Removed external dependencies from autoencoder models

Author: joshthoward

autoencoder/AdditiveGaussianNoiseAutoencoderRunner.py

@@ -4,7 +4,7 @@
 import tensorflow as tf
 from tensorflow.examples.tutorials.mnist import input_data
 
-from autoencoder.autoencoder_models.DenoisingAutoencoder import AdditiveGaussianNoiseAutoencoder
+from autoencoder_models.DenoisingAutoencoder import AdditiveGaussianNoiseAutoencoder
 
 mnist = input_data.read_data_sets('MNIST_data', one_hot = True)
 
@@ -45,7 +45,6 @@ def get_random_block_from_data(data, batch_size):
 
     # Display logs per epoch step
     if epoch % display_step == 0:
-        print "Epoch:", '%04d' % (epoch + 1), \
-            "cost=", "{:.9f}".format(avg_cost)
+        print("Epoch:", '%04d' % (epoch + 1), "cost=", "{:.9f}".format(avg_cost))
 
-print "Total cost: " + str(autoencoder.calc_total_cost(X_test))
+print("Total cost: " + str(autoencoder.calc_total_cost(X_test)))

autoencoder/AutoencoderRunner.py

@@ -4,7 +4,7 @@
 import tensorflow as tf
 from tensorflow.examples.tutorials.mnist import input_data
 
-from autoencoder.autoencoder_models.Autoencoder import Autoencoder
+from autoencoder_models.Autoencoder import Autoencoder
 
 mnist = input_data.read_data_sets('MNIST_data', one_hot = True)
 
@@ -44,7 +44,6 @@ def get_random_block_from_data(data, batch_size):
 
     # Display logs per epoch step
     if epoch % display_step == 0:
-        print "Epoch:", '%04d' % (epoch + 1), \
-            "cost=", "{:.9f}".format(avg_cost)
+        print("Epoch:", '%04d' % (epoch + 1), "cost=", "{:.9f}".format(avg_cost))
 
-print "Total cost: " + str(autoencoder.calc_total_cost(X_test))
+print("Total cost: " + str(autoencoder.calc_total_cost(X_test)))

autoencoder/MaskingNoiseAutoencoderRunner.py

@@ -4,7 +4,7 @@
 import tensorflow as tf
 from tensorflow.examples.tutorials.mnist import input_data
 
-from autoencoder.autoencoder_models.DenoisingAutoencoder import MaskingNoiseAutoencoder
+from autoencoder_models.DenoisingAutoencoder import MaskingNoiseAutoencoder
 
 mnist = input_data.read_data_sets('MNIST_data', one_hot = True)
 
@@ -43,7 +43,6 @@ def get_random_block_from_data(data, batch_size):
         avg_cost += cost / n_samples * batch_size
 
     if epoch % display_step == 0:
-        print "Epoch:", '%04d' % (epoch + 1), \
-            "cost=", "{:.9f}".format(avg_cost)
+        print("Epoch:", '%04d' % (epoch + 1), "cost=", "{:.9f}".format(avg_cost))
 
-print "Total cost: " + str(autoencoder.calc_total_cost(X_test))
+print("Total cost: " + str(autoencoder.calc_total_cost(X_test)))

autoencoder/Utils.py

@@ -4,7 +4,7 @@
 import tensorflow as tf
 from tensorflow.examples.tutorials.mnist import input_data
 
-from autoencoder.autoencoder_models.VariationalAutoencoder import VariationalAutoencoder
+from autoencoder_models.VariationalAutoencoder import VariationalAutoencoder
 
 mnist = input_data.read_data_sets('MNIST_data', one_hot = True)
 
@@ -47,7 +47,6 @@ def get_random_block_from_data(data, batch_size):
 
     # Display logs per epoch step
     if epoch % display_step == 0:
-        print "Epoch:", '%04d' % (epoch + 1), \
-            "cost=", "{:.9f}".format(avg_cost)
+        print("Epoch:", '%04d' % (epoch + 1), "cost=", "{:.9f}".format(avg_cost))
 
-print "Total cost: " + str(autoencoder.calc_total_cost(X_test))
+print("Total cost: " + str(autoencoder.calc_total_cost(X_test)))

autoencoder/VariationalAutoencoderRunner.py

@@ -1,6 +1,4 @@
 import tensorflow as tf
-import numpy as np
-import autoencoder.Utils
 
 class Autoencoder(object):
 
