You are on page 1of 10

W9 Newton’s Method

Suppose you are asked to solve an equation but there is no algebraic method
that will give you the answer?

e.g. Solve the equation 𝑥 𝑥 1 0 (this equation has only one solution)

This is a problem as the equation 𝑥 𝑥 1 0 cannot be solved


algebraically. We will see how Newton’s method can be used to
approximate the solution to this equation to a high level of accuracy.

We know that the solution of the equation 𝑥 𝑥 1 0 has the same


value as the 𝑥 intercept of the graph of 𝑓 𝑥 𝑥 𝑥 1. Hence, if we
can work out the 𝑥 intercept we will know the solution.

Newton’s Method is an iterative process that can approximate the


𝑥 intercept/s of a graph, and hence the solutions to an equation.

Newton’s method involves trying to find the tangent to 𝑓 𝑥 which has the
same 𝑥 intercept as 𝑓 𝑥 .

To understand Newton’s method, the first thing to understand is that the


tangent drawn at the 𝑥 intercept of 𝑓 𝑥 has the same 𝑥 intercept as
𝑓 𝑥 whereas most other tangents will have different x-intercepts. See below.
How To Use Newton’s Method

We start with an initial guess of a point on the graph of 𝑓 𝑥 which is


(hopefully) reasonably close to the 𝑥 intercept.

We then find the equation of the tangent to 𝑓 𝑥 at that point and work out
the 𝑥 intercept of the tangent.

Consider finding the solution to 𝒙𝟓 𝒙 𝟏 𝟎


Iteration 1, using 𝒙 𝟏:
On the graph of 𝑓 𝑥 𝑥 𝑥 1, we can start by choosing the point (1, 1)
and consider the tangent at the point.

The tangent at 𝑥 1, has equaƟon 𝑦 6𝑥 5.

The 𝑥 intercept of this tangent is 𝑥 .


What about the funcƟon’s value?
𝑓 0.235 …

Remember at the 𝑥 intercept 𝑓 𝑥 0, so we are not very close.

So we then use 𝑥 as our new starƟng point and repeat the process
𝟓
IteraƟon 2, using 𝒙 :
𝟔

The tangent at 𝑥 , has equaƟon 𝑦 𝑥 , which has an 𝑥 intercept


of 𝑥

Checking the 𝑦 value:


𝑓 0.025 … We have goƩen closer to the 𝑥 intercept
where 𝑓 𝑥 0
Zoomed in view
The 𝑥 intercept of the 2nd tangent is
much closer to the 𝑥 intercept of 𝑓 𝑥

If we conƟnue this process we get the following


𝑥 int of
IteraƟon 𝑥 𝑓 𝑥
tangent
5
0 1 1
6
5
1 0.23529 … 0.76438 …
6
2 0.76438 … 0.02532 … 0.75502 …
3 0.75502 … 0.00038 … 0.75487 ….
4 0.75487 … 0.00000 …
By the 4th iteraƟon we get 𝑓 𝑥 0.000000093 … which is extremely close to
𝑓 𝑥 0. Hence, 𝑥 0.7549 is a very good approximaƟon for the 𝑥 intercept
of 𝑓 𝑥 𝑥 𝑥 1.

∴ 𝑥 0.7549 (correct to 4 decimal places) is a very good approximaƟon to the


soluƟon of the equaƟon 𝑥 𝑥 1 0.
If we were to repeat the process for another iteraƟon we would get
Tangent
IteraƟon 𝑥 𝑓 𝑥
𝑥 int
4 0.75487 … 0.00000 … 0.75487 …
5 0.75487 …
Because we are so close to the true 𝑥 intercept the next value of 𝑥 we get is the
same as the previous one (to many decimal places).
When going through this iteraƟve process, once we get a repeated value we
stop the process because we can’t get closer to the true intercept.

We can derive a formula for calculating the 𝑥 value at each step to make the
calculations simpler:
At the 𝑛th iteration we can denote the 𝑥 value we are using as 𝑥 , thus the
point on the curve has coordinates 𝑥 , 𝑓 𝑥 and the gradient at this point is
𝑓′ 𝑥 .
Using the formula 𝑦 𝑦 𝑚 𝑥 𝑥 , the tangent has equation:
𝑦 𝑓 𝑥 𝑓′ 𝑥 𝑥 𝑥
𝑦 𝑓 𝑥 𝑓′ 𝑥 𝑥 𝑓′ 𝑥 𝑥
𝑦 𝑓′ 𝑥 𝑥 𝑓′ 𝑥 𝑥 𝑓 𝑥

The 𝑥 intercept of this tangent is 0 𝑓′ 𝑥 𝑥 𝑓′ 𝑥 𝑥 𝑓 𝑥


𝑓′ 𝑥 𝑥 𝑓′ 𝑥 𝑥 𝑓 𝑥
𝑥

𝑥 𝑥

This is the 𝑥 intercept that is used in the next iteration which is the 𝑛 1 𝑡ℎ
iteration.
Newton’s Method:
Using an initial guess of 𝑥 , the 𝑥 intercept of the 𝑛 1 𝑡ℎ tangent is given
by:
𝒇 𝒙𝒏
𝒙𝒏 𝟏 𝒙𝒏 where 𝑛 0,1,2, …
𝒇 𝒙𝒏

This formula is on the formula sheet.


