9
$\begingroup$

Are there infinitely many $n$ such that the decimal expansion $2^n$ begins with $n$?

For example, $2^6=64$ and $2^{10} =1024$.

It can easily be shown that this problem is equivalent to the following.

Are there infinitely many $n$ such that $\{n\log_{10}2-\log_{10}n\}<\log_{10}{(1+\frac{1}{n})}$?

Here, $\{x\}=x-[x]$, where [x] denotes the largest integral number that is smaller than $x$.

More generally, for $\alpha,\beta,\gamma > 0$, are there infinitely many $n$ such that $\{\alpha n - \beta \log n\}<\frac{\gamma}{n}$?

I think it would help if I know something about the distribution of $\{\log n\}$.

Improve this question
$\endgroup$
3
  • 2
    $\begingroup$ Simulposted to m.se, math.stackexchange.com/questions/1998744/… – PLEASE DON'T DO THAT. $\endgroup$
    – Gerry Myerson
    Commented Nov 4, 2016 at 11:36
  • 2
    $\begingroup$ So this is a version of the shrinking target problem for the dynamical system $x\mapsto x+\log_{10}2\bmod 1$ (see math.umd.edu/~dolgop/TorusRev15.pdf for some known results - but which does not answer the current question). In this case, the targets are the sets $[\log_{10}n,\log_{10}(n+1))\bmod 1$. $\endgroup$
    – Anthony Quas
    Commented Nov 4, 2016 at 12:04
  • 1
    $\begingroup$ As noted at the Math.SE post (thanks for pointing to it, Gerry), there is relevant material at OEIS: oeis.org/A100129 $\endgroup$
    – Todd Trimble
    Commented Nov 4, 2016 at 14:09

1 Answer 1

Reset to default
5
$\begingroup$

Some general information on the problem, which is probably open.

As pointed out by Ivan Neretin at math.stackexchange, this is OEIS A100129. There you can find the first 16 numbers with this property.

According to this paper,

the problem goes back to Erdős:

"Recently my attention was drawn to the following observation made by P. Erdős: $2^6=64$ and $2^{10}=1024$. Here we have two examples of the phenomenon that the number $2^n$ starts with the same ordered sequence of digits as the natural number $n$ itself."

The characterization used in van de Lune's calculations is

$$0\leq \{n\log 2\}-\{\log n\}<\log(n+1)-\log n$$

Improve this answer
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .