矩阵分析笔记(二)极大线性无关组

2020-09-17 16:58:17 浏览数 (1)

两个向量组之间的线性表示关系

Vmathbb{F}上的线性空间,alpha_1,alpha_2,...,alpha_pbeta_1,beta_2,...,beta_qV中的两个向量组,betain V

  1. 如果存在p个数k_iin mathbb{F},i=1,2,...,p,使得alpha_1k_1 alpha_2k_2 ··· alpha_pk_p=beta,称向量beta可由向量组alpha_1,alpha_2,...,alpha_p线性表示
  2. 如果每个beta_j都可以由向量组alpha_1,alpha_2,...,alpha_p线性表示,j=1,2,...,q。为了方便,beta_1,beta_2,...,beta_q可由alpha_1,alpha_2,...,alpha_p线性表示,用符号记为{beta_1,beta_2,...,beta_q}≤_{lin}{alpha_1,alpha_2,...,alpha_p}
线性表示关系的传递性

Vmathbb{F}上的线性空间,alpha_1,alpha_2,...,alpha_pbeta_1,beta_2,...,beta_qgamma_1,gamma_2,...,gamma_tV中三个向量组。若

$$ {alpha_1,alpha_2,...,alpha_p}≤_{lin}{beta_1,beta_2,...,beta_q}\ {beta_1,beta_2,...,beta_q}≤_{lin}{gamma_1,gamma_2,...,gamma_t} $$

{alpha_1,alpha_2,...,alpha_p}≤_{lin}{gamma_1,gamma_2,...,gamma_t}

证明:我们利用线性表示关系的矩阵表达。由存在Tin mathbb{F}^{qtimes p},Sin mathbb{F}^{ttimes q}满足

$$ [alpha_1,alpha_2,...,alpha_p]=[beta_1,beta_2,...,beta_q]T\ [beta_1, beta_2,...,beta_q]=[gamma_1,gamma_2,...,gamma_t]S $$

[alpha_1,alpha_2,...,alpha_p]=[gamma_1,gamma_2,...,gamma_t](ST)。由于STin mathbb{F}^{ttimes p}得证

扁(列>行)的齐次线性方程组必有非零解

Ain mathbb{F}^{mtimes n},1≤m<n,则齐次线性方程组Ax=0必有非零解

x=begin{bmatrix} c_1 \ c_2 \ vdots \ c_nend{bmatrix}in mathbb{F}^n,begin{bmatrix} c_1 \ c_2 \ vdots \ c_nend{bmatrix}neq0

