2.6 Computations concerning subspaces

这一节并没有定理内容，完全关于计算。核心是解决以下三个问题：

1. 如何确定${\alpha }_1,\dots ,{\alpha }_m$张成的subspace W的维度？
2. 怎么确定是否$\beta \in span({\alpha }_1,\dots ,{\alpha }_m)$？
3. 如何显式地描述W？也就是W中的向量以最低自由度表示有什么形态？
   第一个问题可以通过化简${\alpha }_1,\dots ,{\alpha }_m$组成的矩阵至row-reduced echelon来解决；第三个问题可以通过代表行变换的可逆矩阵$P$来得到；第二个问题则是将${\alpha }_1,\dots ,{\alpha }_m$作为列向量得到一个方程组，用增广矩阵方法确定方程组是否有解，这个实质是linear combination，比较好理解。
   后面举了两个例子说明此问题。

Exercises

1. Let $s<n$ and $A$ an $s\times n$ matrix with entries in the field $F$. Use Theorem 4 (not its proof) to show that there is a non-zero $X$ in $F^{n\times 1}$ such that $AX=0$.

Solution: Write $A=[{\alpha }_1,\dots ,{\alpha }_n]$, where each ${\alpha }_i\in F^s$, then $AX=0$ is equivalent to ${\sum }_{i=1}^n{x}_i{\alpha }_i$, by Theorem 4, any independent set of vectors in $F^s$ contains no more than $s$ elements, as $s<n$, the set of vectors {α1,…,αn}\{\alpha_1,\dots,\alpha_n \}{α1​,…,αn​} is linearly dependent, thus there must exists $c_1,\dots ,c_n$ not all zero such that ${\sum }_{i=1}^n{c}_i{\alpha }_i=0$, this $(c_1,\dots ,c_n)\in F^{n\times 1}$ is the $X$ we need.

2. Let

$${\alpha }_1=(1,1,-2,1),{\alpha }_2=(3,0,4,-1),{\alpha }_3=(-1,2,5,2).$$

Let

$$\alpha =(4,-5,9,-7),\beta =(3,1,-4,4),\gamma =(-1,1,0,1).$$

( a ) Which of the vectors $\alpha ,\beta ,\gamma$ are in the subspaces of $R^4$ spanned by the ${\alpha }_i$?

( b ) Which of the vectors $\alpha ,\beta ,\gamma$ are in the subspaces of $C^4$ spanned by the ${\alpha }_i$?

( c ) Does this suggest a theorem?

Solution:
( a ) We perform elementary row opeartions on $[{\alpha }_1,{\alpha }_2,{\alpha }_3,\alpha ,\beta ,\gamma ]$.
$$\begin{array}{rl}\left[\begin{array}{cccc}1& 3& -1& y_1\\ 1& 0& 2& y_2\\ -2& 4& 5& y_3\\ 1& -1& 2& y_4\end{array}\right]& \to \left[\begin{array}{cccc}1& 0& 2& y_2\\ 0& 3& -3& y_1-y_2\\ 0& 4& 9& y_3+2y_2\\ 0& -1& 0& y_4-y_2\end{array}\right]\\ & \to \left[\begin{array}{cccc}1& 0& 2& y_2\\ 0& 1& 0& y_2-y_4\\ 0& 0& -3& y_1-y_2+3(y_4-y_2)\\ 0& 0& 9& y_3+2y_2+4(y_4-y_2)\end{array}\right]\\ & \to \left[\begin{array}{cccc}1& 0& 2& y_2\\ 0& 1& 0& y_2-y_4\\ 0& 0& -3& y_1-4y_2+3y_4\\ 0& 0& 0& 3y_1-14y_2+y_3+13y_4\end{array}\right]\end{array}$$
For $\alpha$ we have $12+70+9-91=0$, thus α∈span{α1,α2,α3}\alpha\in span\{\alpha_1,\alpha_2,\alpha_3 \}α∈span{α1​,α2​,α3​}, since $9-14-4+42\ne 0$ and $-3-14+13\ne 0$, both $\beta$ and $\gamma$ is not in the subspace.
( b ) The answer is the same with (a)
( c ) If $F^{\prime }$ is a subspace of $F$, then $U$ is a subspace of $F^{\prime }$ if and only if $U$ is a subspace of $F$.

