(v3.0.4) - Sinergym minor Fixes (#360)

* Controller RBC 5zone: Summer setpoints updated to the default action space

* updated /bin/python3 /workspaces/sinergym/try_env.py

* Updated try_env.py with code comments

* Fixed logger bug

* Added logger warning in documentation

* env.__str__ changed to env.info(), print(env) now print the env type with all wrappers included

* updated Sinergym version from 3.0.3 to 3.0.4

docs/source/pages/output.rst

@@ -86,6 +86,10 @@ present as a wrapper attribute and is called in each timestep and
 in the end of a episode. This class can be substitute by a new one,
 see :ref:`Logger Wrapper personalization/configuration`.
 
+.. warning:: The ``CSVLogger`` requires the info dict with specific keys to 
+             log the information correctly. If you change the info dict structure
+             in Sinergym, you should check this logger or to use a custom one.
+
 .. note:: Normalized observation methods are only used when environment is 
           wrapped with normalization previously.
 

examples/getting_env_information.ipynb

@@ -2595,7 +2595,7 @@
     }
    ],
    "source": [
-    "print(env)"
+    "env.info()"
    ]
   }
  ],

scripts/try_env.py

@@ -2,9 +2,12 @@
 import numpy as np
 
 import sinergym
-from sinergym.utils.wrappers import LoggerWrapper
+from sinergym.utils.wrappers import LoggerWrapper, NormalizeObservation
+from gymnasium.wrappers.normalize import NormalizeReward
 
 env = gym.make('Eplus-demo-v1')
+env = NormalizeObservation(env)
+env = NormalizeReward(env)
 env = LoggerWrapper(env)
 
 for i in range(1):

sinergym/envs/eplus_env.py

@@ -722,8 +722,8 @@ def idd_path(self) -> str:
 
     # -------------------------------- class print ------------------------------- #
 
-    def __str__(self):
-        return """
+    def info(self):
+        print("""
     #==================================================================================#
                                 ENVIRONMENT NAME: {}
     #==================================================================================#
@@ -811,3 +811,4 @@ def __str__(self):
             self.actuator_handlers,
             self.var_handlers,
             self.meter_handlers)
+        )

sinergym/utils/controllers.py

@@ -43,7 +43,7 @@ def __init__(self, env: EplusEnv) -> None:
         self.observation_variables = env.observation_variables
         self.action_varioables = env.action_variables
 
-        self.setpoints_summer = (23, 26.0)
+        self.setpoints_summer = (22.5, 26.0)
         self.setpoints_winter = (20.0, 23.5)
 
     def act(self, observation: List[Any]) -> Sequence[Any]:

sinergym/utils/logger.py

@@ -81,7 +81,7 @@ class CSVLogger(object):
         :param flag: This flag is used to activate (True) or deactivate (False) Logger in real time.
         :param steps_data, rewards, powers, etc: These arrays are used to record steps data to elaborate main data for progress.csv later.
         :param total_timesteps: Current episode timesteps executed.
-        :param total_time_elapsed: Current episode time elapsed (simulation seconds).
+        :param total_time_elapsed: Current episode time elapsed (simulation hours).
         :param comfort_violation_timesteps: Current episode timesteps whose comfort_penalty!=0.
         :param steps_data: It is a array of str's. Each element belong to a step data.
 
@@ -94,6 +94,15 @@ def __init__(
             log_progress_file: str,
             log_file: Optional[str] = None,
             flag: bool = True):
+        """CSVLogger constructor
+
+        Args:
+            monitor_header (str): CSV header for sub_run_N/monitor.csv which record interaction step by step.
+            progress_header (str): CSV header for res_N/progress.csv which record main data episode by episode.
+            log_progress_file (str): log_file path for progress.csv, there will be only one CSV per whole simulation.
+            log_file (Optional[str], optional): log_file path for monitor.csv, there will be one CSV per episode. Defaults to None.
+            flag (bool, optional): Activate (True) or deactivate (False) Logger in real time. Defaults to True.
+        """
 
         self.monitor_header = monitor_header
         self.progress_header = progress_header + '\n'
@@ -157,27 +166,19 @@ def _store_step_information(
 
