Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

$S(x)=\sum_{n=1}^{\infty}a_n \sin(nx) $, $a_n$ is monotonic decreasing $a_n\to 0$, when ${n \to \infty}$. I need to prove that for every $\epsilon >0$, the series is uniformly converges within $[\epsilon, 2\pi - \epsilon]$. Can I use Dirichlet and say that $\sum_{0}^{M} \sin (nx)< M$ for every x in the interval and since $a_n$ is uniformly converges to 0 ( uniformity since it does not depend on $x$), so the series is uniform convergent in this range?

In addition I need to prove that if $\sum_{n=1}^{\infty} a_n^2 = \infty$ so the series is not uniform convergent in $[0, 2 \pi]$, Since I know that from $n_0$ and on $a_n^2< a_n$ I used an inequality, again I'm not sure of that.

In the other hand, maybe I need to use Fourier series somehow.

Thanks for the help!

share|cite|improve this question
up vote 2 down vote accepted

Indeed, the fact that the sequence $\{\sum_{n=1}^N\sin(nx)\}$ is bounded will help us. More precisely, fix $\varepsilon>0$.

  1. Show that we can find a constant $C=C(\varepsilon)$ such that for all $x\in [\varepsilon,2\pi-\varepsilon]$ we have $\left|\sum_{n=1}^N\sin(nx)\right|\leq C$ (to do that compute the sum).

  2. Now we use Abel's tranform. We denote $s_N(x):=\sum_{n=0}^N\sin(nx)$ and $S_N(x):=\sum_{n=1}^Na_n\sin(nx)$. Then show that $$S_N(x)=\sum_{n=1}^Na_n(s_n(x)-s_{n-1}(x))=a_Ns_N(x)-a_1s_0(x)+\sum_{n=1}^{N-1}(a_n-a_{n-1})s_n(x).$$

  3. Prove that the first two terms gives a uniformly convergent sequence (to $0$) whereas the third is a normally convergent sequence on $[\varepsilon,2\pi-\varepsilon]$, hence an uniformly convergent sequence on this interval.

For the second question, I will give you some steps:

  1. Compute for fixed integers $m$ and $n$ $\int_0^{2\pi}S_n(x)S_m(x)dx$.
  2. If we assume that $\{S_N\}$ is uniformly convergent on $[0,2\pi]$, then $$\lim_{n,m\to \infty}\int_0^{2\pi}S_n(x)S_m(x)dx=\int_0^{2\pi}S(x)^2dx.$$
  3. Conclude.
share|cite|improve this answer
Thank you so much Davide! – Jozef Feb 12 '12 at 14:04
You're welcome. – Davide Giraudo Feb 12 '12 at 14:05
Did you manage to solve the problem? – Davide Giraudo Feb 12 '12 at 19:56
Yeah, I believe I did. :) – Jozef Feb 13 '12 at 5:59

For your first question:

You can use Dirichlet's Test$^\dagger$ as long as you can show that $$D_n(x)=\sum\limits_{k=1}^n \sin(kx)$$ is indeed uniformly bounded on $[\epsilon, 2\pi-\epsilon]$.

Dirichet himself did this as follows:

Using the formula $$2\sin(v)\sin(u)=\cos(v-u)-\cos (v+u),$$ for any $n$: $$\eqalign{ 2\sin(x/2) D_n(x)&= \sum_{k=1}^n\bigl[\, 2\sin(x/2)\sin(kx)\,\bigr]\cr &=\sum_{k=1}^n\Bigl[\, \cos \bigl(\,( k-{\textstyle{1\over2}})x\,\bigr) - \cos \bigl(\,(k+{\textstyle{1\over 2}})x\,\bigr)\,\Bigr]\cr &=\cos (x/2)-\cos \bigl(\,(n+\textstyle{1\over2})x\bigr). } $$ So: $$\tag{1} |D_n(x)|= \biggl| {{\cos (x/2)-\cos \bigl(\,(n+{1\over2})x\,\bigr)}\over 2\sin(x/2) } \biggr| \le {{1\over |\sin(x/2)|} }. $$

Now, if $x\in [\epsilon, 2\pi-\epsilon]$, then $x/2\in[\epsilon/2, \pi-\epsilon /2]$ and it follows from inequality $(1)$ that $$ |D_n(x)|\le {1\over \sin(\epsilon/2)}. $$

Dirichlet's Test:

Let $E\subset\Bbb R$ be a non-empty set and let $f_k$, $g_k$ be functions from $E$ to $\Bbb R$.

If $$\biggl|\sum\limits_{k=1}^n f_k(x)\,\biggr|\le M<\infty$$ for all positive integers $n$ and all $x\in E$, and if $g_k\searrow 0$ uniformly on $E$, then $\sum\limits_{k=1}^\infty f_k g_k$ converges uniformly on $E$.

share|cite|improve this answer
Interesting to see how Dirichlet did it. Another way: $\sin(kx)=Im(e^{ikx})$ so $\Sigma_{k=1}^N sin(kx)= Im \left(\sum_{k=1}^N e^{ikx} \right)$ which by Geometric Series formula is $Im ( \frac{1-e^{ix(N+1)}}{1-e^{ix}} )$ which, maximizing the top, is $\le \frac{2}{|1-e^{ix}|}$. Now $|1-e^{ix}|$ is continuous with zeros only at the endpoints of $[0, 2\pi]$. So on compact subsets, it attains a uniform minimum $>0$, which gives us a uniform finite bound $M_\epsilon$ on $[\epsilon, 2\pi-\epsilon]$. – Marcel T. Jul 22 '14 at 22:02

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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