MCPcopy Create free account
hub / github.com/google-deepmind/dm_control / get_observation_components

Method get_observation_components

dm_control/suite/dog.py:415–420  ·  view source on GitHub ↗

Returns the common observations for the Stand task.

(self, physics)

Source from the content-addressed store, hash-verified

413 physics.named.data.qvel['ball_root'][:3] = velocity
414
415 def get_observation_components(self, physics):
416 """Returns the common observations for the Stand task."""
417 obs = super().get_observation_components(physics)
418 obs['ball_state'] = physics.ball_in_head_frame()
419 obs['target_position'] = physics.target_in_head_frame()
420 return obs
421
422 def get_reward_factors(self, physics):
423 """Returns a reward to the agent."""

Callers

nothing calls this directly

Calls 3

ball_in_head_frameMethod · 0.80
target_in_head_frameMethod · 0.80

Tested by

no test coverage detected