Real analysis problems involving proofs with sequences and sets

66 Views Asked by At

The question is as follows:

Suppose $\beta \in \mathbb{R}$ is irrational and the set $S$ is defined by $S \equiv \left \{ m+n\beta : m,n \in \mathbb{Z} \right \}.$

$i)$ Show that $ra+sb \in S$ for all $a,b \in S$ and all $r,z \in \mathbb{Z}.$

$ii)$ Show that there exists a strictly monotone increasing sequence of natural numbers $\left \{ n_{k} \right \}$ such that the sequence $\left \{ n_{k}\beta-\left \lfloor n_{k}\beta \right \rfloor \right \}$ converges.

$iii)$ Show that $S$ is dense in $\mathbb{R}$.

For part $i$, I know that since $a,b \in S$, then $ra+sb$ can be written as $$r(m+n\beta) + s(m'+n'\beta')$$

$$= rm + sm' + rn\beta + sn'\beta'.$$

Then I can say that $x = rm+sm' \in \mathbb{Z}$, but I need to write this in the form $x+y\beta$ where $x,y \in \mathbb{Z}$ and $\beta \in \mathbb{R}$ is irrational. However, I cannot seem to simply this any further to show this expression is indeed in $S$.

For part $ii$, I am pretty sure that I just have to show that the sequence $\left \{ n_{k}\beta-\left \lfloor n_{k}\beta \right \rfloor \right \}$ is strictly monotone increasing and that is bounded. By the Monotone Convergence Theorem, the sequence will converge.

For part $iii$, I know that $$\left \lfloor \frac{k}{\tau } \right \rfloor \leq \frac{k}{\tau } \leq \left \lfloor \frac{k}{\tau } \right \rfloor + 1,$$ but I am not sure how to continue.

I would much appreciate if someone could give me a starting point for each of the problems or help me solve them. Studying even introductory real analysis is very different and tricky for me.

Thanks for reading this long question!

1

There are 1 best solutions below

1
On BEST ANSWER

For your (i), why do you have a $\beta'$ and what is $\alpha$? If $a,b \in S$, there exist $m_a, n_a, m_b, n_b \in \mathbb{Z}$ such that $a = m_a + n_a \beta$ and $b = m_b + n_b \beta$. To show $ra+sb \in S$, you need to find $x,y \in \mathbb{Z}$ such that $ra+sb = x + y \beta$.

(Edited to follow edits to the Question) For your (ii): Generally, you need to say what $n_k$ is. With your edit, $\left( n \sqrt{2} - \lfloor n \sqrt{2}\rfloor\right)_{n \in \mathbb{N}}$ is bounded (by $0$ and $1$), but is not monotonic. Plot of n sqrt(2) - floor(n sqrt(2)) for n=1 .. 100. However, think about $\frac{m}{n}$ being close to $\beta$. Then $m$ is close to $n\beta$. For each $n$, it is automatic that you can pick $m$ such that $\left|\frac{m}{n} - \beta \right| \leq \frac{1}{2n}$. (The fractions with denominator $n$ partition the line into segments of width $\frac{1}{n}$ and $\beta$ is in one of those segments. It can't be more than halfway away from the closest endpoint.) Rather than think of incrementing $n$, perhaps think of choosing an $n$ that exactly subdivides the interval containing $\beta$ more finely.

For (iii): Let $\varepsilon > 0$ and $r \in \mathbb{R}$. Now pretend to solve for $\beta$: $m + n\beta = r$, so $\beta = \frac{r-m}{n}$. Can you find an $n$ large enough (and a matching $m$) so that the gap between $\beta$ and $\frac{r-m}{n}$ is less than $\varepsilon$?