, there are two (2) categories for root-
than 5, do not round up the last significant digit. finding;

E.g. 1. Bracketing Method

2. Open Method
o 43.82 – If you want to 3 significant digits, the last number
is 8. The digit to its right is 2 so, 8 won’t be rounded up.
 In bracketing methods, sign change is important. Sign change STEP 3. Again, sign change is important in bracketing methods.
is when the sign for the left and right values of 𝑦 changes, which is Choose the region with a sign change, and that will be the “new
an indication that the root 𝑥 is in that region. bracket”.

The process will be repeated until either 𝑥𝐿 and 𝑥𝑅 intersects

with the blue line in the illustration. In mathematical expression,
Under bracketing methods, there are two (2) different methods; process will repeat until 𝒇(𝒙𝑴 ) = 𝟎
1. Bisection Method
2. False-Position Method (Regula-Falsi)
Example: Solve for 𝑥 using bisection method: 𝑒 −𝑥 − 𝑥 = 0

i 𝑥𝐿 𝑥𝑅 𝑓(𝑥𝐿 ) 𝑓(𝑥𝑅 ) 𝑥𝑀 𝑓(𝑥𝑀 )

BISECTION METHOD
1 0 1
It is an approximation method to find the roots of the given
equation by repeatedly dividing the intervals. This method will
divide the interval until the resulting interval is found, which is
extremely small.
; where i is the number of iteration; 𝑥𝐿 and 𝑥𝑅 is the chosen bracket.

STEP 1. Select a bracket 𝑥𝐿 and 𝑥𝑅 such that 𝑓(𝑥𝐿 ) ∙ 𝑓(𝑥𝑅 ) < 0

and must be in opposite signs. After choosing bracket values, substitute separately to the given
function 𝑓(𝑥) = 𝑒 −𝑥 − 𝑥
𝑓(0) = 𝑒 −0 − 0
𝑓(𝑥𝐿 ) = 1

𝑓(1) = 𝑒 −1 − 1
𝑓(𝑥𝑅 ) = −0.632
Take note that 𝑥𝐿 < 𝑥𝑅

𝑥𝐿 +𝑥𝑅 i 𝑥𝐿 𝑥𝑅 𝑓(𝑥𝐿 ) 𝑓(𝑥𝑅 ) 𝑥𝑀 𝑓(𝑥𝑀 )

STEP 2. Get the midpoint 𝑥𝑚𝑖𝑑 = = 𝑥𝑀 . Or simply get
2
1 0 1 1 -0.632
the average of 𝑥𝐿 and 𝑥𝑅 to have a guess for the root value.

Now, get the midpoint/average of 𝑥𝐿 and 𝑥𝑀 , and then substitute

again to the function.
0+1
𝑥𝑚𝑖𝑑 = = 0.5
2

𝑓(0.5) = 𝑒 −0.5 − 0.5

Now, that there are three (3) values for 𝑥; 𝑥𝐿 , 𝑥𝑅 , 𝑥𝑀 . There will 𝑓(𝑥𝑀 ) = 0.107
be a new region, hence, it is called the bisection method.

i 𝑥𝐿 𝑥𝑅 𝑓(𝑥𝐿 ) 𝑓(𝑥𝑅 ) 𝑥𝑀 𝑓(𝑥𝑀 )

1 0 1 1 -0.632 0.5 0.107
 In illustration, 𝑥𝑀 is the intersection of x-axis and the line
connecting 𝑥𝐿 and 𝑥𝑅 .

Since there is a sign change in the bracket of 𝑥𝑀 and 𝑥𝑅 , there will

be new values of 𝑥𝐿 and 𝑥𝑅 .

i 𝑥𝐿 𝑥𝑅 𝑓(𝑥𝐿 ) 𝑓(𝑥𝑅 ) 𝑥𝑀 𝑓(𝑥𝑀 ) In mathematical expression, the formula for this is;

1 0 1 1 -0.632 0.5 0.107 𝒚𝑳

𝒙𝑴 = 𝒙𝑳 + (𝒙𝑹 − 𝒙𝑳 )( )
𝒚𝑳 − 𝒚𝑹
2 0.5 1 0.107 -0.632
which is derived from:

Solve, then repeat the process until 𝑓(𝑥𝑀 ) = 0. If 𝑓(𝑥𝐿 ) or

𝑓(𝑥𝑅 ) became 0 at first guess values, then no need for iterations.

