Results 1 to 11 of 11

Math Help - Could someone check the attached document for mistakes?

  1. #1
    Senior Member
    Joined
    Apr 2006
    Posts
    290

    Could someone check the attached document for mistakes?

    I would love someone to please have a look at the Fibonacci document attached and check the proof of the golden ratio and tell me if there are any mistakes.

    Many thanks

    Nat
    Follow Math Help Forum on Facebook and Google+

  2. #2
    Global Moderator

    Joined
    Nov 2005
    From
    New York City
    Posts
    10,616
    Thanks
    10
    That document is not working on my computer.
    Follow Math Help Forum on Facebook and Google+

  3. #3
    Senior Member
    Joined
    Apr 2006
    Posts
    290
    Quote Originally Posted by ThePerfectHacker
    That document is not working on my computer.
    Here is the document (as screenshots) could this be corrected please :-)
    Follow Math Help Forum on Facebook and Google+

  4. #4
    Senior Member
    Joined
    Jun 2005
    Posts
    295
    Awards
    1
    After a quick read-through I think it's OK. There are many many ways of showing that the ratio of successive Fibonacci numbers tends to the Golden Ratio.

    Let's establish some notation. Let \phi = \frac{1+\sqrt5}2 be the Golden Ratio and \bar\phi = \frac{1-\sqrt5}2 the conjugate, so that \phi,\bar\phi are the roots of \Phi^2 - \Phi - 1 with \phi+\bar\phi=1 and \phi\bar\phi = -1.

    One way is to find an explicit expression for F_n. A recurrence relation of the form x_{n+1} = Px_n + Qx_{n-1} has a general solution of the form A\alpha^n + B\beta^n where \alpha,\beta are the roots of the auxiliary polynomial f(t) = t^2 - Pt - Q, and A,B are determined by two initial values. There is an exception when the auxiliary has repeated roots, in which case the general solution is (An+B)\alpha^n: that case doesn't apply here.

    From this is is easy to derive F_n = \frac{\phi^n-\bar\phi^n}{\phi-\bar\phi}, and if you don't want to rely on the assertions just made about the theory of recurrence relations, you can prove it directly by induction (true for cases n=0 and 1, and check that the expression given on the RHS of that formula has the same recurrence relation as F_n). In turn, it's now easy to check that F_{n+1}/F_n \rightarrow \phi.

    Other ways are based on the observation you already made, that if the sequence of ratios tends to a limit at all, that limit must be the Golden Ratio.
    So you could directly consider e_n = F_{n+1}/F_n - \phi. It's not too hard to show that e_{n+1} = F_{n+2}/F_{n+1} - \phi = 1 + F_n/F_{n+1} - \phi = \frac{1}{e_n+\phi} + 1 - \phi = \frac{1}{e_n+\phi} - \frac{1}{\phi} = \frac{\phi - e_n - \phi}{(e_n+\phi)\phi} = \frac{-e_n}{(e_n+\phi)\phi}. In magnitude this is less than \frac{-e_n}{2} at least when n is larger than 6, say. That is, the difference is alternating in sign and diminishing in magnitude by a factor of at least 2 each step, which is easily enough for convergence.
    Last edited by rgep; May 28th 2006 at 05:03 AM.
    Follow Math Help Forum on Facebook and Google+

  5. #5
    Global Moderator

    Joined
    Nov 2005
    From
    New York City
    Posts
    10,616
    Thanks
    10
    The most difficult part of the proof is showing that the sequence is indeed convergent. Then once you have that you can manipulate it to get your result.

    This is why I asked you in your other post if your studied countinued fractions, because all countinued fractions are convergent, thus that messy part is already proven.
    ---------
    This is perhaps an easier way, if you would like. There is a formula for the fibonacci sequence in terms of an exponention function due to Binet. I can provide a proof if you would like.
    Last edited by ThePerfectHacker; May 28th 2006 at 10:44 AM.
    Follow Math Help Forum on Facebook and Google+

  6. #6
    Global Moderator

    Joined
    Nov 2005
    From
    New York City
    Posts
    10,616
    Thanks
    10
    Since you are so interesting in fibonacci sequence. I once was playing around with this sequence and found something interesting Look Here
    Follow Math Help Forum on Facebook and Google+

