Exercise answer and style changes (hunkim#163)

* Small typos corrected.

* Exercise answer

* Remove useless variable

* Change some comments, change some style

lab-09-1-xor.py

@@ -1,8 +1,9 @@
 # Lab 9 XOR
-# This example does not work
 import tensorflow as tf
 import numpy as np
+
 tf.set_random_seed(777)  # for reproducibility
+learning_rate = 0.1
 
 x_data = [[0, 0],
           [0, 1],
@@ -28,7 +29,7 @@
 cost = -tf.reduce_mean(Y * tf.log(hypothesis) + (1 - Y) *
                        tf.log(1 - hypothesis))
 
-train = tf.train.GradientDescentOptimizer(learning_rate=0.1).minimize(cost)
+train = tf.train.GradientDescentOptimizer(learning_rate=learning_rate).minimize(cost)
 
 # Accuracy computation
 # True if hypothesis>0.5 else False

lab-09-2-xor-nn-back_prop.py

@@ -0,0 +1,86 @@
+# Lab 9 XOR-back_prop
+import tensorflow as tf
+import numpy as np
+
+tf.set_random_seed(777)  # for reproducibility
+learning_rate = 0.1
+
+x_data = [[0, 0],
+          [0, 1],
+          [1, 0],
+          [1, 1]]
+y_data = [[0],
+          [1],
+          [1],
+          [0]]
+x_data = np.array(x_data, dtype=np.float32)
+y_data = np.array(y_data, dtype=np.float32)
+
+X = tf.placeholder(tf.float32, [None, 2])
+Y = tf.placeholder(tf.float32, [None, 1])
+
+W1 = tf.Variable(tf.random_normal([2, 2]), name='weight1')
+b1 = tf.Variable(tf.random_normal([2]), name='bias1')
+layer1 = tf.sigmoid(tf.matmul(X, W1) + b1)
+
+W2 = tf.Variable(tf.random_normal([2, 1]), name='weight2')
+b2 = tf.Variable(tf.random_normal([1]), name='bias2')
+hypothesis = tf.sigmoid(tf.matmul(layer1, W2) + b2)
+
+# cost/loss function
+cost = -tf.reduce_mean(Y * tf.log(hypothesis) + (1 - Y) *
+                       tf.log(1 - hypothesis))
+
+def sigmoidGradient(z):
+  return tf.multiply(tf.sigmoid(z), (1 - tf.sigmoid(z)))
+
+diff = hypothesis - Y
+
+d_l2 = tf.multiply(diff, sigmoidGradient(tf.matmul(layer1, W2) + b2))
+d_b2 = d_l2
+d_W2 = tf.matmul(tf.transpose(layer1), d_l2)
+
+d_l1 = tf.multiply(tf.matmul(d_l2, tf.transpose(W2)), sigmoidGradient(tf.matmul(X, W1) + b1))
+d_b1 = d_l1
+d_W1 = tf.matmul(tf.transpose(X), d_l1)
+
+step = [
+  tf.assign(W2, W2 - learning_rate * d_W2),
+  tf.assign(b2, b2 - learning_rate * tf.reduce_mean(d_b2, axis=[0])),
+  tf.assign(W1, W1 - learning_rate * d_W1),
+  tf.assign(b1, b1 - learning_rate * tf.reduce_mean(d_b1, axis=[0]))
+]
+
+# Accuracy computation
+# True if hypothesis > 0.5 else False
+predicted = tf.cast(hypothesis > 0.5, dtype=tf.float32)
+accuracy = tf.reduce_mean(tf.cast(tf.equal(predicted, Y), dtype=tf.float32))
+
+# Launch graph
+with tf.Session() as sess:
+    # Initialize TensorFlow variables
+    sess.run(tf.global_variables_initializer())
+
+    for i in range(10001):
+        sess.run([step, cost], feed_dict={X: x_data, Y: y_data})
+        if i % 1000 == 0:
+            print(i, sess.run(cost, feed_dict={
+                  X: x_data, Y: y_data}), sess.run([W1, W2]))
+
+    # Accuracy report
+    h, c, a = sess.run([hypothesis, predicted, accuracy],
+                       feed_dict={X: x_data, Y: y_data})
+    print("\nHypothesis: ", h, "\nCorrect: ", c, "\nAccuracy: ", a)
+
+
+'''
+Hypothesis:  [[ 0.07884014]
+ [ 0.88706875]
+ [ 0.94088489]
+ [ 0.04933683]]
+Correct:  [[ 0.]
+ [ 1.]
+ [ 1.]
+ [ 0.]]
+Accuracy:  1.0
+'''

lab-09-2-xor-nn.py

@@ -1,7 +1,9 @@
 # Lab 9 XOR
 import tensorflow as tf
 import numpy as np
+
 tf.set_random_seed(777)  # for reproducibility
+learning_rate = 0.1
 
 x_data = [[0, 0],
           [0, 1],
@@ -29,7 +31,7 @@
 cost = -tf.reduce_mean(Y * tf.log(hypothesis) + (1 - Y) *
                        tf.log(1 - hypothesis))
 
-train = tf.train.GradientDescentOptimizer(learning_rate=0.1).minimize(cost)
+train = tf.train.GradientDescentOptimizer(learning_rate=learning_rate).minimize(cost)
 
