1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Simpsons rule error bound.

  1. May 22, 2013 #1
    1. The problem statement, all variables and given/known data
    Calculate the value of n so that the approximation is within 0.0001. b = 2, a = 1. f(x) = 1/x.

    2. Relevant equations
    f4(x) = 24/x^5 (Think this is correct)
    Error <= (b-a)^5/180n^4(MAXx [a,b](f4(x))

    3. The attempt at a solution
    Well, 24/x^5 obtains it's max at x =1. Thus (MAXx [a,b](f4(x)) = 24.
    I subbed in all the given values and keep getting 6 as my answer. The correct answer is 8 though. Could somebody point out where I'm going wrong?
  2. jcsd
  3. May 22, 2013 #2


    Staff: Mentor

    What you wrote is ambiguous.
    Is it ((b - a)5/180) * n4 or
    (b - a)5/(180 * n4)?
  4. May 22, 2013 #3
    Apologies! It's (b-a)^5/(180*n^4).
  5. May 22, 2013 #4


    Staff: Mentor

    I get 6 as well. Is 8 the answer in the back of the book? It's possible they have the wrong answer.

    One way to check is to do Simpson's with n = 6, and compare the answer you get with the integral itself,
    $$\int_1^2 \frac{dx}{x} = ln(2) \approx. .69315$$

    You should have agreement in either 2 or 3 decimal places.
  6. May 22, 2013 #5
    Thanks for this too. There's most likely a mistake alright. I'll be sure to double check it in the morning though.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted