diff --git a/columbus/env.py b/columbus/env.py index 5990075..922a6fb 100644 --- a/columbus/env.py +++ b/columbus/env.py @@ -381,7 +381,7 @@ class ColumbusCandyland(ColumbusEnv): class ColumbusCandyland_Aux10(ColumbusCandyland): def __init__(self, fps=30): super(ColumbusCandyland_Aux10, self).__init__(fps=fps) - self.aux_reward_max = 1 + self.aux_reward_max = 10 class ColumbusEasyObstacles(ColumbusEnv): @@ -389,7 +389,7 @@ class ColumbusEasyObstacles(ColumbusEnv): super(ColumbusEasyObstacles, self).__init__( observable=observable, fps=fps, env_seed=env_seed) self.draw_entities = not hide_map - self.aux_reward_max = 1 + self.aux_reward_max = 10 def setup(self): self.agent.pos = self.start_pos @@ -412,7 +412,7 @@ class ColumbusEasierObstacles(ColumbusEnv): super(ColumbusEasierObstacles, self).__init__( observable=observable, fps=fps, env_seed=env_seed) self.draw_entities = not hide_map - self.aux_reward_max = 1 + self.aux_reward_max = 10 def setup(self): self.agent.pos = self.start_pos @@ -435,7 +435,7 @@ class ColumbusJustState(ColumbusEnv): def __init__(self, observable=observables.StateObservable(), fps=30, num_enemies=0, num_rewards=1, env_seed=None): super(ColumbusJustState, self).__init__( observable=observable, fps=fps) - self.aux_reward_max = 1 + self.aux_reward_max = 10 self.num_enemies = num_enemies self.num_rewards = num_rewards @@ -455,7 +455,7 @@ class ColumbusStateWithBarriers(ColumbusEnv): def __init__(self, observable=observables.StateObservable(coordsAgent=True, speedAgent=False, coordsRelativeToAgent=False, coordsRewards=True, rewardsWhitelist=None, coordsEnemys=True, enemysWhitelist=None, enemysNoBarriers=True, rewardsTimeouts=False, include_rand=True), fps=30, env_seed=3.141, num_enemys=0, num_barriers=3): super(ColumbusStateWithBarriers, self).__init__( observable=observable, fps=fps, env_seed=env_seed) - self.aux_reward_max = 1 + self.aux_reward_max = 10 self.start_pos = (0.5, 0.5) self.num_barriers = num_barriers self.num_enemys = num_enemys