Python gym.utils.EzPickle() Examples

def __init__(self):
        self.multipliers = {
            'body_mass_multiplier': np.ones(5),
            'body_inertia_multiplier': np.ones((5,3)),
            'dof_damping_multiplier': np.ones(6),
            'geom_friction_multiplier': np.ones((5,3))
        }
        MetaMujocoEnv.__init__(self, 'hopper.xml', 4)
        utils.EzPickle.__init__(self)

        self.original_params = {
            'body_mass': self.model.body_mass.copy(),
            'body_inertia': self.model.body_inertia.copy(),
            'dof_damping': self.model.dof_damping.copy(),
            'geom_friction': self.model.geom_friction.copy()
        } 

def __init__(self):
        self.target_obj_sid = 0
        self.S_grasp_sid = 0
        self.obj_bid = 0
        curr_dir = os.path.dirname(os.path.abspath(__file__))
        mujoco_env.MujocoEnv.__init__(self, curr_dir+'/assets/DAPG_relocate.xml', 5)
        
        # change actuator sensitivity
        self.sim.model.actuator_gainprm[self.sim.model.actuator_name2id('A_WRJ1'):self.sim.model.actuator_name2id('A_WRJ0')+1,:3] = np.array([10, 0, 0])
        self.sim.model.actuator_gainprm[self.sim.model.actuator_name2id('A_FFJ3'):self.sim.model.actuator_name2id('A_THJ0')+1,:3] = np.array([1, 0, 0])
        self.sim.model.actuator_biasprm[self.sim.model.actuator_name2id('A_WRJ1'):self.sim.model.actuator_name2id('A_WRJ0')+1,:3] = np.array([0, -10, 0])
        self.sim.model.actuator_biasprm[self.sim.model.actuator_name2id('A_FFJ3'):self.sim.model.actuator_name2id('A_THJ0')+1,:3] = np.array([0, -1, 0])

        self.target_obj_sid = self.sim.model.site_name2id("target")
        self.S_grasp_sid = self.sim.model.site_name2id('S_grasp')
        self.obj_bid = self.sim.model.body_name2id('Object')
        utils.EzPickle.__init__(self)
        self.act_mid = np.mean(self.model.actuator_ctrlrange, axis=1)
        self.act_rng = 0.5*(self.model.actuator_ctrlrange[:,1]-self.model.actuator_ctrlrange[:,0]) 

def __init__(self, reward_type='sparse', terminate_on_success=False):
        initial_qpos = {
            'robot0:slide0': 0.405,
            'robot0:slide1': 0.48,
            'robot0:slide2': 0.0,
            'object0:joint': [1.25, 0.53, 0.4, 1., 0., 0., 0.],
            'object1:joint': [1.25, 0.53, 0.4, 1., 0., 0., 0.],
        }

        FewShotFetchEnv.__init__(
            self, FEW_SHOT_ENV_XML_PATH, has_object=True, block_gripper=False, n_substeps=20,
            gripper_extra_height=0.2, target_in_the_air=True, target_offset=0.0,
            obj_range=0.15, target_range=0.05, distance_threshold=0.05,
            initial_qpos=initial_qpos, reward_type=reward_type, goal_high_prob=1.0,
            min_goal_extra_height=0.15, max_goal_extra_height=0.2,
            min_dist_between_objs=0.1, same_color_radius=0.5,
            terminate_on_success=terminate_on_success
        )
        gym_utils.EzPickle.__init__(self)
        self._max_episode_steps = 65 

def __init__(self, reward_type='sparse', terminate_on_success=False):
        initial_qpos = {
            'robot0:slide0': 0.405,
            'robot0:slide1': 0.48,
            'robot0:slide2': 0.0,
            'object0:joint': [1.25, 0.53, 0.4, 1., 0., 0., 0.],
            'object1:joint': [1.25, 0.53, 0.4, 1., 0., 0., 0.],
        }