 # Accuracy computation
 # True if hypothesis>0.5 else False

lab-09-3-xor-nn-wide-deep.py

@@ -1,7 +1,9 @@
 # Lab 9 XOR
 import tensorflow as tf
 import numpy as np
+
 tf.set_random_seed(777)  # for reproducibility
+learning_rate = 0.1
 
 x_data = [[0, 0],
           [0, 1],
@@ -37,7 +39,7 @@
 cost = -tf.reduce_mean(Y * tf.log(hypothesis) + (1 - Y) *
                        tf.log(1 - hypothesis))
 
-train = tf.train.GradientDescentOptimizer(learning_rate=0.1).minimize(cost)
+train = tf.train.GradientDescentOptimizer(learning_rate=learning_rate).minimize(cost)
 
 # Accuracy computation
 # True if hypothesis>0.5 else False

lab-09-4-xor_tensorboard.py

@@ -1,7 +1,9 @@
 # Lab 9 XOR
 import tensorflow as tf
 import numpy as np
+
 tf.set_random_seed(777)  # for reproducibility
+learning_rate = 0.01
 
 x_data = [[0, 0],
           [0, 1],
@@ -43,7 +45,7 @@
     cost_summ = tf.summary.scalar("cost", cost)
 
 with tf.name_scope("train") as scope:
-    train = tf.train.GradientDescentOptimizer(learning_rate=0.01).minimize(cost)
+    train = tf.train.GradientDescentOptimizer(learning_rate=learning_rate).minimize(cost)
 
 # Accuracy computation
 # True if hypothesis>0.5 else False

lab-12-1-hello-rnn.py

@@ -20,6 +20,7 @@
 hidden_size = 5  # output from the LSTM. 5 to directly predict one-hot
 batch_size = 1   # one sentence
 sequence_length = 6  # |ihello| == 6
+learning_rate = 0.1
 
 X = tf.placeholder(
     tf.float32, [None, sequence_length, input_dim])  # X one-hot
@@ -45,7 +46,7 @@
 sequence_loss = tf.contrib.seq2seq.sequence_loss(
     logits=outputs, targets=Y, weights=weights)
 loss = tf.reduce_mean(sequence_loss)
-train = tf.train.AdamOptimizer(learning_rate=0.1).minimize(loss)
+train = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(loss)
 
 prediction = tf.argmax(outputs, axis=2)
 

lab-12-2-char-seq-rnn.py

@@ -13,6 +13,7 @@
 num_classes = len(char2idx)  # final output size (RNN or softmax, etc.)
 batch_size = 1  # one sample data, one batch
 sequence_length = len(sample) - 1  # number of lstm rollings (unit #)
+learning_rate = 0.1
 
 sample_idx = [char2idx[c] for c in sample]  # char to index
 x_data = [sample_idx[:-1]]  # X data sample (0 ~ n-1) hello: hell
@@ -39,7 +40,7 @@
 sequence_loss = tf.contrib.seq2seq.sequence_loss(
     logits=outputs, targets=Y, weights=weights)
 loss = tf.reduce_mean(sequence_loss)
-train = tf.train.AdamOptimizer(learning_rate=0.1).minimize(loss)
+train = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(loss)
 
 prediction = tf.argmax(outputs, axis=2)
 

lab-12-3-char-seq-softmax-only.py

@@ -13,6 +13,7 @@
 num_classes = len(char2idx)  # final output size (RNN or softmax, etc.)
 batch_size = 1  # one sample data, one batch
 sequence_length = len(sample) - 1  # number of lstm rollings (unit #)
+learning_rate = 0.1
 
 sample_idx = [char2idx[c] for c in sample]  # char to index
 x_data = [sample_idx[:-1]]  # X data sample (0 ~ n-1) hello: hell
@@ -38,7 +39,7 @@
 sequence_loss = tf.contrib.seq2seq.sequence_loss(
     logits=outputs, targets=Y, weights=weights)
 loss = tf.reduce_mean(sequence_loss)  # mean all sequence loss
-train = tf.train.AdamOptimizer(learning_rate=0.1).minimize(loss)
+train = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(loss)
 
 prediction = tf.argmax(outputs, axis=2)
 

lab-12-4-rnn_long_char.py

@@ -16,6 +16,7 @@
 hidden_size = len(char_set)
 num_classes = len(char_set)
 sequence_length = 10  # Any arbitrary number
+learning_rate = 0.1
 
 dataX = []
 dataY = []
@@ -59,7 +60,7 @@
 sequence_loss = tf.contrib.seq2seq.sequence_loss(
     logits=outputs, targets=Y, weights=weights)
 mean_loss = tf.reduce_mean(sequence_loss)
-train_op = tf.train.AdamOptimizer(learning_rate=0.1).minimize(mean_loss)
+train_op = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(mean_loss)
 
 sess = tf.Session()
 sess.run(tf.global_variables_initializer())

lab-12-5-rnn_stock_prediction.py

@@ -42,7 +42,7 @@ def MinMaxScaler(data):
 
 
 # train Parameters
-timesteps = seq_length = 7
+seq_length = 7
 data_dim = 5
 hidden_dim = 10
 output_dim = 1

lab-13-1-mnist_using_scope.py

@@ -1,4 +1,4 @@
-# Lab 7 Learning rate and Evaluation
+# Lab 13 Using Scope
 import tensorflow as tf
 import random
 

lab-13-2-mnist_tensorboard.py

@@ -1,4 +1,4 @@
-# Lab 7 Learning rate and Evaluation
+# Lab 13 Tensorboard
 import tensorflow as tf
 import random
 
@@ -17,7 +17,6 @@
 
 TB_SUMMARY_DIR = './tb/mnist'
 
-
 # input place holders
 X = tf.placeholder(tf.float32, [None, 784])
 Y = tf.placeholder(tf.float32, [None, 10])