04.数组与矩阵运算

生成数组

>>> import numpy as np
>>> np.random.randn(10)
array([ 0.52712347, -1.65888503, -1.00390235,  1.01367036, -0.15752943,
       -2.2986508 , -0.00966312, -0.70276299,  1.03832744, -0.56927384])
>>> np.random.randint(10,size=20).reshape(2,10)
array([[1, 9, 3, 4, 2, 8, 7, 2, 0, 4],
       [5, 0, 5, 7, 5, 3, 2, 6, 8, 3]])
>>> a = np.random.randint(10,size=20).reshape(4,5)
>>> b = np.random.randint(10,size=20).reshape(4,5)
>>> a,b
(array([[4, 3, 0, 6, 0],
       [1, 5, 1, 3, 3],
       [0, 9, 8, 4, 9],
       [4, 9, 5, 5, 2]]), 
 array([[6, 5, 6, 8, 4],
       [0, 7, 4, 9, 9],
       [9, 8, 9, 1, 7],
       [1, 3, 7, 3, 7]]))

数组加减乘除

>>> a+b
array([[10,  8,  6, 14,  4],
       [ 1, 12,  5, 12, 12],
       [ 9, 17, 17,  5, 16],
       [ 5, 12, 12,  8,  9]])
>>> a*b
array([[24, 15,  0, 48,  0],
       [ 0, 35,  4, 27, 27],
       [ 0, 72, 72,  4, 63],
       [ 4, 27, 35, 15, 14]])
>>> a/b
<stdin>:1: RuntimeWarning: divide by zero encountered in true_divide
array([[0.66666667, 0.6       , 0.        , 0.75      , 0.        ],
       [       inf, 0.71428571, 0.25      , 0.33333333, 0.33333333],
       [0.        , 1.125     , 0.88888889, 4.        , 1.28571429],
       [4.        , 3.        , 0.71428571, 1.66666667, 0.28571429]])

矩阵预运算

>>> np.mat([[1,2,3],[4,5,6]])
matrix([[1, 2, 3],
        [4, 5, 6]])
>>> A = np.mat(a)
>>> B = np.mat(b)
>>> A-B
matrix([[-2, -2, -6, -2, -4],
        [ 1, -2, -3, -6, -6],
        [-9,  1, -1,  3,  2],
        [ 3,  6, -2,  2, -5]])
>>> A*B
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "C:UsersMr_waanaconda3libsite-packages
umpymatrixlibdefmatrix.py", line 220, in __mul__
    return N.dot(self, asmatrix(other))
  File "<__array_function__ internals>", line 5, in dot
ValueError: shapes (4,5) and (4,5) not aligned: 5 (dim 1) != 4 (dim 0)
注：矩阵运算法则要求前者的列数=后者的行数

>>> b = np.random.randint(10,size=20).reshape(5,4)
>>> B = np.mat(b)

>>> B
matrix([[5, 2, 8, 5],
[5, 5, 6, 9],
[6, 9, 2, 4],
[8, 6, 2, 2],
[0, 7, 9, 5]])
>>> A*B
matrix([[ 83, 59, 62, 59],
[ 60, 75, 73, 75],
[125, 204, 159, 166],
[135, 142, 124, 141]])

其他

>>> np.unique(a)
array([0, 1, 2, 3, 4, 5, 6, 8, 9])
>>> sum(a)
array([ 9, 26, 14, 18, 14])
>>> sum(a[0])
13
>>> sum(a[:,0]
... )
9
>>> a.max()
9
>>> max(a[0])
6