Prove that the dyadic rationals $D$ are dense in $\mathbb{R}$.
I saw some proofs using showing that $D$ is dense on $[0,1]$.
I proved that if $A$ is dense in $B$ and $B$ is dense in $C$, then $A$ is dense in $C$. So, I want use it. My idea is show that $D$ is dense in $\mathbb{Q}$. Can someone help me? Just a hint, not a solution.
Attempt. Let $D = \left\{\frac{m}{2^{n}}\mid n \in \mathbb{N}\text{ and }m \in \mathbb{Z}\right\}$. Take $I = \left(k-\frac{\epsilon}{2},k+\frac{\epsilon}{2}\right)$ an open interval in $\mathbb{Q}$. We can find a $n \in \mathbb{N}$ such that $\epsilon > \frac{1}{2^{n}}$. If there is no $\frac{m}{2^{n}} \in I$, there is $\frac{m}{2^n}$ and $\frac{m+1}{2^{n}}$ such that $\frac{m}{2^{n}} < k - \frac{\epsilon}{2}$ and $k + \frac{\epsilon}{2} < \frac{m+1}{2^{n}}$. Thus $$\frac{1}{2^{n}} =\frac{m+1}{2^{n}} - \frac{m}{2^n} > \epsilon,$$ a contradiction.
Instead of reducing the statement to the density of $\mathbb{Q}$, just use the same proof idea. Remainder: $\mathbb{Q}$ is dense, because given an interval of length $\varepsilon> 0$, it is possible to pick an integer $n\in \mathbb{N}$ such that $n> \frac{1}{\varepsilon}$. Equivalently, $\varepsilon> \frac{1}{n}$. So given the gap of length $\frac{1}{n}$ (the interval), you cannot step over it if the length of your step is $\frac{1}{n}$. Starting from zero, and walking towards the interval, you will step into it at a point. If this requires $k$ steps, then $\frac{k}{n}$ is in the interval.
Can you modify this proof to make it work for dyadic rationals?