More specifically the question is: If I have a sequence $(u_n)_{n\in\mathbb{N}} \subset \mathbb{R}$ that diverges to infinity. Then it's Cesàro summation sequence $(s_n)_{n\in\mathbb{N}}=(\frac{1}{n+1}\sum_{k=0}^n u_k)_{n\in\mathbb{N}}$ is also divergent to infinity.

This seems true but I can't prove it.

  • 1
    $\begingroup$ it is if $u_n$ diverges to $+\infty$ (so that its mean value "converges to $+\infty$"), but not if it oscillates, for example try with $u_n = (-1)^n \sqrt{n}$ $\endgroup$ – reuns Jun 22 '16 at 20:38
  • $\begingroup$ If the series diverges to infinity then the Césaro summation diverges too. The Césaro summation is "useful" for some non-convergent series that alternate. $\endgroup$ – Masacroso Jun 22 '16 at 20:45

Yes. Let $a\in\Bbb R$. As $u_n\to +\infty$, there exists $N$ such that $u_n>a+1$ for all $n>N$. Then $(n+1)s_n-(N+1)s_N>(n-N)(a+1)$ for $n>N$ and so $$s_n>\frac{n-N}{n+1}\cdot (a+1) +\frac{N+1}{n+1}s_N=a+1+\frac{N+1}{n+1}(s_N-a-1).$$ For $n>\max\{N,(N+1)(s_N-a-1)\}$, this means $s_n>a$.


Let $M$ be any positive real number. Let $s_n = \sum_{k=0}^{n}u_k$. Since $s_n \to \infty$, there is some $N$ such that $s_n > M$ for all $n > N$. Consequently, if $n > N$ we have

$$\sum_{k=0}^{n}s_k = \sum_{k=0}^{N} s_k + \sum_{k=N+1}^{n} s_k > \sum_{k=0}^{N}s_k + (n-N)M$$ and so $$\frac{1}{n+1}\sum_{k=0}^{n}s_k > \frac{1}{n+1}\sum_{k=0}^{N}s_k + \frac{n-N}{n+1}M$$ As $n \to \infty$, the first term on the right hand side converges to zero (since $N$ is fixed), and the second term converges to $M$. Therefore $$\liminf_{n \to \infty}\frac{1}{n+1}\sum_{k=0}^{n}s_k \geq M$$ Since this is true for arbitrary $M$, we conclude that $$\liminf_{n \to \infty}\frac{1}{n+1}\sum_{k=0}^{n}s_k = \infty$$ In other words, the Cesaro sum diverges to infinity.


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.