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

term_next hack actually not necessary

Browse Source
This commit is contained in:
Dominik Moritz Roth 2022-10-14 17:08:12 +02:00
parent 83eb759b9c
commit 61160c8595
1 changed files with 2 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
columbus/env.py
View File

@ -107,7 +107,6 @@ class ColumbusEnv(gym.Env):
self._seed(self.env_seed) self._seed(self.env_seed)
self._init = False self._init = False
self._term_next = False
@property @property
def observation_space(self): def observation_space(self):
@ -218,10 +217,8 @@ class ColumbusEnv(gym.Env):
if self.aux_reward_max: if self.aux_reward_max:
reward += self._get_aux_reward() reward += self._get_aux_reward()
self._steps += 1 self._steps += 1
done = self._term_next or (self.die_on_zero and self.score <= 0) or (self.return_on_score != - done = (self.die_on_zero and self.score <= 0) or (self.return_on_score != -
1 and self.score > self.return_on_score) or self._steps == self.max_steps 1 and self.score > self.return_on_score) or (self._steps == self.max_steps) or (self.terminate_on_reward and gotRew)
# make sure we register the current reward
self._term_next = self.terminate_on_reward and gotRew
info = {'score': self.score, 'reward': reward} info = {'score': self.score, 'reward': reward}
self._rendered = False self._rendered = False
if done: if done:
@ -268,7 +265,6 @@ class ColumbusEnv(gym.Env):
def reset(self): def reset(self):
pygame.init() pygame.init()
self._init = True self._init = True
self._term_next = False
self._steps = 0 self._steps = 0
self._seed(self.env_seed) self._seed(self.env_seed)
self._rendered = False self._rendered = False