参考

https://www.runoob.com/numpy/numpy-tutorial.html
【【莫烦Python】Numpy & Pandas (数据处理教程)】 https://www.bilibili.com/video/BV1Ex411L7oT/?share_source=copy_web&vd_source=9332b8fc5ea8d349a54c3989f6189fd3

创建数组、改变形状、基础运算

import numpy as np

a = np.array([[1, 2, 3], [4, 5, 6]], dtype=np.int32)
# print(a)
# print(a.ndim)
# print(a.shape)
# print(a.size)

# 2
# (2, 3)
# 6

b = np.zeros(6)
print(b)
c = b.reshape(2, 3)
print(c)

d = np.ones((3, 4))
print(d)

f = a.reshape(3, 2)

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

e = np.arange(1, 10, 2)  # [1 3 5 7 9]
print(e)

print(np.max(a))
print(np.sum(a))  # 21
print('axis0:', np.sum(a, axis=0))  # [5 7 9],按列，3列
print('axis1', np.sum(a, axis=1))  # [ 6 15]

print(np.max(a, axis=0))
print('max index:', np.argmax(a))  # 求索引
print(np.argmax(a, axis=0))  # 求索引
print(a.mean())
print(np.median(a))
print(a.cumsum())  # [ 1  3  6 10 15 21],依次累加

print(a.min())

# 矩阵乘法
print(a.dot(f))

print(np.random.random((2, 4)))  # 随机生成0到1的数

t = np.array([[1, 2, 3, 4]])
print(t.min(axis=0))  # [1 2 3 4]

x = a.transpose((1, 0))  # transpose调整数组的维度顺序
print(x)

# [[1 4]
#  [2 5]
#  [3 6]]


y = a.clip(min=4, max=5)  # 截断,超出范围的按某个值算
print(y)

xx = np.arange(24).reshape((3, 4, 2))
print(xx.transpose(2,1,0))

索引、切片、分割

import copy

import numpy as np

# 按索引来访问
a = np.arange(2, 14).reshape(3, 4)

print(a[1])
print(a[1][1])
print(a[1, 1])
print(a[1, ::-1])  # 可以选某个维度进行切片

# [6 7 8 9]
# 7
# 7
# 9

# 展平
f = a.flatten()
print(f)
for it in f:
    print(it)

# 乘号是对应位置逐元素分别相乘
b = np.array([[1, 2, 3], [4, 5, 6]], dtype=np.int32)
c = copy.copy(b)
print(b * c)

# print(np.split(b, 2, axis=1))  # 只能均分，ValueError: array split does not result in an equal division
print(np.split(b, 3, axis=1))