Scribd
Upload
13 views2 pages

Newton - S Method

Newton's Method uses linear approximations to iteratively find roots of complicated functions by taking tangents to the graph at estimated points and finding where they intersect the x-axis. It starts with an initial guess x1 and uses the formula xn+1 = xn - f(xn)/f'(xn) to get closer to the root with each iteration. Two examples are provided to demonstrate finding the square root of 100 and the root of x5 = 5-x.

Uploaded by

limanesoyama
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
13 views2 pages

Newton - S Method

Newton's Method uses linear approximations to iteratively find roots of complicated functions by taking tangents to the graph at estimated points and finding where they intersect the x-axis. It starts with an initial guess x1 and uses the formula xn+1 = xn - f(xn)/f'(xn) to get closer to the root with each iteration. Two examples are provided to demonstrate finding the square root of 100 and the root of x5 = 5-x.

Uploaded by

limanesoyama
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Newton’s Method

Suppose you had a really complicated function of one variable and you were trying to find a root or
zero. You cannot solve it algebraically, but most calculators or computer programs can, most using
Newton’s Method. Newton’s Method uses linear approximations to find a root to a complicated
function.
𝑓(𝑥)

y
(𝑥1 , 𝑓(𝑥1 )

𝑟 𝑥3 𝑥2 𝑥

We want to solve 𝑓(𝑥) = 0, so the roots correspond to the 𝑥 −intercepts of the graph; r labelled in
the above graph. We start with using linear approximation at 𝑥1 , we find the 𝑥 −intercept of the
tangent line, 𝑥2 . In other words, we use a tangent line at (𝑥1 , 𝑓(𝑥1 )) to find 𝒙𝟐 , where the tangent
line intercepts the x-axis: 𝑦 − 𝑓(𝑥1 ) = 𝑓′(𝑥1 )(𝑥 − 𝑥1 )
Now, at 𝑥2 , 𝑦 = 0. ∴ −𝑓(𝑥1 ) = 𝑓′(𝑥1 )(𝑥2 − 𝑥1 )
𝑓(𝑥 )
Re-arranging 𝑥2 = 𝑥1 − 𝑓′(𝑥1 )
1

Now we use a second approximation to get 𝑥3 , and further linear approximations will get you closer
and closer to the root. Therefore, the nth approximation, we obtain
𝑓(𝑥 )
𝑥𝑛+1 = 𝑥𝑛 − 𝑓′(𝑥𝑛 )
𝑛

So as n becomes larger, we can say the sequence converges to r; lim 𝑥𝑛 = 𝑟


𝑛→∞

A note of caution is sometimes the sequence does not converge. For example when an
approximation lies near a turning point.
y 𝑓(𝑥)
(𝑥1 , 𝑓(𝑥1 )

𝑟 𝑥2 𝑥
Example 1
10
Find √100 using Newton’s method.
Solution
10
Let 𝑥 = √100. Then
𝑥 10 = 100 and
𝑥 10 − 100 = 0 Then 𝑓(𝑥) = 𝑥 10 − 100 and 𝑓 ′ (𝑥) = 10𝑥 9
𝑓(𝑥𝑛 ) (𝑥𝑛 )10 − 100
𝑥𝑛+1 = 𝑥𝑛 − = 𝑥𝑛 −
𝑓 ′ (𝑥𝑛 ) 10(𝑥𝑛 )9
How should we choose our first approximation? Well, 110 = 1 and 210 = 1024. So x should lie
between 1 and 2. Choosing 𝑥1 = 1.5
(1.5)10 −100
Then 𝑥2 = 1.5 − 10(1.5)9

= 1.610122949
𝑥3 = 1.586599872
𝑥4 = 1.58490143
𝑥5 = 1.58489319 = 𝑥6 Therefore, the root is 1.58489 correct to 5dpl.
Example 2
Show that the equation 𝑥 5 = 5 − 𝑥 has only one root. Now, using Newton’s Method, find that
root correct to 3dpl.
Solution
We can use Intermediate Value Theorem (IVT) to show there is one root. But a 𝑥5
rough sketch can help us show approximately where it would be. It lies between 5

0 and 5, but closer to 1. Then let 𝑓(𝑥) = 𝑥 5 + 𝑥 − 5, which is continuous on its 5


domain and by IVT: 5-x

𝑓(1.5) = (1.5)5 + 1.5 − 5 = 4.09375


𝑓(1) = (1)5 + 1 − 5 = −3 The sign change indicates there must be a root.
For only one root, show that 𝑓(𝑥) is either increasing always or decreasing always.
The 𝑓 ′(𝑥) = 5(𝑥)4 + 1 > 0 for all x. Therefore 𝑓 is always increasing and has only one root.
𝑓(𝑥 ) (𝑥𝑛 )5 +𝑥𝑛 −5
Newton’s method to find the exact root: 𝑥𝑛+1 = 𝑥𝑛 − 𝑓′ (𝑥𝑛 ) = 𝑥𝑛 − 5(𝑥𝑛 )4 +1
𝑛

Choosing 𝑥1 = 1.2
(1.2)5 +1.2−5
Then 𝑥2 = 1.2 −
5(1.2)4 +1

= 1.315383533
𝑥3 = 1.2995236552
𝑥4 = 1.29915.. Therefore, the root is 1.299 correct to 3dpl.

You might also like