-
Notifications
You must be signed in to change notification settings - Fork 1
/
environments.py
195 lines (157 loc) · 5.13 KB
/
environments.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
import numpy as np
from copy import deepcopy
from matplotlib import pyplot as plt
#env for 1st agent as it needs to set a target
class EnvSetTarget:
#max_time_steps - total num of darsts
#max_lives_lost -total lives
def __init__(self,max_time_steps,max_lives_lost):
self.min_time_steps_elapsed = 0
self.max_time_steps_elapsed=max_time_steps
self.min_lives_lost=0
self.max_lives_lost=max_lives_lost
#self.max_steps = 120
#state variables
self.time_steps_elapsed = 0 #this represents the #darts you have thrown so far(0,119)
self.lives_lost = 0#lives lost so far(min=0,max=9)you have atmost 10 lives
#action sapce
self.action_space={
0:self.__one,
1:self.__two,
2:self.__four,
3:self.__six
}
@property
def time_steps_elapsed(self):
return self.__time_steps_elapsed
@time_steps_elapsed.setter
def time_steps_elapsed(self,time_steps_elapsed):
if time_steps_elapsed >= self.min_time_steps_elapsed and time_steps_elapsed<=self.max_time_steps_elapsed:
self.__time_steps_elapsed = time_steps_elapsed
else:
print("time_steps_elapsed needs to be within ",
self.min_time_steps_elapsed," and ",self.max_time_steps_elapsed)
raise Exception
@property
def lives_lost(self):
return self.__lives_lost
@lives_lost.setter
def lives_lost(self,lives_lost):
if lives_lost >= self.min_lives_lost and lives_lost <= self.max_lives_lost:
self.__lives_lost = lives_lost
else:
print("lives_lost needs to be within ",self.min_lives_lost," and ",self.max_lives_lost)
raise Exception
def set_state(self,state_vector):
"""
Set the state to some arbitrary value
May be helpful in learning for exploratory starts
:param state_vector: array([time_steps_elapsed,lives_lost])
:return:
"""
self.time_steps_elapsed = state_vector[0]
self.lives_lost = state_vector[1]
def reset(self):
"""
reset the environment to start state
:return:
"""
self.time_steps_elapsed = 0
self.lives_lost = 0
return np.array([self.time_steps_elapsed,self.lives_lost])
def step(self,action):
"""
:param action:
:return:
"""
reward = self.action_space[action]()
done=False
self.time_steps_elapsed +=1
if reward == 0:
self.lives_lost +=1
done = self.__all_lives_lost() or self.__time_over()
observation = np.array([self.time_steps_elapsed,self.lives_lost])
info=None
return observation,reward,done,info
def __all_lives_lost(self):
"""
:return: True if all lives lost else returns False
"""
if self.lives_lost == self.max_lives_lost:
return True
else:
return False
def __time_over(self):
"""
Returns true if end of episode due to finishing of time(thrown all the darts)
:return:
"""
if self.time_steps_elapsed == self.max_time_steps_elapsed:
return True
else:
return False
def __one(self):
val = np.random.random()
if val <= 0.95:
return 1
else:
return 0
def __two(self):
val = np.random.random()
if val <= 0.88:
return 2
else:
return 0
def __four(self):
val = np.random.random()
if val <= 0.8:
return 4
else:
return 0
def __six(self):
val = np.random.random()
if val <= 0.6:
return 6
else:
return 0
#second agent
class EnvChaseTarget(EnvSetTarget):
def __init__(self,max_time_steps,max_lives_lost,target):
EnvSetTarget.__init__(self,max_time_steps,max_lives_lost)
self.distance_from_target = target
self.init_target = target
def reset(self,target):
"""
:return:
"""
arr = EnvSetTarget.reset(self)
self.distance_from_target = target
return np.append(arr,self.distance_from_target)
def set_state(self,state_vector):
"""
:param state_vector:
:return:
"""
EnvSetTarget.set_state(self,state_vector)
self.distance_from_target = state_vector[2]
def step(self,action):
"""
:param action:
:return:
"""
observation, reward, done, info = EnvSetTarget.step(self,action)
self.distance_from_target -= reward
observation = np.append(observation,self.distance_from_target)
reward = 0
if done: #if already ended due to time over or loosing of all lives
done = True
reward = -1
if self.__target_achieved():
done = True
reward = +1
return observation, reward, done, info
def __target_achieved(self):
if self.distance_from_target <= 0:
return True
else:
return False