How many steps does Euler's method take to achieve a specific error tolerance?

  • Thread starter Thread starter roam
  • Start date Start date
  • Tags Tags
    Method
roam
Messages
1,265
Reaction score
12

Homework Statement



I need some help with the last part of the following problem:

http://img834.imageshack.us/img834/8366/eulere.jpg

The Attempt at a Solution



My approximation to the solution to the IVP at t=-0.8 using 1 step of the Euler's method was:

x(-0.8)=0.8

Whereas the approximation with 1 step of 4th order Runge-Kutta method was:

x(-0.8)=0.8214

And since the exact solution is

x(-0.8) = e^{-0.8 +1} -2 \times (-0.8) -2 = 0.8214027582

the error in Euler's method would be

|0.8214027582-0.8| =0.0214027582

And the error for Runge-Kutta is

|0.8214027582-0.8214| =2.7582 \times 10^{-6}

I'm stuck here. So how many steps does Euler's method take to produce an answer with an error no larger than 2.7582 x 10-6 (the error of Runge-Kutta)?

I tried to use the following equation:

e_n \leq \frac{k}{n}

Where k is a constant and n is the number of steps and en is the error. I then tried to solve for the constant bu substituting in the values from Euler's method:

0.021402758 = \frac{k}{1} \ \implies k =0.021402758

Then substituting in the new error

2.7582 \times 10^{-6}=\frac{0.021402758}{n} \ \implies n = 7760

But doesn't 7760 steps seem too much? Where did I go wrong? I appreciate it if anyone could help me with this problem.
 
Last edited by a moderator:
Physics news on Phys.org
7760 seems to be OK. You could confirm this using a program or a spreadsheet with Δt = (0.2 / 7760) to see if it corresponds with your answer.
 
Last edited:
But I am wondering if I've even used the correct method for finding the number of steps?
 
roam said:
I tried to use the following equation:
e_n \leq \frac{k}{n}
roam said:
But I am wondering if I've even used the correct method for finding the number of steps?
What was the source of the error equation you used, class notes, a textbook, ... ?
 
rcgldr said:
What was the source of the error equation you used, class notes, a textbook, ... ?

It's from a textbook...

Edit: page 635 of the textbook called "Differential Equations" by Blanchard, Devany and Hall.
 
Last edited:
roam said:
I am wondering if I've even used the correct method for finding the number of steps?
It appears that you have. As mentioned before if you want to check this, you could confirm this using a program or a spreadsheet using Euler method with Δt = (0.2 / 7760) to see if it corresponds with your answer (for the spread sheet you would need to use 7761 rows, the initial state and 7760 steps).
 
Prove $$\int\limits_0^{\sqrt2/4}\frac{1}{\sqrt{x-x^2}}\arcsin\sqrt{\frac{(x-1)\left(x-1+x\sqrt{9-16x}\right)}{1-2x}} \, \mathrm dx = \frac{\pi^2}{8}.$$ Let $$I = \int\limits_0^{\sqrt 2 / 4}\frac{1}{\sqrt{x-x^2}}\arcsin\sqrt{\frac{(x-1)\left(x-1+x\sqrt{9-16x}\right)}{1-2x}} \, \mathrm dx. \tag{1}$$ The representation integral of ##\arcsin## is $$\arcsin u = \int\limits_{0}^{1} \frac{\mathrm dt}{\sqrt{1-t^2}}, \qquad 0 \leqslant u \leqslant 1.$$ Plugging identity above into ##(1)## with ##u...
Back
Top