BMS College of Engineering, Bengaluru-19

Autonomous Institute, Affiliated to VTU, Belagavi

Department Of Mathematics
Course Title Numerical Methods for Engineers Course Code 23MA60ENME
Credits 03 L-T-P 3-0-0
Contact hours 36

UNIT – I

Iterative Methods for System of Equations and Eigenvalues problems

Fixed Point Iteration Method :-

Consider the system of two equation in two variables
i.e., 𝑓(𝑥 , 𝑦) = 0 and 𝑔(𝑥 , 𝑦) = 0.
Let us re-write the given equation as
𝑥 = 𝐹(𝑥 , 𝑦) and 𝑦 = 𝐺(𝑥 , 𝑦) ----------------------- ①
We need to approximate the roots as (𝑥𝑘 , 𝑦𝑘 ) for which we have to choose suitable F and G such
that equation ① gives convergence root.
Let (𝑥𝑘 , 𝑦𝑘 ) be the approximate root then
𝑥𝑘+1 = 𝐹(𝑥𝑘 , 𝑦𝑘 )
}-------------------------------------- ②
𝑦𝑘+1 = 𝐺(𝑥𝑘 , 𝑦𝑘 )
And let ( , ) be exact root then
 = 𝐹( , )
} ------------------------------------- ③
 = 𝐺( , )
Consider ③ - ②
 − 𝑥𝑘+1 = 𝐹( , ) − 𝐹(𝑥𝑘 , 𝑦𝑘 )
 − 𝑦𝑘+1 = 𝐺( , ) − 𝐺(𝑥𝑘 , 𝑦𝑘 )
Let 𝑘 =  − 𝑥𝑘 and 𝛿𝑘 =  − 𝑦𝑘
𝑘+1 = 𝐹(𝑥𝑘 + 𝑘 , 𝑦𝑘 + 𝛿𝑘 ) − 𝐹(𝑥𝑘 , 𝑦𝑘 )
𝛿𝑘+1 = 𝐺(𝑥𝑘 + 𝑘 , 𝑦𝑘 + 𝛿𝑘 ) − 𝐺(𝑥𝑘 , 𝑦𝑘 )
By taylor’s series and after neglecting from 2nd power of  and 𝛿.
𝑘+1 = 𝐹(𝑥𝑘 , 𝑦𝑘 ) + 𝑘 𝐹𝑥 + 𝛿𝑘 𝐹𝑦 − 𝐹(𝑥𝑘 , 𝑦𝑘 )
𝛿𝑘+1 = 𝐺(𝑥𝑘 , 𝑦𝑘 ) + 𝑘 𝐺𝑥 + 𝛿𝑘 𝐺𝑦 − 𝐺(𝑥𝑘 , 𝑦𝑘 )
𝐹𝑥 𝐹𝑦
i.e., 𝐸𝑘+1 = [ ]𝐸
𝐺𝑥 𝐺𝑦 𝑘
 𝐸𝑘+1 = 𝐽(𝐹 , 𝐺). 𝐸𝑘
Where 𝐸 = [ , 𝛿]𝑇
And J → Jacobian of F and G
 |𝐸𝑘+1 | < |𝐸𝑘 | where ‖𝐽‖ < 1
|𝐸𝑘+1 | < ‖𝐽‖|𝐸𝑘 | -------------------------------- ④
Equation ④ is convergent when ‖𝐽‖ < 1.
This is called sufficient condition for the convergence of the root of the system of equations.
Note : If we find norm as absolute maximum value of the different row sum
 𝐹𝑥 𝐹𝑦
i.e., 𝐽 = [ ]
𝐺𝑥 𝐺𝑦
|𝐹𝑥 | + |𝐹𝑦 | < 1
and }
|𝐺𝑥 | + |𝐺𝑦 | < 1
Problems
1. 𝑥 2 + 𝑥𝑦 − 10 = 0 and 𝑦 + 3𝑥𝑦 2 − 57 = 0 approximate the root of these using Fixed point
iteration method or General iteration method near the point 1.5 and 3.5 .
Sol :- Given that 𝑥 2 + 𝑥𝑦 − 10 = 0 and 𝑦 + 3𝑥𝑦 2 − 57 = 0 are the two
equation which can be written as
10−𝑥 2
𝑥= = 𝐹(𝑥 , 𝑦) 𝑦 = −3𝑥𝑦 2 + 57 = 𝐺(𝑥 , 𝑦)
𝑦
The to determine convergence or verify that
|𝐹𝑥 | + |𝐹𝑦 | < 1 |𝐺𝑥 | + |𝐺𝑦 | < 1
2𝑥
i.e., 𝐹𝑥 = 0 − 𝐺𝑥 = −3𝑦 2
𝑦
10−𝑥 2
𝐹𝑦 = − ( ) 𝐺𝑦 = −6𝑥𝑦
𝑦2
𝑥 2 −10
𝐹𝑦 = 𝑦2
2𝑥0 𝑥02 − 10
|𝐹𝑥 (𝑥0 , 𝑦0 )| + |𝐹𝑦 (𝑥0 , 𝑦0 )| = |− |+| |
𝑦0 𝑦02
2(1.5) (1.5)2 −10
= |− |+| (3.5)2
|
3.5
= |0.85714| + |0.632653|
= 1.489793 > 1
so it is not possible,
Hence let’s take another form
i.e., 𝑥 = √10 − 𝑥𝑦 = 𝐹(𝑥 , 𝑦)
1 1
𝐹𝑥 (𝑥 , 𝑦) = 2√10−𝑥𝑦 (−𝑦) 𝐹𝑦 (𝑥 , 𝑦) = 2√10−𝑥𝑦 (−𝑥)
Verify |𝐹𝑥 (𝑥0 , 𝑦0 )| + |𝐹𝑦 (𝑥0 , 𝑦0 )| < 1
|𝐹𝑥 (1.5 , 3.5)| + |𝐹𝑦 (1.5 , 3.5)| < 1
1(−3.5) 1(−1.5)
| |+| |
2√10−(1.5)(3.5) 2√10−(1.5)(3.5)
|−0.802955| + |−0.3441236008|  1.147118608>1
so it is not possible
Hence let’s take one more another form as
𝑥 2 + 𝑥𝑦 = 10
𝑥(𝑥 + 𝑦) = 10
10
𝑥= = 𝑓(𝑥, 𝑦)
𝑥+𝑦
10
𝐹𝑥 = − (1)
(𝑥 + 𝑦)2
|𝐹𝑥 (1.5, 3.5)| + |𝐹𝑦 (1.5, 3.5)| < 1
 10 10
|− (1)| + |− 2 |<1
(𝑥 2 2
+ 𝑦 + 2𝑥𝑦) (𝑥 + 𝑦 2 + 2𝑥𝑦)
|−0.4| + |−0.4| < 1
0.8 < 1
Then we can fix 𝐹(𝑥, 𝑦) = 𝑥 2 + 𝑥𝑦 − 10 as 𝐹(𝑥, 𝑦) = 𝑥(𝑥 + 𝑦) − 10.

2. The system of equations 𝑓(𝑥, 𝑦) = 𝑥 2 + 3𝑥 + 𝑦 − 5 = 0 and 𝑔(𝑥, 𝑦) = 𝑥 2 + 3𝑦 2 − 4 = 0

has a solution (1, 1). Determine the iteration functions 𝐹(𝑥, 𝑦) and 𝐺(𝑥, 𝑦) so that the
sequence of iterates obtained from.

𝑥𝑘+1 = 𝐹(𝑥𝑘 , 𝑦𝑘 )
𝑦𝑘+1 = 𝐺(𝑥𝑘 , 𝑦𝑘 ) , 𝑘 = 0,1, … …
(𝑥0 , 𝑦0 ) = (0.5, 0.5) converges to the root. Perform five iterations.

Solution: We write the given system of equations in an equivalent form as

𝑥 = 𝑥 + 𝛼(𝑥 2 + 3𝑥 + 𝑦 − 5) = 𝐹(𝑥, 𝑦)
𝑦 = 𝑦 + 𝛽(𝑥 2 + 3𝑦 2 − 4) = 𝐺((𝑥, 𝑦)
Where 𝛼 and 𝛽 are arbitrary parameters, which are to be determined. If we use the maximum
absolute row sum norm, we require that.
|𝐹𝑥 (𝑥0 , 𝑦0 ) | + 𝐹𝑦 (𝑥0 , 𝑦0 ) < 1
|𝐺𝑥 (𝑥0 , 𝑦0 ) + |𝐺𝑦 (𝑥0 , 𝑦0 )| < 1. |

And Differentiating 𝐹 and 𝐺 partially with respect to 𝑥 and 𝑦 and evaluating at

(𝑥0 , 𝑦0 ) = (0.5, 0.5)𝑤𝑒 𝑔𝑒𝑡
𝐹(𝑥, 𝑦) = 1 + (2𝑥 + 3)𝛼, 𝐹(0.5, 0.5) = 1 + 4𝛼
𝐹𝑦 (𝑥, 𝑦) = 𝛼, 𝐹𝑦 (0.5 0.5) = 𝛼
𝐺𝑥 (𝑥, 𝑦) = 2𝛽𝑥, 𝐺(0.5 0.5) = 𝛽
𝐺_𝑦 = 𝑥, 𝑦) = 1 + 6𝑦, 𝐺𝑦 (0.5, 0.5) = 1 + 3𝛽

Therefore, the conditions of converges becomes

|1 + 4𝛼| + |𝛼| < 1

|𝛽| + |1 + 3𝛽| < 1
Any values of 𝛼, 𝛽which satisfy the above can be used. Obviously, both 𝛼 and 𝛽 are negative.
1
Taking 𝛼 = − 4 and 𝛽 = −1/6, we obtained the iteration method.
Starting with (𝑥0 , 𝑦0 ) = (0.5, 0.5), we get
(𝑥1 , 𝑦1 ) = (1.1875, 1.0), (𝑥2 , 𝑦2 ) = (0.944336, 0.931641)
(𝑥3 , 𝑦3 ) = (1.030231, 1.015702), (𝑥4 , 𝑦4 ) = (0.988288, 0.989647)
(𝑥5 , 𝑦5 ) = (1.005482, 1.003828)
Newton’s method for solving non linear systems

Approximating the roots of algebraic and transcendal equations

Standard general form of nth degree 𝑎0 𝑥 𝑛 + 𝑎1 𝑥 𝑛−1 + … + 𝑎𝑛−1 𝑥 + 𝑎𝑛 = 0

Intermediate Value Theorem:

If the function 𝑓(𝑥) is continuous in the [𝑎, 𝑏] and 𝑓(𝑎). 𝑓(𝑏) < 0. Then there exists at least
one point in (𝑎, 𝑏) such that 𝑓(𝑐) = 0. Such a points are called roots of the equation.