3. Consider the vectors in $R^4$ defined by

$${\alpha }_1=(-1,0,1,2),{\alpha }_2=(3,4,-2,5),{\alpha }_3=(1,4,0,9).$$

Find a system of homogeneous linear equations for which the space of solutions is exactly the subspace of $R^4$ spanned by the three given vectors.

Solution: First verify if ${\alpha }_1,{\alpha }_2,{\alpha }_3$ is linearly independent:
$$\left[\begin{array}{cccc}-1& 0& 1& 2\\ 3& 4& -2& 5\\ 1& 4& 0& 9\end{array}\right]\to \left[\begin{array}{cccc}1& 0& -1& -2\\ 0& 4& 1& 11\\ 0& 4& 1& 11\end{array}\right]\to \left[\begin{array}{cccc}1& 0& -1& -2\\ 0& 4& 1& 11\\ 0& 0& 0& 0\end{array}\right]$$
thus ${\alpha }_1,{\alpha }_3$ can be a basis for $span({\alpha }_1,{\alpha }_2,{\alpha }_3)$, and one system may be
$$\begin{gathered} 4x_1-x_2+4x_3=0 \\ 8x_1-11x_2+4x_4=0 \end{gathered}$$

4. In $C^3$, let

$${\alpha }_1=(1,0,i),{\alpha }_2=(1+i,1-i,1),{\alpha }_3=(i,i,i).$$

Prove that these vectors form a basis for $C^3$. What are the coordinates of the vectors $(a,b,c)$ in this basis?

Solution: Let $c_1{\alpha }_1+c_2{\alpha }_2+c_3{\alpha }_3=0$, then
$$\begin{gathered} c_1+c_2(1+i)+c_{3}i=0⟹c_1+c_2=0,c_2+c_3=0 \\ {c}_2(1-i)+c_{3}i=0⟹c_2=0⟹c_1=c_3=0 \end{gathered}$$
thus the three vectors are linearly independent.
To show ${\alpha }_1,{\alpha }_2,{\alpha }_3$ may not be a basis for $C^3$, let $x_1{\alpha }_1+x_2{\alpha }_2+x_3{\alpha }_3=(a,b,c)$, then
$$\begin{gathered} \left[\begin{array}{cccc}1& 1+i& i& a\\ 0& 1-i& i& b\\ -i& 1& i& c\end{array}\right]\to \left[\begin{array}{cccc}1& 1+i& i& a\\ 0& 1-i& i& b\\ 0& i& i-1& c+ai\end{array}\right] \\ \to \left[\begin{array}{cccc}1& 2i& 0& a-b\\ 0& 1& 2i-1& b+c+ai\\ 0& 0& 1+2i& c-a+(a-b-c)i\end{array}\right] \end{gathered}$$
If $(a,b,c)=(1+2i,2i,0)$, then from the first two equations we have $x_1=1,x_2=0$, and $x_3=2$,
thus the last equation means $-i+2i=c$, or $i=0$, a contradiction.

