数学代考|Finitely Generated Abelian Groups Structure theorem

Finitely Generated Abelian Groups Structure theorem的大致内容就是任一有限 Abel 群都同构于若干个循环群的直和.
该定理有两种表达形式，设 $G$ 是有限 Abel 群， $n=|G|$ 的素因子分解为 $p_1^{e_1} \cdots p_t^{e_t}$ ，则 (下 面 $Z_d$ 表示 $d$ 阶循环群)：
（1） $G \simeq \bigoplus_{i=1}^t\left(\bigoplus_{j=1}^{k_i} Z_{p_i}\right)$ ，其中 $l_{i j}$ 是一组正整数，满足 $\sum_{j=1}^{k_i} l_{i j}=e_i$ ；
(2) $G \simeq \bigoplus_{i=1}^k Z_{d_i}$ ，其中 $d_i$ 是一组正整数，满足 $d_1\left|d_2\right| \cdots \mid d_k$ 且 $\prod_{i=1}^k d_i=n$. 这些 $d_i$ 称为 $G$ 的不变因子.
部分课程在证明该定理时需要用Module模的性质，所以需要到Module模的部分才会详细介绍。实际上此结构定理有只依赖于群论的证明方法。

UpriviateTA有若干位非常擅长Abstract algebra的老师可以提供抽象代数的代写，代考和辅导服务，可以确保您在学习Abstract algebra的过程中取得满意的成绩。

以下是Duke University Algebraic Structures I Finitely Generated Abelian Groups Structure theorem部分的内容的总结.

如果您在学习Math 500, Abstract Algebra I， 16:640:551 – Abstract Algebra I, The City University of New York MTH339

或类似课程中遇到困难，可以直接联系我们获得帮助。

更多的经典案例请参阅以往案例，关于abstract algebra的更多的以往案例可以参阅相关文章。abstract algebra代写请认准UpriviateTA.

Free abelian groups

The Fundamental Theorem of Finitely Generated Abelian Groups (abbreviated by FTFGAG), which, among other things, provides a complete classification of all finite abelian groups. The proof begins with a study of free abelian groups.

A subset $X \subseteq G$ of an abelian group is called linearly independent if for every finite subset $\left{x_1, x_2, \ldots, x_r\right} \subseteq X$,
$$
c_1 x_1+c_2 x_2+\cdots+c_r x_r=0 \Longrightarrow c_1=c_2=\cdots=c_r=0 .
$$
A basis of an abelian group $G$ is a linearly independent subset $X$ that generates $G$.

free abelian group

An abelian group $(G,+)$ is called a free abelian group if it has a basis.

In particular, $\mathbb{Z}, \mathbb{Z} \oplus \mathbb{Z}$, and more generally $\mathbb{Z}^r$ for a positive integer $r$ are free abelian groups. A free abelian group could have an infinite basis. For example, $\mathbb{Z}[x]$ is a free abelian group with basis $S=\left{1, x, x^2, \ldots\right}$ because every polynomial in $\mathbb{Z}[x]$ is a (finite) linear combination of elements in $S$. If a free abelian group has an infinite basis, every element must still be a finite linear combination of basis elements.

Proof. Let $X=\left{x_1, x_2, \ldots, x_r\right}$ be a finite basis of $G$. Consider the function $\varphi: \mathbb{Z}^r \rightarrow G$ given by
$$
\varphi\left(c_1, c_2, \ldots, c_r\right)=c_1 x_1+c_2 x_2+\cdots+c_r x_r .
$$
This function satisfies
$$
\begin{aligned}
\varphi\left(\left(c_1, c_2, \ldots, c_r\right)+\left(d_1, d_2, \ldots, d_r\right)\right) \
\quad=\varphi\left(c_1+d_1, c_2+d_2, \ldots, c_r+d_r\right) \
\quad=\left(c_1+d_1\right) x_1+\left(c_2+d_2\right) x_2+\cdots+\left(c_r+d_r\right) x_r \
\quad=c_1 x_1+c_2 x_2+\cdots+c_r x_r+d_1 x_1+d_2 x_2+\cdots+d_r x_r \
\quad=\varphi\left(c_1, c_2, \ldots, c_r\right)+\varphi\left(d_1, d_2, \ldots, d_r\right)
\end{aligned}
$$
so it is a homomorphism. Since the basis generates $G$, then $\varphi$ is surjective. Furthermore, since
$$
\begin{aligned}
\operatorname{Ker} \varphi & =\left{\left(c_1, c_2, \ldots, c_r\right) \in \mathbb{Z}^r \mid c_1 x_1+c_2 x_2+\cdots+c_r x_r=0\right} \
& ={(0,0, \ldots, 0)},
\end{aligned}
$$
the homomorphism is also injective. Thus, $\varphi$ is an isomorphism.