Newton Raphson’s method:

Consider an equation in one variable i.e., 𝑓(𝑥) = 0.

By Newton Raphson’s method roots can be approximate by
𝑓(𝑥𝑖 )
𝑥𝑖+1 = 𝑥𝑖 −
𝑓 (𝑥𝑖 )
Remark:
• Convergence of this iterative formula depends on derivative of 𝑓(𝑥) and the initial
approximation to get the convergent root.
• If 𝑓 (𝑥𝑗 ) is not near to zero then the initial approximation is close to the exact solution/root
i,e., |𝑥 − | < 1 where 𝑥 is initial approximation and  is an exact root.
Property:
Any iterative formula is said to be of order ‘P’ as a root of convergence ‘P’ .
If ‘P’ is largest positive real number then there exists a finite constant ‘C’ such that |𝑘+1 | < 𝐶|𝑘 |𝑃
where 𝑘 = |𝑥𝑘 − |

𝑥𝑘 is an approximate root,  is an exact root and 𝐶 is an asymptotic error at every iteration 𝑘.

I. Find the rate of convergence for the Newton Raphson’s method.

Consider an equation 𝑓(𝑥) = 0.
Let 𝑥𝑘 be approximate root and  be the exact root of the equation
 |𝑥𝑘 − | = 𝑘 error at 𝑘 𝑡ℎ iteration.
And by Newton Raphson’s method we have,
𝑓(𝑥𝑖 )
𝑥𝑖+1 = 𝑥𝑖 −
𝑓 (𝑥𝑖 )
𝑓( +)
Then, 𝑘+1 +  = 𝑘 +  − 𝑓 (𝑘 +)
𝑘
𝑓(𝑘 +)
𝑘+1 = 𝑘 − 𝑓 (𝑘 +)
 2
[𝑓()+𝑘 𝑓 ()+ 𝑘 𝑓 ()+⋯…….]
By taylor’s series 𝑘+1 = 𝑘 − 2!
[𝑓 ()+𝑘 𝑓 ()+⋯………….]
ℎ𝑓 (𝑥) ℎ2
∵ 𝑓(𝑥 + ℎ) = 𝑓(𝑥) + + 𝑓 (𝑥) + ⋯.
1! 2!
 −1
𝑘 𝑓 () 2𝐾 𝑓 () 𝑘 𝑓 ()
𝑘+1 = 𝑘 − [ + + ⋯ ] [1 + ( )+ ⋯.]
𝑓 (  ) 2! 𝑓() 𝑓  ( )

∵ (1 + 𝑥)−1 = (1 + 𝑥 + 𝑥 2 + 𝑥 3 + ⋯ )
2𝑘 𝑓 ()
𝑘+1 = 𝑘 − (𝑘 + + ⋯ ) (1 + 𝐹 + 𝐹 2 + ⋯ )
2! 𝑓 ()
𝑘 𝑓 () 2𝑘 𝑓 ()
Where 𝐹 = + + ⋯.
𝑓 (𝑥) 2!𝑓 (𝑥)
𝐶𝑘2 𝑓 ()
𝑘+1 = (𝑘 − 𝑘 + 2! 𝑓 (𝑥)
+ ⋯ ) + 0(𝑘 )3

−𝑓 ()
𝑘+1 = 2!𝑓 () 2𝑘 + 0(𝑘 )3

Neglecting from the cubic power of 𝑘 .

 −𝑓 () 2
𝑘+1= 
2!𝑓 () 𝑘
𝑓 (𝑦)
|𝑘+1 | < 𝐶|𝑘 |2, where constant ‘C’ is equal to 𝐶 =
2!𝑓 ()

Newton Raphson’s method is of order two

II. Derive the Newton’s formula for the system of equations in two variables.

Consider the system of equation as 𝑓(𝑥, 𝑦) and 𝑔(𝑥, 𝑦).

Let (, ) be the exact root.
And let b(𝑥𝑘 , 𝑦𝑘 ) e approximate root.
Also let ∆𝑥 and ∆𝑦 be increments in a x and y direction respectively such that
(𝑥𝑘 + ∆𝑥 , 𝑦𝑘 + ∆𝑦) is exact root.
i.e., 𝑓(𝑥𝑘 + ∆𝑥 , 𝑦𝑘 + ∆𝑦) = 0 and 𝑔(𝑥𝑘 + ∆𝑥 , 𝑦𝑘 + ∆𝑦) = 0
By taylor’s series
𝑓(𝑥𝑘 + ∆𝑥 , 𝑦𝑘 + ∆𝑦) = 𝑓(𝑥𝑘 , 𝑦𝑘 ) + 𝑓𝑥 ∆𝑥 + 𝑓𝑦 ∆𝑦 + ⋯ = 0
𝑔(𝑥𝑘 + ∆𝑥 , 𝑦𝑘 + ∆𝑦) = 𝑔(𝑥𝑘 , 𝑦𝑘 ) + 𝑔𝑥 ∆𝑥 + 𝑔𝑦 ∆𝑦 + ⋯ = 0
Neglecting second powers of ∆.
𝑓(𝑥𝑘 , 𝑦𝑘 ) + 𝑓𝑥 ∆𝑥 + 𝑓𝑦 ∆𝑦 = 0 and 𝑔(𝑥𝑘 , 𝑦𝑘 ) + 𝑔𝑥 ∆𝑥 + 𝑔𝑦 ∆𝑦 = 0
These are the system of equations in terms of ∆𝑥 and ∆𝑦.
𝑓 𝑓 ∆𝑥 𝑓(𝑥𝑘 , 𝑦𝑘 )
In the matrix form [𝑔𝑥 𝑔𝑦 ] [ ] = − [ ]
𝑥 𝑦 ∆𝑦 𝑔(𝑥𝑘 , 𝑦𝑘 )
𝐽. ∆𝑋 = −𝐹(𝑥𝑘 , 𝑦𝑘 )
𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]→Jacobians
𝑥 𝑦
−1
∆𝑋 = −𝐽𝑘 𝐹(𝑥𝑘 , 𝑦𝑘 )
∆𝑥
[ ] = −𝐽𝑘−1 𝐹(𝑥𝑘 , 𝑦𝑘 )
∆𝑦
𝑥𝑘+1 − 𝑥𝑘
[𝑦 ] = −𝐽𝑘−1 𝐹(𝑥𝑘 , 𝑦𝑘 )
𝑘+1 − 𝑦𝑘
𝑥𝑘+1 𝑥𝑘 --------------------①
 [𝑦 ] = [𝑦 ] − 𝐽𝑘−1 𝐹(𝑥𝑘 , 𝑦𝑘 )
𝑘+1 𝑘
in vector form 𝑋𝑘+1 = 𝑋𝑘 − 𝐽𝐾−1 𝐹𝑘 (𝑥𝑘 , 𝑦𝑘 ) --------------------------②

This method given by ② is an extension of the Newton-Raphson method to a system of 2×2

equations.
This method can be easily generalized for solving a system of 𝑛 equations in 𝑛 unknowns
f1 (x1 , x2 , … , xn ) = 0
f2 (x1 , x2 , … , xn ) = 0
… ….
fn (x1 , x2 , … , xn ) = 0
Or F(x) = 0
𝑇 𝑇
where 𝑥 = [x1 , x2 , … , xn ] , 𝐹 = [𝑓1 , 𝑓2 , … . 𝑓𝑛 ] .
(0) (0) (0) 𝑇
If 𝑥 (0) = [𝑥1 , 𝑥2 , … , 𝑥𝑛 ] is an initial approximation to the solution vector 𝑥, then we can write
the method as
𝑥 (𝑘+1) = 𝑥 (𝑘) − 𝐽𝑘−1 𝐹(𝑥 (𝑘) ), 𝑘 = 0,1,2, … … …
Where
𝜕𝑓1 𝜕𝑓1 𝜕𝑓1
…
𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑛
𝜕𝑓2 𝜕𝑓2 𝜕𝑓2
𝐽𝑘 = 𝜕𝑥 …
1 𝜕𝑥2 𝜕𝑥
… …. … …𝑛
𝜕𝑓𝑛 𝜕𝑓𝑛 𝜕𝑓𝑛
….
[𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑛 ](𝑥 (𝑘) )
is the Jacobian matrix of the function 𝑓1 , 𝑓2 , … 𝑓_𝑛 evaluated at 𝑥 (𝑘) .
Note that the matrix 𝐽𝑘−1 is to be evaluated for each iteration. We can also write the method as
𝐽𝑘 (𝑥 (𝑘+1) − 𝑥 (𝑘) ) = −𝐹(𝑥 (𝑘) )
Or 𝐽𝑘 𝜀 (𝑘) = −𝐹(𝑥 (𝑘) )
Where 𝜀 (𝑘) = 𝑥 (𝑘+1) − 𝑥 (𝑘) is the error vector.
We solve it as a linear system of equations (by a direct method if the system is small and by an
iterative method if the system is large) for each iteration.

