-
Notifications
You must be signed in to change notification settings - Fork 5
/
_cartpole.py
316 lines (267 loc) · 12.4 KB
/
_cartpole.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
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
"""
Classic cart-pole system implemented by Rich Sutton et al.
Copied from http://incompleteideas.net/sutton/book/code/pole.c
permalink: https://perma.cc/C9ZM-652R
"""
import math
import gym
from gym import spaces, logger
from gym.utils import seeding
import numpy as np
import os
os.environ["OMP_NUM_THREADS"] = "1"
os.environ["OPENBLAS_NUM_THREADS"] = "1"
class CartPoleEnv(gym.Env):
"""
Description:
A pole is attached by an un-actuated joint to a cart, which moves along
a frictionless track. The pendulum starts upright, and the goal is to
prevent it from falling over by increasing and reducing the cart's
velocity.
Source:
This environment corresponds to the version of the cart-pole problem
described by Barto, Sutton, and Anderson
Observation:
Type: Box(4)
Num Observation Min Max
0 Cart Position -4.8 4.8
1 Cart Velocity -Inf Inf
2 Pole Angle -0.418 rad (-24 deg) 0.418 rad (24 deg)
3 Pole Angular Velocity -Inf Inf
Actions:
Type: Discrete(2)
Num Action
0 Push cart to the left
1 Push cart to the right
Note: The amount the velocity that is reduced or increased is not
fixed; it depends on the angle the pole is pointing. This is because
the center of gravity of the pole increases the amount of energy needed
to move the cart underneath it
Reward:
Reward is 1 for every step taken, including the termination step
Starting State:
All observations are assigned a uniform random value in [-0.05..0.05]
Episode Termination:
Pole Angle is more than 12 degrees.
Cart Position is more than 2.4 (center of the cart reaches the edge of
the display).
Episode length is greater than 200.
Solved Requirements:
Considered solved when the average return is greater than or equal to
195.0 over 100 consecutive trials.
"""
metadata = {
'render.modes': ['human', 'rgb_array'],
'video.frames_per_second': 50
}
def __init__(self, e_max = 200, std = 0.02, seed = 42):
self.gravity = 9.8
self.masscart = 1.0
self.masspole = 0.1
self.total_mass = (self.masspole + self.masscart)
self.length = 0.5 # actually half the pole's length
self.polemass_length = (self.masspole * self.length)
self.force_mag = 10.0
self.tau = 0.02 # seconds between state updates
self.kinematics_integrator = 'euler'
self.e_max = e_max
# self.std = np.repeat(0, 4)
self.std = np.repeat(std, 4)
self.e = 0
# Angle at which to fail the episode
self.theta_threshold_radians = 12 * 2 * math.pi / 360
self.x_threshold = 2.4
# Angle limit set to 2 * theta_threshold_radians so failing observation
# is still within bounds.
high = np.array([self.x_threshold * 2,
np.finfo(np.float32).max,
self.theta_threshold_radians * 2,
np.finfo(np.float32).max],
dtype=np.float32)
self.action_space = spaces.Discrete(2)
self.observation_space = spaces.Box(-high, high, dtype=np.float32)
self.set_seed(seed = seed)
self.seed = seed
self.viewer = None
self.state = None
self.steps_beyond_done = None
def set_seed(self, seed=None):
self.np_random, seed = seeding.np_random(seed)
return [seed]
def step(self, action):
np.random.seed(self.seed)
self.seed += 1
err_msg = "%r (%s) invalid" % (action, type(action))
assert self.action_space.contains(action), err_msg
x, x_dot, theta, theta_dot = self.state
force = self.force_mag if action == 1 else -self.force_mag
costheta = math.cos(theta)
sintheta = math.sin(theta)
# For the interested reader:
# https://coneural.org/florian/papers/05_cart_pole.pdf
temp = (force + self.polemass_length * theta_dot ** 2 * sintheta) / self.total_mass
thetaacc = (self.gravity * sintheta - costheta * temp) / (self.length * (4.0 / 3.0 - self.masspole * costheta ** 2 / self.total_mass))
xacc = temp - self.polemass_length * thetaacc * costheta / self.total_mass
if self.kinematics_integrator == 'euler':
x = x + self.tau * x_dot
x_dot = x_dot + self.tau * xacc
theta = theta + self.tau * theta_dot
theta_dot = theta_dot + self.tau * thetaacc
else: # semi-implicit euler
x_dot = x_dot + self.tau * xacc
x = x + self.tau * x_dot
theta_dot = theta_dot + self.tau * thetaacc
theta = theta + self.tau * theta_dot
errors = np.random.randn(4) * self.std
self.state = (x + errors[0], x_dot + errors[1], theta + errors[2], theta_dot + errors[3])
# self.state = (x, x_dot, theta, theta_dot)
done = bool(
x < -self.x_threshold
or x > self.x_threshold
or theta < -self.theta_threshold_radians
or theta > self.theta_threshold_radians
or self.e > self.e_max
)
self.e += 1
if not done:
""" reward """
reward = self.cal_reward(x, theta) #1.0
elif self.steps_beyond_done is None:
# Pole just fell!
self.steps_beyond_done = 0
""" reward """
reward = self.cal_reward(x, theta) #1.0
else:
if self.steps_beyond_done == 0:
logger.warn(
"You are calling 'step()' even though this "
"environment has already returned done = True. You "
"should always call 'reset()' once you receive 'done = "
"True' -- any further steps are undefined behavior."
