Skip to content

Commit

Permalink
docs: fix #96 (#98)
Browse files Browse the repository at this point in the history
* docs: fix #96

* style: fix dots

* style: minor fix
  • Loading branch information
45gfg9 authored Oct 20, 2024
1 parent a8bae38 commit e7eb4ef
Show file tree
Hide file tree
Showing 6 changed files with 17 additions and 17 deletions.
8 changes: 4 additions & 4 deletions 习题参考答案/专题/4 线性空间的运算.tex
Original file line number Diff line number Diff line change
Expand Up @@ -40,7 +40,7 @@ \section*{4 线性空间的运算}
\end{pmatrix}\]
$\beta_1,\beta_2,e_1,e_2$即是扩张后的基,因此$W$的补空间的一组基为$e_1,e_2$.
\end{enumerate}
\end{enumerate}
\end{enumerate}

\item \begin{enumerate}
\item 先将各向量用坐标表示:
Expand Down Expand Up @@ -226,14 +226,14 @@ \section*{4 线性空间的运算}
\begin{enumerate}
\item 根据定义可知$\tr$是线性的,有 $\tr(\lambda A+\mu B)=\lambda\tr(A)+\mu\tr(B)=0$,则$U$是封闭的,又$W$封闭是显然的,则 $U,W$$V$ 的子空间.

\item 对于$W$,基为$\{E\}$$\dim W=1$. 对于$u$,设$E_{ij}$$a_{ij}=1$,其余元素为0的$n$阶方阵. 则由于$u$是对称矩阵,在非对角线元素上,$u$的基包含$E_{12}+E_{21},E_{13}+E_{31},\ldots ,E_{(n-1)n}+E_{n(n-1)}$. 对于对角线元素 $a_{11}+\cdots +a_{nn}=0$. 方程解为
\item 对于$W$,基为$\{E\}$$\dim W=1$. 对于$U$,设$E_{ij}$$a_{ij}=1$,其余元素为0的$n$阶方阵. 则由于$U$中的矩阵均为对称矩阵,在非对角线元素上,$U$的基包含$E_{12}+E_{21},E_{13}+E_{31},\ldots ,E_{(n-1)n}+E_{n(n-1)}$. 对于对角线元素 $a_{11}+\cdots +a_{nn}=0$. 方程解为
\[\begin{pmatrix}a_{11}\\ \vdots\\ a_{nn} \end{pmatrix}=k_1\begin{pmatrix}1 \\ -1 \\ 0\\ \vdots \\0 \end{pmatrix}=k_2\begin{pmatrix}1 \\ 0 \\ -1\\ \vdots \\0 \end{pmatrix}+\cdots+k_{n-1}\begin{pmatrix}1 \\ 0 \\ 0\\ \vdots \\-1 \end{pmatrix}.\]
则还有$n-1$个基$E_{11}-E_{22},E_{11}-E_{33},\ldots ,E_{11}-E_{nn}$. 故 $\dim U=\dfrac{n^2-n}2+n-2=\dfrac{n^2+n-2}2$.
则还有$n-1$个基$E_{11}-E_{22},E_{11}-E_{33},\ldots ,E_{11}-E_{nn}$. 故 $\dim U=\dfrac{n^2-n}2+n-1=\dfrac{n^2+n-2}2$.

\item$V'=U+W$. 先证明直和,即$U\cap W=\{0\}$. 这是显然的,因为$\tr(\lambda E)=n\lambda$,仅当 $\lambda=0$$\lambda E\in U$$U\cap W=\{0\}$得证. 又$\dim U+\dim W=\dim V'=n=\dim V$,则$V=V'=U\oplus W$得证.
\end{enumerate}

\item 由前 $B$ 组第 8 题的证明可知 $W_0=(W_1\cap W_2)\cup\cdots\cup(W_s\cap W_0)$. 由于$W_1\cap W_2\cdots W_s\cap W_0$都是$W_0$的子空间,根据覆盖定理,必存在$i$,使得$W_0=W_i\cap W_0$,即 $W_0\subseteq W_i$ 得证.
\item 由前 B 组第 8 题的证明可知 $W_0=(W_1\cap W_2)\cup\cdots\cup(W_s\cap W_0)$. 由于$W_1\cap W_0, \ldots, W_s\cap W_0$都是$W_0$的子空间,根据覆盖定理,因为 $W_0$ 是有限维的,必存在$i$,使得$W_0=W_i\cap W_0$,即 $W_0\subseteq W_i$ 得证.
\end{enumerate}