Note: Convergence of the ① and ② formula depends on norms of ‖𝐽𝑘−1 ‖ < 1. Where ‖. ‖is some
suitable norm,
This is called sufficient condition for the convergence of for N-R method.
Necessary and Sufficient condition for the convergence of N-R method is, 𝜌(𝐽𝑘−1 ) < 1 Where
𝜌(𝐴)→Spectral radius of matrix A i.e,. The largest Eigen value of the matrix A in magnitude.

Problems:
Approximate the root of the following system of equations near the given point using Newton’s
method.
1. 𝑥 2 + 𝑥𝑦 + 𝑦 2 = 7 & 𝑥 3 + 𝑦 3 = 9 when (𝑥, 𝑦) = (1.5 , 0.5).
Sol :- By newton’s method we have
𝑥𝑘+1 𝑥𝑘 𝑓(𝑥𝑘 , 𝑦𝑘 )
[𝑦 ] = [𝑦 ] − 𝐽𝑘−1 [ ]------------------------------- ①
𝑘+1 𝑘 𝑔(𝑥𝑘 , 𝑦𝑘 )
𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]
𝑥 𝑦
given, 𝑓(𝑥 , 𝑦) = 𝑥 + 𝑥𝑦 + 𝑦 2 − 7
2

𝑓(𝑥0 , 𝑦0 ) = 𝑥02 + 𝑥0 𝑦0 + 𝑦02 − 7

𝑓(1.5 , 0.5) = (1.5)2 + (1.5)(0.5) + (0.5)2 − 7
𝑓(𝑥0 , 𝑦0 ) = −3.75
𝑓𝑥 = 2𝑥 + 𝑦𝑓𝑥0 = 2(𝑥0 ) + (𝑦0 ) 𝑓𝑥0 = 2(1.5) + (0.5)𝑓𝑥0 = 3.5
𝑓𝑦 = 𝑥 + 2𝑦𝑓𝑦0 = (𝑥0 ) + 2(𝑦0 )𝑓𝑦0 = (1.5) + 2(0.5)𝑓𝑦0 = 2.5

𝑔(𝑥 , 𝑦) = 𝑥 3 + 𝑦 3 − 9
𝑔(𝑥0 , 𝑦0 ) = 𝑥03 + 𝑦03 − 9
𝑔(1.5 , 0.5) = (1.5)3 + (0.5)3 − 9
𝑔(𝑥0 , 𝑦0 ) = −5.5
𝑔𝑥 = 3𝑥 2 𝑔𝑥0 = 3𝑥02 𝑔𝑥0 = 3(1.5)2 𝑔𝑥0 = 6.75
𝑔𝑦 = 3𝑦 2 𝑔𝑦0 = 3𝑦02 𝑔𝑦0 = 3(0.5)2 𝑔𝑦0 = 0.75
𝑓 𝑓
Then 𝐽0 = [𝑔𝑥0 𝑔𝑦0 ]
𝑥0 𝑦0
3.5 2.5
𝐽𝑘 = [ ]
6.75 0.75
−0.052 0.175
𝐽0−1 = [ ]
0.4736 −0.245
For the first iteration 𝑘 = 0 and equation 1 becomes
𝑥0+1 𝑥0 𝑓(𝑥0 , 𝑦0 )
[𝑦 ] = [𝑦 ] − 𝐽0−1 [ ]
0+1 0 𝑔(𝑥0 , 𝑦0 )
𝑥1 1.5 −0.052 0.175 −3.75
[𝑦 ] = [ ] − [ ][ ]
1 0.5 0.4736 −0.245 −5.5
𝑥1 1.5 −0.7675
[𝑦 ] = [ ] − [ ]
1 0.5 −0.428
𝑥1 2.2675
[𝑦 ] = [ ]
1 0.9254
For the second iteration 𝑘 = 1 and equation ① becomes
𝑥1+1 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽𝑘−1 [ ]
1+1 1 𝑔(𝑥1 , 𝑦1 )
𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]
𝑥 𝑦
𝑓(𝑥 , 𝑦) = 𝑥 + 𝑥𝑦 + 𝑦 2 − 7
2

𝑓(𝑥1 , 𝑦1 ) = 𝑥12 + 𝑥1 𝑦1 + 𝑦12 − 7

𝑓(2.2675 , 0.9254) = (2.2675)2 + (2.2675)(0.9254) + (0.9254)2 − 7
𝑓(𝑥1 , 𝑦1 ) = 1.09626591
𝑓𝑥 = 2𝑥 + 𝑦𝑓𝑥1 = 2(𝑥1 ) + (𝑦1 )𝑓𝑥1 = 2(2.2675) + (0.9254)
 𝑓𝑥1 = 5.4604
 𝑓𝑦 = 𝑥 + 2𝑦𝑓𝑦1 = (𝑥1 ) + 2(𝑦1 )𝑓𝑦1 = (2.2675) + 2(0.9254)
 𝑓𝑦1 = 4.1183

𝑔(𝑥 , 𝑦) = 𝑥 3 + 𝑦 3 − 9
𝑔(𝑥1 , 𝑦1 ) = 𝑥13 + 𝑦13 − 9
𝑔(2.2675 , 0.9254) = (2.2675)3 + (0.9254)3 − 9
𝑔(𝑥1 , 𝑦1 ) = 3.450959116
𝑔𝑥 = 3𝑥 2 𝑔𝑋1 = 3𝑥12 𝑔𝑥1 = 3(2.2675)2 𝑔𝑥1 = 15.42466875
𝑔𝑦 = 3𝑦 2 𝑔𝑦1 = 3𝑦12 𝑔𝑦1 = 3(0.9254)2 𝑔𝑦1 = 2.56909548
𝑓 𝑓
Then 𝐽1 = [𝑔𝑥1 𝑔𝑦1 ]
𝑥1 𝑦1
5.4604 4.1183
𝐽1 = [ ]
15.4246 2.5690
−0.0519 0.0832
𝐽1−1 = [ ]
0.3116 −0.11
𝑥1+1 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽1−1 [ ]
1+1 1 𝑔(𝑥1 , 𝑦1 )
𝑥2 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽1−1 [ ]
2 1 𝑔(𝑥1 , 𝑦1 )
𝑥2 2.2675 −0.0519 0.0832 1.0962
[𝑦 ] = [ ]−[ ][ ]
2 0.9254 0.3116 −0.11 3.4509
𝑥2 2.2675 0.2302
[𝑦 ] = [ ]−[ ]
2 0.9254 −0.038
𝑥2 2.0373
[𝑦 ] = [ ]
2 0.9634

2. 𝑥 2 + 3𝑥 + 𝑦 = 5 & 𝑥 2 + 3𝑦 2 = 4 when (𝑥, 𝑦) = (0.5 , 0.5).

Sol :- By newton’s method we have
𝑥𝑘+1 𝑥𝑘 𝑓(𝑥𝑘 , 𝑦𝑘 )
[𝑦 ] = [𝑦 ] − 𝐽𝑘−1 [ ]------------------------------- ①
𝑘+1 𝑘 𝑔(𝑥𝑘 , 𝑦𝑘 )
𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]
𝑥 𝑦
2
Give 𝑓(𝑥 , 𝑦) = 𝑥 + 3𝑥 + 𝑦 − 5
𝑓(𝑥0 , 𝑦0 ) = 𝑥02 + 3𝑥0 + 𝑦0 − 5
𝑓(0.5 , 0.5) = (0.5)2 + 3(0.5) + (0.5) − 5
𝑓(𝑥0 , 𝑦0 ) = −2.75
𝑓𝑥 = 2𝑥 + 3𝑓𝑥0 = 2(𝑥0 ) + 3 𝑓𝑥0 = 2(0.5) + 3𝑓𝑥0 = 4
𝑓𝑦 = 1𝑓𝑦0 = 1

𝑔(𝑥 , 𝑦) = 𝑥 2 + 3𝑦 2 − 4
𝑔(𝑥0 , 𝑦0 ) = 𝑥02 + 3𝑦02 − 4
𝑔(0.5 , 0.5) = (0.5)2 + 3(0.5)2 − 4
𝑔(𝑥0 , 𝑦0 ) = −3
𝑔𝑥 = 2𝑥𝑔𝑥0 = 2𝑥0 𝑔𝑥0 = 2(0.5)𝑔𝑥0 = 1
 𝑔𝑦 = 6𝑦𝑔𝑦0 = 6𝑦0 𝑔𝑦0 = 6(0.5)𝑔𝑦0 = 3
𝑓 𝑓
Then 𝐽0 = [𝑔𝑥0 𝑔𝑦0 ]
𝑥0 𝑦0
4 1
𝐽𝑘 = [ ]
1 3
0.2727 −0.09
Also, 𝐽0−1 = [ ]
−0.09 0.3636