@@ -28,7 +26,8 @@ def __init__(self, n_input, n_hidden, transfer_function=tf.nn.softplus, optimize
 
     def _initialize_weights(self):
         all_weights = dict()
-        all_weights['w1'] = tf.Variable(autoencoder.Utils.xavier_init(self.n_input, self.n_hidden))
+        all_weights['w1'] = tf.get_variable("w1", shape=[self.n_input, self.n_hidden],
+            initializer=tf.contrib.layers.xavier_initializer())
         all_weights['b1'] = tf.Variable(tf.zeros([self.n_hidden], dtype=tf.float32))
         all_weights['w2'] = tf.Variable(tf.zeros([self.n_hidden, self.n_input], dtype=tf.float32))
         all_weights['b2'] = tf.Variable(tf.zeros([self.n_input], dtype=tf.float32))
@@ -46,7 +45,7 @@ def transform(self, X):
 
     def generate(self, hidden = None):
         if hidden is None:
-            hidden = np.random.normal(size=self.weights["b1"])
+            hidden = self.sess.run(tf.random_normal([1, self.n_hidden]))
         return self.sess.run(self.reconstruction, feed_dict={self.hidden: hidden})
 
     def reconstruct(self, X):

autoencoder/autoencoder_models/Autoencoder.py

@@ -1,7 +1,4 @@
 import tensorflow as tf
-import numpy as np
-import autoencoder.Utils
-
 
 class AdditiveGaussianNoiseAutoencoder(object):
     def __init__(self, n_input, n_hidden, transfer_function = tf.nn.softplus, optimizer = tf.train.AdamOptimizer(),
@@ -31,7 +28,8 @@ def __init__(self, n_input, n_hidden, transfer_function = tf.nn.softplus, optimi
 
     def _initialize_weights(self):
         all_weights = dict()
-        all_weights['w1'] = tf.Variable(autoencoder.Utils.xavier_init(self.n_input, self.n_hidden))
+        all_weights['w1'] = tf.get_variable("w1", shape=[self.n_input, self.n_hidden],
+            initializer=tf.contrib.layers.xavier_initializer())
         all_weights['b1'] = tf.Variable(tf.zeros([self.n_hidden], dtype = tf.float32))
         all_weights['w2'] = tf.Variable(tf.zeros([self.n_hidden, self.n_input], dtype = tf.float32))
         all_weights['b2'] = tf.Variable(tf.zeros([self.n_input], dtype = tf.float32))
@@ -53,9 +51,9 @@ def transform(self, X):
                                                        self.scale: self.training_scale
                                                        })
 
-    def generate(self, hidden = None):
+    def generate(self, hidden=None):
         if hidden is None:
-            hidden = np.random.normal(size = self.weights["b1"])
+            hidden = self.sess.run(tf.random_normal([1, self.n_hidden]))
         return self.sess.run(self.reconstruction, feed_dict = {self.hidden: hidden})
 
     def reconstruct(self, X):
@@ -98,7 +96,8 @@ def __init__(self, n_input, n_hidden, transfer_function = tf.nn.softplus, optimi
 
     def _initialize_weights(self):
         all_weights = dict()
-        all_weights['w1'] = tf.Variable(autoencoder.Utils.xavier_init(self.n_input, self.n_hidden))
+        all_weights['w1'] = tf.get_variable("w1", shape=[self.n_input, self.n_hidden],
+            initializer=tf.contrib.layers.xavier_initializer())
         all_weights['b1'] = tf.Variable(tf.zeros([self.n_hidden], dtype = tf.float32))
         all_weights['w2'] = tf.Variable(tf.zeros([self.n_hidden, self.n_input], dtype = tf.float32))
         all_weights['b2'] = tf.Variable(tf.zeros([self.n_input], dtype = tf.float32))
@@ -115,9 +114,9 @@ def calc_total_cost(self, X):
     def transform(self, X):
         return self.sess.run(self.hidden, feed_dict = {self.x: X, self.keep_prob: 1.0})
 
-    def generate(self, hidden = None):
+    def generate(self, hidden=None):
         if hidden is None:
-            hidden = np.random.normal(size = self.weights["b1"])
+            hidden = self.sess.run(tf.random_normal([1, self.n_hidden]))
         return self.sess.run(self.reconstruction, feed_dict = {self.hidden: hidden})
 
     def reconstruct(self, X):

autoencoder/autoencoder_models/DenoisingAutoencoder.py

@@ -1,6 +1,5 @@
 import tensorflow as tf
 import numpy as np
-import autoencoder.Utils
 
 class VariationalAutoencoder(object):
 
@@ -36,8 +35,10 @@ def __init__(self, n_input, n_hidden, optimizer = tf.train.AdamOptimizer()):
 
     def _initialize_weights(self):
         all_weights = dict()
-        all_weights['w1'] = tf.Variable(autoencoder.Utils.xavier_init(self.n_input, self.n_hidden))
-        all_weights['log_sigma_w1'] = tf.Variable(autoencoder.Utils.xavier_init(self.n_input, self.n_hidden))
+        all_weights['w1'] = tf.get_variable("w1", shape=[self.n_input, self.n_hidden],
+            initializer=tf.contrib.layers.xavier_initializer())
+        all_weights['log_sigma_w1'] = tf.get_variable("log_sigma_w1", shape=[self.n_input, self.n_hidden],
+            initializer=tf.contrib.layers.xavier_initializer())
         all_weights['b1'] = tf.Variable(tf.zeros([self.n_hidden], dtype=tf.float32))
         all_weights['log_sigma_b1'] = tf.Variable(tf.zeros([self.n_hidden], dtype=tf.float32))
         all_weights['w2'] = tf.Variable(tf.zeros([self.n_hidden, self.n_input], dtype=tf.float32))