Let $f : \mathbb{R} \to \mathbb{R} $ has the following property:

For every series $\sum_{n=1}^{\infty} a_n$ converges implies $\sum_{n=1}^{\infty} f(a_n)$ converges.

Show that there exist $M>0$ and $\epsilon>0$

$|f(x)|\leq M x$ for every $x\in \mathbb{R}, 0<x<\epsilon $

I know that if $\sum_{n=1}^{\infty} f(a_n)$ is convergent then $f(a_n)$ is convergent as $\lim f(a_n)=0$. i don't have any idea, I need your helps


By contradiction, for any natural number $n$, take $M=n^2$ and $\varepsilon=n^{-2}$. Then there exists $0<x_n<n^{-2}$ such that $|f(x_n)|\geq 1$. You can conclude.

  • $\begingroup$ Contradict to fact that $|f(x_n)|<\epsilon $? $\endgroup$ – SutMar Apr 20 '19 at 14:55
  • $\begingroup$ Contradict to the fact that $\sum x_n$ converges but $\sum f(x_n) $ doesn't $\endgroup$ – elidiot Apr 21 '19 at 15:02

Proof by contradiction. Let's assume that for every $M$ and $\epsilon$ there exist $x\in[0,\epsilon]$ such that $|f(x)| \ge M x$. In particular that means that for every $n\in\mathbb{N}$ there exist $x_n\in[0,\frac{1}{n^2}]$ such that $|f(x_n)| \ge nx_n$.

Let us divide sequence $x_n$ into two subsequences, one for which $f(x_n)>0$ and the other for which $f(x_n)<0$. One of these two subsewuences is guaranteed to be infinite; without a loss of generality we can assume it's the former, i.e. there exists an infinite growing sequence $n(k)$, $k\in\mathbb{N}$ such that $0 < x_{n(k)} < \frac{1}{n(k)^2}$, $f(x_{n(k)})>n(k)x_{n(k)}$. Obviously, it also satisfies $n(k) \ge k$.

Let us construct a seqence $a_n$ in the form $$x_{n(1)}, \dots x_{n(1)}, x_{n(2)}, \dots x_{n(2)}, x_{n(3)}, \dots x_{n(3)}, x_{n(4)}, \dots x_{n(4)}, x_{n(5)}, \dots $$ where number $m_k$ of copies of $x_{n(k)}$ is such that $$ \frac{1}{k^2} < m_k x_{n(k)} < \frac{2}{k^2}$$ Since $0 < x_{n(k)} < \frac{1}{n(k)^2} \le \frac{1}{k^2}$, there's a guarantee that such $m_k$ exists. We have $$\sum_{n=1}^\infty a_n = \sum_{k=1}^\infty m_k x_{n(k)} < \sum_{k=1}^\infty \frac{2}{k^2} $$ so $\sum a_n$ is convergent. However, because $f(x_{n(k)})> n(k)x_{n(k)} \ge kx_{n(k)}$ we have $$ \sum_{n=1}^\infty f(a_n) = \sum_{k=1}^\infty m_k f(x_{n(k)}) > \sum_{k=1}^\infty m_k k x_{n(k)} > \sum_{k=1}^\infty \frac{1}{k}$$ So $\sum_n f(a_n)$ is not convergent. The contradiction proves he theorem.


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.