Use CAS for following quesƟons
Ex 1. Consider the funcƟon 𝑓 𝑥 𝑥 𝑥 5. Using an iniƟal esƟmate of
𝑥 1, use Newton’s method to approximate the 𝑥 intercept of 𝑓 𝑥
correct to 4 decimal places.
Step 1: Define 𝑓 𝑥 on the CAS

Step 2: Define the formula for Newton’s Method formula


(I’ve called it 𝑣 𝑥 but you can call it
whatever you like)

Step 3: Since 𝑥 1, calculate 𝑣 1


This value is 𝑥

Step 4: Calculate 𝑣 𝑥
(either copy the value down or use 𝑣 𝑎𝑛𝑠 )
This value is 𝑥 𝑥
Step 5: Repeat the above step unƟl you get 𝑥
2 answers that are the same to 𝑥
at least 4 decimal places. Once this 𝑥
happens, that answer is the 𝑥
esƟmate of the 𝑥 intercept.

∴ The esƟmate of the 𝑥 intercept is 𝑥 1.3794


1. Consider the funcƟon 𝑓 𝑥 𝑒 𝑥 . Using an iniƟal esƟmate of
𝑥 1, use Newton’s method to approximate the 𝑥 intercept of 𝑓 𝑥
correct to 4 decimal places.

2. Consider the funcƟon 𝑓 𝑥 √𝑥. Using an iniƟal esƟmate of


𝑥 1, use Newton’s method to approximate the 𝑥 intercept of 𝑓 𝑥
correct to 3 decimal places.
Ex 2. Use Newton’s method to esƟmate the soluƟon, correct to 3 decimal places,
to the equaƟon 𝑒 𝑥 7 0 which lies in the interval 2, 3 .

Solving 𝑒 𝑥 7 0 is the same as finding the 𝑥 intercept of the graph


of 𝑦 𝑒 𝑥 7 for 2 𝑥 3. You can use 𝑥 2 or 𝑥 3.
Using 𝑥 2 is shown below

𝒙 𝟐. 𝟐𝟐𝟐

3. Use Newton’s method to esƟmate the soluƟon, correct to 4 decimal places,


to the equaƟon 𝑥 sin 𝑥 0 which lies in the interval 3, 4 .

QuesƟons 4-6 should be done without CAS


4. Newton’s method is used to esƟmate the 𝑥 intercept of the funcƟon
𝑓 𝑥 𝑥 2𝑥 4. Using an iniƟal esƟmate of 𝑥 1, find the value
of 𝑥 .
5. Newton’s method is used to esƟmate the posiƟve 𝑥 intercept of the
funcƟon 𝑓 𝑥 𝑥 2. Using an iniƟal esƟmate of 𝑥 1, find the value
of 𝑥 .

6. Newton’s method is used to esƟmate the 𝑥 intercept of the funcƟon


𝑓 𝑥 𝑥 𝑥 1. Using an iniƟal esƟmate of 𝑥 1, find the value of 𝑥 .
CAS may be used for the remaining quesƟons
7. Newton’s method is being used to approximate the negaƟve 𝑥 intercept
of the funcƟon with equaƟon 𝑓 𝑥 cos 𝑥 𝑥 12. An iniƟal esƟmate
of 𝑥 4 is used.
Which one of the following gives the first esƟmate that would correctly
approximate the intercept to four decimal places?
A. 𝑥
B. 𝑥
C. 𝑥
D. 𝑥
E. 𝑥

8. Newton’s method is being used to approximate the non-zero 𝑥 intercept


of the funcƟon with equaƟon 𝑓 𝑥 √𝑥. An iniƟal esƟmate of 𝑥 1
is used.
Which one of the following gives the first esƟmate that would correctly
approximate the intercept to three decimal places?
A. 𝑥
B. 𝑥
C. 𝑥
D. 𝑥
E. 𝑥

Ex 3. Let 𝑓 𝑥 𝑥 7. Show that Newton’s method gives the iteraƟve


formula 𝑥

𝑥 𝑥

𝑥 Now put this over a common denominator

Note: This is a show that quesƟon, so you


need to show all the steps
Ex 4. If Newton’s method is used to approximate the 𝑥 intercept of the funcƟon
with equaƟon 𝑓 𝑥 𝑥 2𝑥 3, which of the following could be used to
calculate 𝑥 ?
SoluƟon:
A. 𝑥
𝑥 𝑥
B. ↑ CAS may be used for this
Then put this over a common denom:
C. 𝑥 2 Number 7 FracƟon Tools
4 Common Denominator
D.
OpƟon B
E. AlternaƟvely, you can do step 1 by hand, then use CAS to
put over a common denominator

9. Let 𝑓 𝑥 𝑥 5. Show that Newton’s Method gives the iteraƟve


formula 𝑥

10. Let 𝑓 𝑥 𝑥 4𝑥 1. Show that Newton’s Method gives the iteraƟve


formula 𝑥
11. Alina used Newton’s method to approximate the 𝑥 intercept of the
funcƟon with equaƟon 𝑓 𝑥 𝑥 2𝑥 3. Which of the following could
Alina have used to calculate 𝑥 ?
A. 𝑥 𝑥

B. 𝑥

C. 𝑥 𝑥

D. 𝑥

E. 𝑥

Newton’s method does not always work

 Newton’s formula, 𝑥 𝑥 , will only give us a result if the


function is differentiable at 𝑥 and 𝑓 𝑥 0

 SomeƟmes Newton’s method gives an oscillaƟng sequence of values


instead of converging to a single value. To see this, try using Newton’s
method with the funcƟon 𝑓 𝑥 𝑥 5𝑥 and a starƟng value of 1.

 SomeƟmes our iniƟal choice of 𝑥 causes the values of 𝑥 to diverge (get


further apart) instead of converging.

 If we try and use a value of 𝑥 for which the funcƟon has a staƟonary point
Newton’s method won’t work because the tangent there is parallel to the
𝑥 axis so does not have an 𝑥 intercepts, so we cannot get a value of 𝑥 .

You might also like