np.dot()和*的区别

from numpy import *

b = array([[1, 3, 4],
           [7, 6, 5],
           [8, 9, 1]])
c = array([[1, 0, -1],
           [2, 0, -2],
           [1, 0, -1]])
print(c*b)

结果是

[[  1   0  -4]
 [ 14   0 -10]
 [  8   0  -1]]

from numpy import *

b = array([[1, 3, 4],
           [7, 6, 5],
           [8, 9, 1]])
c = array([[1, 0, -1]])
print(c*b)

结果是

[[ 1  0 -4]
 [ 7  0 -5]
 [ 8  0 -1]]

from numpy import *

b = array([[1, 3],
           [7, 6],
           [8, 9]])
c = array([[1, 0, -1],
           [2, 0, -2],
           [1, 0, -1]])
print(c*b)

结果出错，总结*相乘列数必须一样
dot()表示两个矩阵相乘

import numpy as np
from numpy import *

b = array([[1, 3, 4],
           [7, 6, 5],
           [8, 9, 1]])
c = array([[1, 0, -1]])
print(np.dot(c, b))

结果是

[[-7 -6  3]]