fix bug MG_two_storages due to v0.4.2 and v0.4.3

deer/experiment/base_controllers.py

@@ -551,11 +551,11 @@ def onEpochEnd(self, agent):
     def onEnd(self, agent):
         if (self._active == False):
             return
-
         bestIndex = np.argmax(self._validationScores)
-        print("Best neural net obtained after {} epochs, with validation score {}".format(bestIndex+1, self._validationScores[bestIndex]))
-        if self._testID != None:
-            print("Test score of this neural net: {}".format(self._testScores[bestIndex]))
+        if agent.mode() == -1:
+            print("Best neural net obtained after {} epochs, with validation score {}".format(bestIndex+1, self._validationScores[bestIndex]))
+            if self._testID != None:
+                print("Test score of this neural net: {}".format(self._testScores[bestIndex]))
 
         try:
             os.mkdir("scores")

examples/MG_two_storages/MG_two_storages_env.py

@@ -20,15 +20,15 @@
 #from plot_MG_operation import plot_op
 
 class MyEnv(Environment):
-    VALIDATION_MODE = 0
-    TEST_MODE = 1
-
     def __init__(self, rng, reduce_qty_data=None, length_history=None, start_history=None):
         """ Initialize environment
 
         Arguments:
             rng - the numpy random number generator
         """
+        self.VALIDATION_MODE = 0
+        self.TEST_MODE = 1
+
         reduce_qty_data=int(reduce_qty_data) if reduce_qty_data is not None else int(1)
         length_history=int(length_history) if length_history is not None else int(12)
         start_history=int(start_history) if start_history is not None else int(0)
@@ -148,7 +148,7 @@ def reset(self, mode):
             self.production=self.production_train
             self.consumption_norm=self.consumption_train_norm
             self.consumption=self.consumption_train
-        elif mode == MyEnv.VALIDATION_MODE:
+        elif mode == self.VALIDATION_MODE:
             self.production_norm=self.production_valid_norm
             self.production=self.production_valid
             self.consumption_norm=self.consumption_valid_norm

examples/MG_two_storages/run_MG_two_storages.py

@@ -7,7 +7,7 @@
 import numpy as np
 from joblib import hash, dump, load
 import os
-#import matplotlib.pyplot as plt
+import matplotlib.pyplot as plt
 
 import sys
 from os import path
@@ -158,16 +158,16 @@ class Defaults:
     # used to plot the evolution of the validation and test scores (see below) or simply recover the resulting neural 
     # network for your application.
     agent.attach(bc.FindBestController(
-        validationID=MG_two_storages_env.VALIDATION_MODE, 
-        testID=MG_two_storages_env.TEST_MODE,
+        validationID=env.VALIDATION_MODE,
+        testID=env.TEST_MODE,
         unique_fname=fname))
 
     # All previous controllers control the agent during the epochs it goes through. However, we want to interleave a 
     # "validation epoch" between each training epoch (hence the periodicity=1). For each validation epoch, we want also 
     # to  display the sum of all rewards obtained, hence the showScore=True. Finally, we never want this controller to 
     # call the summarizePerformance method of MG_two_storage_env.
     agent.attach(bc.InterleavedTestEpochController(
-        id=MG_two_storages_env.VALIDATION_MODE, 
+        id=env.VALIDATION_MODE,
         epoch_length=parameters.steps_per_epoch,
         periodicity=1, 
         show_score=True,
@@ -178,7 +178,7 @@ class Defaults:
     # want to display the sum of all rewards obtained, hence the showScore=True. Finally, we want to call the 
     # summarizePerformance method of MG_two_storage_env every [parameters.period_btw_summary_perfs] *test* epochs.
     agent.attach(bc.InterleavedTestEpochController(
-        id=MG_two_storages_env.TEST_MODE,
+        id=env.TEST_MODE,
         epoch_length=parameters.steps_per_test,
         periodicity=1,
         show_score=True,