stable_gym.envs.mujoco.humanoid_cost

Modified version of the Humanoid Mujoco environment found in the gymnasium library. This modification was first described by Han et al. 2020. In this modified version:

  • The objective was changed to a velocity-tracking task. To do this, the reward is replaced with a cost. This cost is the squared difference between the Humanoid’s forward velocity and a reference value (error).

  • Three optional variables were added to the observation space; The reference velocity, the reference error (i.e. the difference between the humanoid’s forward velocity and the reference) and the humanoid’s forward velocity. These variables can be enabled using the exclude_reference_from_observation, exclude_reference_error_from_observation and exclude_velocity_from_observation environment arguments.

Submodules

Classes

HumanoidCost

Custom Humanoid gymnasium environment.

Package Contents

class stable_gym.envs.mujoco.humanoid_cost.HumanoidCost(reference_forward_velocity=1.0, randomise_reference_forward_velocity=False, randomise_reference_forward_velocity_range=(0.5, 1.5), forward_velocity_weight=1.0, include_ctrl_cost=False, include_health_penalty=True, health_penalty_size=10, ctrl_cost_weight=0.0001, terminate_when_unhealthy=True, healthy_z_range=(1.0, 2.0), reset_noise_scale=0.01, exclude_current_positions_from_observation=True, exclude_reference_from_observation=False, exclude_reference_error_from_observation=True, exclude_x_velocity_from_observation=False, action_space_dtype=np.float32, observation_space_dtype=np.float64, **kwargs)[source]

Bases: gymnasium.envs.mujoco.humanoid_v4.HumanoidEnv, gymnasium.utils.EzPickle

Custom Humanoid gymnasium environment.

Note

Can also be used in a vectorized manner. See the gym.vector documentation.

Source:

This is a modified version of the Humanoid Mujoco environment found in the gymnasium library. This modification was first described by Han et al. 2020. Compared to the original Humanoid environment in this modified version:

  • The objective was changed to a velocity-tracking task. To do this, the reward is replaced with a cost. This cost is the squared difference between the Humanoid’s forward velocity and a reference value (error). Additionally, also a control cost and health penalty can be included in the cost.

  • Three optional variables were added to the observation space; The reference velocity, the reference error (i.e. the difference between the humanoid’s forward velocity and the reference) and the humanoid’s forward velocity. These variables can be enabled using the exclude_reference_from_observation, exclude_reference_error_from_observation and exclude_velocity_from_observation environment arguments.

The rest of the environment is the same as the original Humanoid environment. Below, the modified cost is described. For more information about the environment (e.g. observation space, action space, episode termination, etc.), please refer to the gymnasium library.

Modified cost:

A cost, computed using the HumanoidCost.cost() method, is given for each simulation step, including the terminal step. This cost is defined as the error between the Humanoid’s forward velocity and a reference value. A control cost and health penalty can also be included in the cost. The cost is computed as:

\[cost = w_{forward\_velocity} \times (x_{velocity} - x_{reference\_x\_velocity})^2 + w_{ctrl} \times c_{ctrl} + p_{health}\]
Solved Requirements:

Considered solved when the average cost is less than or equal to 50 over 100 consecutive trials.

How to use:
import stable_gyms
import gymnasium as gym
env = gym.make("stable_gym:HumanoidCost-v1")
state

The current system state.

Type:

numpy.ndarray

dt

The environment step size. Also available as tau.

Type:

float

reference_forward_velocity

The forward velocity that the agent should try to track.

Type:

float

Initialise a new HumanoidCost environment instance.

Parameters:

  • reference_forward_velocity (float, optional) – The forward velocity that the agent should try to track. Defaults to 1.0.

  • randomise_reference_forward_velocity (bool, optional) – Whether to randomize the reference forward velocity. Defaults to False.

  • randomise_reference_forward_velocity_range (tuple, optional) – The range of the random reference forward velocity. Defaults to (0.5, 1.5).

  • forward_velocity_weight (float, optional) – The weight used to scale the forward velocity error. Defaults to 1.0.

  • include_ctrl_cost (bool, optional) – Whether you also want to penalize the humanoid if it takes actions that are too large. Defaults to False.

  • include_health_penalty (bool, optional) – Whether to penalize the humanoid if it becomes unhealthy (i.e. if it falls over). Defaults to True.

  • health_penalty_size (int, optional) – The size of the unhealthy penalty. Defaults to 10.

  • ctrl_cost_weight (float, optional) – The weight used to scale the control cost. Defaults to 1e-4.

  • terminate_when_unhealthy (bool, optional) – Whether to terminate the episode when the humanoid becomes unhealthy. Defaults to True.

  • healthy_z_range (tuple, optional) – The range of healthy z values. Defaults to (1.0, 2.0).

  • reset_noise_scale (float, optional) – Scale of random perturbations of the initial position and velocity. Defaults to 1e-2.

  • exclude_current_positions_from_observation (bool, optional) – Whether to omit the x- and y-coordinates of the front tip from observations. Excluding the position can serve as an inductive bias to induce position-agnostic behaviour in policies. Defaults to True.

  • exclude_reference_from_observation (bool, optional) – Whether the reference should be excluded from the observation. Defaults to False.

  • exclude_reference_error_from_observation (bool, optional) – Whether the error should be excluded from the observation. Defaults to True.

  • exclude_x_velocity_from_observation (bool, optional) – Whether to omit the x- component of the velocity from observations. Defaults to False.

  • action_space_dtype (union[numpy.dtype, str], optional) – The data type of the action space. Defaults to np.float32.

  • observation_space_dtype (union[numpy.dtype, str], optional) – The data type of the observation space. Defaults to np.float64.

  • **kwargs – Extra keyword arguments to pass to the HumanoidEnv class.

reference_forward_velocity
_randomise_reference_forward_velocity
_randomise_reference_forward_velocity_range
_forward_velocity_weight
_include_ctrl_cost
_include_health_penalty
_health_penalty_size
_exclude_reference_from_observation
_exclude_reference_error_from_observation
_exclude_x_velocity_from_observation
_action_space_dtype
_observation_space_dtype
_action_dtype_conversion_warning = False
state = None
low
high
observation_space
cost(x_velocity, ctrl_cost)[source]

Compute the cost of a given x velocity and control cost.

Parameters:

  • x_velocity (float) – The Humanoid’s x velocity.

  • ctrl_cost (float) – The control cost.

Returns:

tuple containing:

  • cost (float): The cost of the action.

  • info (dict): Additional information about the cost.

Return type:

(tuple)

step(action)[source]

Take step into the environment.

Note

This method overrides the step() method such that the new cost function is used.

Parameters:

action (np.ndarray) – Action to take in the environment.

Returns:

tuple containing:

  • obs (np.ndarray): Environment observation.

  • cost (float): Cost of the action.

  • terminated (bool): Whether the episode is terminated.

  • truncated (bool): Whether the episode was truncated. This value is set by wrappers when for example a time limit is reached or the agent goes out of bounds.

  • info (dict): Additional information about the environment.

Return type:

(tuple)

reset(seed=None, options=None)[source]

Reset gymnasium environment.

Parameters:

  • seed (int, optional) – A random seed for the environment. By default None.

  • options (dict, optional) – A dictionary containing additional options for resetting the environment. By default None. Not used in this environment.

Returns:

tuple containing:

  • obs (numpy.ndarray): Initial environment observation.

  • info (dict): Dictionary containing additional information.

Return type:

(tuple)

property tau
Alias for the environment step size. Done for compatibility with the
other gymnasium environments.
property t
Environment time.
property physics_time
Returns the physics time.