bacow/easy-rl
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge branch 'master' of github.com:datawhalechina/leedeeprl-notes

Browse Source
This commit is contained in:
qiwang067
2020-12-12 11:34:42 +08:00
parent 3f5d997c3f 09f983120a
commit 2bc468d621
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
codes/Q-learning/agent.py
View File

@@ -5,7 +5,7 @@ Author: John
Email: johnjim0816@gmail.com Email: johnjim0816@gmail.com
Date: 2020-09-11 23:03:00 Date: 2020-09-11 23:03:00
LastEditor: John LastEditor: John
LastEditTime: 2020-11-24 20:22:03 LastEditTime: 2020-12-12 10:13:47
Discription: Discription:
Environment: Environment:
''' '''
@@ -56,7 +56,7 @@ class QLearning(object):
action = np.random.choice(self.action_dim) #有一定概率随机探索选取一个动作 action = np.random.choice(self.action_dim) #有一定概率随机探索选取一个动作
return action return action
def predict(self, obs): def predict(self, obs):
'''根据输入观测值，采样输出的动作值，带探索，测试模型时使用 '''根据输入观测值，采样输出的动作值，不带探索，测试模型时使用
''' '''
Q_list = self.Q_table[obs, :] Q_list = self.Q_table[obs, :]
Q_max = np.max(Q_list) Q_max = np.max(Q_list)