I want to show that:

$$ \left\lfloor \frac{n}{2}\right\rfloor + \left\lceil \frac{n}{2} \right\rceil=n$$

That's what I have tried:

  • $ \left\lfloor \frac{n}{2}\right\rfloor=\max \{ m \in \mathbb{Z}: m \leq \frac{n}{2}\}$
  • $\left\lceil \frac{n}{2} \right\rceil= \min \{ m \in \mathbb{Z}: m \geq \frac{n}{2}\}$

If $n=2k,k \in \mathbb{Z}$,then: $\frac{n}{2} \mathbb{Z}$,so $$\left\lfloor \frac{n}{2}\right\rfloor=\max \left\{ \frac{n}{2}, \frac{n-2}{2}, \dots \right\}=\frac{n}{2} \\ \left\lceil \frac{n}{2} \right\rceil=\min \left\{ \frac{n}{2}, \frac{n+2}{2}, \dots \right\}=\frac{n}{2}$$

Therefore, $ \left\lfloor \frac{n}{2}\right\rfloor + \left\lceil \frac{n}{2} \right\rceil=n$.

If $n=2k+1, k \in \mathbb{Z}$,then $\frac{n}{2} \notin \mathbb{Z}$.So:

$$\left\lfloor \frac{n}{2}\right\rfloor=\max \left\{ \frac{n-1}{2}, \frac{n-3}{2}, \dots \right\}=\frac{n-1}{2}\\ \left\lceil \frac{n}{2} \right\rceil=\min \left\{ \frac{n+1}{2}, \frac{n+3}{2}, \dots \right\}=\frac{n+1}{2}$$

Therefore, $ \lfloor \frac{n}{2}\rfloor + \lceil \frac{n}{2} \rceil=\frac{n-1}{2}+\frac{n+1}{2}=n$

Is there also an other way to show the equality or is it the only one?

  • 3
    $\begingroup$ Presumably, for $n$ an integer... $\endgroup$ – Thomas Andrews Jul 26 '14 at 17:08
  • $\begingroup$ Yes,I have to show that the inequality stands for each integer $n$. $\endgroup$ – evinda Jul 26 '14 at 17:11
  • $\begingroup$ @ThomasAndrews Is the way I showed the equality the only possible or is there also an other one? $\endgroup$ – evinda Jul 26 '14 at 23:07
  • $\begingroup$ What do you expect as "an other way" ? $\endgroup$ – Yves Daoust Jul 27 '14 at 8:36

Since the function $f(x)=\{x\}=x-\lfloor x\rfloor$ is periodic with period one, the function: $$ g(n) = \left\lfloor\frac{n}{2}\right\rfloor+\left\lceil\frac{n}{2}\right\rceil -n$$ is periodic with period $2$. Since $g(0)=g(1)=0$, $g(n)=0$ for every $n\in\mathbb{N}$.

| cite | improve this answer | |
  • $\begingroup$ Did you conclude this,because of the fact that $f$ is periodic? $\endgroup$ – evinda Jul 26 '14 at 23:27
  • 1
    $\begingroup$ Yes, it is exactly so. $\endgroup$ – Jack D'Aurizio Jul 26 '14 at 23:28
  • $\begingroup$ I understand..thank you!!! $\endgroup$ – evinda Jul 27 '14 at 12:51

For even $n$, $\lfloor\frac n2\rfloor=\lceil\frac n2\rceil=\frac n2$.

For odd $n$, $\lfloor\frac n2\rfloor=\frac{n-1}2$, and $\lceil\frac n2\rceil=\frac{n+1}2$.

| cite | improve this answer | |
  • $\begingroup$ So,is this the only way to show that the relation is true? $\endgroup$ – evinda Jul 26 '14 at 17:15
  • $\begingroup$ Er, your way is different, its uses bounds on number sets. What else did you expect, this is rather elementary ? $\endgroup$ – Yves Daoust Jul 26 '14 at 17:17
  • $\begingroup$ Could you explain me further which the difference is? :/ $\endgroup$ – evinda Jul 26 '14 at 17:18
  • $\begingroup$ You use bounds on number sets, this is unnecessary. $\endgroup$ – Yves Daoust Jul 26 '14 at 17:20
  • 2
    $\begingroup$ @YvesDaoust how did you prove the equality in your post? $\endgroup$ – Kaster Jul 26 '14 at 17:39

Well, here's another one. In fact, it is possible to prove that for any integer $n$ and any real number $\alpha$, we have $$n=\lfloor (1-\alpha) n\rfloor +\lceil \alpha n\rceil.$$

First, we have $$\alpha n\leq\lceil\alpha n\rceil,$$ therefore, $$n-\lceil\alpha n\rceil\leq (1-\alpha)n,$$ but then, since $n-\lceil\alpha n\rceil$ is an integer, this means $$n-\lceil\alpha n\rceil\leq \lfloor (1-\alpha)n\rfloor.$$

Similarly, we have $$\lfloor (1-\alpha)n\rfloor\leq (1-\alpha)n,$$ therefore, $$\alpha n\leq n-\lfloor (1-\alpha)n\rfloor,$$ but then again, $$\lceil\alpha n\rceil\leq n-\lfloor (1-\alpha)n\rfloor,$$ from the above, we get $$\lfloor (1-\alpha) n\rfloor +\lceil \alpha n\rceil\leq n\leq \lfloor (1-\alpha) n\rfloor +\lceil \alpha n\rceil.$$

| cite | improve this answer | |

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.