##### 线性变换的矩阵表示 - 线性变换的矩阵表示 - **线性变换的矩阵表示**是指在给定[[向量空间的基|基]]下用[[矩阵变换]]表示[[线性变换]], 其[[矩阵]]由变换本身以及两个基决定. 设线性变换 $T: V \to W$, 两空间的基分别为 $\mathcal{B}=\{\mathbf{b}_1, \mathbf{b}_2, \dots, \mathbf{b}_n\}$ 和 $\mathcal{C}=\{\mathbf{c}_1, \mathbf{c}_2, \dots, \mathbf{c}_m\}$, 与[[基的变换]]类似, 线性变换的矩阵表示为 $A_{\mathcal{B}\to \mathcal{C}}$ , 该矩阵每个列向量是基 $\mathbf{b}_j$ 变换后 $T(\mathbf{b}_j)$ 在基 $\mathcal{C}$ 下的坐标向量 $[T(\mathbf{b}_j)]_\mathcal{C}$ . 线性变换关于不同基的矩阵是[[矩阵等价]]的 - $T: V \to W$, $\mathbf{x}\mapsto T(\mathbf{x})$, $[T(\mathbf{x})]_\mathcal{C}= A[\mathbf{x}]_\mathcal{B}$ - $A=\begin{bmatrix} [T(\mathbf{b}_1)]_\mathcal{C} & [T(\mathbf{b}_2)]_\mathcal{C} & \cdots & [T(\mathbf{b}_n)]_\mathcal{C} \end{bmatrix}$ - $A =\begin{bmatrix}a_{11}&a_{12}&\cdots&a_{1n}\\a_{21}&a_{22}&\cdots&a_{2n}\\\vdots&\vdots&\ddots&\vdots\\a_{m1}&a_{m2}&\cdots&a_{mn}\\\end{bmatrix}$ - 设 $\mathbf{x}\in V$ 在基 $\mathcal{B}$ 下的坐标向量为 $(x_1, x_2, \dots, x_n)$, 可以写作线性组合 - $\mathbf{x} = x_1 \mathbf{b}_1 + x_2 \mathbf{b}_2 + \dots + x_n \mathbf{b}_n$ - 线性变换 $T(\mathbf{x})$ 可以展开为 - $\displaystyle T(\mathbf{x}) = T ( \sum_{j=1}^{n} x_j \mathbf{b}_j )=\sum_{j=1}^{n} x_j T(\mathbf{b}_j)$, $j =1,2,\dots,n$ - 每个基向量 $\mathbf{b}_j$​ 变换后 $T(\mathbf{b}_j)$ 是 $W$ 的一个向量, 可以在基 $\mathcal{C}$ 下表示 - $T(\mathbf{b}_j)=a_{1j}\mathbf{c}_1+a_{2j}\mathbf{c}_2+\cdots+a_{mj}\mathbf{c}_m$ - 将 $T(\mathbf{b}_j)$ 的坐标形式代入, 并代入矩阵每一列, 可以写作 - $\displaystyle [T(\mathbf{x})]_{\mathcal{C}} = \sum_{j=1}^{n} x_j [T(\mathbf{b}_j)]_{\mathcal{C}}$ - $[T(\mathbf{x})]_\mathcal{C}= A[\mathbf{x}]_\mathcal{B}$ - $[T(\mathbf{x})]_{\mathcal{C}} = \begin{bmatrix} a_{11} & a_{12} & \dots & a_{1n} \\ a_{21} & a_{22} & \dots & a_{2n} \\ \vdots & \vdots & \ddots & \vdots \\ a_{m1} & a_{m2} & \dots & a_{mn} \end{bmatrix} \begin{bmatrix} x_1 \\ x_2 \\ \vdots \\ x_n \end{bmatrix}$ >[!example]- 从 $P_3(\mathbb{R})$ 到 $P_2(\mathbb{R})$ 的微分映射的矩阵 > - 从 $P_3(\mathbb{R})$ 到 $P_2(\mathbb{R})$ 的微分映射定义为对任意多项式 $p(x) = a_3 x^3 + a_2 x^2 + a_1 x + a_0$​ 进行微分 > - 微分映射 $D$ 将 $p(x)$ 变为其导数 $D(p(x)) = p'(x) = 3a_3 x^2 + 2a_2 x + a_1$ > - 可以看到, 导数 $p'(x)$ 是一个 $P_2(\mathbb{R})$ 的多项式 > - 选择 $P_3(\mathbb{R})$ 和 $P_2(\mathbb{R})$ 的基 > - $P_3(\mathbb{R})$ 的基可以选为 $V=\{1, x, x^2, x^3\}$ > - $P_2(\mathbb{R})$ 的基可以选为 $W=\{1, x, x^2\}$ > - 基变换后的坐标向量 > - $D(1) = 0$ 在基 $W$ 下表示为 $(0,0,0)$ > - $D(x)=1$ 在基 $W$ 下表示为 $(1,0,0)$ > - $D(x2)=2x$ 在基 $W$ 下表示为 $(0,2,0)$ > - $D(x^3)=3x^2$ 在基 $W$ 下表示为 $(0,0,3)$ > - 组装矩阵 > - $D =\begin{bmatrix} [T(\mathbf{v}_1)]_W & [T(\mathbf{v}_2)]_W & \cdots & [T(\mathbf{v}_n)]_W \end{bmatrix}= \begin{bmatrix} 0 & 1 & 0 & 0 \\ 0 & 0 & 2 & 0 \\ 0 & 0 & 0 & 3 \end{bmatrix}$ >[!example]- $T(x, y) = (x+3y,2x +5y,7x +9y)$ > - $V=\{(1,0),(0,1)\}$, $W=\{(1,0,0),(0,1,0),(0,0,1)\}$ > - $[T(\mathbf{v}_1)]_W=(1,2,7)$ > - $[T(\mathbf{v}_2)]_W=(3,5,9)$ > - $A=\begin{bmatrix} 1 &3\\ 2&5 \\ 7&9 \end{bmatrix}$ > - $V=\{(1,0),(0,1)\}$, $W=\{(3,0,0),(0,1,0),(0,0,1)\}$ > - $[T(\mathbf{v}_1)]_W=(\frac{1}{3},2,7)$ > - $[T(\mathbf{v}_2)]_W=(1,5,9)$ > - $A=\begin{bmatrix} \frac{1}{3} &1\\ 2&5 \\ 7&9 \end{bmatrix}$