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

Method before_step

dm_control/rl/control.py:289–299  ·  view source on GitHub ↗

Updates the task from the provided action. Called by `control.Environment` before stepping the physics engine. Args: action: numpy array or array-like action values, or a nested structure of such arrays. Should conform to the specification returned by `self.action_spe

(self, action, physics)

Source from the content-addressed store, hash-verified

287
288 @abc.abstractmethod
289 def before_step(self, action, physics):
290 """Updates the task from the provided action.
291
292 Called by `control.Environment` before stepping the physics engine.
293
294 Args:
295 action: numpy array or array-like action values, or a nested structure of
296 such arrays. Should conform to the specification returned by
297 `self.action_spec(physics)`.
298 physics: Instance of `Physics`.
299 """
300
301 def after_step(self, physics):
302 """Optional method to update the task after the physics engine has stepped.

Callers 1

stepMethod · 0.45

Calls

no outgoing calls

Tested by

no test coverage detected