1
$\begingroup$

I would like to find :

$$\lim_{n \to \infty} n^2 \int_0^1 (1-x)^n \sin(\pi x) \mathrm{d}x $$

We have : $$n^2 \int_0^1 (1-x)^n \sin(\pi x) \mathrm{d}x = \int_0^1 n^2(1-x)^n \sin(\pi x) \mathrm{d}x$$

Moreover we have $\forall x \in [0, 1]$ :

$$\lim_{n \to \infty} n^2(1-x)^n \sin(\pi x) = 0$$

So by the dominated convergence theorem we can deduce that :

$$ \lim_{n \to \infty} \int_0^1 n^2(1-x)^n \sin(\pi x) \mathrm{d}x = 0$$

Yet, here my book say the answer is actually $\pi$, and I don't understand why what I've done is wrong, and how I can actually find that the value is $\pi$.

$\endgroup$
  • $\begingroup$ @TheSimpliFire Yes for all $x \in ]0, 1[$. $\endgroup$ – auhasard Jun 4 '18 at 15:07
  • $\begingroup$ Wait a minute... there seems to be a peak just after $x=0$, say $x_0$ that gets larger and larger as $n$ increases, suggesting that as $n\to\infty$, $x_0\to\infty$? That would contradict your statement. $\endgroup$ – TheSimpliFire Jun 4 '18 at 15:09
  • $\begingroup$ That's actually weird... But it seems to be correct because when I take $x = 1/n$ then we have $n^2(n-1)^n/n^n \to \infty$. But to me this actually very strange. $\endgroup$ – auhasard Jun 4 '18 at 15:18
  • $\begingroup$ @TheSimpliFire So actually there isn't a closed form for the limit of : $n^2 \sin(\pi x) (1-x)^n$ ? $\endgroup$ – auhasard Jun 4 '18 at 15:20
  • $\begingroup$ For the indefinite integral, probably not, seeing as there as so many components to it. I do not know of methods advanced enough to solve your problem. $\endgroup$ – TheSimpliFire Jun 4 '18 at 15:23
4
$\begingroup$

You cannot apply the DCT because there is not an integrable function $g$ (independent of $n$) such that $n^2(1-x)^n\sin(\pi\,x)\le g(x)$.

Integrating by parts twice we get $$ \int_0^1(1-x)^n\sin(\pi\,x)\,dx=\frac{\pi}{(n+1)(n+2)}+\frac{\pi^2}{(n+1)(n+2)}\int_0^1(1-x)^{n+2}\sin(\pi\,x)\,dx. $$ Multiply by $n^2$, let $n\to\infty$ and observe that the DCT can be applied to the integral on the left hand side because $n^2/(n+1)/(n+2)$ is bounded.

$\endgroup$
  • $\begingroup$ Thank you ! Do you agree on the fact that : $n^2(1-x)^n\sin(\pi x) \to 0$ for all $x \in [0, 1]$ ? $\endgroup$ – auhasard Jun 4 '18 at 15:33
  • 2
    $\begingroup$ Yes, but the convergence is pointwise, not uniform. $\endgroup$ – Julián Aguirre Jun 4 '18 at 15:48

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.