Python（16）：Numpy之array数组的数值计算

最新推荐文章于 2025-07-02 22:23:58 发布

原创最新推荐文章于 2025-07-02 22:23:58 发布 · 2.2k 阅读

6 ·

CC 4.0 BY-SA版权

文章标签：

#Python

Python 专栏收录该内容

18 篇文章

订阅专栏

0. 相关文章链接

Python文章汇总

1. 创建Array数组

import numpy as np
demo_array = np.array([[1,2,3],[4,5,6]])
demo_array

array([[1, 2, 3],
       [4, 5, 6]])

2. 基本数值计算

2.1. numpy中的函数

# 使用 numpy 中的函数进行数值计算
print(demo_array)
print("==================")
print(np.sum(demo_array))
print(np.max(demo_array))
print(np.min(demo_array))

[[1 2 3]
 [4 5 6]]
==================
21
6
1

2.2. 数组中的函数

# 直接使用数组的函数进行数值计算
print(demo_array)
print("==================")
print(demo_array.sum())
print(demo_array.max())
print(demo_array.min())

[[1 2 3]
 [4 5 6]]
==================
21
6
1

3. 指定维度进行计算

3.1. numpy中的函数

# 指定要进行的操作是沿着什么轴（维度）
# 会根据 shape 中的维度来统计，以二维数组举例，元组的第一位是有多少行（横轴），第二位是有多少列（竖轴）
# 如果指定根据横轴（axis=0）来计算，那计算方式为将同一个竖轴的元素进行聚合（即横轴上有多少个元素，那返回的数组就有多少元素）
# 如果指定根据竖轴（axis=1）来计算，那计算方式为将同一个横轴的元素进行聚合（即竖轴上有多少个元素，那返回的数组就有多少元素）
# 如果axis=-1，即取shape返回的元组中的最后那个
print(demo_array)
print("==================")
print(np.ndim(demo_array))
print(np.shape(demo_array))
print("==================")
print(np.sum(demo_array,axis=0))
print(np.max(demo_array,axis=0))
print(np.min(demo_array,axis=0))
print("==================")
print(np.sum(demo_array,axis=1))
print(np.max(demo_array,axis=1))
print(np.min(demo_array,axis=1))
print("==================")
print(np.sum(demo_array,axis=-1))
print(np.max(demo_array,axis=-1))
print(np.min(demo_array,axis=-1))

[[1 2 3]
 [4 5 6]]
==================
2
(2, 3)
==================
[5 7 9]
[4 5 6]
[1 2 3]
==================
[ 6 15]
[3 6]
[1 4]
==================
[ 6 15]
[3 6]
[1 4]

3.2. 数组中的函数

# 跟上述类似，只是使用的是数组中的数值计算方法
print(demo_array)
print("==================")
print(demo_array.ndim)
print(demo_array.shape)
print("==================")
print(demo_array.sum(axis = 0))
print(demo_array.max(axis = 0))
print(demo_array.min(axis = 0))
print("==================")
print(demo_array.sum(axis=1))
print(demo_array.max(axis=1))
print(demo_array.min(axis=1))
print("==================")
print(demo_array.sum(axis=-1))
print(demo_array.max(axis=-1))
print(demo_array.min(axis=-1))

[[1 2 3]
 [4 5 6]]
==================
2
(2, 3)
==================
[5 7 9]
[4 5 6]
[1 2 3]
==================
[ 6 15]
[3 6]
[1 4]
==================
[ 6 15]
[3 6]
[1 4]

4. 复杂计算

4.1. 统计乘机

# 求选中范围内数值的乘积
print(demo_array)
print("==================")
print(demo_array.ndim)
print(demo_array.shape)
print("==================")
print(demo_array.prod())
print(demo_array.prod(axis = 0))
print(demo_array.prod(axis = 1))

[[1 2 3]
 [4 5 6]]
==================
2
(2, 3)
==================
720
[ 4 10 18]
[  6 120]

4.2. 获取对应值的索引位置

# 求出对应值的索引位置（多维数值的索引是所有元素放入一起进行计算的）
print(demo_array)
print("==================")
print(demo_array.ndim)
print(demo_array.shape)
print("==================")
print(demo_array.argmin())
print(demo_array.argmin(axis = 0))
print(demo_array.argmin(axis = 1))
print("==================")
print(demo_array.argmax())
print(demo_array.argmax(axis = 0))
print(demo_array.argmax(axis = 1))

[[1 2 3]
 [4 5 6]]
==================
2
(2, 3)
==================
0
[0 0 0]
[0 0]
==================
5
[1 1 1]
[2 2]

4.3. 求平均值

# 求平均值
print(demo_array)
print("==================")
print(demo_array.ndim)
print(demo_array.shape)
print("==================")
print(demo_array.sum())
print(demo_array.size)
print(demo_array.mean())
print("==================")
print(demo_array.sum(axis = 0))
print(np.size(demo_array, axis = 0))
print(demo_array.mean(axis = 0))

[[1 2 3]
 [4 5 6]]
==================
2
(2, 3)
==================
21
6
3.5
==================
[5 7 9]
2
[2.5 3.5 4.5]

4.4. 求标准差

# 求标准差
print(demo_array)
print("==================")
print(demo_array.ndim)
print(demo_array.shape)
print("==================")
print(demo_array.std())
print(demo_array.std(axis = 0))
print(demo_array.std(axis = 1))
print(demo_array.std(axis = -1))

[[1 2 3]
 [4 5 6]]
==================
2
(2, 3)
==================
1.707825127659933
[1.5 1.5 1.5]
[0.81649658 0.81649658]
[0.81649658 0.81649658]

4.5. 求方差

# 求方差
print(demo_array)
print("==================")
print(demo_array.ndim)
print(demo_array.shape)
print("==================")
print(demo_array.var())
print(demo_array.var(axis = 0))
print(demo_array.var(axis = 1))

[[1 2 3]
 [4 5 6]]
==================
2
(2, 3)
==================
2.9166666666666665
[2.25 2.25 2.25]
[0.66666667 0.66666667]

4.6. 取界限值

# 界限值，将该数组中超出该界限值的元素强制赋值成最小和最大值
demo_array.clip(2,5)

array([[2, 2, 3],
       [4, 5, 5]])

4.7. 取整

demo_array = np.array([1.2,3.56,6.41])
print(demo_array)
# 四舍五入取整，不保留小数
print(demo_array.round())
# 四舍五入，保留一位小数
print(demo_array.round(decimals=1))

[1.2  3.56 6.41]
[1. 4. 6.]
[1.2 3.6 6.4]

注：其他Python相关系列文章链接由此进 -> Python文章汇总