这里,m<n

证明:对m用数学归纳法

  1. m=1时,n≥2。根据a_{11}是否为0分两种情况若a_{11}=0,则取begin{bmatrix} c_1 \ c_2 \ vdots \ c_nend{bmatrix}=begin{bmatrix}1\ 0\ vdots \0end{bmatrix},易知其为一非零解若a_{11}neq0,由n≥2,可取begin{bmatrix} c_1 \ c_2 \ vdots \ c_nend{bmatrix}=begin{bmatrix}-frac{a_{12}}{a_{11}}\ 0\ vdots \0end{bmatrix},易知这样取得c为一非零解
  2. 假设m≤p时命题成立,下面证明m=p 1的情形,此时n≥p 2,同样根据x_1的系数,也就是系数矩阵A的第一列alpha_1=begin{bmatrix}a_{11}\a_{21}\ vdots \a_{p 1,1}end{bmatrix}是否为0分两种情况若alpha_1=0,则取c=begin{bmatrix}c_1\ c_2 \ vdots \ c_nend{bmatrix}=begin{bmatrix}1\ 0\ vdots \ 0end{bmatrix}alpha_1neq0,不妨设alpha_{11}neq0,则有与Ax=0同解的线性方程组begin{bmatrix}a_{11} &a_{12}&cdots &a_{1n} \ 0 &0 & cdots &0 \ vdots & vdots & vdots & vdots \ 0 & a_{p 1,2}-frac{a_{p 1,1}}{a_{11}}·a_{12} & cdots & a_{p 1, n}-frac{a_{p 1,1}}{a_{11}}·a_{1n}end{bmatrix}begin{bmatrix}x_1\ x_2 \ vdots \ x_nend{bmatrix}=0p个方程,n-1(≥p 1>p)begin{bmatrix}a_{22}-frac{a_{21}}{a_{11}}·a_{12} & cdots & a_{2n}-frac{a_{21}}{a_{11}}·a_{1n} \ vdots & vdots & vdots \ a_{p 1,2}-frac{a_{p 1,1}}{a_{11}}·a_{12} & cdots & a_{p 1,n}-frac{a_{p 1,1}}{a_{11}}·a_{1n}end{bmatrix}begin{bmatrix}x_2 \ vdots \ x_nend{bmatrix}=0begin{bmatrix}x_2\ vdots \ x_nend{bmatrix}=begin{bmatrix}c_2\ vdots \ c_nend{bmatrix}neq 0,再取x_1=c_1=-frac{1}{a_{11}}(a_{12}c_2 ··· a_{1n}c_n),易知begin{bmatrix}x_1 \ x_2\ vdots \ x_nend{bmatrix}=begin{bmatrix}c_1 \ c_2\ vdots \ c_nend{bmatrix}neq 0是方程组Ax=0的非零解。证毕
线性表示与线性无关性

Vmathbb{F}上的线性空间,alpha_1,alpha_2,...,alpha_pbeta_1, beta_2,...,beta_qV中两个向量组。若alpha_1,alpha_2,...,alpha_p线性无关,且{alpha_1,alpha_2,...,alpha_p}≤_{lin}{beta_1,beta_2,...,beta_q},则p≤q

证明:用反证法,假设p<qTin mathbb{F}^{qtimes p},使得

begin{bmatrix}alpha_1 &alpha_2 & cdots & alpha_pend{bmatrix}=begin{bmatrix}beta_1 &beta_2 & cdots & beta_qend{bmatrix}T

因为扁的齐次方程组必有非零解,所以存在cin mathbb{F}^p非零,使得Tc=0。上述等式两边右乘c

begin{bmatrix}alpha_1 &alpha_2 & cdots & alpha_pend{bmatrix}c=begin{bmatrix}beta_1 &beta_2 & cdots & beta_qend{bmatrix}Tc=0

因为cin mathbb{F}^p非零,此结论与alpha_1,alpha_2,...,alpha_p线性无关矛盾,证毕

以少表多,多必相关

alpha_1, alpha_2,...,alpha_rbeta_1, beta_2,...,beta_s是线性空间V中的两个向量组。若:

  1. alpha_1, alpha_2,...,alpha_r可由beta_1, beta_2,...,beta_s线性表示
  2. r>s

alpha_1, alpha_2,...,alpha_r线性相关

证明:因为{alpha_1,alpha_2,...,alpha_r}≤_{lin}{beta_1,beta_2,...,beta_s},所以exists Ain mathbb{F}^{stimes r},使得

begin{bmatrix}alpha_1 & alpha_2 & cdots & alpha_rend{bmatrix} = begin{bmatrix}beta_1 & beta_2 & cdots & beta_send{bmatrix}A

于是有r(begin{bmatrix}alpha_1 & alpha_2 & cdots & alpha_rend{bmatrix})≤min{r(begin{bmatrix}beta_1 & beta_2 & cdots & beta_send{bmatrix}), r(A)}≤s<r

r(begin{bmatrix}alpha_1 & alpha_2 & cdots & alpha_rend{bmatrix})<ralpha_1, alpha_2,...,alpha_r线性相关

推论:若{alpha_1,alpha_2,...,alpha_r}≤_{lin}{beta_1,beta_2,...,beta_s},且alpha_1, alpha_2,...,alpha_r线性无关,则r≤s

0 人点赞