For the first iteration 𝑘 = 0 and equation ① becomes

𝑥0+1 𝑥0 𝑓(𝑥0 , 𝑦0 )
[𝑦 ] = [𝑦 ] − 𝐽0−1 [ ]
0+1 0 𝑔(𝑥0 , 𝑦0 )
𝑥1 0.5 0.2727 −0.09 −2.75
[𝑦 ] = [ ] − [ ][ ]
1 0.5 −0.09 0.3636 −3
𝑥1 1.5 −0.479925
[𝑦 ] = [ ] − [ ]
1 0.5 −0.8433
𝑥1 0.979925
[𝑦 ] = [ ]
1 1.3433
For the second iteration 𝑘 = 1 and equation ① becomes

𝑥1+1 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽𝑘−1 [ ]
1+1 1 𝑔(𝑥1 , 𝑦1 )
𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]
𝑥 𝑦
2
Give 𝑓(𝑥 , 𝑦) = 𝑥 + 3𝑥 + 𝑦 − 5
𝑓(𝑥1 , 𝑦1 ) = 𝑥12 + 3𝑥1 + 𝑦1 − 5
𝑓(0.9799 , 1.3433) = (0.9799)2 + 3(0.9799) + (1.3433) − 5
𝑓(𝑥1 , 𝑦1 ) = 0.24320401
𝑓𝑥 = 2𝑥 + 3𝑓𝑥1 = 2(𝑥1 ) + 3 𝑓𝑥1 = 2(0.9799) + 3𝑓𝑥1 = 4.9598
𝑓𝑦 = 1𝑓𝑦1 = 1

𝑔(𝑥 , 𝑦) = 𝑥 2 + 3𝑦 2 − 4
𝑔(𝑥1 , 𝑦1 ) = 𝑥12 + 3𝑦12 − 4
𝑔(0.9799 , 1.3433) = (0.9799 )2 + 3(1.3433)2 − 4
𝑔(𝑥1 , 𝑦1 ) = 2.37356868
𝑔𝑥 = 2𝑥𝑔𝑥1 = 2𝑥1 𝑔𝑥1 = 2(0.9799)𝑔𝑥1 = 1.9598
𝑔𝑦 = 6𝑦𝑔𝑦1 = 6𝑦1 𝑔𝑦1 = 6(1.3433)𝑔𝑦1 = 8.0598
𝑓 𝑓
Then 𝐽1 = [𝑔𝑥1 𝑔𝑦1 ]
𝑥1 𝑦1
4.9598 1
𝐽1 = [ ]
1.9598 8.0598
0.2120 −0.026
Also, 𝐽1−1 = [ ]
−0.051 0.1304
𝑥1+1 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽1−1 [ ]
1+1 1 𝑔(𝑥1 , 𝑦1 )
𝑥2 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽1−1 [ ]
2 1 𝑔(𝑥1 , 𝑦1 )
 𝑥2 0.979925 0.2120 −0.026 0.2432
[𝑦 ] = [ ]−[ ][ ]
2 1.3433 −0.051 0.1304 2.3735
𝑥2 0.979925 −0.0101
[𝑦 ] = [ ]−[ ]
2 1.3433 0.2971
𝑥2 0.990025
[𝑦 ] = [ ]
2 1.0462

3. 𝑥 2 𝑦 + 𝑦 3 = 10 & 𝑥𝑦 2 − 𝑥 2 = 3 near (𝑥, 𝑦) = (0.8 , 2.2).

Sol :- By newton’s method we have
𝑥𝑘+1 𝑥𝑘 𝑓(𝑥𝑘 , 𝑦𝑘 )
[𝑦 ] = [𝑦 ] − 𝐽𝑘−1 [ ]------------------------------- ①
𝑘+1 𝑘 𝑔(𝑥𝑘 , 𝑦𝑘 )
𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]
𝑥 𝑦
Give 𝑓(𝑥 , 𝑦) = 𝑥 𝑦 + 𝑦 3 − 10
2

𝑓(𝑥0 , 𝑦0 ) = 𝑥02 𝑦0 + 𝑦03 − 10

𝑓(0.8 , 2.2) = (0.8)2 (2.2) + (2.2)3 − 10
𝑓(𝑥0 , 𝑦0 ) = 2.056
𝑓𝑥 = 2𝑥𝑦𝑓𝑥0 = 2(𝑥0 )(𝑦0 )𝑓𝑥0 = 2(0.8)(2.2)𝑓𝑥0 = 3.52
𝑓𝑦 = 𝑥 2 + 3𝑦 2 𝑓𝑦0 = 𝑥02 + 3𝑦02 𝑓𝑦0 = (0.8)2 + 3(2.2)2 𝑓𝑦0 = 15.16

𝑔(𝑥 , 𝑦) = 𝑥𝑦 2 − 𝑥 2 − 3
𝑔(𝑥0 , 𝑦0 ) = 𝑥0 𝑦02 − 𝑥02 − 3
𝑔(0.8 , 2.2) = (0.8)(2.2)2 − (0.8)2 − 3
𝑔(𝑥0 , 𝑦0 ) = 0.232
𝑔𝑥 = 𝑦 2 − 2𝑥𝑔𝑥0 = 𝑦02 − 2𝑥0 𝑔𝑥0 = (2.2)2 − 2(0.8)𝑔𝑥0 = 3.24
𝑔𝑦 = 2𝑥𝑦𝑔𝑦0 = 2𝑥0 𝑦0 𝑔𝑦0 = 2(0.8)(2.2)𝑔𝑦0 = 3.52
𝑓 𝑓
Then 𝐽0 = [𝑔𝑥0 𝑔𝑦0 ]
𝑥0 𝑦0
3.52 15.16
𝐽𝑘 = [ ]
3.24 3.52
−0.095 0.4127
Also, 𝐽0−1 = [ ]
0.0882 −0.095
For the first iteration 𝑘 = 0 and equation ① becomes
𝑥0+1 𝑥0 𝑓(𝑥0 , 𝑦0 )
[𝑦 ] = [𝑦 ] − 𝐽0−1 [ ]
0+1 0 𝑔(𝑥0 , 𝑦0 )
𝑥1 0.8 −0.095 0.4127 2.056
[𝑦 ] = [ ] − [ ][ ]
1 2.2 0.0882 −0.095 0.232
𝑥1 0.8 −0.0995736
[𝑦 ] = [ ] − [ ]
1 2.2 0.1592992
𝑥1 0.8995736
[𝑦 ] = [ ]
1 2.0407008
For the second iteration 𝑘 = 1 and equation ① becomes
𝑥1+1 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽𝑘−1 [ ]
1+1 1 𝑔(𝑥1 , 𝑦1 )
 𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]
𝑥 𝑦
Give 𝑓(𝑥 , 𝑦) = 𝑥 𝑦 + 𝑦 3 − 10
2

𝑓(𝑥0 , 𝑦0 ) = 𝑥02 𝑦0 + 𝑦03 − 10

𝑓(0.8995 , 2.0407) = (0.8995)2 (2.0407) + (2.0407)3 − 10
𝑓(𝑥0 , 𝑦0 ) = 0.1495372393
𝑓𝑥 = 2𝑥𝑦𝑓𝑥0 = 2(𝑥0 )(𝑦0 )𝑓𝑥0 = 2(0.8995)(2.0407)𝑓𝑥0 = 3.67121
𝑓𝑦 = 𝑥 2 + 3𝑦 2 𝑓𝑦0 = 𝑥02 + 3𝑦02 𝑓𝑦0 = (0.8995)2 + 3(2.0407)2
𝑓𝑦0 = 13.30246972

𝑔(𝑥 , 𝑦) = 𝑥𝑦 2 − 𝑥 2 − 3
𝑔(𝑥0 , 𝑦0 ) = 𝑥0 𝑦02 − 𝑥02 − 3
𝑔(0.8995 , 2.0407) = (0.8995)(2.0407)2 − (0.8995)2 − 3
𝑔(𝑥0 , 𝑦0 ) = −0.06317163725
𝑔𝑥 = 𝑦 2 − 2𝑥𝑔𝑥0 = 𝑦02 − 2𝑥0 𝑔𝑥0 = (2.0407)2 − 2(0.8995)
𝑔𝑥0 = 2.36545649
𝑔𝑦 = 2𝑥𝑦𝑔𝑦0 = 2𝑥0 𝑦0 𝑔𝑦0 = 2(0.8995)(2.0407)𝑔𝑦0 = 3.671219
𝑓 𝑓
Then 𝐽1 = [𝑔𝑥1 𝑔𝑦1 ]
𝑥1 𝑦1
3.6712 13.3024
𝐽1 = [ ]
2.3654 3.6712
−0.2040 0.7395
𝐽1−1 = [ ]
0.1315 −0.204
𝑥1+1 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽1−1 [ ]
1+1 1 𝑔(𝑥1 , 𝑦1 )
𝑥2 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽1−1 [ ]
2 1 𝑔(𝑥1 , 𝑦1 )
𝑥2 0.8995 −0.2040 0.7395 0.1495
[𝑦 ] = [ ]−[ ][ ]
2 2.0407 0.1315 −0.204 −0.0631
𝑥2 0.8995 −0.0771
[𝑦 ] = [ ]−[ ]
2 2.0407 0.0325
𝑥2 0.9766
[𝑦 ] = [ ]
2 2.0082

4. 𝑦𝑐𝑜𝑠(𝑥𝑦) + 1 = 0 & sin(𝑥𝑦) + 𝑥 − 𝑦 = 0 near (𝑥, 𝑦) = (1 , 2).