        FewShotFetchEnv.__init__(
            self, FEW_SHOT_ENV_XML_PATH, has_object=True, block_gripper=False, n_substeps=20,
            gripper_extra_height=0.2, target_in_the_air=True, target_offset=0.0,
            obj_range=0.15, target_range=0.05, distance_threshold=0.05,
            initial_qpos=initial_qpos, reward_type=reward_type, goal_high_prob=1.0,
            min_goal_extra_height=0.15, max_goal_extra_height=0.2,
            min_dist_between_objs=0.1, same_color_radius=0.3,
            terminate_on_success=terminate_on_success
        )
        gym_utils.EzPickle.__init__(self)
        self._max_episode_steps = 65 

def __init__(self):
        self.target_obj_sid = -1
        self.S_grasp_sid = -1
        self.obj_bid = -1
        self.tool_sid = -1
        self.goal_sid = -1
        curr_dir = os.path.dirname(os.path.abspath(__file__))
        mujoco_env.MujocoEnv.__init__(self, curr_dir+'/assets/DAPG_hammer.xml', 5)
        utils.EzPickle.__init__(self)

        # change actuator sensitivity
        self.sim.model.actuator_gainprm[self.sim.model.actuator_name2id('A_WRJ1'):self.sim.model.actuator_name2id('A_WRJ0')+1,:3] = np.array([10, 0, 0])
        self.sim.model.actuator_gainprm[self.sim.model.actuator_name2id('A_FFJ3'):self.sim.model.actuator_name2id('A_THJ0')+1,:3] = np.array([1, 0, 0])
        self.sim.model.actuator_biasprm[self.sim.model.actuator_name2id('A_WRJ1'):self.sim.model.actuator_name2id('A_WRJ0')+1,:3] = np.array([0, -10, 0])
        self.sim.model.actuator_biasprm[self.sim.model.actuator_name2id('A_FFJ3'):self.sim.model.actuator_name2id('A_THJ0')+1,:3] = np.array([0, -1, 0])
        
        self.target_obj_sid = self.sim.model.site_name2id('S_target')
        self.S_grasp_sid = self.sim.model.site_name2id('S_grasp')
        self.obj_bid = self.sim.model.body_name2id('Object')
        self.tool_sid = self.sim.model.site_name2id('tool')
        self.goal_sid = self.sim.model.site_name2id('nail_goal')
        self.act_mid = np.mean(self.model.actuator_ctrlrange, axis=1)
        self.act_rng = 0.5 * (self.model.actuator_ctrlrange[:, 1] - self.model.actuator_ctrlrange[:, 0]) 

def __init__(self):
        self.door_hinge_did = 0
        self.door_bid = 0
        self.grasp_sid = 0
        self.handle_sid = 0
        curr_dir = os.path.dirname(os.path.abspath(__file__))
        mujoco_env.MujocoEnv.__init__(self, curr_dir+'/assets/DAPG_door.xml', 5)
        
        # change actuator sensitivity
        self.sim.model.actuator_gainprm[self.sim.model.actuator_name2id('A_WRJ1'):self.sim.model.actuator_name2id('A_WRJ0')+1,:3] = np.array([10, 0, 0])
        self.sim.model.actuator_gainprm[self.sim.model.actuator_name2id('A_FFJ3'):self.sim.model.actuator_name2id('A_THJ0')+1,:3] = np.array([1, 0, 0])
        self.sim.model.actuator_biasprm[self.sim.model.actuator_name2id('A_WRJ1'):self.sim.model.actuator_name2id('A_WRJ0')+1,:3] = np.array([0, -10, 0])
        self.sim.model.actuator_biasprm[self.sim.model.actuator_name2id('A_FFJ3'):self.sim.model.actuator_name2id('A_THJ0')+1,:3] = np.array([0, -1, 0])

