
Re: PNT lookalikes
Posted:
May 18, 2011 2:16 PM


On May 18, 9:07 am, marcus_b <marcus_bruck...@yahoo.com> wrote: > On May 1, 7:48 am, Han de Bruijn <umum...@gmail.com> wrote: > > I would like to have many _more_ of these PNT lookalikes. Any takers? > > Please do me a favour and leave out the _very_ trivial ones. > OK, here's one I don't know the answer to. Let m be a positive > integer. > Define m to be 'decimalmaximal' if the decimal expansion of 1/m > repeats > after m  1 digits. For example, 1/7 = .142857142... , so 7 is > decimal > maximal. The number 11 is not decimalmaximal. Let M be the set of > decimalmaximal numbers. What is the asymptotic density of M ?
By Fermat's Little Theorem (Carmichael generalization), the number of digits after which 1/m repeats must divide lambda(m), which in turn divides phi(m). It follows that if m is "decimalmaximal" (or "basebmaximal to any base b), it must be _prime_.
http://en.wikipedia.org/wiki/Repeating_decimal http://en.wikipedia.org/wiki/Carmichael_function
So M is a subset of the set P of prime numbers. Therefore the density of M is at most that of P, which is _zero_. QED
Related question: what is rho(M,P) (Katz notation), the density of M _in_P_? I don't know the answer to this one.
> More general: let T3 = the set of numbers m for which the decimal > expansion > of 1/m repeats after 3 digits. For example, 37 is in T3. What is > the > density of T3? Is the set T3 infinite? (Probably easy to prove).
It took me a while to figure this one out, but then I realized that T3 contains not only 37, but 37*10^n for every natural n.
1/37 = .027027027... 1/370 = .0027027027... 1/3700 = .00027027027... etc.
Therefore T3 is infinite. QED