Sol :- By newton’s method we have
𝑥𝑘+1 𝑥𝑘 𝑓(𝑥𝑘 , 𝑦𝑘 )
[𝑦 ] = [𝑦 ] − 𝐽𝑘−1 [ ]------------------------------- ①
𝑘+1 𝑘 𝑔(𝑥𝑘 , 𝑦𝑘 )
𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]
𝑥 𝑦
Give 𝑓(𝑥 , 𝑦) = 𝑦𝑐𝑜𝑠(𝑥𝑦) + 1
𝑓(𝑥0 , 𝑦0 ) = 𝑦0 𝑐𝑜𝑠(𝑥0 𝑦0 ) + 1
𝑓(1 , 2) = (2)(cos (1𝑋2) + 1
𝑓(𝑥0 , 𝑦0 ) = 0.1677063269
 𝑓𝑥 = 𝑦 ∗ (− sin(𝑥𝑦)) ∗ 𝑦𝑓𝑥0 = 𝑦02 ∗ (− sin(𝑥0 𝑦0 ))
𝑓𝑥0 = (2)2 ∗ (− sin(1𝑋2))𝑓𝑥0 = −3.637189707
𝑓𝑦 = 𝑦 ∗ (− sin(𝑥𝑦) ∗ 𝑥) + cos(𝑥𝑦)
 𝑓𝑦0 = 𝑦0 ∗ 𝑥0 ∗ (− sin(𝑥0 𝑦0 )) + cos(𝑥0 𝑦0 )
 𝑓𝑦0 = 2 ∗ 1 ∗ (− sin(1𝑋2)) + cos(1𝑋2)
 𝑓𝑦0 = −2.23474169

𝑔(𝑥 , 𝑦) = sin(𝑥𝑦) + 𝑥 − 𝑦
𝑔(𝑥0 , 𝑦0 ) = sin(𝑥0 𝑦0 ) + 𝑥0 − 𝑦0
𝑔(1 , 2) = sin(1𝑋2) + 1 − 2
𝑔(𝑥0 , 𝑦0 ) = −0.09070257317
𝑔𝑥 = (cos(𝑥𝑦) ∗ 𝑦) + 1𝑔𝑥0 = (cos(𝑥0 𝑦0 ) ∗ 𝑦0 ) + 1
𝑔𝑥0 = (cos(1𝑋2) ∗ 2) + 1𝑔𝑥0 = 0.1677063269
𝑔𝑦 = (cos(𝑥𝑦) ∗ 𝑥) − 1𝑔𝑦0 = (𝑐𝑜𝑠(𝑥0 𝑦0 ) ∗ 𝑥0 ) − 1
𝑔𝑦0 = (cos(1𝑋2) ∗ 1) − 1𝑔𝑦0 = −1.416146837
𝑓 𝑓
Then 𝐽0 = [𝑔𝑥0 𝑔𝑦0 ]
𝑥0 𝑦0
−3.6371 −2.2347
𝐽𝑘 = [ ]
0.1677 −1.4161
−0.2562 0.4044
𝐽0−1 = [ ]
−0.0303 −0.6582
For the first iteration 𝑘 = 0 and equation 1 becomes
𝑥0+1 𝑥0 𝑓(𝑥0 , 𝑦0 )
[𝑦 ] = [𝑦 ] − 𝐽0−1 [ ]
0+1 0 𝑔(𝑥0 , 𝑦0 )
𝑥1 1 −0.2562 0.4044 0.1677
[𝑦 ] = [ ] − [ ][ ]
1 2 −0.0303 −0.6582 −0.0907
𝑥1 1 −0.07964
[𝑦 ] = [ ] − [ ]
1 2 0.05461
𝑥1 1.07964
[𝑦 ] = [ ]
1 1.94539
For the second iteration 𝑘 = 1 and equation 1 becomes
𝑥1+1 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽𝑘−1 [ ]
1+1 1 𝑔(𝑥1 , 𝑦1 )
𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]
𝑥 𝑦
Give 𝑓(𝑥 , 𝑦) = 𝑦𝑐𝑜𝑠(𝑥𝑦) + 1
𝑓(𝑥1 , 𝑦1 ) = 𝑦1 𝑐𝑜𝑠(𝑥1 𝑦1 ) + 1
𝑓(1.0796 , 1.9453) = (1.9453) ∗ (cos (1.0796𝑋1.9453)) + 1
𝑓(𝑥1 , 𝑦1 ) = 0.0176779213
𝑓𝑥 = 𝑦 ∗ (− sin(𝑥𝑦)) ∗ 𝑦  𝑓𝑥1 = 𝑦12 ∗ (− sin(𝑥1 𝑦1 ))
 𝑓𝑥1 = (1.9453)2 ∗ (− sin(1.0796𝑋1.9453))𝑓𝑥1 = −3.266271328
𝑓𝑦 = 𝑦 ∗ (− sin(𝑥𝑦) ∗ 𝑥) + cos(𝑥𝑦)
 𝑓𝑦1 = (1.0796𝑋1.9453) ∗ (− sin(1.0796𝑋1.9453)) +
cos(1.0796𝑋1.9453)
  𝑓𝑦1 = −0.7521959902

𝑔(𝑥 , 𝑦) = sin(𝑥𝑦) + 𝑥 − 𝑦
𝑔(𝑥1 , 𝑦1 ) = sin(𝑥1 𝑦1 ) + 𝑥1 − 𝑦1
𝑔(1.0796 , 1.9453) = sin(1.0796𝑋1.9453) + (1.0796) − (1.9453)
𝑔(𝑥1 , 𝑦1 ) = −2.564289486 𝑋 10−3
𝑔𝑥 = (cos(𝑥𝑦) ∗ 𝑦) + 1𝑔𝑥1 = (cos(𝑥1 𝑦1 ) ∗ 𝑦1 ) + 1
 𝑔𝑥1 = (cos(1.0796𝑋1.9453) ∗ 1.9453) + 1𝑔𝑥1 = 0.0176779213
𝑔𝑦 = (cos(𝑥𝑦) ∗ 𝑥) − 1𝑔𝑦1 = (𝑐𝑜𝑠(𝑥1 𝑦1 ) ∗ 𝑥1 ) − 1
𝑔𝑦1 = (cos(1.0796𝑋1.9453) ∗ 1.0796) − 1𝑔𝑦1 = −1.545167797
𝑓𝑥 𝑓𝑦
Then 𝐽1 = [𝑔 1 𝑔 1 ]
𝑥1 𝑦1
−3.2662 −0.7521
𝐽1 = [ ]
0.0176 −1.5451
−0.3053 0.1486
𝐽1−1 = [ −3 ]
−3.4783 𝑋 10 −0.6455
𝑥1+1 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽1−1 [ ]
1+1 1 𝑔(𝑥1 , 𝑦1 )
𝑥2 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽1−1 [ ]
2 1 𝑔(𝑥1 , 𝑦1 )
𝑥2 1.0796 −0.3053 0.1486 0.0176
[𝑦 ] = [ ]−[ ][ ]
2 1.9453 −3.4783 𝑋 10 −3
−0.6455 −2.5642 𝑋 10−3
𝑥2 0.8995 −3
[𝑦 ] = [ ] − [−5.7543 𝑋 10−3 ]
2 2.0407 1.5939 𝑋 10
𝑥2 0.9052
[𝑦 ] = [ ]
2 2.0391

5. log(𝑥 2 + 𝑦) − 1 + 𝑦 = 0 & √𝑥 + 𝑥𝑦 = 0 near (𝑥, 𝑦) = (2.4 , −0.6).

Sol :- By newton’s method we have
𝑥𝑘+1 𝑥𝑘 𝑓(𝑥𝑘 , 𝑦𝑘 )
[𝑦 ] = [𝑦 ] − 𝐽𝑘−1 [ ]------------------------------- ①
𝑘+1 𝑘 𝑔(𝑥 𝑘 , 𝑦𝑘 )
𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]
𝑥 𝑦
Give 𝑓(𝑥 , 𝑦) = log(𝑥 2 + 𝑦) − 1 + 𝑦
𝑓(𝑥0 , 𝑦0 ) = log(𝑥02 + 𝑦0 ) − 1 + 𝑦0
𝑓(2.4 , −0.6) = log((2.4)2 − 0.6) − 1 + (−0.6)
𝑓(2.4 , −0.6) = log(5.16) − 1 + (−0.6)
𝑓(𝑥0 , 𝑦0 ) = −0.8873502984
1
𝑓𝑥 = ((𝑥 2+𝑦) ∗ 2𝑥)
1 1
𝑓𝑥0 = ((𝑥 2 +𝑦 ) ∗ 2𝑥0 ) 𝑓𝑥0 = ((2.4)2+(−0.6) ∗ 2(2.4))
0 0