\clearpage
2 changes: 1 addition & 1 deletion 讲义/专题/1 预备知识.tex
Original file line number Diff line number Diff line change
Expand Up @@ -149,7 +149,7 @@ \section{基本代数结构}
\begin{enumerate}
\item 由于 $\mathbf{F}$ 是一个数域,它至少包含乘法单位元 $1 \neq 0$.通过交换律和结合律,我们可以反复使用加法构造正整数:
\[
1, \ 1 + 1 = 2, \ 1 + 1 + 1 = 3, \dots
1, \enspace 1 + 1 = 2, \enspace 1 + 1 + 1 = 3, \ldots
\]
因此,所有正整数都属于 $\mathbf{F}$.对于负整数,由于加法逆元的存在性,对于每个正整数 $n$$-n \in \mathbf{F}$,即负整数也在数域 $\mathbf{F}$ 中.因此,所有整数 $\mathbf{Z}$ 都在 $\mathbf{F}$ 中.
\item 由于 $\mathbf{F}$ 是一个数域,所以它对乘法和除法(除数不为 $0$)封闭,故 $\forall p,q \in \mathbf{Z}, p,q \in \mathbf{F}, \dfrac{p}{q} \in \mathbf{F}$,故 $\mathbf{Q} \subseteq \mathbf{F}$.
Expand Down
10 changes: 5 additions & 5 deletions 讲义/专题/2 线性空间.tex
Original file line number Diff line number Diff line change
Expand Up @@ -591,30 +591,30 @@ \section{线性表示 \quad 线性扩张}

\begin{answer}
\begin{enumerate}
\item 对于集合 \( W = \{(x_1, \dots, x_n) \in F^n \mid a_1x_1 + \cdots + a_nx_n = 0\} \)
\item 对于集合 \( W = \{(x_1, \ldots, x_n) \in F^n \mid a_1x_1 + \cdots + a_nx_n = 0\} \)

首先,我们来判断该集合是否为线性空间的子空间. 判断一个集合是否为线性空间的子空间需要满足以下三个条件:

\begin{enumerate}
\item \text{零向量在其中}:

\( x_1 = x_2 = \cdots = x_n = 0 \) 时,方程 \( a_1x_1 + \cdots + a_nx_n = 0 \) 显然成立,因此零向量 \( (0, 0, \dots, 0) \) 属于 \( W \).
\( x_1 = x_2 = \cdots = x_n = 0 \) 时,方程 \( a_1x_1 + \cdots + a_nx_n = 0 \) 显然成立,因此零向量 \( (0, 0, \ldots, 0) \) 属于 \( W \).

\item \text{封闭性(加法)}:

假设 \( \mathbf{v} = (x_1, \dots, x_n) \)\( \mathbf{w} = (y_1, \dots, y_n) \) 属于 \( W \),即:
假设 \( \mathbf{v} = (x_1, \ldots, x_n) \)\( \mathbf{w} = (y_1, \ldots, y_n) \) 属于 \( W \),即:
\[
a_1x_1 + \cdots + a_nx_n = 0 \quad \text{且} \quad a_1y_1 + \cdots + a_ny_n = 0.
\]
那么,对于 \( \mathbf{v} + \mathbf{w} = (x_1 + y_1, \dots, x_n + y_n) \),有:
那么,对于 \( \mathbf{v} + \mathbf{w} = (x_1 + y_1, \ldots, x_n + y_n) \),有:
\[
a_1(x_1 + y_1) + \cdots + a_n(x_n + y_n) = (a_1x_1 + \cdots + a_nx_n) + (a_1y_1 + \cdots + a_ny_n) = 0 + 0 = 0.
\]
因此,\( \mathbf{v} + \mathbf{w} \in W \).

