numpy的一些常用方法

# -*- coding: utf-8 -*-
import numpy as np


def demo_1():
    '''
    维度
    :return:
    '''
    a_1 = np.array([1, 2, 3, 4, 5, 6, 7, 8])
    print(a_1.shape)  # 8
    a_2 = np.array([[1], [2], [3], [4], [5], [6], [7], [8]])
    print(a_2.shape)  # 8, 1

    a_3 = np.array([[[1], [2], [3], [4], [5], [6], [7], [8]]])
    print(a_3.shape)  # (1,8,1)

    a_4 = np.array([[[1, 2, 3, 4, 5, 6, 7, 8]]])
    print(a_4.shape)  # (1,1,8)

    b_1 = np.floor(10 * np.random.rand(8, 1))  # 二维数组   由8个1维数组组成 每个一维数组1个元素
    b_2 = np.floor(10 * np.random.rand(8, 2))  # 二维数组   由8个1维数组组成 每个一维数组2个元素
    c_3 = np.floor(10 * np.random.rand(8, 3, 2))  # 三维数组   由8个2维数组组成 每个二维数组由3个一维数组组成 每个一维数组2个元素
    d_4 = np.floor(10 * np.random.rand(5))  # 1维数组  5个元素


def demo_2():
    '''
    切片
    :return:
    '''
    x = np.array([[6, 7, 8, 9, 10],
                  [11, 12, 13, 14, 15],
                  [16, 17, 18, 19, 20],
                  [21, 22, 23, 24, 25],
                  [26, 27, 28, 29, 30],
                  [31, 32, 33, 34, 35]])
    print(x.shape)  # 6,5

    print(x[1])  # 取第二行数据 序号从0开始 结果: [11 12 13 14 15]
    print("\n")

    # 下面四个都是等价的
    print(x[:1])  # 截取[0:1)  结果: [[ 6  7  8  9 10]]
    print(x[0:1])  # 截取[0:1) 结果: [[ 6  7  8  9 10]]
    print(x[0:1, ])
    print(x[0:1, :])
    print("\n")

    # 取某一列数据
    print(x[:, 1])  # [ 7 12 17 22 27 32]

    # 截取 下面两个是等价的
    print(x[:, :1])  # 截取[0:1)列 [[6][11][16][21][26][31]]
    print(x[:, 0:1])
    print(x[:, 1:3])  # 截取[1,3)列
    # 取中间的4行3列的的数据
    print(x[1:5, 1:4])  # 截取[1,5)行 [1,4)列

    # 按索引切片
    # 行 下标 [1 5] 按步长 2切片
    print(x[1:5:2])  # 即第2行第4行
    # 列
    print(x[:, 1:5:2])  # 即第2列第4列


def demo_3():
    '''
    合并
    :return:
    '''
    x = np.array([[6, 7, 8, 9, 10],
                  [11, 12, 13, 14, 15],
                  [16, 17, 18, 19, 20],
                  [21, 22, 23, 24, 25],
                  [26, 27, 28, 29, 30],
                  [31, 32, 33, 34, 35]])
    a = x[0:1]  # 第一行
    b = x[5:6]  # 第六行
    print(np.concatenate((a, b), axis=0))  # 按照0 轴合并 [[ 6  7  8  9 10][31 32 33 34 35]]
    print(np.concatenate((a, b), axis=1))  # 按照1 轴合并 [[ 6  7  8  9 10 31 32 33 34 35]]
    c = x[:, 1:2]  # 第一列
    d = x[:, 3:4]  # 第4列
    print(np.concatenate((c, d), axis=0))
    print(np.concatenate((c, d), axis=1))

    e = x[1:4, 0:4]  # shape 3,4
    f = x[1:4, :]  # shape (3, 5)
    g = x[:, 1:5]  # shape (6, 4)
    # 合并
    print(np.concatenate((e, f), axis=1))
    print(np.concatenate((e, g), axis=0))


def demo_4():
    '''
    排序
    功能寥寥
    :return:
    '''
    x = np.array([[6, 7, 11, 9, 10],
                  [26, 27, 28, 29, 30],
                  [11, 12, 11, 14, 15],
                  [16, 12, 18, 19, 20],
                  [13, 22, 23, 24, 25],
                  [31, 32, 33, 34, 35]])
    x_1 = x.copy()
    x_1.sort(axis=0)  # 将每一列的数据进行从小到大排序
    print(x_1)
    print('\n')
    print(x_1[:: -1])  # 按列翻转 即从小到大 排列
    print('\n')

    x_2 = x.copy()
    x_2.sort(axis=1)  # 将每一行的数据进行从小到大排序
    print(x_2)
    print(x_2[:, :-1])  # 按行翻转 即从小到大 排列


def demo_5():
    '''
    常用计算
    :return:
    '''
    x = np.array([[6, 65, 11, 9, 10],
                  [26, 27, 1, 29, 30],
                  [11, 12, 11, 4, 15],
                  [16, 12, 18, 19, 20],
                  [13, 22, 23, 24, 25],
                  [31, 32, 33, 34, 35]])
    # 每一列中的 最小值
    print(np.amin(x, axis=0))
    # 每一列中的最大值
    print(np.amax(x, axis=0))

    # 每一行中的 最小值
    print(np.amin(x, axis=1))
    # 每一行中的最大值
    print(np.amax(x, axis=1))
    print("\n")

    # 列中极值
    print(np.ptp(x, axis=0))
    # 行中极值
    print(np.ptp(x, axis=1))

    # 均值
    # 列均值
    print(np.mean(x, axis=0))
    # 行均值
    print(np.mean(x, axis=1))

    print(np.amin(x))  # 全局最小值
    print(np.amax(x))  # 全局最大值
    print(np.mean(x))  # 全局平均值

    pass


def demo_6():
    '''
    大于等于15的值置换成30 小于15的置换成0
    :return:
    '''
    x = np.array([[6, 65, 11, 9, 10],
                  [26, 27, 1, 29, 30],
                  [11, 12, 11, 4, 15],
                  [16, 12, 18, 19, 20],
                  [13, 22, 23, 24, 25],
                  [31, 32, 33, 34, 35]])
    print(x)
    print('\n')
    b = np.where(x < 15, 0, x)
    b = np.where(b >= 15, 30, b)
    print(b)
    # 删除3行
    print(np.delete(x, 2, axis=0))
    print('\n')
    # 删除2列
    print(np.delete(x, 1, axis=1))
    print("\n")
    # 删除第三列 第5列
    print(np.delete(x, [2, 4], axis=1))


def demo_7():
    # 变形
    x = np.array([[6, 7, 8, 9, 10], [31, 32, 33, 34, 35]])
    print(x.shape)  # (2,5)
    '''
    reshape(1,-1)转化成1行：
    reshape(2,-1)转换成两行：
    reshape(-1,1)转换成1列：
    reshape(-1,2)转化成两列
    '''
    print(np.reshape(x, newshape=[1, -1]))  # 数据降为1维

    arr = np.arange(12)
    print(arr)
    re_1 = np.reshape(arr, newshape=[2, 6])
    print(re_1)

    # 三维数组 中 有 3个二维数组
    # 每个二维数组中 2个 1维数组
    # 每个一维数组中 2个元素 最后一个中括号有几个元素
    print(np.reshape(arr, newshape=[3, 2, 2]))

    #
    print(np.reshape(re_1, (6, 2)))
    print(np.reshape(re_1, (12, 1)))


if __name__ == '__main__':
    demo_7()