0
$\begingroup$

I was thinking about this problem for some time which can be stated in plain words as follows:

Can we plug a linear sequence of real numbers into the sine function such that the resulting sequence will be a one-to-one\injective sequence?

Translating it into more mathematical language we get that the problem can be stated as:

Can you find real numbers $a,b\in\mathbb{R}$ such that plugging the linear sequence $\{x_n\}_{n=1}^\infty$ defined as $\forall n\in\mathbb{N}, x_n:=an+b$ into the sine function will produce a new sequence $\{y_n\}_{n=1}^\infty$ defined as $\forall n\in\mathbb{N}, y_n:=\sin(x_n)$ that will satisfy $\forall m,n\in\mathbb{N}, y_m=y_n\longrightarrow m=n$
?


I’ve tried to attack the question by first showing that the condition:

$\forall m,n\in\mathbb{N}, y_m=y_n\longrightarrow m=n$

is equivalent to:

$\forall m,n\in\mathbb{N},\Big( (\exists k_1\in\mathbb{Z},a(n-m)=2\pi k_1) \lor (\exists k_2\in\mathbb{Z}, a(n+m)+2b=\pi+2\pi k_2 )\Big)\longrightarrow m=n$


Let $m,n\in\mathbb{N}$, Then:

$\begin{gather} y_m = y_n \\ \Updownarrow \\ \sin(x_m)=\sin(x_n) \\ \Updownarrow \\ \sin(x_n)-\sin(x_m)=0 \\ \Updownarrow \\ 2\sin\frac{x_n-x_m}{2}\cos\frac{x_n+x_m}{2} = 0 \\ \Updownarrow \\ \sin\frac{x_n-x_m}{2}\cos\frac{x_n+x_m}{2} = 0 \\ \Updownarrow \\ \sin\frac{(an+b)-(am+b)}{2}\cos\frac{(an+b)+(am+b)}{2} = 0\\ \Updownarrow \\ \sin\frac{a(n-m)}{2}\cos\frac{a(n+m)+2b}{2} = 0 \\ \Updownarrow \\ \sin\frac{a(n-m)}{2}=0 \lor \cos\frac{a(n+m)+2b}{2}=0 \\ \Updownarrow \\ \Big(\exists k_1\in\mathbb{Z}, \frac{a(n-m)}{2}=\pi k_1\Big) \lor \Big(\exists k_2\in\mathbb{Z},\frac{a(n+m)+2b}{2}=\frac{\pi}{2}+\pi k_2\Big) \\ \Updownarrow \\ \Big(\exists k_1\in\mathbb{Z}, a(n-m)=2 \pi k_1\Big) \lor \Big(\exists k_2\in\mathbb{Z},a(n+m)+2b=\pi+2\pi k_2 \Big) \end{gather}$

And therefore, the condition:

$\forall m,n\in\mathbb{N}, y_m=y_n\longrightarrow m=n$

is equivalent to:

$\forall m,n\in\mathbb{N}, \Big((\exists k_1\in\mathbb{Z},a(n-m)=2\pi k_1) \lor (\exists k_2\in\mathbb{Z}, a(n+m)+2b=\pi+2\pi k_2 )\Big)\longrightarrow m=n$


Thanks...

Note: A sequence $\{x_n\}_{n=1}^\infty$ is called linear if and only if $\exists a,b\in\mathbb{R},\forall n\in\mathbb{N},x_n=an+b$

$\endgroup$
1
$\begingroup$

What about $a=b=1$? You have$$\sin m=\sin n\iff m=n.$$

| cite | improve this answer | |
$\endgroup$
0
$\begingroup$

Well, The answer is yes and based on Santos comment, we can just take $a=1$ and $b=0$, and we will have $\forall m,n\in\mathbb{N}, \sin m=\sin n \longrightarrow m=n$:

Proof:

Let $m,n\in\mathbb{N}$ be such that $\sin m=\sin n$, Then by what was just shown above, We get that there are two possibilities: (1) $\exists k_1\in\mathbb{Z},n-m=2\pi k_1$ or (2) $\exists k_2\in\mathbb{Z}, n+m=\pi+2\pi k_2$.

We will show that case (2) cannot happen, Because if we suppose by contradiction that case (2) happens, The we get that $\exists k_2\in\mathbb{Z}, n+m=\pi+2\pi k_2$, But since $\pi+2\pi k_2$ is an irrational number for any $k_2\in\mathbb{Z}$, We get that $n+m$ is also irrational which contradicts the fact that $n+m$ is rational as it is the sum of two integers. Therefore case (2) cannot happen, And thus it must be the case that case (1) happens and we conclude that $\exists k_1\in\mathbb{Z},n-m=2\pi k_1$, We will show that it must be the case that $k_1=0$ because if we suppose by contradiction that $k_1\neq 0$, The we get that $2\pi k_1$ is an irrational number and thus $n-m$ is also an irrational number which contradicts the fact that $n-m$ is a rational number as it is the difference between two integers. Thus it must be the case that $n-m=0$, And we can conclude that $n=m$ as was to be shown.

| cite | improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.