  7. #7
    Senior Member
    Joined
    Apr 2006
    Posts
    290
    Quote Originally Posted by ThePerfectHacker
    Since you are so interesting in fibonacci sequence. I once was playing around with this sequence and found something interesting Look Here
    Thanks for the offer but as I have never studied Binet, therefore I don't think I should use it. I'm trying hard on the other thread though the one on 'Proof by induction' to get an answer. To fill the gaps basically and it is due in tomorrow :-(
    Follow Math Help Forum on Facebook and Google+

  8. #8
    Global Moderator

    Joined
    Nov 2005
    From
    New York City
    Posts
    10,616
    Thanks
    10
    Quote Originally Posted by Natasha1
    ...never studied Binet, ....
    I am assuming you think that 'Binet' means some kind of theory/branch in math. No it is just a formula involving fibonacci numbers, (and its proof is based on induction). It is very simple nothing complicated if that was what frightened you.
    Follow Math Help Forum on Facebook and Google+

  9. #9
    Senior Member
    Joined
    Apr 2006
    Posts
    290
    Well in that case, and if you don't mind I would love to see it :-). Thank you!
    Follow Math Help Forum on Facebook and Google+

  10. #10
    Global Moderator

    Joined
    Nov 2005
    From
    New York City
    Posts
    10,616
    Thanks
    10
    Quote Originally Posted by Natasha1
    Well in that case, and if you don't mind I would love to see it :-). Thank you!
    As you know the 'divine proprtion' satisfies,
    x^2=x+1
    Now this equation has two roots,
    \phi=\frac{1+\sqrt{5}}{2} called 'the divine proprtion'
    and,
    \psi=\frac{1-\sqrt{5}}{2} no special interest.
    ----
    Therefore,
    \phi^2=\phi+1
    Multiply both sides by \phi thus,
    \phi^3=\phi^2+\phi
    but, \phi^2=\phi+1 thus,
    \phi^3=(\phi+1)+\phi=2\phi+1
    Multiply both sides by \phi thus,
    \phi^4=2\phi^2+\phi
    but, \phi^2=\phi+1 thus,
    \phi^4=2(\phi+1)+\phi=3\phi+2
    Multiply both sides by \phi thus,
    \phi^5=3\phi^2+2\phi
    but, \phi^2=\phi+1 thus,
    \phi^5=3(\phi+1)+2\phi=5\phi+3 thus,
    -----
    Did you notice something?
    Firstly that,
    \phi^n=A\phi+B for some integers A,B, that you probably saw. But that you see that,
    A=F(n)\mbox{ and }B=F(n-1)?
    I will leave that as an excerise to you that prove that this fibonacci pattern still holds (this is induction type proof).

    Similarly we have that,
    \psi^n=F(n)\psi+F(n-1)
    because it is developed in the same manner as for \phi.
    ---------
    As a result we have two equations, for n>0.
    \left\{ \begin{array}{c}\phi^n=F(n)\phi+F(n-1) \mbox{     (1)}\\\psi^n=F(n)\psi+F(n-1)\mbox{     (2)}

    From (1) subtract (2) to get,
    \phi^n-\psi^n=F(n)(\phi-\psi)
    Thus,
    F(n)=\frac{\phi^n-\psi^n}{\phi-\psi}
    Thus,
    F(n)=\frac{ \left(\frac{1+\sqrt{5}}{2}\right)^n - \left( \frac{1-\sqrt{5}}{2}\right)^n}{\sqrt{5}}-this is called binet's formula.

    Anyways, you wish to demonstrate that the fibonaccis converge to \phi you can do this by showing that,
    \lim_{n\to\infty}\frac{F(n+1)}{F(n)}=\frac{\phi^{n  +1}-\psi^{n+1}}{\phi^n-\psi^n}=\phi

    Though this looks like a complicated limit it is not. Divide the numerator and denominator by \phi^n thus,
    \lim_{n\to\infty}\frac{\phi-\psi\left(\frac{\psi}{\phi}\right)^n}{1-\left(\frac{\psi}{\phi}\right)^n}
    But,
    -1<\frac{\psi}{\phi}<1 therefore as n\to\infty} we have that, \left(\frac{\psi}{\phi}\right)^n=0
    Thus,
    \lim_{n\to\infty}\frac{\phi-\psi\left(\frac{\psi}{\phi}\right)^n}{1-\left(\frac{\psi}{\phi}\right)^n}=\frac{\phi-\psi\cdot 0}{1-0}=\phi

