minor updates

Author: qlan3

components/replay.py

@@ -43,8 +43,8 @@ class FiniteReplay(object):
   '''
   def __init__(self, memory_size, keys=None):
     if keys is None:
-      keys = []
-    self.keys = keys + ['action', 'reward', 'mask']
+      keys = ['action', 'reward', 'mask']
+    self.keys = keys
     self.memory_size = int(memory_size)
     self.clear()
 

envs/env.py

@@ -2,20 +2,24 @@
 import gym_pygame
 import gym_minatar
 import gym_exploration
-try:
-  import pybullet
-  import pybullet_envs
-except ImportError:
-  pass
 from gym.wrappers.time_limit import TimeLimit
 
 from envs.wrapper import *
 
 
 def make_env(env_name, max_episode_steps, episode_life=True):
-  env = gym.make(env_name)
-  env_group_title = get_env_group_title(env)
-  # print(env_group_title, env_name)
+  if 'DMC' in env_name:
+    import dmc2gym
+    domain, task, _ = env_name.split('-')  # reacher-hard-DMC
+    env = dmc2gym.make(domain_name=domain, task_name=task)
+    env_group_title = 'dmc'
+  else:
+    if 'BulletEnv' in env_name:
+      import pybullet
+      import pybullet_envs
+    env = gym.make(env_name)
+    env_group_title = get_env_group_title(env)
+  
   if env_group_title == 'gym_minatar':
     env = make_minatar(env, max_episode_steps, scale=False)
     if len(env.observation_space.shape) == 3:
@@ -33,7 +37,7 @@ def make_env(env_name, max_episode_steps, episode_life=True):
     if len(env.observation_space.shape) == 3:
       env = TransposeImage(env)
     env = FrameStack(env, 4)
-  elif env_group_title in ['classic_control', 'box2d', 'gym_pygame', 'gym_exploration', 'pybullet', 'mujoco', 'robotics']:
+  elif env_group_title in ['classic_control', 'box2d', 'gym_pygame', 'gym_exploration', 'pybullet', 'mujoco', 'robotics', 'dmc']:
     if max_episode_steps > 0: # Set max episode steps
       env = TimeLimit(env.unwrapped, max_episode_steps)
   return env

run.py

@@ -18,7 +18,7 @@ def main(argv):
     # Job time
     'time': '0-10:00:00',
     # GPU/CPU type
-    '--cpus-per-task': 1,
+    'cpus-per-task': 1,
     # Memory
     'mem-per-cpu': '2000M',
     # Email address

utils/logger.py

@@ -7,7 +7,7 @@ class Logger(object):
   def __init__(self, logs_dir, file_name='log.txt', filemode='w'):
     logging.basicConfig(
       format='%(asctime)s - %(levelname)s: %(message)s',
-      filename=f'{logs_dir}/{file_name}',
+      filename=f'{logs_dir}{file_name}',
       filemode=filemode
     )
     logger = logging.getLogger()

utils/plotter.py

@@ -72,7 +72,7 @@ def merge_index(self, config_idx, mode, processed, exp=None):
         result_list[i] = result_list[i][:n]
         result_list[i].loc[:, self.x_label] = new_x
         result_list[i].loc[:, self.y_label] = new_y
-    else:
+    elif processed == False:
       # Cut off redundant results
       n = min(len(result) for result in result_list)
       for i in range(len(result_list)):
@@ -381,7 +381,7 @@ def find_key_value(config_dict, key):
       return config_dict[k]
     elif type(v) == dict:
       value = find_key_value(v, key)
-      if value is not '/':
+      if value != '/':
         return value
   return '/'
 

utils/submitter.py

@@ -49,4 +49,4 @@ def submit(self):
           if len(self.job_list) == 0:
             print("Finish submitting all jobs!")
             exit(1)
-      time.sleep(self.check-time-interval)
+      time.sleep(self.check_time_interval)