Results 1 to 3 of 3

Math Help - Maclaurin series for x/(1- cos x)?

  1. #1
    Junior Member
    Joined
    Nov 2009
    Posts
    29

    Maclaurin series for x/(1- cos x)?

    Okay, so I have a function defined like this:
    f(x) = x^2 / (1 - cos x), when x is not 0
    f(0) = 2

    I'm supposed to find Maclaurin series for this, but I can't figure out how I should do it. All derivatives in the point x=0 have the value 0, so I'm only left with 2 + 0 + 0 + 0...
    I really doubt that this is the answer I'm looking for.


    Follow Math Help Forum on Facebook and Google+

  2. #2
    MHF Contributor undefined's Avatar
    Joined
    Mar 2010
    From
    Chicago
    Posts
    2,340
    Awards
    1
    Quote Originally Posted by Koaske View Post
    Okay, so I have a function defined like this:
    f(x) = x^2 / (1 - cos x), when x is not 0
    f(0) = 2

    I'm supposed to find Maclaurin series for this, but I can't figure out how I should do it. All derivatives in the point x=0 have the value 0, so I'm only left with 2 + 0 + 0 + 0...
    I really doubt that this is the answer I'm looking for.


    Hi Koaske.

    My understanding is that to get f'(0), you can't use the formula for f'(x) obtained by using the quotient rule, etc., because this only applies when x \neq 0. Instead, you need to look at right- and left-derivatives, defined by

    lim_{h->0^+}(f(a+h)-f(a))/h

    and

    lim_{h->0^-}(f(a+h)-f(a))/h

    respectively, at the point x=a=0. If both exist and are equal, then that is the value of the derivative at that point. I worked some of it on paper and had to use L'Hopital's rule, then got lazy and switched to Mathematica, anyway I found that f'(0)=0, but f''(0)=1/3, and I didn't try to find out the higher order derivatives.

    Hope that helps.

    EDIT: It seems that f(x) is infinitely differentiable, which makes consideration of right- and left-derivatives unnecessary. So to get that f''(0)=1/3, it is enough to evaluate lim_{x->0}f''(x)=1/3.

    EDIT 2: Did the problem require you to characterize all terms of the Maclaurin series, or just some of the first terms? Because even if I'm completely lazy and let Mathematica do all the work for me, I still don't see a pattern to characterize the terms...

    \{f^{(n)}\} = 0, \dfrac{1}{3}, 0, \dfrac{1}{5}, 0, \dfrac{5}{21}, 0, \dfrac{7}{15}, 0, \dfrac{15}{11}, 0, \dfrac{7601}{1365}, ... \ \text{for}\ n = 1, 2, 3, ...
    Last edited by undefined; April 6th 2010 at 09:33 PM.
    Follow Math Help Forum on Facebook and Google+

  3. #3
    MHF Contributor chisigma's Avatar
    Joined
    Mar 2009
    From
    near Piacenza (Italy)
    Posts
    2,162
    Thanks
    5
    Let's write...

    \frac{x^{2}}{1-\cos x} = \sum_{n=0}^{\infty} a_{n}\cdot x^{n} (1)

    ... and then we search a way to find the a_{n}. Remembering that is...

    \cos x = \sum_{k=0}^{\infty} (-1)^{k}\cdot \frac{x^{2k}}{(2k)!} (2)

    ... with simple steps we obtain...

    \frac{1 - \cos x}{x^{2}} = \sum_{k=0}^{\infty} (-1)^{k}\cdot \frac{x^{2k}}{(2k+2)!} = \frac{1}{2} - \frac{x^{2}}{24} + \frac{x^{4}}{720} - \dots (3)

    Now the a_{n} are found imposing the conditions ...

    \sum_{n=0}^{\infty} a_{n}\cdot x^{n} \cdot \sum_{k=0}^{\infty} (-1)^{k}\cdot \frac{x^{2k}}{(2k+2)!} = 1 (4)

    ... and solving for the a_{n}...

    Kind regards

    \chi \sigma
    Follow Math Help Forum on Facebook and Google+

Similar Math Help Forum Discussions

  1. Replies: 0
    Last Post: January 26th 2010, 09:06 AM
  2. Replies: 2
    Last Post: September 16th 2009, 08:56 AM
  3. Binomial Series to find a Maclaurin Series
    Posted in the Calculus Forum
    Replies: 4
    Last Post: July 21st 2009, 08:15 AM
  4. Multiplying power series - Maclaurin series
    Posted in the Calculus Forum
    Replies: 4
    Last Post: March 8th 2009, 12:24 AM
  5. Replies: 1
    Last Post: May 5th 2008, 10:44 PM

Search Tags


/mathhelpforum @mathhelpforum