Proof. Suppose that $G$ has a basis $X$ with $r$ elements. Then $G$ is isomorphic to $\mathbb{Z}^r$. The subgroup $2 G={g+g \mid g \in G}$ is isomorphic to $(2 \mathbb{Z})^r$ so by Exercise 2.4.7,
$$
G / 2 G=(\mathbb{Z} \oplus \mathbb{Z} \oplus \cdots \oplus \mathbb{Z}) /(2 \mathbb{Z} \oplus 2 \mathbb{Z} \oplus \cdots \oplus 2 \mathbb{Z}) \cong Z_2^r .
$$
Thus, $|G / 2 G|=2^r$.
First, assume that $G$ also has a finite basis with $s \neq r$ elements. Then $|G / 2 G|=2^s \neq 2^r$, a contradiction.

Second, assume that $G$ has an infinite basis $Y$. Let $y_1, y_2 \in Y$. Assume also that $\bar{y}_1=\bar{y}_2$ in $G / 2 G$, then $y_1-y_2 \in 2 G$, so $y_1-y_2$ is a finite linear combination of elements in $Y$ (with even coefficients). In particular, $Y$ is a linearly dependent set, which contradicts $Y$ being a basis. Thus, in the quotient group $G / 2 G$, the elements ${\bar{y} \mid y \in Y}$ are all distinct. Since $Y$ is an infinite set, so is $G / 2 G$. This contradicts $|G / 2 G|=2^r$.

Consequently, if $G$ has a basis of $r$ elements, then every other basis is finite and has $r$ elements.

If $G$ is a finitely generated free abelian group, then the common number $r$ of elements in a basis is called the rank. The rank is also called the Betti number of $G$ and is denoted by $\beta(G)$.

Proof. We prove the theorem by starting from a basis of $G$ and repeatedly adjusting it

By the well-ordering of the integers, there exists a minimum value $n_1$ in the set
$\left{c_1 \in \mathbb{N}^* \mid c_1 y_1+c_2 y_2+\cdots+c_s y_s \in H\right.$ for any basis $\left{y_1, y_2, \ldots, y_s\right}$ of $\left.G\right}$.
Let $z_1 \in H$ be an element that instantiates this minimum value, and write $z_1=n_1 y_1+c_2 y_2+\cdots+c_s y_s$. By integer division, for all $i \geq 2$, we can write $c_i=n_1 q_i+r_i$ with $0 \leq r_i<n_1$. Set $x_1=y_1+q_2 y_2+\cdots+q_s y_s$. $\left{x_1, y_2, \ldots, y_s\right}$ is a basis of $G$. Furthermore,
$$
z_1=n_1 x_1+r_2 y_2+\cdots+r_s y_s .
$$
However, since $n_1$ is the least positive coefficient that occurs in any linear combination over any basis of $G$ and, since $0 \leq r_i<n_1$, we have $r_2=\cdots=$ $r_s=0$. So, in fact $z_1=n_1 x_1$.

If $\left{n_1 x_1\right}$ generates $H$, then we are done and $\left{n_1 x_1\right}$ is a basis of $H$. If not, then there is a least positive value of $c_2$ for linear combinations
$$
z_2=a_1 x_1+c_2 y_2+\cdots+c_s y_s \in H,
$$