0.5 + 1
𝑥𝑚𝑖𝑑 = = 0.75
2

𝑓(0.75) = 𝑒 −0.75 − 0.75

𝑓(𝑥𝑀 ) = −0.278

i 𝑥𝐿 𝑥𝑅 𝑓(𝑥𝐿 ) 𝑓(𝑥𝑅 ) 𝑥𝑀 𝑓(𝑥𝑀 )

1 0 1 1 -0.632 0.5 0.107 𝑥𝑀 − 𝑥𝐿 𝑥𝑅 − 𝑥𝐿
=
2 0.5 1 0.107 -0.632 0.75 -0.278 −𝑦𝐿 𝑦𝑅 − 𝑦𝐿
𝑥𝑅 − 𝑥𝐿
𝑥𝑀 − 𝑥𝐿 = (−𝑦𝐿 )
𝑦𝑅 − 𝑦𝐿
Without having to graph, always consider 𝒙𝑴 , then choose
one from 𝑥𝐿 and 𝑥𝑅 which has to be in opposite sign from 𝑥𝑀 .In 𝑦𝐿
𝑥𝑀 = 𝑥𝐿 + (𝑥𝑅 − 𝑥𝐿 )( )
this case, 𝑥𝑀 is negative. Therefore, 𝑥𝐿 should remain because it’s 𝑦𝐿 − 𝑦𝑅
positive.

Example: Solve for 𝑥 using false-position method: 𝑒 −𝑥 − 𝑥 = 0

i 𝑥𝐿 𝑥𝑅 𝑓(𝑥𝐿 ) 𝑓(𝑥𝑅 ) 𝑥𝑀 𝑓(𝑥𝑀 )
i 𝑥𝐿 𝑥𝑅 𝑦𝐿 𝑦𝑅 𝑥𝑀 𝑦𝑀
1 0 1 1 -0.632 0.5 0.107
1 0 1 1 -0.632 0.613 -0.071
2 0.5 1 0.107 -0.632 0.75 -0.278
2 0 0.613 1 -0.071 0.572 -0.0076
3 0.5 0.75 0.107 -0.278 0.625 -0.0897
3 0 0.572 1 -0.0076 0.5677 -0.0009
4 0.5 0.625 0.107 -0.0897 0.5625 0.007

1
𝑥𝑀 = 0 + (1 − 0) ( ) = 0.613
Since 𝑓(𝑥𝑀 ) is now approaching zero 0. It’s safe to conclude 1 + 0.632
that the value of x, that will make the function equivalent to 0, is 1
appriximately equal to 0.5625. 𝑥𝑀 = 0 + (0.613 − 0) ( ) = 0.572
1 + 0.071
1
𝑥𝑀 = 0 + (0.572 − 0) ( ) = 0.5677
1 + 0.0076
FALSE-POSITION METHOD
This method has the same process with bisection method excpet
on how to get the new 𝑥𝑀 . This method uses Linear Interpolation to With 𝑥𝑀 approaching 0, root value is approximately 0.5677.
get 𝑥𝑀 .
Under open methods, there are two (2) different methods; SECANT METHOD
1. Newton-Raphson Method
In this method, there are 2 guess values; 𝑥𝑎 and 𝑥𝑏 (depends on us)
2. Secant Method
3. Incremental Search Method 𝑓(𝑥𝑎 )(𝑥𝑎 − 𝑥𝑏 )
4. Fixed-Point Iteration Method 𝑥𝑛𝑒𝑤 = 𝑥𝑎 −
𝑓(𝑥𝑎 ) − 𝑓(𝑥𝑏 )

NEWTON-RAPHSON METHOD E.g. 𝑥 2 − 𝑥 − 1 = 0 Guess value: 𝑥𝑎 = 3 𝑥𝑎 = 2

In this method, there’s only 1 guess value.
i 𝑥𝑎 𝑥𝑏 𝑓(𝑥𝑎 ) 𝑓(𝑥𝑏 ) 𝜀
𝑓(𝑥𝑜𝑙𝑑 )
𝑥𝑛𝑒𝑤 = 𝑥𝑜𝑙𝑑 − 1 3 2 5 1
𝑓′(𝑥𝑜𝑙𝑑 )
2 2
Relative Error:
𝑥𝑛𝑒𝑤 − 𝑥𝑜𝑙𝑑
𝜀=| | × 100 5(3 − 1)
𝑥𝑛𝑒𝑤 𝑥𝑛𝑒𝑤 = 3 − = 1.75
5−1

E.g. 𝑥 2 − 𝑥 − 1 = 0
i 𝑥𝑎 𝑥𝑏 𝑓(𝑥𝑎 ) 𝑓(𝑥𝑏 ) 𝜀
Guess value: 𝑥 = 1
1 3 2 5 1
𝑓(𝑥) = 𝑥 2 − 𝑥 − 1
2 2 1.75 1 0.3125 50%
𝑓 ′ (𝑥) = 2𝑥 − 1
… … … … … …
i 𝑥𝐿 𝑓(𝑥𝑜𝑙𝑑 ) 𝑓′(𝑥𝑜𝑙𝑑 )
1.61803 1.61803 0 0 0.000%
1 1 -1 1

−1
𝑥𝑛𝑒𝑤 = 1 − =2
1

i 𝑥𝐿 𝑓(𝑥𝑜𝑙𝑑 ) 𝑓′(𝑥𝑜𝑙𝑑 )

1 1 -1 1

2 2 1 3

3 1.667 0.112 2.334

4 1.619 0.002 2.238

5 1.618 -0.000076 2.236

 INCREMENTAL METHOD i 𝑥 𝑔(𝑥)
1 0 -1
In this method, we start with an initial guess for the root (𝑥𝑜 ).
We then increment the value of the independent variable (usually x)
by a small step. That value of g(x) is now the new value of x in the next iteration.
𝑥𝑛𝑒𝑤 = 𝑥𝑜 + ∆𝑥
The process will go on until an interval where the
i 𝑥 𝑔(𝑥)
function changes sign happens. The root lies somewhere within
that interval. 1 0 -1

2 -1 -0.5

E.g. 𝑥 2 − 𝑥 − 1 = 0, let 𝑥𝑜 = 1.5, ∆𝑥 = 0.005 … … …

n … -0.618
i 𝑥 𝑓(𝑥𝑜 )
1 1.5 -0.25
The process will be repeated once a desired accuracy is
2 1.505 -0.239975 achieved. In this case, if you substitute -0.618 to thhe formed
3 1.51 -0.2299 equation, you’ll get the same thing, -0.618, hence the definition of
accuracy.
4 1.515 -0.219775
Therefore, 𝑥 = −0.618
5 1.52 -0.2096

… … …

n 1.61 -0.0179 But in this method, the roots is dependent on how 𝑔(𝑥) is
expressed. There could be many forms of a function.
… 1.615 -0.006775
Let’s say, if;
… 1.62 0.0044
𝑥2 − 𝑥 − 1
𝑥2 = 𝑥 + 1
Now, that there is a change from 1.615 to 1.62, that means that there
is a root between that interval. Therefore, we can say that 1
𝑥 = 1+
𝑥 ≈ 1.6185 𝑥

This value of x is a guess value within the interval because any value i 𝑥 𝑔(𝑥)
within that can be the answer. 1 1 2

2 2 1.5

… … …

n … 1.618

FIXED-POINT METHOD
Let’s form another 𝑥 = 𝑔(𝑥)
In this method, there is an initial guess value 𝑥𝑜 and a form:
𝑥2 − 𝑥 − 1
𝑥 = 𝑔(𝑥)
𝑥 = 𝑥2 − 1
𝑥𝑛𝑒𝑤 = 𝑔(𝑥𝑜 )
i 𝑥 𝑔(𝑥)
1 1 0
E.g. 𝑥 2 − 𝑥 − 1 = 0
2 0 -1
STEP 1. Express the equation into 𝑥𝑜 = 𝑔(𝑥) 3 -1 0

𝑥2 − 𝑥 − 1 4 0 -1

𝑥2 − 𝑥 = 1
If this is the case, then don’t use this function.
𝑥(𝑥 − 1) = 1
1
𝑥=
𝑥−1
1
STEP 2. Let’s say 𝑥𝑜 = 0, substitute that to 𝑔(𝑥) which is 𝑥−1
 MATRICES AND DETERMINANTS 7. ZERO or NULL MATRIX is a matrix wherein all elements
are zero.
0 0 0
[0 0 0]
MATRIX 0 0 0
A rectangular array of numbers enclosed within brackets or quantity
and arranged in “m” rows and “n” columns. Each number in the
matrix is called an element. 8. TRIANGULAR MATRIX is a square matrix whose
elements above the principal diagonal (or below the
principal diagonal) are zero.
1 7 9
[0 2 4]
0 0 4
1 0 0
The number 𝑎11 ........... 𝑎𝑚𝑛 are called the elements of the [4 2 0]
matrix. The horizontal lines are called rows and the vertical lines are 1 2 4
called columns of the matrix. Thus, 𝑎𝑖𝑗 is the elements in the ith row
and the jth column of A. Size: 𝒎 𝒙 𝒏

PROPERTIES OF MATRIX
TYPES OF MATRICES EQUALITY OF MATRICES. Two matrices A and B are equal if they
1. SQUARE MATRIX is a matrix in which the number of rows are of the same order and corresponding elements are equal. When m
equals the number of column. = n, it is a square matrix. An n x n or a square matrix is said to be of
1 3 4 order “n”. The number associated with a square matrix is called a
[5 2 4] determinants.
1 9 6

2. IDENTITY MATRIX is a square matrix in which the ADDITION/SUBRTRICTION OF MATRICES

diagonal elements are “one” and all the off-diagonal
elements are zero. 1 2 5 6 1+5 6+2 6 8
[ ]+[ ]=[ ]=[ ]
1 0 0 3 4 7 8 3+7 4+8 10 12
[0 1 0]
0 0 1
MULTIPLICATION OF MATRICES

3. ROW MATRIX is a matrix having only one row and “n” To multiply matrices, the inner dimension must be equal.
columns. It is called a“row vector”. ⬚ ⬚
𝐴 = [⬚ ⬚] 𝐵 = [⬚ ⬚ ⬚]
[2 4 7] ⬚ ⬚ ⬚
⬚ ⬚
𝐴: 3 𝑥 2 𝐵: 2 𝑥 3
4. COLUMN MATRIX is a matrix with “m” rows and only one
column. It is also called a column vector.
Their inner dimensions are both 2. Thefore, A and B can be multiplied.
2
The final dimension would be their outer dimensions. Therefore, it’s
[4 ]
7 3𝑥3
1 2
7 8 2
𝐴 = [3 4] 𝐵=[ ]
5. DIAGONAL MATRIX is a square matrix for which every 9 10 1
5 6
term off the main diagonal is zero.
1 0 0 Multiply row 1 and column 1 since it’s position in the output matrix is
[0 3 0] on row 1 column 1
0 0 4 1(7) + 2(9) = 25

6. SCALAR MATRIX is a square matrix for which all terms on

the main diagonal are equal. 25 28 4
𝐴𝐵 = [57 64 10]
3 0 0
[0 3 0] 89 100 16
0 0 3 Note: 𝐴𝐵 ≠ 𝐵𝐴
MULTIPLICATION OF A MATRIX BY A SCALAR E.g #2 (Without 1)
1 2 3 −2 2 5
𝐴=[ ]
3 4 −2 −2 0 3
𝐴=[ ]
4 0 −2 6
If A is to be multiplied by scalar 𝜆 = 2, then
−3 3 2 4
2(1) 2(2) 1 4
2𝐴 = [ ]=[ ]
2(3) 2(4) 6 8
• Factor out/Divide a number to a chosen row or column to
get 1.
DIVISION OF MATRICES
3 −2 2 5 ⬚
𝐴 −2 −2 0 3 ⬚
= 𝐴𝐵 −1 𝐴=[ ]
𝐵 4 0 −2 6 −2
−3 3 2 4 ⬚

MATRIX DETERMINANTS
3 −2 2 5
1 4 −2 −2 0 3
𝑑𝑒𝑡𝐴 𝑜𝑟 |𝐴| = | | = 𝑎𝑑 − 𝑏𝑐 𝐴=[ ]
6 8 −2 0 1 −3
= 1(8) − 4(6) = −16 −3 3 2 4
• Proceed to the process.

For 3 𝑥 3 matrices
1 3 4 MINORS OF A MATRIX
𝐴 = |5 2 4| E.g. Solve for 𝑚22 (minor of row 2 col 2)
1 9 6
1 3 4
1 3 4 1 3 1 3 4 1 3
𝐴=5 2 4
5 2 4 5 2 − 5 2 4 5 2
1 9 6
1 9 6 1 9 1 9 6 1 9
Delete row 2 and col 2
1 3 4
[1(2)(6) + 3(4)(1) + 4(5)(9)] 1 4
𝐴=5 2 4=
− [4(2)(1) + 1(4)(9) + 3(5)(6)] 1 6
1 9 6
= 204 − 134 = 70
𝑚22 = 1(6) − 4(1) = 2