        utils.EzPickle.__init__(self)
        ob = self.reset_model()
        self.act_mid = np.mean(self.model.actuator_ctrlrange, axis=1)
        self.act_rng = 0.5*(self.model.actuator_ctrlrange[:,1]-self.model.actuator_ctrlrange[:,0])
        self.door_hinge_did = self.model.jnt_dofadr[self.model.joint_name2id('door_hinge')]
        self.grasp_sid = self.model.site_name2id('S_grasp')
        self.handle_sid = self.model.site_name2id('S_handle')
        self.door_bid = self.model.body_name2id('frame') 

def __init__(self):
        utils.EzPickle.__init__(self)
        mujoco_env.MujocoEnv.__init__(self, 'reacher.xml', 2)

    # def _step(self, a):
    #     # x = self._get_obs()[None]
    #     # assert np.allclose(self.get_body_com("fingertip")[:2], get_fingertips(x)),\
    #     # str(self.get_body_com("fingertip")) + " "+ str(get_fingertips(x))
    #     vec = self.get_body_com("fingertip")-self.get_body_com("target")
    #     reward_dist = - np.linalg.norm(vec[:2])
    #     reward_ctrl = - np.square(a).sum()*0.01
    #     reward = reward_dist + reward_ctrl
    #     self.do_simulation(a, self.frame_skip)
    #     ob = self._get_obs()
    #     done = False
    #     return ob, reward, done, dict(reward_dist=reward_dist, reward_ctrl=reward_ctrl) 

def __init__(self, reward_type='sparse'):
        initial_qpos = {
            'robot0:slide0': 0.405,
            'robot0:slide1': 0.48,
            'robot0:slide2': 0.0,
            'table0:slide0': 1.05,
            'table0:slide1': 0.4,
            'table0:slide2': 0.0,
            'object0:joint': [1.25, 0.53, 0.4, 1., 0., 0., 0.],
        }
        fetch_env.FetchEnv.__init__(
            self, 'fetch/pick_and_place.xml', has_object=True, block_gripper=False, n_substeps=20,
            gripper_extra_height=0.2, target_in_the_air=True, target_offset=0.0,
            obj_range=0.15, target_range=0.15, distance_threshold=0.05,
            initial_qpos=initial_qpos, reward_type=reward_type)
        utils.EzPickle.__init__(self) 

def __init__(self, reward_type='sparse'):
        initial_qpos = {
            'robot0:slide0': 0.405,
            'robot0:slide1': 0.48,
            'robot0:slide2': 0.0,
            'table0:slide0': 1.05,
            'table0:slide1': 0.4,
            'table0:slide2': 0.0,
            'object0:joint': [1.25, 0.53, 0.4, 1., 0., 0., 0.],
        }
        fetch_env.FetchEnv.__init__(
            self, 'fetch/push.xml', has_object=True, block_gripper=True, n_substeps=20,
            gripper_extra_height=0.0, target_in_the_air=False, target_offset=0.0,
            obj_range=0.15, target_range=0.15, distance_threshold=0.05,
            initial_qpos=initial_qpos, reward_type=reward_type)
        utils.EzPickle.__init__(self) 

def __init__(self, reward_type='sparse'):
        initial_qpos = {
            'robot0:slide0': 0.405,
            'robot0:slide1': 0.48,
            'robot0:slide2': 0.0,
            'object0:joint': [1.25, 0.53, 0.4, 1., 0., 0., 0.],
            'object1:joint': [1.25, 0.53, 0.4, 1., 0., 0., 0.],
        }

