[DOC] Numpy basic usage collection

import numpy as np


print(' 1.Create '.center(66, '='))

a0 = np.array([2, 32, 13, 24, 15, 6, 9, 8, 7, 12])
a1 = np.zeros([3, 4])
a2 = np.ones([4, 5])
a3 = np.arange(50)
a4 = a3.reshape([5, 10])
a5 = np.random.random([4, 5])
a6 = a0.copy()  # Deep copy, (a6=a0)是浅拷贝
print(a4)


print('\n', ' 2.Operation '.center(66, '='))

print(a4[3][5], a4[3, 5])
print(a4[:, 5:-1])

for col in a4.T:  # 迭代 列
    print(col)

b0 = a4 ** 2
b1 = np.tan(a0)
b2 = a2 * a5
b3 = np.dot(a2.T, a5)
print(b2)

# 合并 array
c0 = np.array([1, 2, 3])
c1 = np.array([4, 5, 6])
c2 = np.vstack([c0, c1])  # 垂直合并 Vertical stack
c3 = np.hstack([c0, c1])  # 水平合并 Horizon stack
c4 = np.concatenate([c0, c1], axis=0)  # 合并 综合了上两个
print(c3)

# 分割 array
d0 = np.split(a5, 5, axis=1)
d1 = np.array_split(a5, 5, axis=0)  # 可不等分
d2 = np.vsplit(a5, 4)
d3 = np.hsplit(a5, 5)
print(d1)

# 删除 行、列
d4 = np.delete(a4, -2, axis=1)

print('\n', ' 3.MO '.center(66, '='))

print(np.argmin(a0))  # 返回所在位置的索引
print(np.argmax(a0))

print(np.min(a0))
print(np.max(a0))
print(np.sum(a0))

print(np.mean(a0))
print(np.median(a0))
print(np.std(a0))  # 标准差

print(np.nonzero(a4))  # 所有非零的坐标
print(np.sort(a5))  # 排序
print(np.transpose(a5), '\n', a5.T)  # 转向
print(np.nonzero(a4))  # 所有非零的坐标
print(np.clip(a0, 5, 20))  # v<min:v=min，v>max:v=max
print(a4.flatten())  # 将多维降为一维 a4.flat是iter用于for

 

PS: I didn't write comments because I used some simple ones myself.