2
$\begingroup$

I am trying to come up with a proof that numbers of the form $2^{-n}m$, where $~m \in \Bbb Z,~n \in \Bbb Z^+$ are a Euclidean ring/domain. The division property is easily verifiable (although the quotient and remainder are not unique), but I am having trouble coming up with an appropriate norm for the ring. One should, besides satisfying the axioms, in my opinion have the following features:

  1. Since this ring contains $\Bbb Z$ as a subring, the norm when restricted to it should become the absolute value,
  2. It should produce the same value for any representative of a given equivalence class of rationals (i.e. the same value on $\frac 1 2$ and $\frac 2 4$),
  3. It should produce a nonnegative integral result on the purely fractional elements of the ring (ones of the form $2^{-n}$).

I started off with simply trying $|m|$ or $|m+n|$, but neither clearly satisfies the 2nd property. I then tried something a little more complicated, namely $|2^{log_2m-n}|$, which seems to satisfy the 1st and 2nd properties, but not the 3rd, as I now realise. I am now trying to find a way to make it evaluate to $|2^n|$, when $m=1$, and the previous formula otherwise. Also, having considered this norm, I am not even sure that it satisfies either of the required axioms anyway, namely that $N(ab)>N(a)$ and that given a division $a=qb+r$, $N(r)<N(b)$.

Am I on the right track or do I err in trying to build a norm like this? Is there a much simpler one out there that satisfies the axioms and has the aforementioned features?

$\endgroup$
  • $\begingroup$ Hint: you can reformulate your ring as $2^nm$ where $m \in \mathbb Z$ is odd and $n \in \mathbb Z$. Then, |m| or similar should do it. $\endgroup$ – lisyarus Dec 2 at 23:17
  • $\begingroup$ Hmm, that does make things much easier than before, clearly the odd part doesn't change when the numerator and denominator are scaled by $2$. But what about restriction to $\Bbb Z$, wouldn't, for example, $N(6)=3$ in this case, as opposed to $6$ itself? $\endgroup$ – V.Ch. Dec 2 at 23:34
  • 1
    $\begingroup$ See Are all subrings of the rationals Euclidean domains?. $\endgroup$ – CopyPasteIt Dec 3 at 0:10
1
$\begingroup$

If $R$ is an euclidean domain with euclidean function $N$ then $R[a^{-1}]$ is an euclidean domain with euclidean function $$N(\frac{b}{a^k}) = \min \{ N(\frac{b}{a^m}),\frac{b}{a^m}\in R\}$$

Here $R[a^{-1}]=\Bbb{Z}[2^{-1}]$ and $$N(2^k(2m+1)) =N(\frac{2m+1}{2^k}) = |2m+1| = \# \Bbb{Z}[2^{-1}]/(2^k(2m+1))$$

$\endgroup$
  • $\begingroup$ Please refer to my comment above, doesn't this definition of norm conflict with the absolute value when restricted to $\Bbb Z$, i.g. $N(6)=3$, as opposed to $6$, as when viewed as an element of $\Bbb Z$? $\endgroup$ – V.Ch. Dec 3 at 18:57

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.