def main():
environment = RacecarGymEnv(renders=True, isDiscrete=isDiscrete)
environment.reset()
targetVelocitySlider = environment._p.addUserDebugParameter("wheelVelocity",-1,1,0)
steeringSlider = environment._p.addUserDebugParameter("steering",-1,1,0)
while (True):
targetVelocity = environment._p.readUserDebugParameter(targetVelocitySlider)
steeringAngle = environment._p.readUserDebugParameter(steeringSlider)
if (isDiscrete):
discreteAction = 0
if (targetVelocity<-0.33):
discreteAction=0
else:
if (targetVelocity>0.33):
discreteAction=6
else:
discreteAction=3
if (steeringAngle>-0.17):
if (steeringAngle>0.17):
discreteAction=discreteAction+2
else:
discreteAction=discreteAction+1
action=discreteAction
else:
action=[targetVelocity,steeringAngle]
state, reward, done, info = environment.step(action)
obs = environment.getExtendedObservation()
print("obs")
print(obs)
def main():
env = RacecarGymEnv(renders=True,isDiscrete=True)
act = deepq.load("racecar_model.pkl")
print(act)
while True:
obs, done = env.reset(), False
print("===================================")
print("obs")
print(obs)
episode_rew = 0
while not done:
env.render()
obs, rew, done, _ = env.step(act(obs[None])[0])
episode_rew += rew
print("Episode reward", episode_rew)
def main():
environment = RacecarGymEnv(renders=True, isDiscrete=isDiscrete)
environment.reset()
targetVelocitySlider = environment._p.addUserDebugParameter("wheelVelocity", -1, 1, 0)
steeringSlider = environment._p.addUserDebugParameter("steering", -1, 1, 0)
while (True):
targetVelocity = environment._p.readUserDebugParameter(targetVelocitySlider)
steeringAngle = environment._p.readUserDebugParameter(steeringSlider)
if (isDiscrete):
discreteAction = 0
if (targetVelocity < -0.33):
discreteAction = 0
else:
if (targetVelocity > 0.33):
discreteAction = 6
else:
discreteAction = 3
if (steeringAngle > -0.17):
if (steeringAngle > 0.17):
discreteAction = discreteAction + 2
else:
discreteAction = discreteAction + 1
action = discreteAction
else:
action = [targetVelocity, steeringAngle]
state, reward, done, info = environment.step(action)
obs = environment.getExtendedObservation()
print("obs")
print(obs)
def main():
environment = RacecarGymEnv(renders=True, isDiscrete=isDiscrete)
environment.reset()
targetVelocitySlider = environment._p.addUserDebugParameter(
"wheelVelocity", -1, 1, 0)
steeringSlider = environment._p.addUserDebugParameter("steering", -1, 1, 0)
y2z = environment._p.getQuaternionFromEuler([0, 0, 1.57])
meshScale = [1, 1, 1]
visualShapeId = environment._p.createVisualShape(
shapeType=environment._p.GEOM_MESH,
fileName="domino/domino.obj",
rgbaColor=[1, 1, 1, 1],
specularColor=[0.4, .4, 0],
visualFrameOrientation=y2z,
meshScale=meshScale)
boxDimensions = [0.5 * 0.00635, 0.5 * 0.0254, 0.5 * 0.0508]
collisionShapeId = environment._p.createCollisionShape(
environment._p.GEOM_BOX, halfExtents=boxDimensions)
while (True):
targetVelocity = environment._p.readUserDebugParameter(
targetVelocitySlider)
steeringAngle = environment._p.readUserDebugParameter(steeringSlider)
if (isDiscrete):
discreteAction = 0
if (targetVelocity < -0.33):
discreteAction = 0
else:
if (targetVelocity > 0.33):
discreteAction = 6
else:
discreteAction = 3
if (steeringAngle > -0.17):
if (steeringAngle > 0.17):
discreteAction = discreteAction + 2
else:
discreteAction = discreteAction + 1
action = discreteAction
else:
action = [targetVelocity, steeringAngle]
qKey = ord('a')
keys = environment._p.getKeyboardEvents()
if qKey in keys and keys[qKey] & environment._p.KEY_WAS_TRIGGERED:
objid = environment._p.createMultiBody(
baseMass=0.025,
baseCollisionShapeIndex=collisionShapeId,
baseVisualShapeIndex=visualShapeId,
basePosition=[0.04, 0.05, 0.06],