    \mathbb{Q}.\mathbb{E}.\mathbb{D}
    Follow Math Help Forum on Facebook and Google+

  11. #11
    Senior Member
    Joined
    Apr 2006
    Posts
    290
    Quote Originally Posted by ThePerfectHacker
    As you know the 'divine proprtion' satisfies,
    x^2=x+1
    Now this equation has two roots,
    \phi=\frac{1+\sqrt{5}}{2} called 'the divine proprtion'
    and,
    \psi=\frac{1-\sqrt{5}}{2} no special interest.
    ----
    Therefore,
    \phi^2=\phi+1
    Multiply both sides by \phi thus,
    \phi^3=\phi^2+\phi
    but, \phi^2=\phi+1 thus,
    \phi^3=(\phi+1)+\phi=2\phi+1
    Multiply both sides by \phi thus,
    \phi^4=2\phi^2+\phi
    but, \phi^2=\phi+1 thus,
    \phi^4=2(\phi+1)+\phi=3\phi+2
    Multiply both sides by \phi thus,
    \phi^5=3\phi^2+2\phi
    but, \phi^2=\phi+1 thus,
    \phi^5=3(\phi+1)+2\phi=5\phi+3 thus,
    -----
    Did you notice something?
    Firstly that,
    \phi^n=A\phi+B for some integers A,B, that you probably saw. But that you see that,
    A=F(n)\mbox{ and }B=F(n-1)?
    I will leave that as an excerise to you that prove that this fibonacci pattern still holds (this is induction type proof).

    Similarly we have that,
    \psi^n=F(n)\psi+F(n-1)
    because it is developed in the same manner as for \phi.
    ---------
    As a result we have two equations, for n>0.
    \left\{ \begin{array}{c}\phi^n=F(n)\phi+F(n-1) \mbox{     (1)}\\\psi^n=F(n)\psi+F(n-1)\mbox{     (2)}

    From (1) subtract (2) to get,
    \phi^n-\psi^n=F(n)(\phi-\psi)
    Thus,
    F(n)=\frac{\phi^n-\psi^n}{\phi-\psi}
    Thus,
    F(n)=\frac{ \left(\frac{1+\sqrt{5}}{2}\right)^n - \left( \frac{1-\sqrt{5}}{2}\right)^n}{\sqrt{5}}-this is called binet's formula.

    Anyways, you wish to demonstrate that the fibonaccis converge to \phi you can do this by showing that,
    \lim_{n\to\infty}\frac{F(n+1)}{F(n)}=\frac{\phi^{n  +1}-\psi^{n+1}}{\phi^n-\psi^n}=\phi

    Though this looks like a complicated limit it is not. Divide the numerator and denominator by \phi^n thus,
    \lim_{n\to\infty}\frac{\phi-\psi\left(\frac{\psi}{\phi}\right)^n}{1-\left(\frac{\psi}{\phi}\right)^n}
    But,
    -1<\frac{\psi}{\phi}<1 therefore as n\to\infty} we have that, \left(\frac{\psi}{\phi}\right)^n=0
    Thus,
    \lim_{n\to\infty}\frac{\phi-\psi\left(\frac{\psi}{\phi}\right)^n}{1-\left(\frac{\psi}{\phi}\right)^n}=\frac{\phi-\psi\cdot 0}{1-0}=\phi

    \mathbb{Q}.\mathbb{E}.\mathbb{D}
    That's very neat. Thanks so much :-)
    Follow Math Help Forum on Facebook and Google+

Similar Math Help Forum Discussions

  1. Preview Document?
    Posted in the LaTeX Help Forum
    Replies: 2
    Last Post: August 2nd 2010, 06:32 AM
  2. Starting A Document
    Posted in the LaTeX Help Forum
    Replies: 2
    Last Post: May 19th 2010, 08:34 AM
  3. where are my mistakes? (basic derivatives)
    Posted in the Calculus Forum
    Replies: 1
    Last Post: October 23rd 2009, 07:26 AM
  4. Two mistakes in my tutorial sheet?
    Posted in the Business Math Forum
    Replies: 6
    Last Post: July 9th 2008, 10:04 AM
  5. Dilusion...Did I make any mistakes?
    Posted in the Calculus Forum
    Replies: 0
    Last Post: September 27th 2007, 01:56 AM

Search Tags


/mathhelpforum @mathhelpforum