For 4 𝑥 4 matrices
COFACTORS OF A MATRIX
PIVOTAL METHOD
The cofactor of a matrix with elements is equal to the minors
1 3 6 −2 multiplied by (−1)𝑟+𝑐
3 2 7 −5
𝐴=[ ] 𝑪𝒎𝒏 = (−𝟏)𝒎+𝒏 (𝒎𝒎𝒏 )
4 5 5 4
6 −3 −3 6
𝑷𝒊𝒗𝒐𝒕𝒂𝒍 𝑬𝒍𝒆𝒎𝒆𝒏𝒕 = 𝟏 1 4
𝑚22 = =2
1 6

1 3 6 −2 𝐶22 = (−1)2+2 (2) = 2

3 2 7 −5
𝐴=[ ]
4 5 5 4
6 −3 −3 6

TRANSPOSE OF A MATRIX
2 − 3(3) 7 − 3(6) −5 − 3(−2)
1+1 It is the interchange of rows and columns of matrix, denoted by 𝐴𝑇
= (−1) | 5 − 4(3) 5 − 4(6) 4 − 4(−2) |
−3 − 6(3) −3 − 6(6) 6 − 6(−2) 𝑎 𝑏 𝑐 𝑎 𝑑 𝑔
𝐴=𝑑 𝑒 𝑓 𝐴𝑇 = 𝑏 𝑒 ℎ
𝑔 ℎ 𝑖 𝑐 𝑓 𝑖
−7 −11 1
= [ −7 −19 12]
−21 −39 18
ADJOINT OF A MATRIX SOLUTIONS TO SYSTEMS OF LINEAR
Example: EQUATIONS
3 6 2 𝐶11 𝐶12 𝐶13 CRAMER’S RULE
𝐴 = 4 7 −1 𝐴𝑑𝑗 𝐴 = 𝐶21 𝐶22 𝐶23
5 1 3 𝐶31 𝐶32 𝐶33 det 𝐴𝑥𝑛
𝑥𝑛 =
det 𝐴
Example:
7 −1
𝐶11 = (−1)1+1 [ ]
1 3 2𝑥 + 𝑦 − 𝑧 = 1
𝐶11 = (−1)1+1 [7(3) − 1(−1)] = 22 𝐴 = [ 3𝑥 − 𝑦 + 𝑧 = 4 ]
2𝑥 + 3𝑦 + 𝑧 = 3

22 𝐶12 𝐶13
STEP 1. Transform to augmented
𝐴𝑑𝑗 𝐴 = 𝐶21 𝐶22 𝐶23
𝐶31 𝐶32 𝐶33 2 1 −1 1
𝐴 = 3 −1 1 4
2 3 1 3
After finding all the cofactors, transpose the matrix and that will be
𝐴𝑥 = 𝐵
the adjoint of A.

STEP 2. Get the determinant of 𝐴

INVERSE OF A MATRIX
det 𝐴 = [−2 + 2 + (−9)] − [2 + 6 + 3] = −20
𝐴𝑑𝑗 𝐴
𝐴−1 =
det 𝐴
E.g. STEP 3. Replace 𝐴𝑥 by 𝐵 where 𝐴𝑥 is the first column of 𝐴, then
3 5 get the determinant
𝐴=[ ]
−2 4 1 1 −1
det 𝐴 = 12 − (−10) = 22 𝐴 = 4 −1 1
3 3 1
det 𝐴𝑥 = [−1 + 3 + (−12)] − [3 + 3 + 4] = −20
4 2
𝐴=[ ]
−5 3
𝐶11 = (−1)1+1 [4] = 4 3.1. Replace 𝐴𝑦 by 𝐵 where 𝐴𝑦 is the second column of 𝐴, then
get the determinant
𝐶12 = (−1)1+2 [−2] = 2
2 1 −1
𝐶21 = (−1)2+1 [5] = −5 𝐴=3 4 1
2 3 1
𝐶22 = (−1)2+2 [3] = 3
det 𝐴𝑦 = [8 − 9 + 2] − [−8 + 3 + 6] = 0

4 −5
𝐴𝑇 = [ ]
2 3 3.1. Replace 𝐴𝑧 by 𝐵 where 𝐴𝑧 is the third column of 𝐴, then get
the determinant
2 1 1
4 −5
[ ] 𝐴 = 3 −1 4
𝐴−1 = 2 3 = [0.18 −0.22
]
0.09 0.14 2 3 3
22
det 𝐴𝑧 = [−6 + 9 + 8] − [−2 + 9 + 24] = −20

det 𝐴𝑥 −20
𝑥= = =1
det 𝐴 −20
det 𝐴𝑦 0
𝑦= = =0
det 𝐴 −20
det 𝐴𝑧 −20
𝑧= = =1
det 𝐴 −20
GAUSS-JORDAN METHOD 1 0 0 1
𝐴=0 1 0 2
Example: 0 0 1 3
𝑥 + 3𝑦 + 𝑧 = 10
𝐴 = [ 𝑥 − 2𝑦 − 𝑧 = −6 ]
2𝑥 + 𝑦 + 2𝑧 = 10 Example 2:
𝑥+𝑦+𝑧=4
𝐴 = [ 2𝑥 − 3𝑦 + 𝑧 = 2 ]
STEP 1. Transform to augmented −𝑥 + 2𝑦 − 𝑧 = −1
1 3 1 10
𝐴=1 −2 −1 −6
2 1 2 10 1 1 1 4
𝐴 = 2 −3 1 2
𝐴𝑥 = 𝐵 −1 2 −1 −1
𝑅2′ = 𝑅2 − 2𝑅1
STEP 2. 𝐴 must be an identity matrix. Formulate an equation to do 𝑅3′ = 𝑅3 + 𝑅1
that.
1 3 1 10
𝐴=1 −2 −1 −6 1 1 1 4
2 1 2 10 𝐴=0 −5 −1 −6
0 3 0 3
Solve the first column first
𝑅2′ = −𝑅1 + 𝑅2
Interchange 𝑅2 and 𝑅3
𝑅3′ = −2𝑅1 + 𝑅3
1 1 1 4
𝐴=0 3 0 3
0 −5 −1 −6
1 3 1 10
𝐴 = 0 −5 −2 −16 𝑅2 ′
𝑅2′′ =
0 −5 0 −10 3

Interchange 𝑅2 and 𝑅3 1 1 1 4
𝐴=0 1 0 1
1 3 1 10 0 −5 −1 −6
𝐴 = 0 −5 0 −10
0 −5 −2 −16 𝑅1 ′ = 𝑅1 − 𝑅2 ′′
𝑅2 ′ 𝑅3′′ = 𝑅3 ′ + 5𝑅2 ′′
𝑅2′′ =
−5

1 0 1 3
1 3 1 10 𝐴=0 1 0 1
𝐴=0 1 0 2 0 0 −1 −1
0 −5 −2 −16
𝑅3 ′′
𝑅3′′ = 𝑅3 ′ + 5𝑅2 ′′ 𝑅3′′′ =
−1
𝑅1 ′ = 𝑅1 − 3𝑅2 ′′
1 0 1 3
𝐴=0 1 0 1
1 0 1 4 0 0 1 1
𝐴=0 1 0 2
0 0 −2 −6 𝑅1 ′′ = 𝑅1 ′ − 𝑅3 ′′′
𝑅3 ′′
𝑅3′′′ =
−2 1 0 0 2
𝐴=0 1 0 1
0 0 1 1
1 0 1 4
𝐴=0 1 0 2
0 0 1 3
𝑅1 ′′ = 𝑅1 ′ − 𝑅3 ′′′
Example 3:
2𝑥 + 𝑦 − 𝑧 = 1
𝐴 = [ 3𝑥 − 𝑦 + 𝑧 = 4 ]
2𝑥 + 3𝑦 + 𝑧 = 3

2 1 −1 1
𝐴 = 3 −1 1 4
2 3 1 3
𝑅1
𝑅1′ =
2

1 1 1
1 −
𝐴= 2 2 2
3 −1 1 4
2 3 1 3
𝑅2′ = 𝑅2 − 3𝑅1
𝑅3′ = 𝑅3 − 2𝑅1

1 1 1
1 −
2 2 2
𝐴= 5 5 5
0 −
2 2 2
0 2 2 2
−2
𝑅2′′ = 𝑅 ′
5 2

1 1 1
1 −
𝐴= 2 2 2
0 1 −1 −1
0 2 2 2
1
𝑅1′′ = 𝑅1 ′ − 𝑅2 ′′
2
𝑅3′′ = 𝑅3 ′ − 2𝑅2 ′′

1 0 0 1
𝐴 = 0 1 −1 −1
0 0 4 4
𝑅3 ′′
𝑅3′′′ =
4

1 0 0 1
𝐴 = 0 1 −1 −1
0 0 1 1
𝑅2′′′ = 𝑅2′′ + 𝑅3′′′

1 0 0 1
𝐴=0 1 0 0
0 0 1 1