visualizers.viser_app.controllers.sampling.predictive_sampling#

Module Contents#

class visualizers.viser_app.controllers.sampling.predictive_sampling.PredictiveSamplingConfig#

Bases: jacta.visualizers.viser_app.controllers.sampling_base.SamplingBaseConfig

Configuration for predictive sampling. .. py:attribute:: sigma

type:

float

value:

0.05

noise_ramp: float = 1.0#
class visualizers.viser_app.controllers.sampling.predictive_sampling.PredictiveSampling(task: jacta.visualizers.viser_app.tasks.task.Task, config: PredictiveSamplingConfig, reward_config: jacta.visualizers.viser_app.tasks.task.TaskConfig)#

Bases: jacta.visualizers.viser_app.controllers.sampling_base.SamplingBase

Predictive sampling planner.

Parameters:

  • config (PredictiveSamplingConfig) – configuration object with hyperparameters for planner.

  • model – mujoco model of system being controlled.

  • data – current configuration data for mujoco model.

  • reward_func – function mapping batches of states/controls to batches of rewards.

  • task (jacta.visualizers.viser_app.tasks.task.Task) –

  • reward_config (jacta.visualizers.viser_app.tasks.task.TaskConfig) –

update_action(curr_state: numpy.ndarray, curr_time: float, additional_info: dict[str, Any]) → None#

Performs rollouts + reward computation from current state.

Parameters:

  • curr_state (numpy.ndarray) –

  • curr_time (float) –

  • additional_info (dict[str, Any]) –

Return type:

None

action(time: float) → numpy.ndarray#

Current best action of policy.

Parameters:

time (float) –

Return type:

numpy.ndarray