python label 与 one-hot 之间的互相转换
有时候需要label,比如强化学习的离散动作空间,输出动作索引;有时候需要one-hot,比如训练数据或者输入上一个状态的动作,简单的互相转换还是重要的。通过快速生成:one-hot 转labelnumpy可以通过实现,pytorch 可以通过实现:...
·
前言
有时候需要label,比如强化学习的离散动作空间,输出动作索引;有时候需要one-hot,比如训练数据或者输入上一个状态的动作,简单的互相转换还是重要的。
label 转 one-hot
通过 np.eye(action_dims)[actions]
快速生成:
>>> import numpy as np
>>> label = [1,2,2,3]
>>> np.eye(4)[label]
array([[0., 1., 0., 0.],
[0., 0., 1., 0.],
[0., 0., 1., 0.],
[0., 0., 0., 1.]])
one-hot 转label
numpy可以通过 np.argmax(onehot, 1)
实现,pytorch 可以通过 torch.topk(one_hot, 1)[1].squeeze(1)
实现:
>>> import torch
>>> onehot
array([[0., 1., 0., 0.],
[0., 0., 1., 0.],
[0., 0., 1., 0.],
[0., 0., 0., 1.]])
>>> np.argmax(onehot,1)
array([1, 2, 2, 3], dtype=int64)
>>> torch.topk(torch.tensor(onehot), 1)[1].squeeze(1)
tensor([1, 2, 2, 3])
更多推荐
所有评论(0)