2
$\begingroup$

This is a homework problem so no need to give me an immediate answer.

My general plan has been to try to prove that f(x) cannot have both an upper bound and a lower bound.

One strategy I tried was using induction on the degree of f. The base case for $f(x) = a_0 + a_1x$ is very doable to show that it cannot have both upper & lower bounds. It's when you get to $deg(f) = n$ that things start to get tricky.

The chapter of the book I'm working with, Contemporary Abstract Algebra, 8th edition, has the theorem for the division algorithm for polynomials, and as a corollary we see that f(a) is the remainder of f(x) when dividing by (x - a). So I am assuming they want me to try using that, but I'm not sure.

I feel as though I could brute force this problem by

  1. Starting with f(b) = lower bound and f(c) = upper bound
  2. Checking if the leading term has an even/odd power
  3. Checking if the leading coefficient is positive/negative

Once all those are fixed, I would have to come up with a clever value for $b$ or $c$, and then use that to show that we can't have both an upper bound and a lower bound.

What I would really love is if there is a more elegant solution using the image of the polynomial f(Z) and showing some contradiction if it's finite. But nothing I've come across fits that.

$\endgroup$
  • $\begingroup$ But induction doesn't seems so hard, every $n+1$ degree polynomial can be written as $b+x(\sum_{i=0}^na_ix^i)$, the constant doesn't matter, so you have unbounded expression times $x$, so unless $x\to0$ you have $n+1$ degree also unbounded $\endgroup$ – Holo May 12 '18 at 22:10
4
$\begingroup$

If $f(X)\in\Bbb{Z}[X]$ takes on only finitely many values in $\Bbb{Z}$, then it must take one of these values (say $c$) infinitely many times. But then $f(X)-c$ has infinitely many zeroes in $\Bbb{Z}$...

$\endgroup$
  • $\begingroup$ The pigeon hole principle strikes again! Yea, I thought I was looking in the wrong direction while I was doing it. Thanks! $\endgroup$ – farleyknight May 12 '18 at 22:12

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.