常用数组操作

将值为1的索引以列表的形式取出

mislabeled_indices = np.asarray(np.where(a == 1))

令<=0的元素为0

 # When z <= 0, you should set dz to 0 as well. 
dZ[Z <= 0] = 0

打乱数组

#第一步：打乱顺序
permutation = list(np.random.permutation(m)) #它会返回一个长度为m的随机数组，且里面的数是0到m-1
shuffled_X = X[:,permutation]   #将每一列的数据按permutation的顺序来重新排列。
shuffled_Y = Y[:,permutation].reshape((1,m))

合并列表和数组

列表合并

a = [[1,2,3],[4,5,6]]
b = [[1,1,1],[2,2,2]]
a+b
>>>[[1, 2, 3], [4, 5, 6], [1, 1, 1], [2, 2, 2]]

a.extend(b)
a
>>>[[1, 2, 3], [4, 5, 6], [1, 1, 1], [2, 2, 2]]

数组合并

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

纵向合并
c = np.vstack((a,b))
c
>>>array([[1, 2, 3],
       [4, 5, 6],
       [1, 1, 1],
       [2, 2, 2]])
       
c = np.r_[a,b]
c
>>>array([[1, 2, 3],
       [4, 5, 6],
       [1, 1, 1],
       [2, 2, 2]])
       
数组横向合并
d = np.hstack((a,b))
d
>>>array([[1, 2, 3, 1, 1, 1],
       [4, 5, 6, 2, 2, 2]])
       
d = np.c_[a,b]
d
>>>array([[1, 2, 3, 1, 1, 1],
       [4, 5, 6, 2, 2, 2]])

a = np.array([[1, 2], [3, 4]])
b = np.array([[5, 6]])              
np.concatenate((a, b), axis=0)
>>>array([[1, 2],
       [3, 4],
       [5, 6]])
np.concatenate((a, b.T), axis=1)
>>>array([[1, 2, 5],
       [3, 4, 6]])

判断有无相等的值

[[1,2],[1,2]]==[1,2]    列表会作为一个整体进行比较
>>>False  

np.asarray([[1,2],[1,2]])==np.asarray([1,1])
>>>array([[ True, False],
       [ True, False]])

np.where(np.asarray([[1,2],[1,2]])==np.asarray([1,1]))
>>>(array([0, 1]), array([0, 0]))

numpy.where() 用法详解
注意：索引得是tuple

创建多维随机数组

np.random.rand(2, 3)
>>>array([[0.1733986 , 0.98218523, 0.00832304],
       [0.73289522, 0.39713299, 0.72641348]])
np.random.random((2,3))
>>>array([[0.27718037, 0.84159993, 0.40267099],
       [0.82867263, 0.79519155, 0.18532267]])
np.random.randint(0,10,size=(2,3))
>>>array([[7, 4, 6],
       [3, 9, 1]])

调整数组列元素次序

import torch as t
a = t.randn((1,5))
a
>>>tensor([[-0.5184, -0.4428, -1.1630, -1.3399,  0.6762]])

a[:, [0, 2, 1, 4, 3]]
>>>tensor([[-0.5184, -1.1630, -0.4428,  0.6762, -1.3399]])

扩展数组

import torch as t
a = t.randn((3,))
b = t.randn((3,3))

a.unsqueeze(0)
>>>tensor([[-1.0501, -1.9088,  1.8951]])
a.unsqueeze(1)
>>>tensor([[-1.0501],
        [-1.9088],
        [ 1.8951]])

a[None, :]
>>>tensor([[-1.0501, -1.9088,  1.8951]])
a[:, None]
>>>tensor([[-1.0501],
        [-1.9088],
        [ 1.8951]])

a.expand_as(b)
>>>tensor([[-1.0501, -1.9088,  1.8951],
        [-1.0501, -1.9088,  1.8951],
        [-1.0501, -1.9088,  1.8951]])

数组挑选元素（不保持原数组维度）

a = np.random.randint(0,10,size=[2,3])
a
>>>array([[1, 6, 9],
       [1, 1, 7]])
       
mask = a > 1
mask
>>>array([[False,  True,  True],
       [False, False,  True]])
       
a[mask]
>>>array([6, 9, 7])  维度已经变了

划分网格操作

 p_n_x, p_n_y = torch.meshgrid(
            torch.arange(-1, 2),
            torch.arange(-2, 3)
 )
print(torch.arange(-1, 2))
print("=="*20)
print(torch.arange(-2, 3))
>>>tensor([-1,  0,  1])
========================================
tensor([-2, -1,  0,  1,  2])

print(p_n_x)
print("=="*20)
print(p_n_y)
>>>tensor([[-1, -1, -1, -1, -1],
        [ 0,  0,  0,  0,  0],
        [ 1,  1,  1,  1,  1]])
========================================
tensor([[-2, -1,  0,  1,  2],
        [-2, -1,  0,  1,  2],
        [-2, -1,  0,  1,  2]])

去除重复元素

打乱数据集顺序

#第一步：打乱顺序
permutation = list(np.random.permutation(m)) #它会返回一个长度为m的随机数组，且里面的数是0到m-1
shuffled_X = X[:,permutation]   #将每一列的数据按permutation的顺序来重新排列。
shuffled_Y = Y[:,permutation].reshape((1,m))

参考文献