New results et surtout travail sur les graphes

code/wilshire_5000/nn.py

@@ -14,8 +14,8 @@ def sinus_cosinus(x):
 def swish(x):
     return(x*tf.math.sigmoid(x))
 
-#activations = [tf.keras.activations.relu,swish,sinus_cosinus,sinus,snake]
-activations = [snake]
+activations = [tf.keras.activations.relu,swish,sinus_cosinus,sinus,snake]
+#activations = [snake]
 models = []
 errors_train,errors_test = [],[]
 mean_y_train,mean_y_test,std_y_test=[],[],[]
@@ -42,7 +42,7 @@ for activation in activations :
     y_test_5=[]
     errors_train_5=[]
     errors_test_5=[]
-    for k in range(1):
+    for k in range(2):
 
         model =  tf.keras.Sequential()
 
@@ -73,7 +73,7 @@ for activation in activations :
     # y_preds_test.append(y_pred_test)
 
 
-x = np.arange(df_train.shape[0]+df_test.shape[0]+908)
+x = np.arange(9000)
 x = x / maximum
 future_preds = model.predict(x)  ## Calculated with a website the number of working days between 01-06-2020 and 01-01-2024
     
@@ -85,7 +85,7 @@ def plot_total(x_train,y_train,y_pred_train,x_test,y_test,y_pred_test):
     plt.figure()
     plt.plot(x,y_true,label="True data")
     plt.plot(x,y_pred,label="Predictions")
-    plt.vlines([5240,5326])
+    plt.vlines([6545,6629])
     plt.legend()
     plt.show()
 
@@ -99,6 +99,9 @@ x_cut = np.arange(df_train.shape[0]+df_test.shape[0])
 plt.figure()
 plt.plot(x_cut,y_true,label="True data")
 plt.plot(x,future_preds,label="Predictions")
-plt.vlines([5240,5326],ymin=0,ymax=1)
+plt.xticks(range(0, 9000, 250), range(1995, 2030, 1))
+plt.xlabel("Années")
+plt.ylabel("Index Willshire5000 normalisé")
+plt.vlines([6545,6629],ymin=0,ymax=1)
 plt.legend()
 plt.show()

code/wilshire_5000/results.txt

@@ -0,0 +1,8 @@
+100 epochs adam mse 2_runs 4_batch_size
+
+[[0.5401026606559753, 0.0], [0.5401026606559753, 0.0], [0.18415303528308868, 0.009042516350746155], 
+[0.1577790528535843, 0.0004640519618988037], [0.007161237532272935, 0.0007062254007905722]]
+
+
+[[0.0681375540792942, 0.11859790716458475], [0.0113261666148901, 0.002322317928070604], [nan, nan], 
+[0.008264125511050224, 0.0013573866481290681], [0.008175850845873356, 0.001503577316727166]]