1
$\begingroup$

Is there anyone who knows, and want to help, how to show that this is true $\sum_{k=0}^{\infty}\frac{x^{3k}}{(3k)!} = \frac{1}{3} e^x + \frac{2}{3} e^{-\frac{x}{2}} \cos\left(\frac{\sqrt{3}}{2} x\right)$ ?

I know that $\sum_{k=0}^{\infty}\frac{x^{k}}{k!}= e^x$, but how to use it I don't know. I can't find my lecture notes so it is like that.

I'm thankful for your help.

$\endgroup$
3
  • $\begingroup$ Use $\cos y=\frac12(e^y+e^{-y})$. $\endgroup$ – Angina Seng Jan 5 '18 at 16:55
  • 1
    $\begingroup$ @LordSharktheUnknown you forgot the $i$'s $\endgroup$ – Shashi Jan 5 '18 at 16:56
  • $\begingroup$ @LordSharktheUnknown that looks more like $\cosh(x)$ $\endgroup$ – Henry Jan 5 '18 at 16:57
2
$\begingroup$

Outline: Differentiate three times and notice that the series satisfies $y'''=y$. Then solve that differential equation, and match with the conditions $y(0)=1$ and $y'(0)=y''(0)=0$, that you get by looking at the coefficients of the series.

$\endgroup$
5
  • $\begingroup$ $\sum_{k=0}^{\infty}[\frac{1}{x^3 (3k-3)!}-\frac{1}{(3k)!} ]x^{3k} -1 = 0$ is this in the right direction? This is what I get when trying to solve y'''-y=0 $\endgroup$ – 123 Jan 5 '18 at 17:30
  • $\begingroup$ If not then I try again. $\endgroup$ – 123 Jan 5 '18 at 17:32
  • $\begingroup$ You probably have solved that kind of differential equation before. It is linear and homogeneous. Use the method of characteristic equation. $\endgroup$ – mickep Jan 5 '18 at 17:41
  • 1
    $\begingroup$ @potete I think using the characteristic polynomial will be better than that. Just solve $\lambda^3-1=0$ you will get three different lambda's and your solution will be $$y=Ae^{\lambda_1 x}+Be^{\lambda_2 x} +Ce^{\lambda_3 x} $$ don't forget to make your solution real. Using series is circular. $\endgroup$ – Shashi Jan 5 '18 at 17:45
  • $\begingroup$ You need to find the complex roots. $\endgroup$ – mickep Jan 5 '18 at 18:05
4
$\begingroup$

Outline: Let $\omega = \exp(2\pi i/3) = -1/2 + i\sqrt3/2$. Apply the exponential series to $\exp(x) + \exp(\omega x) + \exp(\omega^2x)$. Since $\omega^3 = 1$ and $1 + \omega + \omega^2 = 0$, the result is $3$ times the series you're interested in.

$\endgroup$
1
  • $\begingroup$ My friend, I don't see why you apply the first thing to exp(x)+exp(wx)+exp(w^2 x) $\endgroup$ – 123 Jan 5 '18 at 17:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.