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!

Homework Help: Stirling's Formula: How does n! relate to n^n.

  1. Nov 17, 2012 #1
    Hello all.
    I am doing a school paper on Stirling's formula and I want first to show how the factor e comes in to place. So I found somewhere on the net the definition n! = (1-1/2)×(1-1/3)2... × (1-1/n)n-1] × nn.

    I will then use this result and the definition of e^x since we are talking for large n to proceed.

    1. The problem statement, all variables and given/known data
    Problem is, how do I prove it? Although I haven't tried induction yet, I was told that I should find an algebraic method.

    I don't want a complete solution, just a hint to guide me through the proof.

    Thank you
    Last edited: Nov 17, 2012
  2. jcsd
  3. Nov 17, 2012 #2
    First represent all the 1-.5 and 1-1/3 by (1-1/n+1)^n, we note this is (n/(n+1))^n.
    Now note what happens as we multiply this for different n's:
    (1/2)^1 * (2/3)^2 * (3/4)^3... = (1^1 * 2^2 * 3^3...)/(2^1 * 3^2 * 4^3...).
    Do you notice what is happening? Try canceling things up, in problems like this, I recommend you put things in variables like I did (n/(n+1)) in our case.

    Thanks, Bonaparte
    Last edited: Nov 17, 2012
  4. Nov 17, 2012 #3
    Thank you for your response. I got it now, the fraction reduces to n!/(n+1)^n = Product from 1 to n of (n/(n+1))^n which leads to n!=(n+1)^n * the Product, but then the last term of the product cancels out with (n+1)^n and so comes the required result.

    However, I am still interested on how you derived it. I mean, suppose we had absolutely no clue on how does n^n relate to n!. How would we then, think of this relationship? What steps would we have to take?
  5. Nov 17, 2012 #4
    The relationship between n^n and n! is ruled by their definition. Now that I think about it a clearer way to have stated this is instead of n/n+1 (where we have to change n^n to (n+1)^n) is to state it in their way (1-1/n)^(n-1).
    Then we have ((n-1)/n)^(n-1). Then proceed as before, although now it will be clearer. Considering your writing a paper about it, I recommend the second way.

    As for your question, I derived it by look at the numbers, seeing the pattern (which is not even necessary considering its given) and implementing it into numbers.

    Thanks and good luck, Bonaparte
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook