gym_ignition_environments.randomizers¶

gym_ignition_environments.randomizers.cartpole¶

class gym_ignition_environments.randomizers.cartpole.CartpoleEnvRandomizer(env, num_physics_rollouts=0)¶

Bases: gym_ignition.randomizers.gazebo_env_randomizer.GazeboEnvRandomizer, gym_ignition_environments.randomizers.cartpole.CartpoleRandomizersMixin

Concrete implementation of cartpole environments randomization.

class gym_ignition_environments.randomizers.cartpole.CartpoleRandomizersMixin(randomize_physics_after_rollouts=0)¶

Bases: gym_ignition.randomizers.abc.TaskRandomizer, gym_ignition.randomizers.abc.PhysicsRandomizer, gym_ignition.randomizers.abc.ModelDescriptionRandomizer, abc.ABC

Mixin that collects the implementation of task, model and physics randomizations for cartpole environments.

get_engine()¶

Return the physics engine to use for the rollout.

Note

Supported physics engines:

scenario.bindings.gazebo.PhysicsEngine_dart

Returns: The desired physics engine to set in the world.

randomize_model_description(task, **kwargs)¶

Randomize the model description.

Parameters: task (Union[CartPoleDiscreteBalancing, CartPoleContinuousSwingup, CartPoleContinuousBalancing]) – The task that operates on the model description to randomize.
Return type: str
Returns: A string with the randomized model description.

randomize_physics(task, **kwargs)¶

Method that insert and configures the physics of a Task’s world.

By default this method loads a plugin that uses DART with no randomizations. Randomizing physics engine parameters or changing physics engine backend could be done by redefining this method and passing it to GazeboRuntime.

Parameters: task (Union[CartPoleDiscreteBalancing, CartPoleContinuousSwingup, CartPoleContinuousBalancing]) – A task containing a world object without physics.
Return type: None

randomize_task(task, **kwargs)¶

Randomize a Task instance.

Parameters: task (Union[CartPoleDiscreteBalancing, CartPoleContinuousSwingup, CartPoleContinuousBalancing]) – the task to randomize.

Note

Note that each task has a world property that provides access to the simulated scenario.bindings.core.World.

Return type: None

gym_ignition_environments.randomizers.cartpole_no_rand¶

class gym_ignition_environments.randomizers.cartpole_no_rand.CartpoleEnvNoRandomizations(env)¶

Bases: gym_ignition.randomizers.gazebo_env_randomizer.GazeboEnvRandomizer

Dummy environment randomizer for cartpole tasks.

Check CartpoleRandomizersMixin for an example that randomizes the task, the physics, and the model.

randomize_task(task, **kwargs)¶

Prepare the scene for cartpole tasks. It simply removes the cartpole of the previous rollout and inserts a new one in the default state. Then, the active Task will reset the state of the cartpole depending on the implemented decision-making logic.

Return type: None