如何通俗理解施密特正交化(图文版)
马同学图解数学
编辑于 2024年05月22日 12:13
收录于文集
共4篇

施密特正交化是将一个向量组正交化的重要方法,是求向量空间的规范正交基的一个重要步骤。我们之前发布过数形结合的讲解视频,最近看的同学挺多,现将图文版也发布出来,搭配视频学习可以更好的理解。

如果%5Cboldsymbol%7Bx%7D_1%2C%5Cboldsymbol%7Bx_2%7D%2C%5Ccdots%5Cboldsymbol%7Bx_n%7D 是某向量空间的基,那么可通过下列做法找到该向量空间中的n 个两两正交的向量%5Cboldsymbol%7Bv%7D_1%2C%5Cboldsymbol%7Bv_2%7D%2C%5Ccdots%5Cboldsymbol%7Bv_n%7D :

%5Cboldsymbol%7Bx_1%7D%2C%5Ccdots%2C%5Cboldsymbol%7Bx_n%7D%0A%5Cxrightarrow%7B%5Cquad%5Ctext%7B%E6%96%BD%E5%AF%86%E7%89%B9%E6%AD%A3%E4%BA%A4%E5%8C%96%7D%5Cquad%7D%0A%5Cbegin%7Bcases%7D%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_1%7D%3D%5Cboldsymbol%7Bx_1%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_2%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_3%7D%3D%5Cboldsymbol%7Bx_3%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%7B%5Cboldsymbol%7Bv_2%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%5Cboldsymbol%7Bv_2%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cqquad%5Cqquad%5Cvdots%5C%5C%0A%2C%2C%2C%2C%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_n%7D%3D%5Cboldsymbol%7Bx_n%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_n%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D-%5Ccdots-%5Cfrac%7B%5Cboldsymbol%7Bx_n%7D%5Ccdot%5Cboldsymbol%7Bv_%7Bn-1%7D%7D%7D%7B%5Cboldsymbol%7Bv_%7Bn-1%7D%7D%5Ccdot%5Cboldsymbol%7Bv_%7Bn-1%7D%7D%7D%5Cboldsymbol%7Bv_%7Bn-1%7D%7D%0A%5Cend%7Bcases%7D%0A

该方法称为施密特正交化(Gram–Schmidt process)。

施密特正交化的几何意义是,比如已知%5Cmathbb%7BR%7D%5E3 中的某向量空间(下图中的蓝色平面)的基为%5Cboldsymbol%7Bx%7D_1%2C%5Cboldsymbol%7Bx%7D_2 :

那么通过施密特正交化,可借助%5Cboldsymbol%7Bx%7D_1%2C%5Cboldsymbol%7Bx%7D_2 得到%5Cboldsymbol%7Bv%7D_1%2C%5Cboldsymbol%7Bv%7D_2 , %5Cboldsymbol%7Bv%7D_1%2C%5Cboldsymbol%7Bv%7D_2 就是该向量空间的一个正交基:

%5Cboldsymbol%7Bx_1%7D%2C%5Cboldsymbol%7Bx_2%7D%0A%5Cxrightarrow%7B%5Cquad%5Ctext%7B%E6%96%BD%E5%AF%86%E7%89%B9%E6%AD%A3%E4%BA%A4%E5%8C%96%7D%5Cquad%7D%0A%5Cbegin%7Bcases%7D%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_1%7D%3D%5Cboldsymbol%7Bx_1%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_2%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D%0A%5Cend%7Bcases%7D%0A

下面来解释下施密特正交化是如何推导出来的。

1 二维平面

先来讲解下如何寻找二维向量空间。

先从特殊的二维向量空间%5Cmathbb%7BR%7D%5E2 说起。比如知道%5Cmathbb%7BR%7D%5E2 的一组基,也就是下图中的两个向量:

只要将其中一个向量对另外一个向量进行投影,就可以得到%5Cmathbb%7BR%7D%5E2 的正交基:

下面来进行代数推导,假设基为%5Cboldsymbol%7Bx_1%7D%2C%5Cboldsymbol%7Bx_2%7D :

任选其一作为%5Cboldsymbol%7Bv_1%7D ,比如选%5Cboldsymbol%7Bx_1%7D :

