找回密码
 快速注册
搜索
查看: 83|回复: 5

[数论] $ℤ\left[1+\sqrt{-19}\over2\right]$不是 ED

[复制链接]

3149

主题

8386

回帖

6万

积分

$\style{scale:11;fill:#eff}꩜$

积分
65391
QQ

显示全部楼层

hbghlyj 发表于 2023-2-13 02:11 |阅读模式
本帖最后由 hbghlyj 于 2023-2-15 17:40 编辑 设$θ=\frac{1}{2}(1+\sqrt{-19})$, $ℤ\left[θ\right]=\Set{m+nθ:m,n∈ℤ}$,
证明:不存在函数 $d :ℤ\left[θ\right]\setminus\{0\}→ \Bbb N$,满足:
  • 若 $a$ 整除 $b$,则 $d ( a ) ≤ d ( b )$。
  • 若 $a$ 不整除 $b$,则存在 $q , r ∈ ℤ\left[θ\right]$ 使得 $a = b q + r$,而且 $d ( r ) < d ( b )$。

来自Algebraic structures I, Spring 2011
证明
We choose $m \in R$ such that $d(m)$ is as small as possible, subject to $m$ not being 0 or a unit. First we divide 2 by $m$, and get a quotient and remainder:
$$
2=m q+r \text { with } d(r)<d(m) \text { or } r=0 .
$$
This means $r=0,1$, or $-1$. If $r=0$, then $m \mid 2$, which means $m=\pm 2$, since 2 is irreducible, and $m$ is not a unit. Similarly, if $r=-1$, then $m=\pm 3$. The case $r=1$ cannot happen, for if it did, then $m \mid 1$ so $m$ is a unit.
Next we divide $\theta$ by $m$ in the same way, getting
$$
\theta=m q^{\prime}+r^{\prime} \text { with } d\left(r^{\prime}\right)<d(m) \text { or } r^{\prime}=0 .
$$
Again we have $r^{\prime}=0,1$, or $-1$. So one of $\theta, \theta+1$, or $\theta-1$ is divisible by $m$. But $m=\pm 2$ or $\pm 3$, and it is easy to see that none of these quotients is in $R$. This contradiction tells us that $R$ is not a Euclidean domain for any Euclidean function.

点评

这是好题目/好反例的其中一半, 也就是证明这个环不是 ED(Euclidean domain), 另一部分是证明这个环是 PID(principle ideal domain).  发表于 2023-2-13 14:42

3149

主题

8386

回帖

6万

积分

$\style{scale:11;fill:#eff}꩜$

积分
65391
QQ

显示全部楼层

 楼主| hbghlyj 发表于 2023-2-13 16:12
θ 可以换成其它的吗

点评

有很多类型构造以及序列, 直接维基找  发表于 2023-2-13 18:02

3149

主题

8386

回帖

6万

积分

$\style{scale:11;fill:#eff}꩜$

积分
65391
QQ

显示全部楼层

 楼主| hbghlyj 发表于 2023-4-12 22:14
扩大环到 $\mathbb Q[\sqrt{-19}]$ 是否仍然正确

3149

主题

8386

回帖

6万

积分

$\style{scale:11;fill:#eff}꩜$

积分
65391
QQ

显示全部楼层

 楼主| hbghlyj 发表于 2023-4-12 22:16
$ℤ\left[1+\sqrt{-19}\over2\right]$也是另一个问题的反例:一般线性群是由初等矩阵生成的吗?
这两个问题有关系吗?他们为什么有相同的反例?
Remark 6.1.
We write $\text{GL}_n (R)$ for the group $U (M_n (R))$, and $\text{GE}_n (R)$ for the subgroup of $\text{GL}_n (R)$ generated by the transvections, dilations, and interchanges.

In general $\text{GL}_2 (R) \neq \text{GE}_2 (R)$, though this can be hard to show. An example, taken from [Coh66, p23], is the ring $\mathbb{Z}[\theta]$ where $\theta^2 - \theta + 5 = 0$. Here the matrix
\[ A ≔ \left(\begin{array}{cc}
     3 - \theta & 2 + \theta\\
     - 3 - 2 \theta & 5
     - 2 \theta
   \end{array}\right) \]
is in $\text{GL}_2 (\mathbb{Z}[\theta])$ but not in $\text{GE}_2(\mathbb{Z}[\theta])$.

手机版|悠闲数学娱乐论坛(第3版)

GMT+8, 2025-3-4 15:56

Powered by Discuz!

× 快速回复 返回顶部 返回列表