5. Given an explicit description of the type (2-25) for the vectors $\beta =(b_1,b_2,b_3,b_4,b_5)$ in $R^5$ which are linear combinations of the vectors

$$\begin{gathered} {\alpha }_1=(1,0,2,1,-1),{\alpha }_2=(-1,2,-4,2,0), \\ {\alpha }_3=(2,-1,5,2,1){\alpha }_4=(2,1,3,5,2). \end{gathered}$$
Solution: We have
$$\begin{array}{rl}\left[\begin{array}{ccccc}1& -1& 2& 2& b_1\\ 0& 2& -1& 1& b_2\\ 2& -4& 5& 3& b_3\\ 1& 2& 2& 5& b_4\\ -1& 0& 1& 2& b_5\end{array}\right]& \to \left[\begin{array}{ccccc}1& -1& 2& 2& b_1\\ 0& 2& -1& 1& b_2\\ 0& -2& 1& -1& b_3-2b_1\\ 0& 3& 0& 3& b_4-b_3\\ 0& -1& 3& 4& b_1+b_5\end{array}\right]\\ & \to \left[\begin{array}{ccccc}1& -1& 2& 2& b_1\\ 0& -1& 3& 4& b_1+b_5\\ 0& 3& 0& 3& b_4-b_3\\ 0& 2& -1& 1& b_2\\ 0& 0& 0& 0& b_3+b_2-2b_1\end{array}\right]\\ & \to \left[\begin{array}{ccccc}1& -1& 2& 2& b_1\\ 0& -1& 3& 4& b_5+b_1\\ 0& 0& 5& 9& 2b_1+b_2+2b_5\\ 0& 0& 9& 15& 3b_1-b_3+b_4+3b_5\\ 0& 0& 0& 0& b_3+b_2-2b_1\end{array}\right]\end{array}$$
thus those vectors which are linear combinations of ${\alpha }_1,{\alpha }_2,{\alpha }_3,{\alpha }_4$ are those $\beta$ which satisfy $b_3+b_2=2b_1$.

6. Let $V$ be the real vector space spaned by the rows of the matrix

$$A=\left[\begin{array}{ccccc}3& 21& 0& 9& 0\\ 1& 7& -1& -2& -1\\ 2& 14& 0& 6& 1\\ 6& 42& -1& 13& 0\end{array}\right]$$

( a ) Find a basis for $V$.

( b ) Tell which vectors $(x_1,x_2,x_3,x_4,x_5)$ are elements of $V$.

( c ) If $(x_1,x_2,x_3,x_4,x_5)$ is in $V$ what are its coordinates in the basis chosen in part (a)?

Solution:
( a ) We have
$$\left[\begin{array}{ccccc}3& 21& 0& 9& 0\\ 1& 7& -1& -2& -1\\ 2& 14& 0& 6& 1\\ 6& 42& -1& 13& 0\end{array}\right]\to \left[\begin{array}{ccccc}1& 7& 0& 3& 0\\ 0& 0& -1& -5& -1\\ 0& 0& 0& 0& 1\\ 0& 0& -1& -5& 0\end{array}\right]\to \left[\begin{array}{ccccc}1& 7& 0& 3& 0\\ 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 1\\ 0& 0& 1& 5& 0\end{array}\right]$$
So a basis for $V$ is $(1,7,0,3,0),(0,0,0,0,1),(0,0,1,5,0)$.
( b ) Those $(x_1,x_2,x_3,x_4,x_5)$ which satisfy $x_2=7x_1$ and $x_4=3x_1+5x_3$.
( c ) The coordinates are $(x_1,x_5,x_3)$.

7. Let $A$ be an $m\times n$ matrix over the field $F$, and consider the system of equations $AX=Y$. Prove that this system of equations has a solution if and only if the row rank of $A$ is equal to the row rank of the augmented matrix of the system.

Solution: Perform elementary row operations on $A$ to get a row-reduced echelon matrix R, the same operations will lead $A^{\prime }=[A,Y]$ to a matrix $[R,Y^{\prime }]$, the system has a solution if and only if the nonzero rows of $[R,Y^{\prime }]$ is the same as the nonzero rows of $R$, which means their row ranks are the same. The conclusion follows since row rank of $A$ equals row rank of $R$, and row rank of $A^{\prime }$ equals row rank of $[R,Y^{\prime }]$, by Theorem 9.