作出%5Cboldsymbol%7Bx_2%7D 在%5Cboldsymbol%7Bv_1%7D 上的投影%5Coverline%7B%5Cboldsymbol%7Bx_2%7D%7D ,其垂线向量%5Cboldsymbol%7Bx_2%7D-%5Coverline%7B%5Cboldsymbol%7Bx_2%7D%7D 就是要求的%5Cboldsymbol%7Bv_2%7D ,即%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_2%7D-%5Coverline%7B%5Cboldsymbol%7Bx_2%7D%7D :

因为%5Cboldsymbol%7Bx_2%7D%E3%80%81%5Cboldsymbol%7Bv_2%7D 和%5Coverline%7B%5Cboldsymbol%7Bx_2%7D%7D 构成三角形,所以根据向量减法的几何意义有

%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_2%7D-%5Coverline%7B%5Cboldsymbol%7Bx_2%7D%7D

又投影%5Coverline%7B%5Cboldsymbol%7Bx_2%7D%7D 和%5Cboldsymbol%7Bv_1%7D 在一条直线上,两者线性相关,所以可假设

%5Coverline%7B%5Cboldsymbol%7Bx_2%7D%7D%3Dk_1%5Cboldsymbol%7Bv_1%7D

因此:

%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_2%7D-%5Coverline%7B%5Cboldsymbol%7Bx_2%7D%7D%3D%5Cboldsymbol%7Bx_2%7D-k_1%5Cboldsymbol%7Bv_1%7D

因为%5Cboldsymbol%7Bv_2%7D 和%5Cboldsymbol%7Bv_1%7D 正交,所以:

%0A%5Cbegin%7Baligned%7D%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%3D0%0A%2C%2C%2C%2C%2C%2C%2C%2C%26%5Cimplies%2C(%5Cboldsymbol%7Bx_2%7D-k_1%5Cboldsymbol%7Bv_1%7D)%5Ccdot%5Cboldsymbol%7Bv_1%7D%3D0%5C%5C%0A%2C%2C%2C%2C%2C%2C%2C%2C%26%5Cimplies%2C%5Cboldsymbol%7Bx_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D-k_1%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%3D0%5C%5C%0A%2C%2C%2C%2C%2C%2C%2C%2C%26%5Cimplies%2Ck_1%3D%5Cfrac%7B%5Cboldsymbol%7Bx_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%0A%5Cend%7Baligned%7D%0A

所以:

%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_2%7D-k_1%5Cboldsymbol%7Bv_1%7D%3D%5Cboldsymbol%7Bx_2%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D

这样就得到了%5Cmathbb%7BR%7D%5E2 的一组正交基%5Cboldsymbol%7Bv%7D_1%2C%5Cboldsymbol%7Bv%7D_2 :

上述方法就是二维空间中的施密特正交化,可以总结如下:

%5Cboldsymbol%7Bx_1%7D%2C%5Cboldsymbol%7Bx_2%7D%0A%5Cxrightarrow%7B%5Cquad%5Ctext%7B%E6%96%BD%E5%AF%86%E7%89%B9%E6%AD%A3%E4%BA%A4%E5%8C%96%7D%5Cquad%7D%0A%5Cbegin%7Bcases%7D%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_1%7D%3D%5Cboldsymbol%7Bx_1%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_2%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D%0A%5Cend%7Bcases%7D%0A

上述推导过程并没有被限制在%5Cmathbb%7BR%7D%5E2 中,所以它也可以完成开头提到的在三维空间中的平面上寻找正交基的任务:

再来看看如何寻找三维向量空间的正交基。

2.1 思路

还是以特殊的三维向量空间%5Cmathbb%7BR%7D%5E3 为例。比如知道%5Cmathbb%7BR%7D%5E3 的一组基,也就是下图中的三个向量:

先按照二维平面的方法,将其中任意两个向量正交化:

然后向这两个正交向量的张成空间作垂线,从而得到三个正交向量,也就是%5Cmathbb%7BR%7D%5E3 的一组正交基:

2.2 代数

下面来进行代数推导,假设基为%5Cboldsymbol%7Bx_1%7D 、%5Cboldsymbol%7Bx_2%7D 和%5Cboldsymbol%7Bx_3%7D :

任选两个向量,按照上一节介绍的方法将其中任意两个向量正交化,得到%5Cboldsymbol%7Bv_1%7D 和%5Cboldsymbol%7Bv_2%7D :