𝑓𝑥0 = 0.9302325581
1
𝑓𝑦 = ((𝑥 2 +𝑦) ∗ 𝑦) + 1
 1
 𝑓𝑦0 = ((𝑥 2 +𝑦 ) ∗ 𝑦0 ) + 1
0 0

1
 𝑓𝑦0 = ((2.4)2+(−0.6) ∗ (−0.6)) + 1

 𝑓𝑦0 = 1.19379845

𝑔(𝑥 , 𝑦) = √𝑥 + 𝑥𝑦
𝑔(𝑥0 , 𝑦0 ) = √𝑥0 + 𝑥0 𝑦0
𝑔(2.4 , −0.6) = √2.4 + (2.4 ∗ −0.6)
𝑔(𝑥0 , 𝑦0 ) = 0.1091933385
1 1
𝑔𝑥 = (2 𝑥) + 𝑦  𝑔𝑥0 = (2 ) + 𝑦0
√ √ 𝑥0
1
 𝑔𝑥0 = (2√2.4) + (−0.6)  𝑔𝑥0 = −0.27772513878
𝑔𝑦 = 𝑥  𝑔𝑦0 = 𝑥0  𝑔𝑦0 = (2.4)
 𝑔𝑦0 = 2.4
𝑓 𝑓
Then 𝐽0 = [𝑔𝑥0 𝑔𝑦0 ]
𝑥0 𝑦0
0.9302 1.1937
𝐽𝑘 = [ ]
−0.2777 2.4
0.9363 −0.4657
Also, 𝐽0−1 = [ ]
0.1080 0.3629
For the first iteration 𝑘 = 0 and equation 1 becomes
𝑥0+1 𝑥0 𝑓(𝑥0 , 𝑦0 )
[𝑦 ] = [𝑦 ] − 𝐽0−1 [ ]
0+1 0 𝑔(𝑥0 , 𝑦0 )
𝑥1 2.4 0.9363 −0.4657 −0.8873
[𝑦 ] = [ ]−[ ][ ]
1 −0.6 0.1080 0.3629 0.1091
𝑥1 2.4 −0.8815
[𝑦 ] = [ ]−[ ]
1 −0.6 −0.0562
𝑥1 3.2815
[𝑦 ] = [ ]
1 −0.5438
For the second iteration 𝑘 = 1 and equation 1 becomes
𝑥1+1 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽𝑘−1 [ ]
1+1 1 𝑔(𝑥1 , 𝑦1 )
𝑓 𝑓
Where 𝐽𝑘 = [𝑔𝑥 𝑔𝑦 ]
𝑥 𝑦
Give 𝑓(𝑥 , 𝑦) = log(𝑥 2 + 𝑦) − 1 + 𝑦
𝑓(𝑥1 , 𝑦1 ) = log(𝑥12 + 𝑦1 ) − 1 + 𝑦1
𝑓(3.2815 , −0.5438) = log((3.2815)2 − 0.5438) − 1 + (−0.5438)
𝑓(3.2815 , −0.5438) = log(10.2244) − 1 + (−0.5438)
𝑓(𝑥1 , 𝑦1 ) = −0.5341603737
1
𝑓𝑥 = ((𝑥 2+𝑦) ∗ 2𝑥)
1 1
𝑓𝑥1 = ((𝑥 2 +𝑦 ) ∗ 2𝑥1 ) 𝑓𝑥1 = ((3.2815)2+(−0.5438) ∗ 2(3.2815))
1 1
 𝑓𝑥1 = 0.6418932045
1
𝑓𝑦 = ((𝑥 2 +𝑦) ∗ 𝑦) + 1
1
 𝑓𝑦1 = ((𝑥 2 +𝑦 ) ∗ 𝑦1 ) + 1
1 1

1
 𝑓𝑦1 = ((3.2815)2+(−0.5438) ∗ (−0.5438)) + 1

 𝑓𝑦1 = 1.24521809

𝑔(𝑥 , 𝑦) = √𝑥 + 𝑥𝑦
𝑔(𝑥1 , 𝑦1 ) = √𝑥1 + 𝑥1 𝑦1
𝑔(3.2815 , −0.5438) = √3.2815 + (3.2815 ∗ −0.5438)
𝑔(𝑥1 , 𝑦1 ) = 0.02701139852
1 1
𝑔𝑥 = (2 𝑥) + 𝑦  𝑔𝑥1 = (2 ) + 𝑦1
√ √ 𝑥1
1
 𝑔𝑥1 = (2√3.2815) + (−0.5438)  𝑔𝑥1 = −0.2677842909
𝑔𝑦 = 𝑥  𝑔𝑦1 = 𝑥1  𝑔𝑦1 = (3.2815)
 𝑔𝑦0 = 3.2815
𝑓 𝑓
Then 𝐽1 = [𝑔𝑥1 𝑔𝑦1 ]
𝑥1 𝑦1
0.6418 1.2452
𝐽1 = [ ]
−0.2677 3.2815
1.3452 −0.5104
𝐽1−1 = [ ]
0.1097 0.2630
𝑥1+1 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽1−1 [ ]
1+1 1 𝑔(𝑥1 , 𝑦1 )
𝑥2 𝑥1 𝑓(𝑥1 , 𝑦1 )
[𝑦 ] = [𝑦 ] − 𝐽1−1 [ ]
2 1 𝑔(𝑥1 , 𝑦1 )
𝑥2 3.2815 1.3452 −0.5104 −0.5341
[𝑦 ] = [ ]−[ ][ ]
2 −0.5438 0.1097 0.2630 0.0270
𝑥2 3.2815 −0.7322
[𝑦 ] = [ ]−[ ]
2 −0.5438 −0.0514
𝑥2 4.0137
[𝑦 ] = [ ]
2 −0.4924

Approximate eigen value and corresponding eigen value using Jacobi’s method

Problems:

1. Solve by using Jacobian’s method

𝟏 √𝟐 𝟒
𝐀 = [√𝟐 𝟑 √𝟐]
𝟒 √𝟐 𝟏
 The largest off diagonal elements |𝑎𝑖𝑘 | = |𝑎13 | = |4|.
Here 𝑖 = 1 and 𝑘 = 3.
2𝑎𝑖𝑘
Then to determine  we have tan(2) = .
𝑎𝑖𝑖 −𝑎𝑘𝑘
2(4)
tan(2) = .
1−1
8
 tan(2) = 0.
8 
tan−1 ( )
 = 0
 = 2
.
2 2

  = 4.
𝑎𝑖𝑖 𝑎𝑖𝑘 𝑎11 𝑎13 𝑐𝑜𝑠 −𝑠𝑖𝑛
Then the submatrix is [𝑎 𝑎 ] = [𝑎 𝑎33 ] = [ 𝑠𝑖𝑛 𝑐𝑜𝑠 ].
𝑘𝑖 𝑘𝑘 31
𝑐𝑜𝑠 0 𝑠𝑖𝑛
i.e,. 𝑅1 = [ 0 1 0 ]
𝑠𝑖𝑛 0 𝑐𝑜𝑠
 
cos (4) 0 −sin (4)
 𝑅1 = [ 0 1 0 ].
 
sin (4) 0 cos (4)
0.7071 0 −0.7170
 𝑅1 = [ 0 1 0 ].
0.7071 0 0.7071
Now let us denote 𝐷1 = 𝑅1−1 𝐴𝑅1 .
 𝐷1 = 𝑅1−1 𝐴𝑅1 = 𝑅1𝑇 𝐴𝑅1 .
0.7071 0 0.7170 1 √2 4 0.7071 0 −0.7170
 𝐷1 = [ 0 1 0 ] ∗ [√2 3 √2] ∗ [ 0 1 0 ]
−0.7071 0 0.7071 4 √2 1 0.7071 0 0.7071
5 2 0
 𝐷1 = [2 3 0 ]. This is not diagonal matrix.
0 0 −3
So repeat the process. i.e,.. 𝐷2 = 𝑅1−1 𝐷1 𝑅1.
5 2 0
Let us consider 𝐷1 = [2 3 0 ].
0 0 −3
The largest off diagonal elements |𝑎𝑖𝑘 | = |𝑎12 | = |2|.
Here 𝑖 = 1 and 𝑘 = 2.
2𝑎𝑖𝑘
Then to determine  we have tan(2) = .
𝑎𝑖𝑖 −𝑎𝑘𝑘
2(2)
tan(2) = .
5 −3
4
 tan(2) = 2.
tan−1 (2) 1.1071
 =  = = 0.5535.
2 2
  = 0.5535.
𝑎𝑖𝑖 𝑎𝑖𝑘 𝑎11 𝑎12 𝑐𝑜𝑠 −𝑠𝑖𝑛
Then the submatrix is [𝑎 𝑎𝑘𝑘 ] = [𝑎21 𝑎22 ] = [ 𝑠𝑖𝑛 𝑐𝑜𝑠 ].
𝑘𝑖
 𝑐𝑜𝑠 −𝑠𝑖𝑛 0
i.e,. 𝑅2 = [ 𝑠𝑖𝑛 𝑐𝑜𝑠 0]
0 0 1
cos(0.5535) − sin(0.5535) 0
 𝑅2 = [ sin(0.5535) cos(0.5535) 0]
0 0 1
0.8506 −0.5256 0
 𝑅2 = [0.5256 0.8506 0].
0 0 1
Now let us denote 𝐷2 = 𝑅2−1 𝐷1 𝑅2 .
 𝐷2 = 𝑅2−1 𝐷1 𝑅2 = 𝑅2𝑇 𝐷1 𝑅2 .
0.8506 0.5256 0 5 2 0 0.8506 −0.5256 0
 𝐷2 = [−0.5256 0.8506 0] ∗ [2 3 0 ] ∗ [0.5256 0.8506 0].
0 0 1 0 0 −3 0 0 1
−3
6.2243 1.25𝑋10 0
 𝐷2 = [1.25𝑋10−3 1.76 0].
0 0 1
for our convenience let us consider 𝐷2 as follow.
6.2243 0 0
 𝐷2 = [ 0 1.76 0] .
0 0 1
 𝐷2 is now a diagonal matrix .
Then the diagonal elements so obtained are the approximate eigen value.
i.e,. 1 = 6.2243, 2 = 1.76 and 3 = 1 are the approximate eigen value.
Now let us determine eigen vector for the corresponding egien value,
To determine eigen vector we have to multiply 𝑅1 & 𝑅2 .
So simply we call 𝑅 = 𝑅1 ∗ 𝑅2
0.7071 0 −0.7170 0.8506 −0.5256 0
Then 𝑅 = [ 0 1 0 ] ∗ [0.5256 0.8506 0]
0.7071 0 0.7071 0 0 1
0.6014 −0.371 −0.7071
 𝑅 = [0.5256 0.8506 0 ].
0.6014 −0.371 0.7071
It means 1 = 𝑋1 , 2 = 𝑋2 and 3 = 𝑋3 are the corresponding eigen vector for eigen value  in
matrix R.
Where 𝑋1 , 𝑋2 & 𝑋3 represents first, second & third columns respectively.
0.6014 −0.371 −0.7071
 𝑋1 = [0.5256], 𝑋2 = [ 0.8506 ] and 𝑋3 = [ 0 ].
0.6014 −0.371 0.7071
Are the corresponding eigen vector for ′𝑠.