 
         """
-        # In reset, info this keys are not available
-        if info.get('reward'):
-            self.episode_data['rewards'].append(info.get('reward'))
-        if info.get('abd_energy'):
-            self.episode_data['powers'].append(info.get('abs_energy'))
-        if info.get('comfort_term'):
-            self.episode_data['comfort_penalties'].append(
-                info.get('comfort_term'))
-        if info.get('energy_term') is not None:
-            self.episode_data['power_penalties'].append(
-                info.get('energy_term'))
-        if info.get('abs_comfort') is not None:
-            self.episode_data['abs_comfort'].append(
-                info.get('abs_comfort'))
-        if info.get('abs_energy') is not None:
-            self.episode_data['abs_energy'].append(
-                info.get('abs_energy'))
-        if info.get('comfort_term') != 0:
-            self.episode_data['comfort_violation_timesteps'] += 1
-        self.episode_data['total_timesteps'] = info.get('timestep')
-        self.episode_data['total_time_elapsed'] = info.get('time_elapsed')
+        # In reset (timestep=1), some keys are not available in info
+        if info['timestep'] > 1:
+
+            self.episode_data['rewards'].append(info['reward'])
+            self.episode_data['powers'].append(info['abs_energy'])
+            self.episode_data['comfort_penalties'].append(info['comfort_term'])
+            self.episode_data['power_penalties'].append(info['energy_term'])
+            self.episode_data['abs_comfort'].append(info['abs_comfort'])
+            self.episode_data['abs_energy'].append(info['abs_energy'])
+            if info['comfort_term'] != 0:
+                self.episode_data['comfort_violation_timesteps'] += 1
+            self.episode_data['total_time_elapsed'] = info['time_elapsed(hours)']
+            self.episode_data['total_timesteps'] = info['timestep']
 
     def _reset_logger(self) -> None:
         """Reset relevant data to next episode summary in progress.csv.

sinergym/utils/wrappers.py

@@ -210,7 +210,7 @@ def __init__(
         super(LoggerWrapper, self).__init__(env)
         # Headers for csv logger
         monitor_header_list = monitor_header if monitor_header is not None else ['timestep'] + env.observation_variables + env.action_variables + [
-            'time (hours)', 'reward', 'power_penalty', 'comfort_penalty', 'abs_comfort', 'abs_energy', 'terminated']
+            'time (hours)', 'reward', 'energy_penalty', 'comfort_penalty', 'abs_comfort', 'abs_energy', 'terminated']
         self.monitor_header = ''
         for element_header in monitor_header_list:
             self.monitor_header += element_header + ','
@@ -229,8 +229,8 @@ def __init__(
             'mean_comfort_violation',
             'std_comfort_violation',
             'cumulative_comfort_violation',
-            'length(timesteps)',
-            'time_elapsed(seconds)']
+            'length (timesteps)',
+            'time_elapsed (hours)']
         self.progress_header = ''
         for element_header in progress_header_list:
             self.progress_header += element_header + ','

sinergym/version.txt

@@ -1 +1 @@
-3.0.3
+3.0.4

try_env.py

@@ -2,28 +2,38 @@
 import numpy as np
 
 import sinergym
-from sinergym.utils.wrappers import LoggerWrapper
+from sinergym.utils.wrappers import LoggerWrapper, NormalizeObservation
 
-env = gym.make('Eplus-office-hot-continuous-v1')
+# Creating environment and applying wrappers for normalization and logging
+env = gym.make('Eplus-5zone-hot-continuous-stochastic-v1')
+env = NormalizeObservation(env)
 env = LoggerWrapper(env)
 
-for i in range(1):
+# Execute interactions during 3 episodes
+for i in range(3):
+    # Reset the environment to start a new episode
     obs, info = env.reset()
     rewards = []
     terminated = False
     current_month = 0
     while not terminated:
+        # Random action control
         a = env.action_space.sample()
+        # Read observation and reward
         obs, reward, terminated, truncated, info = env.step(a)
         rewards.append(reward)
+        # If this timestep is a new month start
         if info['month'] != current_month:  # display results every month
             current_month = info['month']
+            # Print information
             print('Reward: ', sum(rewards), info)
+    # Final episode information print
     print(
         'Episode ',
         i,
         'Mean reward: ',
         np.mean(rewards),
         'Cumulative reward: ',
         sum(rewards))
+# Close the environment
 env.close()