%5Cboldsymbol%7Bx_1%7D%2C%5Cboldsymbol%7Bx_2%7D%0A%5Cxrightarrow%7B%5Cquad%5Ctext%7B%E6%96%BD%E5%AF%86%E7%89%B9%E6%AD%A3%E4%BA%A4%E5%8C%96%7D%5Cquad%7D%0A%5Cbegin%7Bcases%7D%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_1%7D%3D%5Cboldsymbol%7Bx_1%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_2%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D%0A%5Cend%7Bcases%7D%0A

作出%5Cboldsymbol%7Bx_3%7D 在%5Cboldsymbol%7Bv_1%7D%2C%5Cboldsymbol%7Bv_2%7D 张成平面上的投影%5Coverline%7B%5Cboldsymbol%7Bx_3%7D%7D ,连接%5Cboldsymbol%7Bx_3%7D 和%5Coverline%7B%5Cboldsymbol%7Bx_3%7D%7D 就得到要求的垂线向量%5Cboldsymbol%7Bv_3%7D :

因为%5Cboldsymbol%7Bx_3%7D%E3%80%81%5Cboldsymbol%7Bv_3%7D 和%5Coverline%7B%5Cboldsymbol%7Bx_3%7D%7D 构成三角形,所以根据向量减法的几何意义有

%5Cboldsymbol%7Bv_3%7D%3D%5Cboldsymbol%7Bx_3%7D-%5Coverline%7B%5Cboldsymbol%7Bx_3%7D%7D

又投影%5Coverline%7B%5Cboldsymbol%7Bx_3%7D%7D 在%5Cboldsymbol%7Bv_1%7D%2C%5Cboldsymbol%7Bv_2%7D 的张成平面上,所以%5Coverline%7B%5Cboldsymbol%7Bx_3%7D%7D%E6%98%AF%5Cboldsymbol%7Bv_1%7D%2C%5Cboldsymbol%7Bv_2%7D 的线性组合,可假设

%5Coverline%7B%5Cboldsymbol%7Bx_3%7D%7D%3Dk_1%5Cboldsymbol%7Bv_1%7D%2Bk_2%5Cboldsymbol%7Bv_2%7D

因此:

%5Cboldsymbol%7Bv_3%7D%3D%5Cboldsymbol%7Bx_3%7D-%5Coverline%7B%5Cboldsymbol%7Bx_3%7D%7D%3D%5Cboldsymbol%7Bx_3%7D-k_1%5Cboldsymbol%7Bv_1%7D-k_2%5Cboldsymbol%7Bv_2%7D

因为%5Cboldsymbol%7Bv_3%7D 垂直于%5Cboldsymbol%7Bv_1%7D%2C%5Cboldsymbol%7Bv_2%7D 的张成平面,所以%5Cboldsymbol%7Bv_3%7D 必然垂直于%5Cboldsymbol%7Bv_1%7D 和%5Cboldsymbol%7Bv_2%7D ,所以有:

%0A%5Cbegin%7Bcases%7D%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_3%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%3D(%5Cboldsymbol%7Bx_3%7D-k_1%5Cboldsymbol%7Bv_1%7D-k_2%5Cboldsymbol%7Bv_2%7D)%5Ccdot%5Cboldsymbol%7Bv_1%7D%3D0%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_3%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%3D(%5Cboldsymbol%7Bx_3%7D-k_1%5Cboldsymbol%7Bv_1%7D-k_2%5Cboldsymbol%7Bv_2%7D)%5Ccdot%5Cboldsymbol%7Bv_2%7D%3D0%5C%5C%0A%5Cend%7Bcases%7D%0A

注意到%5Cboldsymbol%7Bv_1%7D 和%5Cboldsymbol%7Bv_2%7D 正交,即有%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%3D0 ,根据上面的方程组可以分别推出:

%0A%5Cbegin%7Baligned%7D%0A%2C%2C%2C%2C(%5Cboldsymbol%7Bx_3%7D-k_1%5Cboldsymbol%7Bv_1%7D-k_2%5Cboldsymbol%7Bv_2%7D)%5Ccdot%5Cboldsymbol%7Bv_1%7D%3D0%0A%2C%2C%2C%2C%26%5Cimplies%2C%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D-k_1%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D-k_2%5Cboldsymbol%7Bv_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%3D0%5C%5C%0A%2C%2C%2C%2C%26%5Cimplies%2C%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D-k_1%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%3D0%5C%5C%0A%2C%2C%2C%2C%26%5Cimplies%2Ck_1%3D%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%0A%5Cend%7Baligned%7D%0A

