dodox/fancy_gym
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

hacky codes, only learn the weight basis params

Browse Source
This commit is contained in:
Hongyi Zhou 2022-10-26 22:46:53 +02:00
parent be6137ec81
commit e49d1563fe
1 changed files with 8 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
fancy_gym/black_box/black_box_wrapper.py
View File

@ -62,7 +62,11 @@ class BlackBoxWrapper(gym.ObservationWrapper):
# spaces
self.return_context_observation = not (learn_sub_trajectories or self.do_replanning)
self.traj_gen_action_space = self._get_traj_gen_action_space()
self.action_space = self._get_action_space()
# self.action_space = self._get_action_space()
tricky_action_upperbound = [np.inf] * (self.traj_gen_action_space.shape[0] - 7)
tricky_action_lowerbound = [-np.inf] * (self.traj_gen_action_space.shape[0] - 7)
self.action_space = spaces.Box(np.array(tricky_action_lowerbound), np.array(tricky_action_upperbound), dtype=np.float32)
self.observation_space = self._get_observation_space()
# rendering
@ -145,6 +149,9 @@ class BlackBoxWrapper(gym.ObservationWrapper):
def step(self, action: np.ndarray):
""" This function generates a trajectory based on a MP and then does the usual loop over reset and step"""
## tricky part, only use weights basis
weights_basis = action.reshape(-1, 7)
# TODO remove this part, right now only needed for beer pong
mp_params, env_spec_params = self.env.episode_callback(action, self.traj_gen)
position, velocity = self.get_trajectory(mp_params)