6
$\begingroup$

I am reading Fulton's book Algebraic Curves. Currently I am working on a specific problem (2.43), and I have doubts about my work and would appreciate another opinion(s).

Assume $p$ is the origin in $\mathbb{A}^n$ and $\mathcal O_p(\mathbb{A}^n)$ is the set of all rational functions defined on $\mathbb A^n$ and $m_p (\mathbb{A} ^n)$ is the set of non units. Show $I\mathcal O_p = m_p$ so $I^r\mathcal O_p = m_p^r$ where $I$ is the ideal generated by $x_1,...,x_n$.

My proof seems too simple and that is what bothers me.

$(\supset)$ Let $\phi \in m_p$ thus $\phi = \frac{f}{g} $ such that $f(p)=0$. Well $I \mathcal O_p$ is the set generated by $\frac{r}{u}$ where $r \in I$. Thus because $f(p)=0$ this implies it is in $\mathcal I(V(p))$ thus $f \in I$

$(\subset)$ Let $\phi \in I\mathcal O_p$ thus $\phi= \frac{k}{h}$ where $k \in I =<x_1,x_2,..x_n>$ thus $\phi \in m_p(\mathbb{A}^n)$

I am not very sure on the second part, but how does this look?

$\endgroup$
1
  • 1
    $\begingroup$ Fulton's book is a tough one, I've already read it. Your proof seems correct to me. In the first part, remember only to use the fact that $P=0\in\mathbb{A}^n$. At the second part, remember to use the fact that, in any ring, any ideal is contained in a maximal ideal (and the inclusion follows because there's only one maximal ideal in the local ring). By the way, once you are about to start section $2.9$, prepare yourself: proposition $6$ was the most difficult one to me. $\endgroup$
    – Larara
    Feb 26 '16 at 7:20
2
+50
$\begingroup$

The idea of the proof is correct. The proof is simple, but conveys the important relationship between $\mathcal{I}(p)$, the maximum ideal of polynomials vanishing on $p$, and $m_p$, the maximum ideal of rational functions vanishing on $p$. They are based on the same property of vanishing on $p$, which is not affected by denominators!

One important fact to state is that $\mathcal{I}(p) = I$. With this, the logical flow of the end of the first and second part of your proof is clear: $f(p) = 0$ implies $f\in\mathcal{I}(p) = I$ (note: not $\mathcal{I}(V(p))$!) for the first part, and $k\in I = \mathcal{I}(p)$ implies $k(p) = 0$ for the second.

In fact, the proof can be concisely written as follows:

$$ \phi= \frac{f}{g}\in m_p \Leftrightarrow \frac{f(p)}{g(p)} = 0 \Leftrightarrow f(p) = 0 \Leftrightarrow f\in \mathcal{I}(p) = I\Leftrightarrow \phi = \frac{f}{g}\in I\mathcal{O}_p $$

$\endgroup$

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.