2

Does there exist a polynomial with rational coefficients $P(x)$ with the property $\frac {P(n+1)}{P(n)} \in \mathbb{N}$ for every $n \in \mathbb{N}_+$? If so how can I construct such a polynomial?

Kinheadpump
  • 1,331

2 Answers2

5

For every polynomial $P(x)$, one has $$\lim_{x \rightarrow \infty} \dfrac{P(x+1)}{P(x)}=1,$$ and so if a polynomial satisfies the desired property, one must have $P(n+1)=P(n)$ for all $n$ large enough, hence constant polynomials $P(x)=c$ are the only polynomials with that property.

Marco
  • 2,733
  • 6
  • 15
  • Nice! Also worth noting that it suffices, by cancelling the common factor, to assume that $P(x) \in \mathbb{Z}[x]$, from which it is easier to see that $P(n+1)=P(n)$ for $n$ large. – Dzoooks Aug 12 '19 at 21:51
1

Assume, for sake of contradiction, that $P(x)=\sum_{i=0}^d a_i x^i$ is such a polynomial. Assume that $d \geq 1$ and that the $a_i \in \mathbb{N}$ have no common factor.

Note that $$\frac{P(3)}{P(1)} = \frac{P(3)}{P(2)}\cdot\frac{P(2)}{P(1)} \in \mathbb{N},$$ and similarly $\frac{P(n)}{P(1)} \in \mathbb{N}$ for all $n$.

Now take $n=\sum_{i=0}^d a_i=P(1)$. Then we have $\sum_{i=0}^d a_i \mid a_0,$ a contradiction.

Dzoooks
  • 2,125
  • 1
  • 11
  • 20