def __init__(self):
self.horizon = 400
self.gamma = .95
self._theta = 0.
self._theta_dot = 0.
self._g = 9.8
self._m = 2.
self._M = 8.
self._l = .5
self._alpha = 1. / (self._m + self._M)
self._noise = 10.
self._angle_max = np.pi / 2.
self.max_velocity = np.inf
self._dt = 0.1
# gym attributes
self.viewer = None
high = np.array([self._angle_max, self.max_velocity])
self.observation_space = spaces.Box(low=-high, high=high)
self.action_space = fqispaces.DiscreteValued([-50, 0, 50], decimals=0)
# initialize state
self.seed()
self.reset()
def __init__(self):
self.horizon = 100
self.gamma = .95
self.max_action = 5
self._g = 9.81
self._M1 = self._M2 = 1
self._L1 = self._L2 = 1
self._mu1 = self._mu2 = .01
self._dt = .1
# gym attributes
self.viewer = None
high = np.array([np.inf, np.inf, np.inf, np.inf])
self.observation_space = spaces.Box(low=-high, high=high)
self.action_space = fqispaces.DiscreteValued([-5, 5], decimals=0)
# evaluation initial states
self.initial_states = np.zeros((41, 4))
self.initial_states[:, 0] = np.linspace(-2, 2, 41)
# initialize state
self.seed()
self.reset()
def __init__(self):
self.horizon = 100
self.gamma = 0.95
self.max_pos = 1.
self.max_velocity = 3.
self._g = 9.81
self._m = 1
self._dt = .1
# gym attributes
self.viewer = None
high = np.array([self.max_pos, self.max_velocity])
self.observation_space = spaces.Box(low=-high, high=high)
self.action_space = fqispaces.DiscreteValued([-4., 4.], decimals=0)
# evaluation initial states
self.initial_states = np.zeros((289, 2))
cont = 0
for i in range(-8, 9):
for j in range(-8, 9):
self.initial_states[cont, :] = [0.125 * i, 0.375 * j]
cont += 1
# initialize state
self.seed()
self.reset()
def __init__(self, **kwargs):
self.horizon = 100
self.gamma = 0.9
self._m = 1.
self._l = 1.
self._g = 9.8
self._mu = 0.01
self._dt = 0.02
# gym attributes
self.viewer = None
high = np.array([np.inf, np.inf])
self.observation_space = spaces.Box(low=-high, high=high)
self.action_space = fqispaces.DiscreteValued([-5, 0, 5], decimals=5)
# initialize state
self.seed()
self.reset()
def __init__(self):
self.horizon = 300
self.gamma = 0.95
self.max_pos = 1.
self.max_velocity = 3.
self._g = 9.81
self._m = 1
self._dt = .1
# gym attributes
self.viewer = None
high = np.array([self.max_pos, self.max_velocity])
self.observation_space = spaces.Box(low=-high, high=high)
self.action_space = fqispaces.DiscreteValued([-4., 4.], decimals=0)
# initialize state
self.seed()
self.reset()