From 61e1534751db6338656274184a6ce70241be2a68 Mon Sep 17 00:00:00 2001
From: Sergey Kolpinskiy <skolpinsky@bk.ru>
Date: Sun, 1 Oct 2023 09:41:28 +0300
Subject: [PATCH] clean lab02_lib code

---
 labworks/LW2/lab02_lib.py | 32 +++++++++++++++-----------------
 1 file changed, 15 insertions(+), 17 deletions(-)

diff --git a/labworks/LW2/lab02_lib.py b/labworks/LW2/lab02_lib.py
index ac7ea56..ec90383 100644
--- a/labworks/LW2/lab02_lib.py
+++ b/labworks/LW2/lab02_lib.py
@@ -18,7 +18,7 @@ from matplotlib import pyplot
 from pandas import DataFrame
 import numpy as np
 import matplotlib.pylab as plt
-#import aes_lib as aes
+
 import tensorflow.keras
 import numpy as np
 import math
@@ -29,9 +29,7 @@ from pandas import DataFrame
 from sklearn.metrics import precision_score, recall_score, f1_score, confusion_matrix
 from tensorflow.keras.models import Sequential
 from tensorflow.keras.layers import Dense, Activation
-#from tensorflow.keras.utils import to_categorical
-#from sklearn.metrics import zero_one_loss
-#from sklearn.metrics import confusion_matrix
+
 
 
 visual = True
@@ -104,35 +102,39 @@ def create_fit_save_ae(cl_train, ae_file, irefile, epohs, verbose_show, patience
         # Ниже строки читать входные данные пользователя с помощью функции map ()
         ae_arch = list(map(int, input("Задайте архитектуру скрытых слоёв автокодировщика, например, в виде 3 1 3 : ").strip().split()))[:n]
         ae = tensorflow.keras.models.Sequential()
+        
+        # input layer
         ae.add(tensorflow.keras.layers.Dense(size))
         ae.add(tensorflow.keras.layers.Activation('tanh'))
+        
+        # hidden layers
         for i in range(len(ae_arch)):
             ae.add(tensorflow.keras.layers.Dense(ae_arch[i]))
             ae.add(tensorflow.keras.layers.Activation('tanh'))
+        
+        # output layer
         ae.add(tensorflow.keras.layers.Dense(size))
         ae.add(tensorflow.keras.layers.Activation('linear'))
     else:
         ae = tensorflow.keras.models.Sequential()
+        
+        # input layer
         ae.add(tensorflow.keras.layers.Dense(size))
         ae.add(tensorflow.keras.layers.Activation('tanh'))
+        
+        # hidden layers
         ae.add(tensorflow.keras.layers.Dense(3))
         ae.add(tensorflow.keras.layers.Activation('tanh'))
-        #ae.add(tensorflow.keras.layers.Dense(4))
-        #ae.add(tensorflow.keras.layers.Activation('tanh'))
-        #ae.add(tensorflow.keras.layers.Dense(5))
-        #ae.add(tensorflow.keras.layers.Activation('tanh'))
         ae.add(tensorflow.keras.layers.Dense(2))
         ae.add(tensorflow.keras.layers.Activation('tanh'))
         ae.add(tensorflow.keras.layers.Dense(1))
         ae.add(tensorflow.keras.layers.Activation('tanh'))
         ae.add(tensorflow.keras.layers.Dense(2))
         ae.add(tensorflow.keras.layers.Activation('tanh'))
-        #ae.add(tensorflow.keras.layers.Dense(5))
-        #ae.add(tensorflow.keras.layers.Activation('tanh'))
-        #ae.add(tensorflow.keras.layers.Dense(4))
-        #ae.add(tensorflow.keras.layers.Activation('tanh'))
         ae.add(tensorflow.keras.layers.Dense(3))
         ae.add(tensorflow.keras.layers.Activation('tanh'))
+        
+        # output layer
         ae.add(tensorflow.keras.layers.Dense(size))
         ae.add(tensorflow.keras.layers.Activation('linear'))
 
@@ -176,11 +178,7 @@ def test(y_pred, Y_test):
 def predict_ae(nn, x_test, threshold):
     x_test_predicted = nn.predict(x_test)
     ire = ire_array(x_test, x_test_predicted)
-    # Расчет ошибки при нормализации: иначе закоментировать и раскоментировать 81 и 82
-    #x_test_norm = norm_array(x_test, 0)
-    #x_test_predicted_norm = nn.predict(x_test_norm)
-    #x_test_predicted = norm_array(x_test_predicted_norm, 1)
-    #ire = ire_array(x_test, x_test_predicted)
+
     predicted_labels = (ire > threshold).astype(float)
     predicted_labels = predicted_labels.reshape((predicted_labels.shape[0], 1))
     ire = np.transpose(np.array([ire]))