Merge pull request #46 from jajimer/log

V1.0.0

doc/source/pages/output.rst

@@ -26,6 +26,6 @@ When a simulation is run, this generate a directory called `Eplus-env-<env_name>
 	- A copy of **socket.cfg** and **utilSocket.idf** which are being used in order to communication interface with Energyplus during simulation.
 	- **monitor.csv**: This records all interactions Agent-Enviroment during the episode timestep by timestep, the format is: *timestep, observation_values, action_values, simulation_time (seconds), reward, done*.
 	- **output/**: This directory has EnergyPlus environment output.
-- **progress.csv**: This file has information about general simulation results. There is a row per episode and it records most important data. Currently, the format is: *episode,mean_reward,cumulative_reward,num_timestep,time_elapsed*.
+- **progress.csv**: This file has information about general simulation results. There is a row per episode and it records most important data. Currently, the format is: *episode_num,cumulative_reward,mean_reward,cumulative_power_consumption,mean_power_consumption,cumulative_comfort_penalty,mean_comfort_penalty,cumulative_power_penalty,mean_power_penalty,comfort_violation (%),length(timesteps),time_elapsed(seconds)*.
 
 .. note:: For more information about EnergyPlus output, visit `EnegyPlus documentation <https://energyplus.net/documentation>`__.

energym/envs/eplus_env.py

@@ -15,9 +15,10 @@
 
 from opyplus import Epm, WeatherData
 
-from ..utils.common import get_current_time_info, parse_variables, create_variable_weather, parse_observation_action_space, CSVLogger
+from ..utils.common import get_current_time_info, parse_variables, create_variable_weather, parse_observation_action_space, setpoints_transform, CSVLogger
 from ..simulators import EnergyPlus
 from ..utils.rewards import SimpleReward
+from pprint import pprint
 
 
 class EplusEnv(gym.Env):
@@ -92,13 +93,24 @@ def __init__(
 
         # Action space
         self.flag_discrete = discrete_actions
+
+        # Discrete
         if self.flag_discrete:
             self.action_mapping = discrete_action_def
             self.action_space = gym.spaces.Discrete(len(discrete_action_def))
+        # Continuous
         else:
+            # Defining action values setpoints (one per value)
+            self.action_setpoints = []
+            for i in range(len(self.variables['action'])):
+                # action_variable --> [low,up]
+                self.action_setpoints.append([
+                    continuous_action_def[0][i], continuous_action_def[1][i]])
+
             self.action_space = gym.spaces.Box(
-                low=np.array(continuous_action_def[0]),
-                high=np.array(continuous_action_def[1]),
+                # continuous_action_def[2] --> shape
+                low=np.repeat(-1, continuous_action_def[2][0]),
+                high=np.repeat(1, continuous_action_def[2][0]),
                 dtype=continuous_action_def[3]
             )
 
@@ -110,12 +122,12 @@ def __init__(
         # Headers for csv loggers
         monitor_header_list = ['timestep,month,day,hour']+self.variables['observation'] + \
             self.variables['action']+['time (seconds)', 'reward',
-                                      'total_power_no_units', 'comfort_penalty', 'done']
+                                      'power_penalty', 'comfort_penalty', 'done']
         self.monitor_header = ''
         for element_header in monitor_header_list:
             self.monitor_header += element_header+','
         self.monitor_header = self.monitor_header[:-1]
-        self.progress_header = 'episode,cumulative_reward,mean_reward,mean_power_consumption,comfort_violation (%),num_timesteps,time_elapsed'
+        self.progress_header = 'episode_num,cumulative_reward,mean_reward,cumulative_power_consumption,mean_power_consumption,cumulative_comfort_penalty,mean_comfort_penalty,cumulative_power_penalty,mean_power_penalty,comfort_violation (%),length(timesteps),time_elapsed(seconds)'
 
         # Create simulation logger, by default is active (flag=True)
         self.logger = CSVLogger(monitor_header=self.monitor_header, progress_header=self.progress_header,
@@ -150,10 +162,12 @@ def step(self, action):
                 else:
                     setpoints = action
             else:
-                print("ERROR: ", action)
+                print("ERROR: ", type(action))
             action_ = list(setpoints)
         else:
-            action_ = list(action)
+            # transform action to setpoints simulation
+            action_ = setpoints_transform(
+                action, self.action_space, self.action_setpoints)
 
         # Send action to the simulator
         self.simulator.logger_main.debug(action_)
@@ -190,7 +204,8 @@ def step(self, action):
             'total_power_no_units': terms['reward_energy'],
             'comfort_penalty': terms['reward_comfort'],
             'temperatures': temp_values,
-            'out_temperature': obs_dict['Site Outdoor Air Drybulb Temperature (Environment)']
+            'out_temperature': obs_dict['Site Outdoor Air Drybulb Temperature (Environment)'],
+            'action_': action_
         }
 
         # Record action and new observation in simulator's csv