Duplicate: Perfect set without rationals

My approach: We consider the set $[e, \pi]$. I am trying to "cover" the rationals by enclosing each one of them by open intervals with irrational endpoints, then remove them.

Let $\{x_1,x_2,x_3,...\}$ be the enumeration of the rationals in $[e, \pi]$.

Let $x_n$ be any rational in the interval and let the minimum of the distances of $x_n$ from the endpoints be $r_n$ . We can enclose $x_n$ by $I_n=(a_n,b_n)=\displaystyle(x_n-\frac{r_n}{2^{100n}},x_n+\frac{r_n}{2^{100n}})\subsetneq [e, \pi] $.

[Note: The irrationality of the endpoints will be maintained for every $n \in \mathbb{N}$, since $r_n$ is irrational].

Now, $\sum_{n=1}^\infty |I_n| < 1.33639×10^{-30}$ as well as $A=\displaystyle[e, \pi] \setminus\{\bigcup_{n=1}^\infty I_n\}$ is closed.

Now, how do I show that any point $c \in A$ is a limit point of $A$? The goal is to find out an irrational that lies arbitrarily close to $c$ and is not "trapped" by any $I_n$.

Is my approach even valid?


Let $c$ be an element in $A \subset [e,\pi]$, which is an irrational number. Assume for contradiction that $c$ is not a limit point of $A \subset [e,\pi]$. Then there is a small real number $\delta > 0$ such that $(c-\delta,c+\delta) \subset [e,\pi]$ and contains no point of $A \subset [e,\pi]$ other than $c$. By the construction of $A$, this means that there exists 2 connected subsets, $(c-\delta,c)$ and $(c,c+\delta)$, of rational numbers $\mathbb{Q}$, which is a contradiction to the fact that rational numbers $\mathbb{Q}$ is totally-disconnected.


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.