where $y_2, \ldots, y_s \in G$ such that $\left{x_1, y_2, \ldots, y_s\right}$ is a basis of $G$. Call this least positive integer $n_2$. Note that we must have $n_1 \mid a_1$, because otherwise, since $n_1 x_1 \in H$ by subtracting a suitable multiple of $n_1 x_1$ from $a_1 x_1+c_2 y_2+\cdots+$ $c_s y_s$ we would obtain a linear combination of $\left{x_1, y_2, \ldots, y_s\right}$ that is in $H$ and has a lesser positive coefficient for $x_1$, which would contradict the minimality of $n_1$. Again, if we take the integer division of $c_i$ by $n_2, c_i=n_2 q_i+r_i$ with $0 \leq r_i0$ would contradict the minimality of $n_2$. Thus, $z_2=a_1 x_1+n_2 x_2 \in H$ and also $n_2 x_2=z_2-\left(a_1 / n_1\right) n_1 x_1 \in H$. By Lemma 2.5.8, $\left{x_1, x_2, y_3, \ldots, y_s\right}$ is a basis of $G$. Furthermore, if we consider the integer division of $n_2$ by $n_1$, written as $n_2=n_1 q+r$ with $0 \leq r<n_1$, then
$$
n_1 x_1+n_2 x_2=n_1\left(x_1+q x_2\right)+r x_2 \in H .
$$
But then, since $r<n_1$, by the minimal positive condition on $n_1$, we must have $r=0$. Hence, $n_1 \mid n_2$.

If $\left{n_1 x_2, n_2 x_2\right}$ generates $H$, then we are done because the set is linearly independent since by construction $\left{x_1, x_2, y_3, \ldots, y_s\right}$ is a basis of $G$ so $\left{x_1, x_2\right}$ is linearly independent and $\left{n_1 x_1, n_2 x_2\right}$ is then a basis of $H$. The pattern continues and only terminates when it results in a basis
$$
\left{x_1, \ldots, x_t, y_{t+1}, \ldots, y_s\right}
$$
of $G$ such that $\left{n_1 x_1, n_2 x_2, \ldots, n_t x_t\right}$ is a basis of $H$ for some positive integers $n_i$ such that $n_i \mid n_{i+1}$ for $1 \leq i \leq t-1$.

This proof is not constructive since it does not provide a procedure to find the $n_i$, which is necessary to construct the $x_1, x_2$, and so on. We merely know the $n_i$ exist by the well-ordering of integers. In some instances it is easy to find a basis of the subgroup as in the following example.

Example

Consider the free abelian group $G=\mathbb{Z}^3$ and the subgroup $H=\left{(x, y, z) \in \mathbb{Z}^3 \mid x+2 y+3 z=0\right}$. If we considered the equation $x+$ $2 y+3 z=0$ in $\mathbb{Q}^3$, then the Gauss-Jordan elimination algorithm gives $H$ as the span of $\operatorname{Span}({(-2,1,0),(-3,0,1)}$. Taking only integer multiples of these two vectors does give all points $(x, y, z)$ with $y$ and $z$ taking on every pair of integers. Hence, ${(-2,1,0),(-3,0,1)}$ is a basis of $H$ and we see clearly that $H$ is a free abelian group of rank 2.

Invariant Factors Decomposition

The difficult work in the proof of above Theorem leads to the Fundamental Theorem for Finitely Generated Abelian Groups.

Proof. Since $G$ is finitely generated, there is a finite subset $\left{g_1, g_2, \ldots, g_s\right}$ that generates $G$. Define the function $h: \mathbb{Z}^s \rightarrow G$ by
$$
h\left(n_1, n_2, \ldots, n_s\right)=n_1 g_1+n_2 g_2+\cdots+n_s g_s .
$$
By the same reasoning as the proof of Theorem 2.5.5, $h$ is a surjective homomorphism. Then Ker $h$ is a subgroup of $\mathbb{Z}^s$ and by the First Isomorphism Theorem, since $h$ is surjective, $\mathbb{Z}^s /(\operatorname{Ker} h) \cong G$. By Theorem 2.5 .9 , there exists a basis $\left{x_1, x_2, \ldots, x_s\right}$ for $\mathbb{Z}^s$ and positive integers $n_1, n_2, \ldots, n_t$, where $n_i$ divides $n_{i+1}$ for all $1 \leq i \leq t-1$ such that $\left{n_1 x_1, n_2 x_2, \ldots, n_t x_t\right}$ is a basis of Ker $h$. Then
$$
\begin{aligned}
G & \cong(\mathbb{Z} \oplus \mathbb{Z} \oplus \cdots \oplus \mathbb{Z}) /\left(n_1 \mathbb{Z} \oplus \cdots \oplus n_t \mathbb{Z} \oplus{0} \oplus \cdots \oplus{0}\right) \
& \cong Z_{n_1} \oplus Z_{n_2} \oplus \cdots \oplus Z_{n_t} \oplus \mathbb{Z} \oplus \cdots \oplus \mathbb{Z} .
\end{aligned}
$$
Now if $n_i=1$, then $Z_{n_i} \cong{0}$, the trivial group. Let $k$ be the number of indices such that $n_i>1$. The theorem follows after setting $d_i=n_{t+1-i}$ for $1 \leq i \leq k$.
Definition 2.5.12
As with free groups, the integer $r$ is called the rank or the Betti number of $G$. It is sometimes denoted by $\beta(G)$. The integers $d_1, d_2, \ldots, d_k$ are called the invariant factors of $G$ and the expression (2.7) is called the invariant factors decomposition of $G$.
It is interesting to note that the proof of Theorem 2.5.11 is not constructive in the sense that it does not provide a method to find specific elements in $G$ whose orders are the invariant factors of $G$. The invariant factors exist by virtue of the well-ordering principle of the integers.

Applied to finite groups (which obviously are finitely generated), Theorem 2.5.11 gives us an effective way to describe all abelian groups of a given order $n$. If $G$ is finite, the $\operatorname{rank}$ of $G$ is 0 . Then we must find all finite sequences of integers $d_1, d_2, \ldots, d_k$ such that

• $d_i \geq 2$ for $1 \leq i \leq k$;
• $d_{i+1} \mid d_i$ for $1 \leq i \leq k-1$;
• $n=d_1 d_2 \cdots d_k$.

The first two conditions are explicit in the above theorem. The last condition follows from the fact that $d_i=\left|x_i\right|$ where the $\left{x_1, x_2, \ldots, x_k\right}$ is a list of corresponding generators of $G$. Then every element in $G$ can be written uniquely as
$$
g=\alpha_1 x_1+\alpha_2 x_2+\cdots+\alpha_k x_k
$$
for $0 \leq \alpha_i<d_i$. Thus, the order of the group is $n=d_1 d_2 \cdots d_k$.
We list a few examples of abelian groups of a given order where we find the invariant factors decomposition. (Note that in the notation $Z_d$ for a cyclic group, we assume the operation is multiplication but that is irrelevant to the theorem.)

A Few Comments on Partitions of Integers

Partitions of integers form a vast area of research and find applications in many areas of higher mathematics. See [19, Chapter XIX] or [1] for an introduction to partitions of integers.

Considering a partition $\alpha$ as any finite non-increasing sequence of positive integers, we use the notation
$$
|\alpha| \stackrel{\text { def }}{=} \alpha_1+\alpha_2+\cdots+\alpha_k
$$
and call this the content of $\alpha$. The integer $k$ is the length (or the number of parts) of $\alpha$ and is written $\ell(\alpha)=k$. In the terminology of partitions, each summand $\alpha_i$ of a partition is called a part of $\alpha$. For example, if $\alpha=(5,2,2,1)$, then the content of $\alpha$ is $|\alpha|=10$, so $\alpha$ is a partition of 10 with 4 parts.

We often represent a partition by its so-called Young diagram in which each part $\alpha_i$ is represented by $\alpha_i$ boxes that are left aligned, descending on the page as $i$ increases. The Young diagram for $(5,2,2,1)$ is:

Like a matrix, we call the main diagonal of the Young diagram the set of $i$ th boxes in the $i$ th row.

The conjugate of a partition $\alpha$ is the partition $\alpha^{\prime}$ obtained by reflecting the Young diagram of $\alpha$ through its main diagonal. Algebraically, the values of the conjugate partition are
$$
\alpha_j^{\prime}=\left|\left{1 \leq i \leq k \mid \alpha_i \geq j\right}\right| .
$$
With $\alpha=(5,2,2,1)$, the conjugate partition is $\alpha^{\prime}=(4,3,1,1,1)$. The intuition of the Young diagram makes it clear that $\left|\alpha^{\prime}\right|=|\alpha|$, so if $\alpha$ is a partition of $n$, then so is $\alpha^{\prime}$.

abstract algebra代写请认准UpriviateTA. UpriviateTA为您的留学生涯保驾护航。

更多内容请参阅另外一份Galois代写.