update evaluate

evaluate.py

@@ -161,18 +161,21 @@ def eval_from_file(path):
     valence_p = []
     arousal_t = []
     arousal_p = []
+    ratings = [[], [], [], [], [], [], [], []]
 
     with open(path, 'r') as csvfile:
         reader = csv.reader(csvfile, delimiter=',')
         for row in reader:
             if row[0] == '' or row[0] == 'id':
                 continue
-            true_l.append(int(row[2]))
-            pred_l.append(int(row[5]))
-            probs = []
-            for i in range(6, 13):
-                probs.append(float(row[i]))
-            pred_r.append(probs)
+
+            if int(row[1]) != 1 and int(row[1]) != 3:
+                true_l.append(int(row[2]))
+                pred_l.append(int(row[5]))
+                probs = []
+                for i in range(6, 13):
+                    probs.append(float(row[i]))
+                pred_r.append(probs)
 
             if row[15] == ' 0.00.0' or row[15] == '':
                 valence_t.append(0)
@@ -185,6 +188,8 @@ def eval_from_file(path):
             else:
                 arousal_t.append(float(row[16]))
             arousal_p.append(float(row[4]))
+
+            ratings[int(row[2])].append(int(float(row[14])))
     true_r = to_categorical(true_l, num_classes=8)
 
     print('ACC'.ljust(20), ACC(true_l, pred_l))
@@ -193,12 +198,19 @@ def eval_from_file(path):
     print('ALPHA'.ljust(20), ALPHA(true_l, pred_l))
     print('AUCPR'.ljust(20), AUCPR(true_r, pred_r))
     print('AUC'.ljust(20), AUC(true_r, pred_r))
+    print(confusion_matrix(true_l, pred_l))
     print('')
     print(''.ljust(20), 'VALENCE'.ljust(20), 'AROUSAL')
     print('RMSE'.ljust(20), str(RMSE(valence_t, valence_p)).ljust(20), RMSE(arousal_t, arousal_p))
     print('CORR'.ljust(20), str(CORR(valence_t, valence_p)).ljust(20), CORR(arousal_t, arousal_p))
     print('SAGR'.ljust(20), str(SAGR(valence_t, valence_p)).ljust(20), SAGR(arousal_t, arousal_p))
     print('CCC'.ljust(20), str(CCC(valence_t, valence_p)).ljust(20), CCC(arousal_t, arousal_p))
+    print('')
+    t = (0, 0)
+    for i, r in enumerate(ratings):
+        print(str(i).ljust(5), np.mean(r))
+        t = (t[0] + np.sum(r), t[1] + len(r))
+    print('Total', t[0] / t[1])
 
 
 if __name__ == '__main__':