2. Solve by to determine eigen value and corresponding eigen vector using Jacobian’s method.
𝟓 𝟎 𝟏
𝑨 = [𝟎 −𝟐 𝟎]
𝟏 𝟎 𝟓

The largest off diagonal elements |𝑎𝑖𝑘 | = |𝑎13 | = |1|.

Here 𝑖 = 1 and 𝑘 = 3.
 2𝑎𝑖𝑘
Then to determine  we have tan(2) = .
𝑎𝑖𝑖 −𝑎𝑘𝑘
2(4)
tan(2) = .
1−1
8
 tan(2) = 0.
8 
tan−1 ( )
 = 0
 = 2
.
2 2

  = 4.
𝑎𝑖𝑖 𝑎𝑖𝑘 𝑎11 𝑎13 𝑐𝑜𝑠 −𝑠𝑖𝑛
Then the submatrix is [𝑎 𝑎𝑘𝑘 ] = [ 𝑎31 𝑎33 ] = [ 𝑠𝑖𝑛 𝑐𝑜𝑠 ].
𝑘𝑖
𝑐𝑜𝑠 0 𝑠𝑖𝑛
i.e,. 𝑅1 = [ 0 1 0 ]
𝑠𝑖𝑛 0 𝑐𝑜𝑠
 
cos (4) 0 −sin (4)
 𝑅1 = [ 0 1 0 ].
 
sin (4) 0 cos (4)
0.7071 0 −0.7170
 𝑅1 = [ 0 1 0 ].
0.7071 0 0.7071
Now let us denote 𝐷1 = 𝑅1−1 𝐴𝑅1 .
 𝐷1 = 𝑅1−1 𝐴𝑅1 = 𝑅1𝑇 𝐴𝑅1 .
0.7071 0 0.7170 5 0 1 0.7071 0 −0.7170
 𝐷1 = [ 0 1 0 ] ∗ [0 −2 0] ∗ [ 0 1 0 ]
−0.7071 0 0.7071 1 0 5 0.7071 0 0.7071
6 0 0
 𝐷1 = [0 −2 0]. This is diagonal matrix.
0 0 4
 𝐷1 is now a diagonal matrix .
Then the diagonal elements so obtained are the approximate eigen value.
i.e,. 1 = 6 , 2 = −2 and 3 = 4 are the approximate eigen value and the corresponding eigen
vector is given by.
0.7071 0 −0.7170
𝑅= [ 0 1 0 ]
0.7071 0 0.7071
It means 1 = 𝑋1 , 2 = 𝑋2 and 3 = 𝑋3 are the corresponding eigen vector for eigen value  in
matrix R.
Where 𝑋1 , 𝑋2 & 𝑋3 represents first, second & third columns respectively.
0.7071 0 −0.7071
 𝑋1 = [ 0 ], 𝑋2 = [1] and 𝑋3 = [ 0 ].
0.7071 0 0.7071
Are the corresponding eigen vector for ′𝑠.
𝟑 𝟐 𝟐
3. 𝐀 = [𝟐 𝟓 𝟐]
𝟐 𝟐 𝟑
The largest off diagonal elements |𝑎𝑖𝑘 | = |𝑎13 | = |2|.
Here 𝑖 = 1 and 𝑘 = 3.
 2𝑎𝑖𝑘
Then to determine  we have tan(2) = .
𝑎𝑖𝑖 −𝑎𝑘𝑘
2(4)
tan(2) = .
1−1
8
 tan(2) = 0.
8 
tan−1 ( )
 = 0
 = 2
.
2 2

  = 4.
𝑎𝑖𝑖 𝑎𝑖𝑘 𝑎11 𝑎13 𝑐𝑜𝑠 −𝑠𝑖𝑛
Then the submatrix is [𝑎 ] = [ 𝑎33 ] = [ 𝑠𝑖𝑛 𝑐𝑜𝑠 ].
𝑘𝑖 𝑎𝑘𝑘 𝑎31
𝑐𝑜𝑠 0 𝑠𝑖𝑛
i.e,. 𝑅1 = [ 0 1 0 ]
𝑠𝑖𝑛 0 𝑐𝑜𝑠
 
cos (4) 0 −sin (4)
 𝑅1 = [ 0 1 0 ].
 
sin (4) 0 cos (4)
0.7071 0 −0.7170
 𝑅1 = [ 0 1 0 ].
0.7071 0 0.7071
Now let us denote 𝐷1 = 𝑅1−1 𝐴𝑅1 .
 𝐷1 = 𝑅1−1 𝐴𝑅1 = 𝑅1𝑇 𝐴𝑅1 .
0.7071 0 0.7170 3 2 2 0.7071 0 −0.7170
 𝐷1 = [ 0 1 0 ] ∗ [2 5 2] ∗ [ 0 1 0 ]
−0.7071 0 0.7071 2 2 3 0.7071 0 0.7071
5 2.8284 0
 𝐷1 = [2.8284 5 0]. This is not a diagonal matrix.
0 0 1
So repeat the process. i.e,.. 𝐷2 = 𝑅1−1 𝐷1 𝑅1.
5 2.8284 0
Let us consider 𝐷1 = [2.8284 5 0].
0 0 1
The largest off diagonal elements |𝑎𝑖𝑘 | = |𝑎12 | = |2.8284|.
Here 𝑖 = 1 and 𝑘 = 2.
2𝑎𝑖𝑘
Then to determine  we have tan(2) = .
𝑎𝑖𝑖 −𝑎𝑘𝑘
2(2.8284)
tan(2) = .
5−5
5.6568
 tan(2) = .
0
tan−1 (2) 
 =   = 4.
2
𝑎𝑖𝑖 𝑎𝑖𝑘 𝑎11 𝑎12 𝑐𝑜𝑠 −𝑠𝑖𝑛
Then the submatrix is [𝑎 𝑎𝑘𝑘 ] = [ 𝑎21 𝑎22 ] = [ 𝑠𝑖𝑛 𝑐𝑜𝑠 ].
𝑘𝑖
 
cos (4) − sin (4) 0
i.e,. 𝑅2 = [ sin () cos (4)

0]
4
0 0 1
 0.7071 −0.7170 0
 𝑅2 = [0.7071 0.7071 0].
0 0 1
−1
Now let us denote 𝐷2 = 𝑅2 𝐷1 𝑅2 .
 𝐷2 = 𝑅2−1 𝐷1 𝑅2 = 𝑅2𝑇 𝐷1 𝑅2 .
0.7071 0 0.7170 5 2.8284 0 0.7071 0 −0.7170
 𝐷2 = [ 0 1 0 ] ∗ [2.8284 5 0] ∗ [ 0 1 0 ].
−0.7071 0 0.7071 0 0 1 0.7071 0 0.7071

7.8284 0 0
 𝐷2 = [ 0 2.1715 0].
0 0 1
 𝐷2 is now a diagonal matrix .
Then the diagonal elements so obtained are the approximate eigen value.
i.e,. 1 = 7.8284 , 2 = 2.1715 and 3 = 1 are the approximate eigen value.
Now let us determine eigen vector for the corresponding egien value,
To determine eigen vector we have to multiply 𝑅1 & 𝑅2 .
So simply we call 𝑅 = 𝑅1 ∗ 𝑅2
0.7071 0 −0.7170 0.7071 −0.7170 0
Then 𝑅 = [ 0 1 0 ] ∗ [0.7071 0.7071 0]
0.7071 0 0.7071 0 0 1
0.4999 −0.499 −0.707
 𝑅 = [0.7071 0.7071 0 ].
0.4999 −0.499 0.7071
It means 1 = 𝑋1 , 2 = 𝑋2 and 3 = 𝑋3 are the corresponding eigen vector for eigen value  in
matrix R.
Where 𝑋1 , 𝑋2 & 𝑋3 represents first, second & third columns respectively.
0.4999 −0.499 −0.707
 𝑋1 = [0.7071], 𝑋2 = [ 0.7071 ] and 𝑋3 = [ 0 ].
0.4999 −0.499 0.7071
Are the corresponding eigen vector for ′𝑠.
𝟏 √𝟑 𝟒
4. 𝐀 = [√𝟑 𝟓 √𝟑]
𝟒 √𝟑 𝟏
The largest off diagonal elements |𝑎𝑖𝑘 | = |𝑎13 | = |4|.
Here 𝑖 = 1 and 𝑘 = 3.
2𝑎𝑖𝑘
Then to determine  we have tan(2) = .
𝑎𝑖𝑖 −𝑎𝑘𝑘
2(4)
tan(2) = .
1−1
8
 tan(2) = 0.
8 
tan−1 ( )
 = 0
 = 2
.
2 2

