Numpy的使用

从原有数据装换为数组

a_list = list(range(20))
b = np.array(a_list)
print(b)

[ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19]

生成数组

a = np.zeros(20,dtype=int)
print(type(a))
print(a.dtype)

<class 'numpy.ndarray'>
int32

生成多维数组

a = np.zeros([5,4],dtype=float)
print(type(a))
print(a)

<class 'numpy.ndarray'>
[[0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]]

生产带有初始化值的多维数组

a = np.full([3,3],9.9)
print(a)

[[9.9 9.9 9.9]
 [9.9 9.9 9.9]
 [9.9 9.9 9.9]]

仿照生成数组

b = np.zeros_like(a)
print(b)

b = np.ones_like(a)
print(b)

b = np.full_like(a,3.3)
print(b)

[[9.9 9.9 9.9]
 [9.9 9.9 9.9]
 [9.9 9.9 9.9]]
[[0. 0. 0.]
 [0. 0. 0.]
 [0. 0. 0.]]
[[1. 1. 1.]
 [1. 1. 1.]
 [1. 1. 1.]]
[[3.3 3.3 3.3]
 [3.3 3.3 3.3]
 [3.3 3.3 3.3]]

生成随机数组

import random
rad = random.randint(1,50)
print(rad)

nprad = np.random.random((5,5))
print(nprad)

[[0.88665299 0.15469618 0.92227034 0.5504737  0.04441958]
 [0.19598922 0.19546181 0.93763495 0.5003918  0.06118416]
 [0.25083197 0.63792372 0.25348255 0.54956645 0.80567947]
 [0.07546286 0.16186728 0.04644008 0.09530208 0.72491862]
 [0.46018823 0.14911618 0.41774599 0.68126355 0.22812571]]

生成一个多维数组的随机数

nprad = np.random.randint(0,10 ,(5,5))
print(nprad)

[[1 0 3 9 3]
 [7 7 9 8 0]
 [6 7 8 9 3]
 [6 9 9 5 9]
 [2 2 9 2 4]]

范围取值

l = list(range(0,10))
print(l)

npl = np.arange(0,20,3)
print(npl)

npr = np.linspace(0,3,10)
print(npr)

[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
[ 0  3  6  9 12 15 18]
[0.         0.33333333 0.66666667 1.         1.33333333 1.66666667
 2.         2.33333333 2.66666667 3.        ]

生成单位矩阵(5*5维单位矩阵)

e = np.eye(5)
print(e)

[[1. 0. 0. 0. 0.]
 [0. 1. 0. 0. 0.]
 [0. 0. 1. 0. 0.]
 [0. 0. 0. 1. 0.]

访问数组中的元素

嵌套列表中的数组访问

var = [[1,2,3],[4,5,6],[7,8,9]]
print(var)
print(var[0][0])

[[1, 2, 3], [4, 5, 6], [7, 8, 9]]
1

访问数组中的元素

var1 = np.full((4,4),8.9,dtype=float)
print(var1)
print(var1[0][0])
print(var1[0,0])

[[8.9 8.9 8.9 8.9]
 [8.9 8.9 8.9 8.9]
 [8.9 8.9 8.9 8.9]
 [8.9 8.9 8.9 8.9]]
8.9
8.9

数组切片访问

var2 = np.array(var)
print(var2)
print(var2[:2,:2])

[[1 2 3]
 [4 5 6]
 [7 8 9]]
[[1 2]
 [4 5]]

数组属性

#数组维度
print(var2.ndim)
#shape
print(var2.shape)
#大小
print(var2.size)

#数组类型
print(var2.dtype)
#每个元素所占的字节数
print(var2.itemsize)

2
(3, 3)
9
int32
4

数组运算 + - * /

var = np.array(list(range(10)))
print(var)

var1 = var +10
print(var1)

var2 = var -10
print(var2)

var = np.full((3,3),2.3,dtype=float)
print( var * 10)

[0 1 2 3 4 5 6 7 8 9]
[10 11 12 13 14 15 16 17 18 19]
[-10  -9  -8  -7  -6  -5  -4  -3  -2  -1]
[[23. 23. 23.]
 [23. 23. 23.]
 [23. 23. 23.]]

统计类型运算

#求和

数组一维求和

var = np.full(10,3.6)
s = np.sum(var)
print(s)

36.0

数组多维求和

var = np.array([[1,2,3],[4,5,6],[7,8,9]])
s = np.sum(var)#效率很高，推荐使用
s = sum(var)
print(s)

[12 15 18]

求最大者

m = np.max(var)
print(m)

9

比较

var = np.array(range(10))
print(var)
print(var>3)
print(var!=3)

[0 1 2 3 4 5 6 7 8 9]
[False False False False  True  True  True  True  True  True]
[ True  True  True False  True  True  True  True  True  True]

数组中所有的元素大于-1

print(np.all(var >-1))

True

生产一个3*10的数组

var = np.full((2,10),1,dtype=float)
print(var)

[[1. 1. 1. 1. 1. 1. 1. 1. 1. 1.]
 [1. 1. 1. 1. 1. 1. 1. 1. 1. 1.]]

变形

print(var.reshape(4,5))

[[1. 1. 1. 1. 1.]
 [1. 1. 1. 1. 1.]
 [1. 1. 1. 1. 1.]
 [1. 1. 1. 1. 1.]]
 

aaa