3

$$ 2n^3 + 5 | 2n^4 +2n +2 - 2n^4 - 5n$$

$$= 2n^3+5 | 2-3n$$

$$3n-2≥2n^3 + 5$$

is this correct? Is there a more efficient way?

Helen
  • 455
  • 2
    Three lines seems pretty efficient to me – Ross Millikan Oct 06 '19 at 02:20
  • I don't think you mean "=" in your second line, but rather mean equivalence. If so, how are you getting that second line? Are you subtracting off some number of copies of $2n^3+5$ from $2n^4 +2n +2 - 2n^4 - 5n$? If so, this is a reasonable step and a reasonable way of doing this. A full answer would show some of the work between the first and second line. – JoshuaZ Oct 06 '19 at 02:19
  • 2
    @JoshuaZ: I think the OP is merely using the fact that $2n^4+2n+2-2n^4-5n=2-3n$. It looks they got the $2n^4+2n+2-2n^4-5n$ from $2(n^4+n+1)-n(2n^3+5)$. If $2n^3+5$ divides $n^4+n+1$, then it divides that combination as well. – Barry Cipra Oct 06 '19 at 02:25
  • @BarryCipra Oh!, I just copied the polynomial they wrote down without even noticing that it had terms which canceled. That makes complete sense. – JoshuaZ Oct 06 '19 at 02:37

5 Answers5

3

Using the Extended Euclidean Algorithm (which can be applied to $\mathbb{Q}[x]$ as well as $\mathbb{Z}$), we get $$ \left(36n^2+24n+16\right)\left(n^4+n+1\right)-\left(18n^3+12n^2+8n-27\right)\left(2n^3+5\right)=151 $$ if $2n^3+5\mid n^4+n+1$, we must also have $2n^3+5\mid151$. Since $151$ is prime and there is no integer $n$ so that $2n^3+5=\pm1$ or $2n^3+5=\pm151$, there is no integer $n$ so that $2n^3+5\mid n^4+n+1$.

robjohn
  • 345,667
  • One should at least mention where the first equation comes from. Pulling things out of a hat is for magic - not math. – Bill Dubuque Oct 06 '19 at 14:08
  • To help possibly perplexed readers I added an answer showing simpler modular derivation (Bezout coef's need not apply!). – Bill Dubuque Oct 06 '19 at 14:31
2

Mod $(2n^3+5)$, you have that $2n^3\equiv-5$. Now suppose $(2n^3+5)$ divides $n^4+n+1$. Then mod $(2n^3+5)$:

$$ \begin{align} n^4&\equiv-n-1\\ 2n^4&\equiv-2n-2\\ n(2n^3)&\equiv-2n-2\\ -5n&\equiv-2n-2\\ -3n&\equiv-2\\ 2-3n&\equiv0 \end{align} $$

So $2n^3+5$ must divide $2-3n$. You can then test values of $n$ to arrive at the inequality.

Axion004
  • 10,056
2

Worth emphasis: it is much easier via mod arithmetic (vs. extended Euclidean algorithm), e.g.

$\!\bmod 2n^3\!+\!5\!:\,\ \color{#c00}{2n^3\equiv -5}\ $ so $\ 0\equiv 2(n^4\!+\!n\!+\!1)\equiv \color{#c00}{2n^3}n\!+\!2n\!+\!2\equiv \color{#0a0}{-3n\!+\!2}$

so $\,\color{#0a0}{3n\equiv 2}\,\Rightarrow\, 0\equiv 3^3(2n^3\!+\!5)\equiv 2(\color{#0a0}{3n)}^3\!+5(27)\equiv 151,\ $ so $\ 2n^3\!+\!5\mid 151$

Remark $ $ We are essentially employing divisibility mod reduction. The above inferences are actually bidirectional since $\,2\,$ is invertible, by $\,(2,2n^3\!+\!5) = (2,5)=1$.

Such scalings by the lead coef of the divisor (to eliminate fractions in the division algorithm) are modular forms of the Nonmonic Division Algorithm, e.g. see here for another example using this idea and the Euclidean algorithm.

