numpy模块（2）

傷城~ 2022-01-29 02:43 365阅读 0赞

1、利用布尔值来取元素

import numpy as np
    mask = np.array([1,0,1],dtype=bool) # 1表示取对应的元素，0表示不取
    arr = np.array([[1,2,3],[4,5,6],[7,8,9]])
    print(arr[mask,2])
    print(arr[:,mask])

运行结果：

[3 9]
    [[1 3]
     [4 6]
     [7 9]]

一维数组取值

arr = np.array([1,2,3,2,4,5,6,4,6])
    mask = arr>3
    print(arr[mask])
    mask1 = arr==2
    print(arr[mask1])

运行结果：

![20190520192459545.png][]

2、变换数组形态

（1）展平数组 ravel 和 flatten

# 一维数组变为二维数组
    arr = np.arange(1,10,1).reshape(3,3)
    print(arr)
    # 二维数组变为一维数组
    arr1 = np.array([[1,2,3],[4,5,6],[7,8,9]])
    print(arr1)
    arr1.ravel()
    print(arr1)

运行结果：

![20190520193619704.png][]

3、展平数组 flatten

# 展平方式 flatten
    arr1 = np.array([[1,2,3],[4,5,6],[7,8,9]])
    print(arr1)
    print('横向展平：',arr1.flatten())
    print('纵向展平：',arr1.flatten('F'))

运行结果：

![20190520194206837.png][]

4、拼接 hstack 和 vstack

arr1 = np.array([[1,2,3],[4,5,6],[7,8,9]])
    arr2 = np.ones((3,3))
    print(arr2)
    # 横向拼接
    print('横向拼接:\n',np.hstack((arr1,arr2)))
    # 纵向拼接
    print('纵向拼接:\n',np.vstack((arr1,arr2)))

运行结果：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70][]

5、横向变换与纵向变换

# concatenate  横向变换与纵向变换
    print('横向变换:\n',np.concatenate((arr1,arr2),axis=1))
    print('纵向变换:\n',np.concatenate((arr1,arr2),axis=0))

运行结果：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 1][]

6、横向分割与纵向分割 hsplit 与 vsplit

arr = np.diag([1,2,3,4])
    print(arr)
    print('横向分割:\n',np.hsplit(arr,2))
    print('纵向分割:\n',np.vsplit(arr,2))
    
    print('横向分割:\n',np.split(arr,2,axis=1))
    print('纵向分割:\n',np.split(arr,2,axis=0))

运行结果：

E:\Anaconda3\envs\Pachong\python.exe E:/Project/爬虫阶段/day-05-20/test.py
    [[1 0 0 0]
     [0 2 0 0]
     [0 0 3 0]
     [0 0 0 4]]
    横向分割:
     [array([[1, 0],
           [0, 2],
           [0, 0],
           [0, 0]]), array([[0, 0],
           [0, 0],
           [3, 0],
           [0, 4]])]
    纵向分割:
     [array([[1, 0, 0, 0],
           [0, 2, 0, 0]]), array([[0, 0, 3, 0],
           [0, 0, 0, 4]])]
    横向分割:
     [array([[1, 0],
           [0, 2],
           [0, 0],
           [0, 0]]), array([[0, 0],
           [0, 0],
           [3, 0],
           [0, 4]])]
    纵向分割:
     [array([[1, 0, 0, 0],
           [0, 2, 0, 0]]), array([[0, 0, 3, 0],
           [0, 0, 0, 4]])]
    
    Process finished with exit code 0

7、创建矩阵 mat, matrix, bmat

matr1 = np.mat('1 2 3;4 5 6;7 8 9')
    print(matr1)
    matr2 = np.matrix([[1,2,3],[4,5,6],[7,8,9]])
    print(matr2)
    # 合并矩阵bmat
    arr = np.eye(3)
    arr1 = np.diag([1,2,3])
    print('合并之后的大矩阵为:\n',np.bmat('arr arr1;arr1 arr'))

运行结果是：

E:\Anaconda3\envs\Pachong\python.exe E:/Project/爬虫阶段/day-05-20/test.py
    [[1 2 3]
     [4 5 6]
     [7 8 9]]
    [[1 2 3]
     [4 5 6]
     [7 8 9]]
    合并之后的大矩阵为:
     [[1. 0. 0. 1. 0. 0.]
     [0. 1. 0. 0. 2. 0.]
     [0. 0. 1. 0. 0. 3.]
     [1. 0. 0. 1. 0. 0.]
     [0. 2. 0. 0. 1. 0.]
     [0. 0. 3. 0. 0. 1.]]
    
    Process finished with exit code 0

8、矩阵的四则运算

（1）矩阵的转置与求逆

matr1 = np.mat('1 2 3;4 5 6;7 8 9')
    # 专置
    print('转置之后的矩阵为:\n',matr1.T)
    # 求逆
    matr2 = np.mat('1 0;3 2')
    # 如果原有矩阵不可逆，会重新创建一个伪逆矩阵
    matr21 = matr2.I
    print('求逆之后的矩阵为:\n',matr2.I)
    print(matr2*matr21)

运行结果：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 2][]

矩阵的其他算法：加减乘

matr1 = np.mat('1 2 3;4 5 6;7 8 9')
    matr2 = np.eye(3,3)
    print(matr1+matr2)
    print(matr1-matr2)
    print(matr1*matr2)

运行结果：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 3][]

9、ufunc函数 通用函数 ，对数组中所有元素进行操作的函数

x = np.array([1,2,3])
    y = np.array([2,2,4])
    print(x+y)
    print(x-y)
    print(x*y)
    print(x/y)
    
    # 比较运算，返回布尔值
    print(x>y)
    print(x<y)
    print(x>=y)
    print(x<=y)
    print(x==y)
    print(x!=y)
    
    # 逻辑运算
    print(np.all(x==y)) # 与
    print(np.any(x==y)) # 或

运算结果：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 4][]

10、广播机制：

如果两个数组的**后缘维度(从末尾开始算起的维度)**的**轴长度相符**或**其中一方的长度为1**，则认为它们是广播兼容的。广播会在**缺失维度**和(或)**轴长度为1的维度**上进行自动补全。

arr1 = np.array([[1,2,3],[4,5,6],[7,8,9]])
    arr2 = np.array([1,1,1])
    print(arr1+arr2)
    
    arr3 = ([1],[2],[3])
    print(arr1+arr3)

运行结果：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 5][]

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 6][]

[20190520192459545.png]: /images/20220129/a7a99d3c505b4cd78b33b08f7b1195bf.png
[20190520193619704.png]: /images/20220129/d7e215ebf4ff48b5893266fe1737ebf3.png
[20190520194206837.png]: /images/20220129/aca0a22e0da147d299524e4f48a32615.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70]: /images/20220129/cd81d7a334614762886f9ecc5e52a447.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 1]: /images/20220129/539fb1d3b83e4b3d9affbb91560ae381.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 2]: /images/20220129/af81dd3612014113af07f7bead931b7e.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 3]: /images/20220129/23b527ab5f034869962958f95b9983e0.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 4]: /images/20220129/3e210020eaa04031bf857daff5742273.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 5]: /images/20220129/f0c2d2751fc7481aa122d5547b2780f1.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80NDIzOTU0MQ_size_16_color_FFFFFF_t_70 6]: /images/20220129/7c59af61791240198e310afb2df64f17.png