![](/img/trans.png)
[英]AssertionError: The algorithm only supports <class 'gym.spaces.box.Box'> as action spaces but Box(-1.0, 1.0, (3,), float32) was provided
[英]Is there something similar as gym.spaces.Box in numpy?
在OpenAI健身房中,您可以執行以下操作:
from gym import spaces
low = [1, 2, 3]
high = [4, 4, 4]
box = spaces.Box(np.array(low), np.array(high))
observation = np.array([2, 2, 2])
if not box.contains(observation):
print("This is invalid!")
它基本上檢查每個維度
def contains(self, obs):
n = len(obs) # == len(low) == len(high)
for i in range(n):
if not (self.low[i] <= obs[i] <= self.high[i]):
return False
return True
numpy是否也帶有spaces.Box類之類的東西?
我不知道函數,但是編寫自己很容易。
import numpy as np
np.all(np.less_equal(low, observation)) and np.all(np.greater_equal(observation, high))
這將檢查所有觀察值是否在指定范圍內。 如果省略np.all
,則可以看到問題所在的維度。
聲明:本站的技術帖子網頁,遵循CC BY-SA 4.0協議,如果您需要轉載,請注明本站網址或者原文地址。任何問題請咨詢:yoyou2525@163.com.