Proof that a certain subset of the reals is not a ring

Let $A = \{x \sin x : x \in \mathbb{Z}\} \subset \mathbb{R}$. Is $A$ a ring under the usual addition and multiplication operations of $\mathbb{R}$? It looks like it's not, but I can't find something concrete to justify this.

-
Under what operations ? The usual ones in $\;\Bbb R\;$ ? –  DonAntonio Feb 25 at 16:36
Is this even closed under negation? After all, $(-x)\sin(-x) = x\sin x$ because $\sin x = -\sin (-x)$... –  fgp Feb 25 at 16:49
It's not immediately clear why there isn't some $y\in\mathbb{Z}$ such that $y\sin{y}=-x\sin{x}$ though (I agree that this is unlikely). –  Matt Pressland Feb 25 at 16:55
To take things even further (and actually possibly point to a proof): is there even one non-trivial identity among some finite number of members of $A$? –  Steven Stadnicki Feb 25 at 17:16
Another identity is that $f(1 \sin 1, 2 \sin 2) = 0$ where $$f(x,y) = y^2 - 8 x^2 (1-x^2) (1-2x^2)^2$$ –  Hurkyl Feb 25 at 17:32

Suppose $A$ were a ring; then in particular, we have $2\sin 1=n\sin n$ for some $n$. Now, use the duplication formulas for $\sin$ to write $\sin n$ as $(\sin 1)\cdot P_n(\cos 1)$, where $P_n(x)$ is a polynomial of degree $n$ (this can be shown straightforwardly via induction; these polynomials are known as the Chebyshev polynomials of the second kind). This yields an identity of the form $nP_n(\cos 1)-2=0$, contradicting the fact that $\cos 1$ is transcendental. In fact, this shows that no non-trivial linear relation can hold among any finite number of members of $A$.

-
Could you at least hint at a justification of the last sentence? –  Manny Reyes Feb 25 at 17:28
@MannyReyes I should correct that to 'additive identity', actually (I was thinking of $A$ as a group, not a ring). The proof in that case is almost identical: use the $\sin n$ identity to write each of the terms in the form $(\sin 1)\cdot P_n(\cos 1)$, sum, and derive a contradiction. –  Steven Stadnicki Feb 25 at 17:31
Thanks, the "additive identity" helps a lot. So it seems that you're actually proving that $\{\sin(n) \mid n \in \mathbb{N}\}$ is a $\mathbb{Z}$-linearly independent subset of $\mathbb{R}$. Very nice! –  Manny Reyes Feb 25 at 17:35
(Note that similarly to a now-deleted answer from @MarcinŁoś, this uses the Lindemann-Weirstraß theorem to derive its conclusion - in my case, it's implicit in the assertion that $\cos 1$ is transcendental.) –  Steven Stadnicki Feb 25 at 20:39

There is no 1 (multiplicative identity) so it is not a ring. x sin(x) is either 0 (when x=0) or is a transcendental number due to Lindemann–Weierstrass theorem when x <> 0

-
Depends on your definition of a ring. Rings are not usually required to have an identity though. –  the_fox Mar 5 at 17:31
True that is unclear from the question if they mean ring with identity or not. The other answer also assumes there is a 1 in the ring when it say there is an n such that 2 sin 1 = n sin n. Because the 2 here is an element 2 in the ring to multiply 1 sin 1 by. Hence it assumes there is a 1 in the ring to make the element 2 from. –  Michael Smith Mar 5 at 21:09
Now I am curious if you assume that there is not a 1 in the ring if you can still prove the statement :-) –  Michael Smith Mar 5 at 21:15
I think that Steven's answer tackles closure of addition, i.e. whether $1\sin 1 + 1 \sin 1 = 2\sin 1$ is an element of this set or not. –  the_fox Mar 6 at 1:36
Aha I had missed that point. Thanks for explaining! So given we are using the 1 from R here in regular way it seems like we must be assuming the ring has identity. –  Michael Smith Mar 6 at 10:48