        FewShotReachEnv.__init__(
            self, FEW_SHOT_ENV_XML_PATH, has_object=True, block_gripper=False, n_substeps=20,
            gripper_extra_height=0.2, target_in_the_air=True, target_offset=0.0,
            obj_range=0.15, target_range=0.0, distance_threshold=0.05,
            initial_qpos=initial_qpos, reward_type=reward_type, goal_high_prob=1.0,
            min_goal_extra_height=0.05, max_goal_extra_height=0.05,
            min_dist_between_objs=0.1, same_color_radius=0.3
        )
        gym_utils.EzPickle.__init__(self)
        self._max_episode_steps = 30 

def __init__(self, episode_len=499):
        # # 1 x 1 x 8 x 2
        # self.valid_targets = np.array(
        #     [[[
        #         [8.0, 0.0],
        #         [0.0, 8.0],
        #         [-8.0, 0.0],
        #         [0.0, -8.0],
        #     ]]]
        # )

        self.timestep = 0.0
        self.episode_len = episode_len
        # self.init_xy = init_pos.copy()

        xml_path = os.path.join(os.path.dirname(__file__), "assets", 'low_gear_ratio_ant.xml')
        # xml_path = os.path.join(os.path.dirname(__file__), "assets", 's_maze.xml')
        MujocoEnv.__init__(self, xml_path, 5)
        utils.EzPickle.__init__(self) 

def __init__(
        self,
        direction=1,
        maze_length=0.6,
        sparse_reward=False,
        no_reward=False,
        include_vel=False,
        episode_length=100,
    ):
        utils.EzPickle.__init__(self)
        self.sparse_reward = sparse_reward
        self.no_reward = no_reward
        self.include_vel = include_vel
        self.max_episode_length = episode_length
        self.direction = direction
        self.length = maze_length

        self.episode_length = 0

        model = point_mass_maze(direction=self.direction, length=self.length)
        with model.asfile() as f:
            mujoco_env.MujocoEnv.__init__(self, f.name, 5) 

def __init__(self, use_rel_pos_obs=False, terminate_near_target=False):
        self.terminate_near_target = terminate_near_target
        self.terminate_radius = 0.5
        self.use_rel_pos_obs = use_rel_pos_obs
        # 1 x 1 x 8 x 2
        self.valid_targets = np.array(
            [[[
                # [2.0, 0.0],
                # # [1.41, 1.41],
                # [0.0, 2.0],
                # # [-1.41, 1.41],
                # [-2.0, 0.0],
                # # [-1.41, -1.41],
                # [0.0, -2.0],
                # # [1.41, -1.41]

                [4.0, 0.0],
                [0.0, 4.0],
                [-4.0, 0.0],
                [0.0, -4.0],
            ]]]
        )

        xml_path = os.path.join(os.path.dirname(__file__), "assets", 'low_gear_ratio_ant.xml')
        MujocoEnv.__init__(self, xml_path, 5)
        utils.EzPickle.__init__(self) 

def __init__(self, max_timesteps=1000, disabled=False, gear=150):
        #mujoco_env.MujocoEnv.__init__(self, 'ant.xml', 5)
        utils.EzPickle.__init__(self)
        self.timesteps = 0
        self.max_timesteps=max_timesteps

        if disabled:
            model = angry_ant_crippled(gear=gear)
        else:
            model = ant_env(gear=gear)
        with model.asfile() as f:
            mujoco_env.MujocoEnv.__init__(self, f.name, 5) 

def __init__(self, game='pong', obs_type='ram', frameskip=(2, 5), repeat_action_probability=0.):
        """Frameskip should be either a tuple (indicating a random range to
        choose from, with the top value exclude), or an int."""

        utils.EzPickle.__init__(self, game, obs_type, frameskip, repeat_action_probability)
        assert obs_type in ('ram', 'image')

        self.game_path = atari_py.get_game_path(game)
        if not os.path.exists(self.game_path):
            raise IOError('You asked for game %s but path %s does not exist'%(game, self.game_path))
        self._obs_type = obs_type
        self.frameskip = frameskip
        self.ale = atari_py.ALEInterface()
        self.viewer = None

        # Tune (or disable) ALE's action repeat:
        # https://github.com/openai/gym/issues/349
        assert isinstance(repeat_action_probability, (float, int)), "Invalid repeat_action_probability: {!r}".format(repeat_action_probability)
        self.ale.setFloat('repeat_action_probability'.encode('utf-8'), repeat_action_probability)

        self.seed()

        self._action_set = self.ale.getMinimalActionSet()
        self.action_space = spaces.Discrete(len(self._action_set))

