Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.



Re: An interesting set
Posted:
Apr 20, 2013 9:12 AM


In article <k7b3n8pok9afdn2jbjpi3f9oid3263bb4t@4ax.com>, <dullrich@sprynet.com> wrote:
> On Fri, 19 Apr 2013 08:10:55 0700 (PDT), apoorv <skjshr@gmail.com> > wrote: > > >I am not sure whether this set has been discussed earlier.Let the set > >X contain alll binary decimals x, between 0 and 1 , > >such that every initial portion of the decimal Contains not more than > >1/3 '1st'. For example , .00100... would be in the set. > >Then as far as I can make out,X is uncountable and so is [0,1]X. Both > >are dense in each other. > >What would be the measure of these sets? > > I'm pretty sure it's pretty clear that X has measure zero. > > Given N, let x(N) be the number of Nbit sequences with > at most N/3 1's. Then it's not hard to show, from the > Central Limit Theorem if not by something simpler, > that x(N)/2^N > 0 as N > infinity. > >
Yes. Almost all reals in [0,1] have digits with limiting frequency 1/2 for digit 1. So Law of Large Numbers is enough, Central Limit Theorem is overkill.
 G. A. Edgar http://www.math.ohiostate.edu/~edgar/