  = 4.
 𝑎𝑖𝑖 𝑎𝑖𝑘 𝑎11 𝑎13 𝑐𝑜𝑠 −𝑠𝑖𝑛
Then the submatrix is [𝑎 𝑎𝑘𝑘 ] = [ 𝑎31 𝑎33 ] = [ 𝑠𝑖𝑛 𝑐𝑜𝑠 ].
𝑘𝑖
𝑐𝑜𝑠 0 𝑠𝑖𝑛
i.e,. 𝑅1 = [ 0 1 0 ]
𝑠𝑖𝑛 0 𝑐𝑜𝑠
 
cos (4) 0 −sin (4)
 𝑅1 = [ 0 1 0 ].
 
sin (4) 0 cos (4)
0.7071 0 −0.7170
 𝑅1 = [ 0 1 0 ].
0.7071 0 0.7071
Now let us denote 𝐷1 = 𝑅1−1 𝐴𝑅1 .
 𝐷1 = 𝑅1−1 𝐴𝑅1 = 𝑅1𝑇 𝐴𝑅1 .
0.7071 0 0.7170 1 √3 4 0.7071 0 −0.7170
 𝐷1 = [ 0 1 0 ] ∗ [√3 5 √3] ∗ [ 0 1 0 ]
−0.7071 0 0.7071 4 √3 1 0.7071 0 0.7071
5 2.4494 0
 𝐷1 = [2.4494 5 0 ]. This is not a diagonal matrix.
0 0 −3
So repeat the process. i.e,.. 𝐷2 = 𝑅1−1 𝐷1 𝑅1.
5 2.4494 0
Let us consider 𝐷1 = [2.4494 5 0 ].
0 0 −3
The largest off diagonal elements |𝑎𝑖𝑘 | = |𝑎12 | = |2.4494|.
Here 𝑖 = 1 and 𝑘 = 2.
2𝑎𝑖𝑘
Then to determine  we have tan(2) = .
𝑎𝑖𝑖 −𝑎𝑘𝑘
2(2.4494)
tan(2) = .
5−5
4.8988
 tan(2) = .
0
tan−1 (2) 
 =   = 4.
2
𝑎𝑖𝑖 𝑎𝑖𝑘 𝑎11 𝑎12 𝑐𝑜𝑠 −𝑠𝑖𝑛
Then the submatrix is [𝑎 𝑎𝑘𝑘 ] = [ 𝑎21 𝑎22 ] = [ 𝑠𝑖𝑛 𝑐𝑜𝑠 ].
𝑘𝑖
 
cos (4) − sin (4) 0
i.e,. 𝑅2 = [ sin () 
cos ( ) 0]
4 4
0 0 1
0.7071 −0.7170 0
 𝑅2 = [0.7071 0.7071 0].
0 0 1
−1
Now let us denote 𝐷2 = 𝑅2 𝐷1 𝑅2 .
 𝐷2 = 𝑅2−1 𝐷1 𝑅2 = 𝑅2𝑇 𝐷1 𝑅2 .
0.7071 0 0.7170 5 2.4494 0 0.7071 0 −0.7170
 𝐷2 = [ 0 1 0 ] ∗ [2.4494 5 0 ]∗[ 0 1 0 ].
−0.7071 0 0.7071 0 0 −3 0.7071 0 0.7071
 7.4494 0 0
 𝐷2 = [ 0 2.5505 0 ].
0 0 −3
 𝐷2 is now a diagonal matrix .
Then the diagonal elements so obtained are the approximate eigen value.
i.e,. 1 = 7.4494 , 2 = 2.5505 and 3 = −3 are the approximate eigen value.
Now let us determine eigen vector for the corresponding egien value,
To determine eigen vector we have to multiply 𝑅1 & 𝑅2 .
So simply we call 𝑅 = 𝑅1 ∗ 𝑅2
0.7071 0 −0.7170 0.7071 −0.7170 0
Then 𝑅 = [ 0 1 0 ] ∗ [0.7071 0.7071 0]
0.7071 0 0.7071 0 0 1
0.4999 −0.499 −0.707
 𝑅 = [0.7071 0.7071 0 ].
0.4999 −0.499 0.7071
It means 1 = 𝑋1 , 2 = 𝑋2 and 3 = 𝑋3 are the corresponding eigen vector for eigen value  in
matrix R.
Where 𝑋1 , 𝑋2 & 𝑋3 represents first, second & third columns respectively.
0.4999 −0.499 −0.707
 𝑋1 = [0.7071], 𝑋2 = [ 0.7071 ] and 𝑋3 = [ 0 ].
0.4999 −0.499 0.7071
Are the corresponding eigen vector for ′𝑠.
𝟏 √𝟐 𝟐
5. 𝐀 = [√𝟐 𝟑 √𝟐]
𝟐 √𝟐 𝟏

The largest off diagonal elements |𝑎𝑖𝑘 | = |𝑎13 | = |2|.

Here 𝑖 = 1 and 𝑘 = 3.
2𝑎𝑖𝑘
Then to determine  we have tan(2) = .
𝑎𝑖𝑖 −𝑎𝑘𝑘
2(2)
tan(2) = .
1−1
4
 tan(2) = 0.
8 
tan−1 ( )
 = 0
 = 2
.
2 2

  = 4.
𝑎𝑖𝑖 𝑎𝑖𝑘 𝑎11 𝑎13 𝑐𝑜𝑠 −𝑠𝑖𝑛
Then the submatrix is [𝑎 𝑎𝑘𝑘 ] = [𝑎31 𝑎33 ] = [ 𝑠𝑖𝑛 𝑐𝑜𝑠 ].
𝑘𝑖
𝑐𝑜𝑠 0 𝑠𝑖𝑛
i.e,. 𝑅1 = [ 0 1 0 ]
𝑠𝑖𝑛 0 𝑐𝑜𝑠
 
cos (4) 0 −sin (4)
 𝑅1 = [ 0 1 0 ].
 
sin (4) 0 cos (4)
 0.7071 0 −0.7170
 𝑅1 = [ 0 1 0 ].
0.7071 0 0.7071
Now let us denote 𝐷1 = 𝑅1−1 𝐴𝑅1 .
 𝐷1 = 𝑅1−1 𝐴𝑅1 = 𝑅1𝑇 𝐴𝑅1 .
0.7071 0 0.7170 1 √2 2 0.7071 0 −0.7170
 𝐷1 = [ 0 1 0 ] ∗ [√2 3 √2] ∗ [ 0 1 0 ]
−0.7071 0 0.7071 2 √2 1 0.7071 0 0.7071
3 2 0
 𝐷1 = [2 3 0 ]. This is not a diagonal matrix.
0 0 −1
So repeat the process. i.e,.. 𝐷2 = 𝑅1−1 𝐷1 𝑅1.
3 2 0
Let us consider 𝐷1 = [2 3 0 ].
0 0 −1
The largest off diagonal elements |𝑎𝑖𝑘 | = |𝑎12 | = |2.4494|.
Here 𝑖 = 1 and 𝑘 = 2.
2𝑎𝑖𝑘
Then to determine  we have tan(2) = .
𝑎𝑖𝑖 −𝑎𝑘𝑘
2(2.4494)
tan(2) = .
5−5
4.8988
 tan(2) = .
0
tan−1 (2) 
 =   = 4.
2
𝑎𝑖𝑖 𝑎𝑖𝑘 𝑎11 𝑎12 𝑐𝑜𝑠 −𝑠𝑖𝑛
Then the submatrix is [𝑎 𝑎𝑘𝑘 ] = [ 𝑎21 𝑎22 ] = [ 𝑠𝑖𝑛 𝑐𝑜𝑠 ].
𝑘𝑖
 
cos (4) − sin (4) 0
i.e,. 𝑅2 = [ sin () 
cos (4) 0]
4
0 0 1
0.7071 −0.7170 0
 𝑅2 = [0.7071 0.7071 0].
0 0 1
−1
Now let us denote 𝐷2 = 𝑅2 𝐷1 𝑅2 .
 𝐷2 = 𝑅2−1 𝐷1 𝑅2 = 𝑅2𝑇 𝐷1 𝑅2 .
0.7071 0 0.7170 3 2 0 0.7071 0 −0.7170
 𝐷2 = [ 0 1 0 ] ∗ [2 3 0 ] ∗ [ 0 1 0 ].
−0.7071 0 0.7071 0 0 −1 0.7071 0 0.7071
5 0 0
 𝐷2 = [0 1 0 ].
0 0 −1
 𝐷2 is now a diagonal matrix .
Then the diagonal elements so obtained are the approximate eigen value.
i.e,. 1 = 5 , 2 = 1 and 3 = −1 are the approximate eigen value.
Now let us determine eigen vector for the corresponding egien value,
To determine eigen vector we have to multiply 𝑅1 & 𝑅2 .
So simply we call 𝑅 = 𝑅1 ∗ 𝑅2
 0.7071 0 −0.7170 0.7071 −0.7170 0
Then 𝑅 = [ 0 1 0 ] ∗ [ 0.7071 0.7071 0]
0.7071 0 0.7071 0 0 1

0.4999 −0.499 −0.707

 𝑅 = [0.7071 0.7071 0 ].
0.4999 −0.499 0.7071

It means 1 = 𝑋1 , 2 = 𝑋2 and 3 = 𝑋3 are the corresponding eigen vector for eigen value  in
matrix R.

Where 𝑋1 , 𝑋2 & 𝑋3 represents first, second & third columns respectively.

0.4999 −0.499 −0.707

 𝑋1 = [0.7071], 𝑋2 = [ 0.7071 ] and 𝑋3 = [ 0 ].
0.4999 −0.499 0.7071

Are the corresponding eigen vector for ′𝑠.