%0A%5Cbegin%7Baligned%7D%0A%2C%2C%2C%2C(%5Cboldsymbol%7Bx_3%7D-k_1%5Cboldsymbol%7Bv_1%7D-k_2%5Cboldsymbol%7Bv_2%7D)%5Ccdot%5Cboldsymbol%7Bv_2%7D%3D0%0A%2C%2C%2C%2C%26%5Cimplies%2C%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D-k_1%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D-k_2%5Cboldsymbol%7Bv_2%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%3D0%5C%5C%0A%2C%2C%2C%2C%26%5Cimplies%2C%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D-k_2%5Cboldsymbol%7Bv_2%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%3D0%5C%5C%0A%2C%2C%2C%2C%26%5Cimplies%2Ck_2%3D%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%7B%5Cboldsymbol%7Bv_2%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%0A%5Cend%7Baligned%7D%0A

所以:

%5Cboldsymbol%7Bv_3%7D%3D%5Cboldsymbol%7Bx_3%7D-k_1%5Cboldsymbol%7Bv_1%7D-k_2%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_3%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%7B%5Cboldsymbol%7Bv_2%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%5Cboldsymbol%7Bv_2%7D

%5Cboldsymbol%7Bv_3%7D%3D%5Cboldsymbol%7Bx_3%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%7B%5Cboldsymbol%7Bv_2%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%5Cboldsymbol%7Bv_2%7D

这样就得到了%5Cmathbb%7BR%7D%5E3 的一组正交基%5Cboldsymbol%7Bv%7D_1%2C%5Cboldsymbol%7Bv_2%7D%2C%5Cboldsymbol%7Bv_3%7D :

2.3 总结

上述方法就是三维空间中的施密特正交化,可以总结如下:

%5Cboldsymbol%7Bx_1%7D%2C%5Cboldsymbol%7Bx_2%7D%2C%5Cboldsymbol%7Bx_3%7D%0A%5Cxrightarrow%7B%5Cquad%5Ctext%7B%E6%96%BD%E5%AF%86%E7%89%B9%E6%AD%A3%E4%BA%A4%E5%8C%96%7D%5Cquad%7D%0A%5Cbegin%7Bcases%7D%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_1%7D%3D%5Cboldsymbol%7Bx_1%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_2%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_3%7D%3D%5Cboldsymbol%7Bx_3%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%7B%5Cboldsymbol%7Bv_2%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%5Cboldsymbol%7Bv_2%7D%0A%5Cend%7Bcases%7D%0A

3 更高维度

更高维度的情况以此类推,从而得到

%5Cboldsymbol%7Bx_1%7D%2C%5Ccdots%2C%5Cboldsymbol%7Bx_n%7D%0A%5Cxrightarrow%7B%5Cquad%5Ctext%7B%E6%96%BD%E5%AF%86%E7%89%B9%E6%AD%A3%E4%BA%A4%E5%8C%96%7D%5Cquad%7D%0A%5Cbegin%7Bcases%7D%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_1%7D%3D%5Cboldsymbol%7Bx_1%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_2%7D%3D%5Cboldsymbol%7Bx_2%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_2%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_3%7D%3D%5Cboldsymbol%7Bx_3%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_3%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%7B%5Cboldsymbol%7Bv_2%7D%5Ccdot%5Cboldsymbol%7Bv_2%7D%7D%5Cboldsymbol%7Bv_2%7D%5C%5C%0A%2C%2C%2C%2C%5Cquad%5C%5C%0A%2C%2C%2C%2C%5Cqquad%5Cqquad%5Cvdots%5C%5C%0A%2C%2C%2C%2C%5C%5C%0A%2C%2C%2C%2C%5Cboldsymbol%7Bv_n%7D%3D%5Cboldsymbol%7Bx_n%7D-%5Cfrac%7B%5Cboldsymbol%7Bx_n%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%7B%5Cboldsymbol%7Bv_1%7D%5Ccdot%5Cboldsymbol%7Bv_1%7D%7D%5Cboldsymbol%7Bv_1%7D-%5Ccdots-%5Cfrac%7B%5Cboldsymbol%7Bx_n%7D%5Ccdot%5Cboldsymbol%7Bv_%7Bn-1%7D%7D%7D%7B%5Cboldsymbol%7Bv_%7Bn-1%7D%7D%5Ccdot%5Cboldsymbol%7Bv_%7Bn-1%7D%7D%7D%5Cboldsymbol%7Bv_%7Bn-1%7D%7D%0A%5Cend%7Bcases%7D%0A

内容选自《马同学图解线性代数》