)
self.steps_beyond_done += 1
reward = 0.0
return np.array(self.state), reward, done, {}
def cal_reward(self, x, theta, u = 0):
# https://towardsdatascience.com/infinite-steps-cartpole-problem-with-variable-reward-7ad9a0dcf6d0
# If x and θ represents cart position and pole angle respectively, we define the reward as:
reward = (1 - (x ** 2) / 11.52 - (theta ** 2) / 288)
return reward
# def angle_normalize(x):
# return (((x+np.pi) % (2*np.pi)) - np.pi)
# # angle_normalise((th)**2 +.1*thdot**2 + .001*(action**2))
# costs = angle_normalize(th)**2 + .1*thdot**2 + .001*(u**2)
#return -costs
def reset(self):
self.state = self.np_random.uniform(low=-0.05, high=0.05, size=(4,))
self.steps_beyond_done = None
self.e = 0
return np.array(self.state)
##################################################################################################
##################################################################################################
def reset_multiple(self, N):
self.states = self.np_random.uniform(low=-0.05, high=0.05, size=(4, N))
self.steps_beyond_done = np.repeat(None, N)
self.dones = np.repeat(False, N)
self.e = np.repeat(0, N)
return np.array(self.states)
def step_multiple(self, actions):
np.random.seed(self.seed)
self.seed += 1
N = len(actions)
assert self.action_space.contains(actions[0]), err_msg
x, x_dot, theta, theta_dot = self.states # [B, 4] -> [4, N]
force = np.array([self.force_mag if action == 1 else -self.force_mag for action in actions])
costheta = np.cos(theta)
sintheta = np.sin(theta)
# For the interested reader:
# https://coneural.org/florian/papers/05_cart_pole.pdf
temp = (force + self.polemass_length * theta_dot ** 2 * sintheta) / self.total_mass
thetaacc = (self.gravity * sintheta - costheta * temp) / (self.length * (4.0 / 3.0 - self.masspole * costheta ** 2 / self.total_mass))
xacc = temp - self.polemass_length * thetaacc * costheta / self.total_mass
if self.kinematics_integrator == 'euler':
x = x + self.tau * x_dot
x_dot = x_dot + self.tau * xacc
theta = theta + self.tau * theta_dot
theta_dot = theta_dot + self.tau * thetaacc
else: # semi-implicit euler
x_dot = x_dot + self.tau * xacc
x = x + self.tau * x_dot
theta_dot = theta_dot + self.tau * thetaacc
theta = theta + self.tau * theta_dot
""" next state """
errors = (np.random.randn(4, N).T * self.std).T
self.old_states = self.states.copy()
self.states = np.array([x + errors[0], x_dot + errors[1], theta + errors[2], theta_dot + errors[3]])
self.states[:, self.dones] = self.old_states[:, self.dones]
self.dones = np.logical_or.reduce((x < -self.x_threshold
, x > self.x_threshold, theta < -self.theta_threshold_radians
, theta > self.theta_threshold_radians, self.e > self.e_max)).astype(bool)
self.e += 1
rewards = self.cal_reward(x, theta) #1.0
# elif self.steps_beyond_done is None:
# # Pole just fell!
# self.steps_beyond_done = 0
# """ reward """
# reward = self.cal_reward(x, theta) #1.0
# else:
# if self.steps_beyond_done == 0:
# logger.warn(
# "You are calling 'step()' even though this "
# "environment has already returned done = True. You "
# "should always call 'reset()' once you receive 'done = "
# "True' -- any further steps are undefined behavior."
# )
# self.steps_beyond_done += 1
# reward = 0.0
return np.array(self.states), rewards, self.dones, {}
# def render(self, mode='human'):
# screen_width = 600
# screen_height = 400
# world_width = self.x_threshold * 2
# scale = screen_width/world_width
# carty = 100 # TOP OF CART
# polewidth = 10.0
# polelen = scale * (2 * self.length)
# cartwidth = 50.0
# cartheight = 30.0
# if self.viewer is None:
# from gym.envs.classic_control import rendering
# self.viewer = rendering.Viewer(screen_width, screen_height)
# l, r, t, b = -cartwidth / 2, cartwidth / 2, cartheight / 2, -cartheight / 2
# axleoffset = cartheight / 4.0
# cart = rendering.FilledPolygon([(l, b), (l, t), (r, t), (r, b)])
# self.carttrans = rendering.Transform()
# cart.add_attr(self.carttrans)
# self.viewer.add_geom(cart)
# l, r, t, b = -polewidth / 2, polewidth / 2, polelen - polewidth / 2, -polewidth / 2
# pole = rendering.FilledPolygon([(l, b), (l, t), (r, t), (r, b)])
# pole.set_color(.8, .6, .4)
# self.poletrans = rendering.Transform(translation=(0, axleoffset))
# pole.add_attr(self.poletrans)
# pole.add_attr(self.carttrans)
# self.viewer.add_geom(pole)
# self.axle = rendering.make_circle(polewidth/2)
# self.axle.add_attr(self.poletrans)
# self.axle.add_attr(self.carttrans)
# self.axle.set_color(.5, .5, .8)
# self.viewer.add_geom(self.axle)
# self.track = rendering.Line((0, carty), (screen_width, carty))
# self.track.set_color(0, 0, 0)
# self.viewer.add_geom(self.track)
# self._pole_geom = pole
# if self.state is None:
# return None
# # Edit the pole polygon vertex
# pole = self._pole_geom
# l, r, t, b = -polewidth / 2, polewidth / 2, polelen - polewidth / 2, -polewidth / 2
# pole.v = [(l, b), (l, t), (r, t), (r, b)]
# x = self.state
# cartx = x[0] * scale + screen_width / 2.0 # MIDDLE OF CART
# self.carttrans.set_translation(cartx, carty)
# self.poletrans.set_rotation(-x[2])
# return self.viewer.render(return_rgb_array=mode == 'rgb_array')
# def close(self):
# if self.viewer:
# self.viewer.close()
# self.viewer = None