在pycharm中的setting安装numpy，或者在cmd里面通过pip install方法安装均可
# 代码
代码语言：javascript复制from numpy import *
# 构造一个4x4的随机数组
print("数组:n",random.rand(4, 4))
# 调用mat()函数将数组转化为矩阵
randMat = mat(random.rand(4, 4))
# .I操作符实现了矩阵求逆的运算
print("矩阵求逆:n",randMat.I)
# 矩阵乘以逆矩阵
invRandMat = randMat.I
print("矩阵乘以逆矩阵:n",randMat*invRandMat)
# 函数eye(4)创建一个4x4的单位矩阵，下面是算误差
myEye = randMat*invRandMat
print("计算误差:n",myEye-eye(4))# 运行结果
代码语言：javascript复制数组:
[[ 0.46089523 0.94167701 0.76910005 0.96892095]
[ 0.38894192 0.88828275 0.28144475 0.91774255]
[ 0.13997437 0.7435124 0.89709322 0.68924161]
[ 0.41461243 0.19911356 0.64856312 0.49160523]]
矩阵求逆:
[[-0.45014095 1.92871242 0.28217777 -0.76959459]
[ 1.09829851 -0.9829715 -1.78955478 2.70643663]
[ 0.95057662 -1.40675696 0.67345673 -0.70029109]
[-1.11436955 -0.35496085 1.88297962 -0.76459429]]
矩阵乘以逆矩阵:
[[ 1.00000000e 00 7.20736257e-17 -3.04329603e-17 -1.66886258e-17]
[ -2.27212878e-16 1.00000000e 00 2.93001892e-17 -1.63043911e-16]
[ 8.73583235e-18 -3.06881469e-17 1.00000000e 00 -3.50839733e-17]
[ -4.04804848e-17 1.45085950e-16 -1.23671214e-16 1.00000000e 00]]
计算误差:
[[ -1.11022302e-16 7.20736257e-17 -3.04329603e-17 -1.66886258e-17]
[ -2.27212878e-16 2.22044605e-16 2.93001892e-17 -1.63043911e-16]
[ 8.73583235e-18 -3.06881469e-17 -2.22044605e-16 -3.50839733e-17]
[ -4.04804848e-17 1.45085950e-16 -1.23671214e-16 2.22044605e-16]]