\item \text{封闭性(数乘)}:

对于任意 \( \mathbf{v} = (x_1, \dots, x_n) \in W \) 和任意 \( \lambda \in F \),有:
对于任意 \( \mathbf{v} = (x_1, \ldots, x_n) \in W \) 和任意 \( \lambda \in F \),有:
\[
a_1(\lambda x_1) + \cdots + a_n(\lambda x_n) = \lambda (a_1x_1 + \cdots + a_nx_n) = \lambda \cdot 0 = 0.
\]
Expand Down
4 changes: 2 additions & 2 deletions 讲义/专题/24 多重线性映射与张量的计算.tex
Original file line number Diff line number Diff line change
Expand Up @@ -49,7 +49,7 @@ \section{多重线性映射}
\]
则它就是一个 $V_2 \to W$ 的线性映射. 设它所对应的矩阵为 $M_i$,考虑 $V_1$ 中的向量 $v_1 = \lambda_1 e_1 + \lambda_2 e_2 + \cdots \lambda_n e_n$,其中 $n = \dim V_1$,则
\[
f(v_1, v_2) = f(\lambda_1 e_1 + \lambda_2 e_2 + \dots + \lambda_n e_n, v_2) = \lambda_1 f_1 (v_2) + \lambda_2 f_2 (v_2) + \cdots + \lambda_n f_n (v_2)
f(v_1, v_2) = f(\lambda_1 e_1 + \lambda_2 e_2 + \ldots + \lambda_n e_n, v_2) = \lambda_1 f_1 (v_2) + \lambda_2 f_2 (v_2) + \cdots + \lambda_n f_n (v_2)
\]
将线性映射翻译成矩阵,我们就有:
\[
Expand Down Expand Up @@ -85,7 +85,7 @@ \section{张量积}
实际上,这个命题的证明无非就是把我们在最开始所做的关于多重线性映射的表示的讨论重述一遍.

\begin{proof}
$V$ 有一组基 $\{e_1, e_2, \dots, e_n\}$$f \in \mathcal{L}(V, W^*; \R)$. 记
$V$ 有一组基 $\{e_1, e_2, \ldots, e_n\}$$f \in \mathcal{L}(V, W^*; \R)$. 记
\[
f_i (\rho) = f(e_i, \rho), \forall \rho \in W^*
\]
Expand Down
8 changes: 4 additions & 4 deletions 讲义/专题/4 线性空间的运算.tex
Original file line number Diff line number Diff line change
Expand Up @@ -645,7 +645,7 @@ \subsection{仿射子集与商空间}
假设 $r = k-1$ 时也成立, 即存在 $\alpha \in V$
使得 $\alpha$ 同时不属于 $V_i (i=1,2,\ldots,k-1)$
下面证明 $r=k$ 时也成立. 显然当此 $\alpha \notin V_k$$\alpha$ 即为所求.
$\alpha \in V_k$, 由于 $V_k$$V$ 的真子空间, 故有 $\beta \notin V_k$, 易知对任意 $V_i (i=1,2,\ldots,k-1)$ 都至多有一个 $k_i$, 使得 $\beta + k_i \alpha \notin V_i; i=1,2,\ldots,k-1$, 我们取异于 $k_i$ 的数 $k$, 则必有 $\beta + k \alpha \notin V_i; i=1,2,\ldots,k-1$. 又由于 $\alpha \in V_k, \beta \notin V_k$ 故有 $\beta + k \alpha \notin V_k$, 故 $\beta + k \alpha \notin V_k (i = 1,2,\dots,k)$.
$\alpha \in V_k$, 由于 $V_k$$V$ 的真子空间, 故有 $\beta \notin V_k$, 易知对任意 $V_i (i=1,2,\ldots,k-1)$ 都至多有一个 $k_i$, 使得 $\beta + k_i \alpha \notin V_i; i=1,2,\ldots,k-1$, 我们取异于 $k_i$ 的数 $k$, 则必有 $\beta + k \alpha \notin V_i; i=1,2,\ldots,k-1$. 又由于 $\alpha \in V_k, \beta \notin V_k$ 故有 $\beta + k \alpha \notin V_k$, 故 $\beta + k \alpha \notin V_k \enspace (i = 1,2,\ldots,k)$.
\end{answer}
\end{exgroup}