        (screen_width,screen_height) = self.ale.getScreenDims()
        if self._obs_type == 'ram':
            self.observation_space = spaces.Box(low=0, high=255, dtype=np.uint8, shape=(128,))
        elif self._obs_type == 'image':
            self.observation_space = spaces.Box(low=0, high=255, shape=(screen_height, screen_width, 3), dtype=np.uint8)
        else:
            raise error.Error('Unrecognized observation type: {}'.format(self._obs_type)) 

def __init__(self):
        utils.EzPickle.__init__(self)
        mujoco_env.MujocoEnv.__init__(self, 'reacher.xml', 2) 

def __init__(self, use_relative_pos=False):
        self.ACCEPT_RADIUS = 0.5
        self.true_label = 0
        self.task_hyperplane = np.ones(4) / 2.0
        self.first_sample = np.random.uniform(size=4)
        self.second_sample = np.random.uniform(size=4)
        self.targets = np.array(
            [
                [1.41, 1.41],
                [-1.41, 1.41]
            ]
        )
        self.use_relative_pos = use_relative_pos
        MetaMujocoEnv.__init__(self, 'low_gear_ratio_ant.xml', 5)
        utils.EzPickle.__init__(self) 

def __init__(self):
        self.goal_direction = np.array([1.0, 0.0])
        # MetaMujocoEnv.__init__(self, 'ant.xml', 5)
        MetaMujocoEnv.__init__(self, 'low_gear_ratio_ant.xml', 5)
        # MetaMujocoEnv.__init__(self, 'low_gear_ratio_ant.xml', 5)
        utils.EzPickle.__init__(self) 

def __init__(self):
        mujoco_env.MujocoEnv.__init__(self, 'humanoid.xml', 5)
        utils.EzPickle.__init__(self) 

def __init__(self):
        self.goal_pos = np.array([14.1, 14.1])
        self.target_distance = 25.0

        MetaMujocoEnv.__init__(self, 'low_gear_ratio_ant_8_star.xml', 5)
        utils.EzPickle.__init__(self) 

def __init__(self):
        mujoco_env.MujocoEnv.__init__(self, "walker2d.xml", 4)
        utils.EzPickle.__init__(self) 

def __init__(self):
        utils.EzPickle.__init__(self)
        self._ball_hit_ground = False
        self._ball_hit_location = None
        mujoco_env.MujocoEnv.__init__(self, 'thrower.xml', 5) 

def __init__(self):
        utils.EzPickle.__init__(self)
        self._striked = False
        self._min_strike_dist = np.inf
        self.strike_threshold = 0.1
        mujoco_env.MujocoEnv.__init__(self, 'striker.xml', 5) 

def __init__(self):
        mujoco_env.MujocoEnv.__init__(self, 'swimmer.xml', 4)
        utils.EzPickle.__init__(self) 

def __init__(self):
        mujoco_env.MujocoEnv.__init__(self, "walker2d.xml", 4)
        utils.EzPickle.__init__(self) 

def __init__(self):
        utils.EzPickle.__init__(self)
        mujoco_env.MujocoEnv.__init__(self, 'pusher.xml', 5) 

def __init__(self):
        mujoco_env.MujocoEnv.__init__(self, 'ant.xml', 5)
        utils.EzPickle.__init__(self) 

def __init__(self):
        mujoco_env.MujocoEnv.__init__(self, 'humanoidstandup.xml', 5)
        utils.EzPickle.__init__(self) 

def __init__(self):
        mujoco_env.MujocoEnv.__init__(self, 'half_cheetah.xml', 5)
        utils.EzPickle.__init__(self) 

def __init__(self):
        mujoco_env.MujocoEnv.__init__(self, 'inverted_double_pendulum.xml', 5)
        utils.EzPickle.__init__(self)