let $S=\{\sin (n)n \in N\}$. We can prove $S$ is dense in $[1,1]$. So is the set $\{\sin( n^2)n \in N\}$; but the set $\{\sin (n^3) n \in N\}$ is not dense in $[1,1]$. How to prove this?
3

4$\begingroup$ Given your previous question: is this from a set of exercises? a problem in a book you are reading? a claim in a paper you are studying? Knowing these things would help people (a) decide if your question is appropriate (b) give more precise and helpful answers $\endgroup$ – Yemon Choi Oct 4 '11 at 8:23

3$\begingroup$ Why aren't they all dense? It's just a question of whether $(n^k/2\pi)$ is equidistributed modulo 1, and that will be so for any fixed k since $\pi$ is irrational. You'd use Weyl's inequality for a rigorous proof. $\endgroup$ – Ben Green Oct 4 '11 at 12:48

$\begingroup$ As far as I remember, H. Weyl proves in 'Weyl, H. (1916). "Ueber die Gleichverteilung von Zahlen mod. Eins,". Math. Ann. 77 (3): 313352.' that for each polynomial $P(n)$ with integer coefficients the sequence $exp(iP(n))$ is equidistributed in the unit circle $\mathbb{S}^{1}$\ and this implies of course that $sin(P(n))$ is dense in $[1,1]$. So the answer to your question should be  no, you cannot prove this. $\endgroup$ – t22 Oct 4 '11 at 19:05
Add a comment

$\begingroup$
$\endgroup$
1
If you read On the behaviour of $\sin(n!\pi x)$ when $x$ is irrational. carefully, you will be enlightened.

$\begingroup$ how to download the Weyle'such article "Ueber die Gleichverteilung von Zahlen mod. Eins,". Math. Ann. 77 (3): 313352.' $\endgroup$ – gubo Oct 12 '11 at 13:58