Expand Down Expand Up @@ -673,7 +673,7 @@ \subsection{仿射子集与商空间}
0 & 0 & 0 & 0 \\
0 & 0 & 0 & 0
\end{pmatrix}\]
则解向量$u_1=\left(\dfrac {3}{17},\dfrac{19}{17},1,0\right)^T,u_2=\left(\dfrac{13}{17},-\dfrac{20}{17},0,1\right)^T$. $u_1,u_2$$V_1\cap V_2$的基,则其维数为2.
则解向量$u_1=\left(\frac {3}{17},\frac{19}{17},1,0\right)^\mathrm{T},u_2=\left(\frac{13}{17},-\frac{20}{17},0,1\right)^\mathrm{T}$. $u_1,u_2$$V_1\cap V_2$的基,则其维数为2.

\item 易得$\dim V_1=2,\dim V_2=2$,则由维数公式$\dim (V_1+V_2)=\dim V_1+\dim V_2-\dim (V_1\cap V_2)= 2$$V_1\cap V_2\subseteq V_1+V_2$,且二者维数相等. 则$V_1\cap V_2=V_1+V_2$,亦即$V_1=V_2$. 所以$V_1+V_2$的基也是$u_1,u_2$,同$V_1\cap V_2$.
\end{enumerate}
Expand Down Expand Up @@ -791,7 +791,7 @@ \subsection{仿射子集与商空间}

\item 受三个有限集之并集的元素数量公式的启发,你可能会这样猜测:如果 $V_1,V_2,V_3$ 是一有限维向量空间的子空间,那么有
\begin{align*}
\dim(V_1+V_2+V_3) =\ &\dim V_1+\dim V_2+\dim V_3 \\
\dim(V_1+V_2+V_3) ={} &\dim V_1+\dim V_2+\dim V_3 \\
&-\dim(V_1\cap V_2)-\dim(V_1\cap V_3)-\dim(V_2\cap V_3) \\
&+\dim(V_1\cap V_2\cap V_3).
\end{align*}
Expand All @@ -803,7 +803,7 @@ \subsection{仿射子集与商空间}
\item 证明:如果 $V_1,V_2,V_3$ 是一有限维向量空间的子空间,那么
\begin{align*}
&\dim(V_1+V_2+V_3) \\
&= \dim V_1+\dim V_2+\dim V_3 \\
={} & \dim V_1+\dim V_2+\dim V_3 \\
&-\dfrac{\dim(V_1\cap V_2)+\dim(V_1\cap V_3)+\dim(V_2\cap V_3)}{3} \\
&+\dfrac{\dim((V_1 + V_2) \cap V_3) - \dim((V_1 + V_3) \cap V_2) + \dim((V_2 + V_3) \cap V_1)}{3}.
\end{align*}
Expand Down
2 changes: 1 addition & 1 deletion 讲义/专题/8 相抵标准形.tex
Original file line number Diff line number Diff line change
Expand Up @@ -611,7 +611,7 @@ \subsection{从初等变换到相抵标准形}
定理的证明很简单,只需对各个初等变换逐一通过计算验证即可,具体证明如下:

\begin{proof}
初等列变换和初等行变换的证明是完全类似的,所以我们仅就初等行变换的情形给以证明. 设 $m\times n$ 矩阵 $A$$m$ 个行向量为 $\{ a_1, a_2, \dots, a_m \}$.
初等列变换和初等行变换的证明是完全类似的,所以我们仅就初等行变换的情形给以证明. 设 $m\times n$ 矩阵 $A$$m$ 个行向量为 $\{ a_1, a_2, \ldots, a_m \}$.
\begin{enumerate}
\item$A$ 的第 $i, j$ 行对换得到 $B$,则 $B$$A$ 的行向量组相同,故 $A, B$ 的行秩相等.
\item$A$ 的第 $i$ 行乘非零常数 $c$ 得到 $B$,则 $B$ 的行向量组:$\beta_i = c\alpha_i, \beta_k = \alpha_k \enspace (k \neq i)$. 因此 $B$$A$ 的行向量组可以互相线性表示. 所以 $A$$B$ 的行秩相等.
Expand Down

0 comments on commit e7eb4ef

Please sign in to comment.