Bill Dubuque
  • 272,048
1

You can use polynomial long division instead:

$$ \require{enclose} \begin{array}{rll} \frac{1}{2}n && \\[-3pt] 2n^3+5 \enclose{longdiv}{n^4+0n^3+0n^2+n+1}\kern-.2ex \\[-3pt] \underline{- (n^4 \quad \quad \quad \quad \ \ + \frac{5}{2}n) \ \ \ } \\[-3pt] - \frac{3}{2}n+ 1 \\[-3pt] \end{array} $$

Therefore we have: $$\frac{n^4+n+1}{2n^3+5} = \frac{1}{2}n + \frac{-\frac{3}{2}n+1}{2n^3+5}$$

and if $2n^3+5 | n^4+n+1$, then $2n^3+5 | 2n^4 + 2n + 2$:

$$\frac{2n^4+2n+2}{2n^3+5} = n + \frac{-3n+2}{2n^3+5}$$

At $n=0$, $-3(0)+2 = 2$ and $2(0)^3+5 = 5$. However, since $n^3$ is monotonically increasing (equivalent to $(n+1)^3-n^3 > 0$), and $-3x+2$ is monotonically decreasing, then $\frac{-3n+2}{2n^3+5} = 1$ does not have solutions in the positive reals. You can repeat this argument for any positive integer $k$.

For negative integers $-k$, we have $-3n+2 = -k(2n^3+5) \Rightarrow 3n-2 = (2k)n^3 - 5k$. At $n=0$, $3(0)-2=-2$ while $2k(0)^3 + 5k = 5k$. Then when $n$ is positive, cubic functions grow much faster than linear solutions, so there are no solutions. The proof of this fact relies on limits.

Toby Mak
  • 16,827
  • 1
    Can you prove the notion about if its an integer it must equal $-1$ or $1$ – Aops Vol. 2 Oct 06 '19 at 02:33
  • thank you, my doubt was this. Is there any theorem about the last part of what you said? does the fraction need to have module 1? – Helen Oct 06 '19 at 02:40
  • 1
    I'm working on that part now since it wasn't very well expressed. I have an algebra background, so you might find Axion004's proof nicer and simpler because it uses modular algebra. – Toby Mak Oct 06 '19 at 02:41
  • I can't get over this part, because the fact that functions don't intersect doesn't prevent the fraction from assuming whole values. or prevent? – Helen Oct 06 '19 at 03:06
  • 1
    If $\frac{-3n+2}{2n^3+5} = 1$, there must be some $n$ where $-3n+2 = 2n^3+5$. If $\frac{-3n+2}{2n^3+5} =2$, there must be some $n$ where $-3n+2 = 2(2n^3+5)$. Do you get my point now? – Toby Mak Oct 06 '19 at 03:21
  • @Helen I added an answer which highlights how much simpler it is to use modular arithmetic. – Bill Dubuque Oct 06 '19 at 14:38
1

What you've done is entirely correct, I daresay even ingenious, you just haven't taken it all the way to its natural conclusion. Also, it could be written out in a clearer fashion; as is, it reads more like scratchwork, where you've jotted down the key ideas, with the logical connections between them clear to you, but less so to readers, except those who are able to read your mind. Here's one way to express things more fully, including the denouement:

Since

$$2(n^4+n+1)-n(2n^3+5)=2n^4+2n+2-2n^4-5n=2-3n$$

we have

$$(2n^3+5)\mid(n^4+n+1)\implies(2n^3+5)\mid(2-3n)\implies|2n^3+5|\le|2-3n|$$

Since we are only looking for positive integer solutions, we can remove the absolute values to obtain $2n^3+5\le3n-2$, which is not satisfied by any positive integer $n$.

(Remark: If you want to rule out negative integer solutions as well, you can write $n=-m$ with $m\ge0$ and remove the absolute values to obtain $2m^3-5\le3m+2$, which is satisfied only by $m=1$. But $n=-1$ is not a solution, since $2(-1)^3+5=3$ does not divide $(-1)^4+(-1)+1=1$.)

Barry Cipra
  • 79,832