useMaximalCoordinates=True)
environment._p.resetBaseVelocity(objid, linearVelocity=[2, 2, 1])
state, reward, done, info = environment.step(action)
obs = environment.getExtendedObservation()
print("obs")
print(obs)
def main():
env = RacecarGymEnv(renders=False, isDiscrete=True)
model = deepq.models.mlp([64])
act = deepq.learn(env,
q_func=model,
lr=1e-3,
max_timesteps=10000,
buffer_size=50000,
exploration_fraction=0.1,
exploration_final_eps=0.02,
print_freq=10,
callback=callback)
print("Saving model to racecar_model.pkl")
act.save("racecar_model.pkl")
def main():
env = RacecarGymEnv(renders=True, isDiscrete=True)
act = deepq.load("racecar_model.pkl")
print(act)
while True:
obs, done = env.reset(), False
print("===================================")
print("obs")
print(obs)
episode_rew = 0
while not done:
env.render()
obs, rew, done, _ = env.step(act(obs[None])[0])
episode_rew += rew
print("Episode reward", episode_rew)
config['trainBatchSize'] = 32
config['gamma'] = 0.9
config['learningRate'] = 0.001
config['netGradClip'] = 1
config['logFlag'] = True
config['logFileName'] = 'StabilizerOneDLog/traj'
config['logFrequency'] = 100
config['priorityMemoryOption'] = False
config['netUpdateOption'] = 'doubleQ'
config['netUpdateFrequency'] = 1
config['priorityMemory_absErrUpper'] = 5
import gym
from pybullet_envs.bullet.racecarGymEnv import RacecarGymEnv
env = RacecarGymEnv(renders=True, isDiscrete=True)
N_S = env.observation_space.shape[0]
N_A = env.action_space.n
netParameter = dict()
netParameter['n_feature'] = N_S
netParameter['n_hidden'] = [100]
netParameter['n_output'] = N_A
policyNet = MultiLayerNetRegression(netParameter['n_feature'],
netParameter['n_hidden'],
netParameter['n_output'])
targetNet = deepcopy(policyNet)
optimizer = optim.Adam(policyNet.parameters(), lr=config['learningRate'])
"""
#add parent dir to find package. Only needed for source code build, pip install doesn't need it.
import os, inspect
currentdir = os.path.dirname(os.path.abspath(inspect.getfile(inspect.currentframe())))
parentdir = os.path.dirname(os.path.dirname(currentdir))
os.sys.path.insert(0, parentdir)
from pybullet_envs.bullet.racecarGymEnv import RacecarGymEnv
isDiscrete = False
environment = RacecarGymEnv(renders=True, isDiscrete=isDiscrete)
environment.reset()
targetVelocitySlider = environment._p.addUserDebugParameter("wheelVelocity", -1, 1, 0)
steeringSlider = environment._p.addUserDebugParameter("steering", -1, 1, 0)
while (True):
targetVelocity = environment._p.readUserDebugParameter(targetVelocitySlider)
steeringAngle = environment._p.readUserDebugParameter(steeringSlider)
if (isDiscrete):
discreteAction = 0
if (targetVelocity < -0.33):
discreteAction = 0
else:
if (targetVelocity > 0.33):
discreteAction = 6
# 乱数の種固定
np.random.seed(args.seed)
torch.manual_seed(args.seed)
# GPU or CPU
device_name = 'cpu' if args.cuda < 0 else "cuda:{}".format(args.cuda)
device = torch.device(device_name)
set_device(device)
# logのcsvファイル確保
score_file = os.path.join(args.log, 'progress.csv')
logger.add_tabular_output(score_file)
# Gymのenviromentを生成
from pybullet_envs.bullet.racecarGymEnv import RacecarGymEnv
env = RacecarGymEnv(renders=False, isDiscrete=False)
# 観測と行動の次元
observation_space = env.observation_space
action_space = env.action_space
# Q-Network
qf_net = QNet(observation_space, action_space, args.h1, args.h2)
qf = DeterministicSAVfunc(
observation_space, action_space, qf_net,
data_parallel=args.data_parallel) # 決定的行動状態価値関数?q-netの出力の形を少し整える
# target Q network theta1
targ_qf1_net = QNet(observation_space, action_space, args.h1, args.h2)
targ_qf1_net.load_state_dict(qf_net.state_dict()) # model（重み）をロード(q-netからコピー)
targ_qf1 = CEMDeterministicSAVfunc(
observation_space,