pytorch常用小结

 1 import numpy as np
 2 
 3 features = np.array([[1,2,3],[4,5,6],[7,8,9]])
 4 components = [True,True,True]
 5 features = features[:, components]
 6 print(features)
 7 fea = np.array([[1,2,3],[4,5,6],[7,8,9]])
 8 components = [True,False,True]
 9 fea = fea[:, components]
10 print(fea)

结果：

[[1 2 3]

[4 5 6]

[7 8 9]]

[[1 3]

[4 6]

[7 9]]

2.生成均值为0，方差为1的离散正态分布

 1 import torch
 2 
 3 batch_size = 5
 4 embedding_dim = 2
 5 mean = torch.zeros(batch_size, embedding_dim)
 6 print(mean)
 7 std = mean + 1
 8 print(std)
 9 fakez = torch.normal(mean=mean, std=std) # 500*128，均值为0，方差为1的离散正态分布
10 print(fakez)

orch.randn(*sizes, out=None) → Tensor

返回一个张量，包含了从标准正态分布（均值为0，方差为1，即高斯白噪声）中抽取的一组随机数。

torch.rand(*sizes, out=None) → Tensor

返回一个张量，包含了从区间[0, 1)的均匀分布中抽取的一组随机数。

3.torch中的repeat函数：之后的维度，对应维度相乘

1 alpha = torch.rand( 3, 1, 1)

2 print(alpha)

3 alpha = alpha.repeat(1, 2, 3)

4 print(alpha)

结果：

tensor([[[0.0029]],

[[0.0730]],

[[0.7339]]])

tensor([[[0.0029, 0.0029, 0.0029],

[0.0029, 0.0029, 0.0029]],

[[0.0730, 0.0730, 0.0730],

[0.0730, 0.0730, 0.0730]],

[[0.7339, 0.7339, 0.7339],

[0.7339, 0.7339, 0.7339]]])

将1改成2：

alpha = torch.rand( 3, 1, 1)
print(alpha)
alpha = alpha.repeat(2, 2, 3)
print(alpha)

结果：6个两行三列

tensor([[[0.8134]],

[[0.6626]],

[[0.5064]]])

tensor([[[0.8134, 0.8134, 0.8134],

[0.8134, 0.8134, 0.8134]],

[[0.6626, 0.6626, 0.6626],

[0.6626, 0.6626, 0.6626]],

[[0.5064, 0.5064, 0.5064],

[0.5064, 0.5064, 0.5064]],

[[0.8134, 0.8134, 0.8134],

[0.8134, 0.8134, 0.8134]],

[[0.6626, 0.6626, 0.6626],

[0.6626, 0.6626, 0.6626]],

[[0.5064, 0.5064, 0.5064],

[0.5064, 0.5064, 0.5064]]])