Added datasets

README.md

@@ -1,13 +1,23 @@
-# Lab code (WIP)
-This is work in progress. Please do not use them, since they may have many bugs and trial code. We will let you know when it's done.
+# Lab code (WIP), but call for comments
+This is code for labs covered in TensorFlow basic tutorials (in Korean) at https://www.youtube.com/watch?v=BS6O0zOGX4E&list=PLlMkM4tgfjnLSOjrEJN31gZATbcj_MpUm. 
+(We also have a plan to record videos in English.)
 
-## Naming rule:
+This is work in progress, and may have bugs. 
+However, we call for your comments and pull requests. Check out our style guide line:
 
-* klab-XX-X-[name].py: Keras labs
-* lab-XX-X-[name].py: regular tensorflow labs
+* More TF (1.0) style: use more recent and decent TF APIs.
+* More Pythonic: fully leverage the powe of python
+* Readability (over efficiency): Since it's for instruction purposes, we prefer *readability* over others.  
+* Understandability (over everything): Understanding TF key concepts is the main goal of this code.
+
+## File naming rule:
+
+* klab-XX-X-[name].py: Keras labs code
+* lab-XX-X-[name].py: TensorFlow lab code
 
 
 ## Run test and autopep8
+TODO: Need to add more test cases
 
 ```bash
 python -m unittest discover -s tests;

klab-04-3-file_input_linear_regression.py

@@ -1,7 +1,7 @@
 from keras.models import Sequential
 from keras.layers import Dense
 import numpy as np
-xy = np.loadtxt('data.csv', delimiter=',')
+xy = np.loadtxt('data-01-linear.csv', delimiter=',')
 x_data = xy[:, 0:-1]
 y_data = xy[:, [-1]]
 

klab-04-4-stock_linear_regression.py

@@ -3,7 +3,7 @@
 import numpy as np
 from sklearn.preprocessing import MinMaxScaler
 
-xy = np.loadtxt('stock_daily.csv', delimiter=',')
+xy = np.loadtxt('data-02-data-02-stock_daily.csv', delimiter=',')
 
 # very important. It does not work without it.
 scaler = MinMaxScaler(feature_range=(0, 1))

klab-12-3-rnn_prediction.py

@@ -17,7 +17,7 @@
 
 import matplotlib.pyplot as plt
 # Open,High,Low,Close,Volume
-xy = np.loadtxt('stock_daily.csv', delimiter=',')
+xy = np.loadtxt('data-02-data-02-stock_daily.csv', delimiter=',')
 xy = xy[::-1]  # reverse order (chronically ordered)
 
 # very important. It does not work without it.

klab-12-4-rnn_2_prediction.py

@@ -18,7 +18,7 @@
 
 import matplotlib.pyplot as plt
 # Open,High,Low,Close,Volume
-xy = np.loadtxt('stock_daily.csv', delimiter=',')
+xy = np.loadtxt('data-02-data-02-stock_daily.csv', delimiter=',')
 xy = xy[::-1]  # reverse order (chronically ordered)
 
 # very important. It does not work without it.

lab-04-3-file_input_linear_regression.py

@@ -2,7 +2,7 @@
 import tensorflow as tf
 import numpy as np
 
-xy = np.loadtxt('data.csv', delimiter=',', dtype=np.float32)
+xy = np.loadtxt('data-01-linear.csv', delimiter=',', dtype=np.float32)
 x_data = xy[:, 0:-1]
 y_data = xy[:, [-1]]
 

lab-05-1-logistic_regression_classifier.py

@@ -2,9 +2,8 @@
 import tensorflow as tf
 import numpy as np
 
-xy = np.loadtxt('data.csv', delimiter=',')
-x_data = xy[:, 0:-1]
-y_data = xy[:, [-1]]
+x_data = np.array([[1, 2], [2, 3], [3, 1], [4, 3], [5, 3], [6, 2]])
+y_data = np.array([[0], [0], [0], [1], [1], [1]])
 
 print(x_data.shape, x_data, len(x_data))
 print(y_data.shape, y_data)
@@ -14,28 +13,32 @@
 Y = tf.placeholder(tf.float32)
 
 W = tf.Variable(tf.random_uniform(
-    shape=[3, 1], minval=-1.0, maxval=1.0, dtype=tf.float32))
-# Hypothesis
-h = tf.matmul(X, W)
-hypothesis = tf.div(1., 1. + tf.exp(-h))
+    shape=[2, 1], minval=-1.0, maxval=1.0, dtype=tf.float32))
 
-# cost function
+# Hypothesis using sigmoid: tf.div(1., 1. + tf.exp(tf.matmul(X, W)))
+hypothesis = tf.sigmoid(tf.matmul(X, W))
+
+# Cost function
 cost = -tf.reduce_mean(Y * tf.log(hypothesis) + (1 - Y)
                        * tf.log(1 - hypothesis))
 
-# Minimize
-optimizer = tf.train.GradientDescentOptimizer(learning_rate=0.0001)
-train = optimizer.minimize(cost)
+train = tf.train.GradientDescentOptimizer(learning_rate=0.01).minimize(cost)
 
-# Initialize variable
-init = tf.global_variables_initializer()
+# Accuracy computation
+predicted = tf.cast(hypothesis > 0.5, dtype=tf.float32) # True if hypothesis>0.5 else False
+accuracy = tf.reduce_mean(tf.cast(tf.equal(predicted, Y), dtype=tf.float32))
 
 # Launch graph
-sess = tf.Session()
-sess.run(init)
-
-for step in range(2001):
-    sess.run(train, feed_dict={X: x_data, Y: y_data})
-    if step % 20 == 0:
-        print(step, sess.run(cost, feed_dict={
-              X: x_data, Y: y_data}), sess.run(W))
+with tf.Session() as sess:
+    # Initialize tensorflow variables
+    sess.run(tf.global_variables_initializer())
+
+    for step in range(2001):
+        sess.run(train, feed_dict={X: x_data, Y: y_data})
+        if step % 200 == 0:
+            print(step, sess.run(cost, feed_dict={
+                  X: x_data, Y: y_data}), sess.run(W))
+
+    # 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)

lab-06-1-softmax_classifier.py

@@ -11,6 +11,7 @@
 Y = tf.placeholder("float", [None, 3])
 
 W = tf.Variable(tf.zeros([3, 3]))
+
 # Softmax
 hypothesis = tf.nn.softmax(tf.matmul(X, W))
 # tf.nn.softmax computes softmax activations

lab-06-2-softmax_zoo_classifier.py

@@ -0,0 +1,59 @@
+# Lab 6 Softmax Classifier
+import tensorflow as tf
+import numpy as np
+
+x_data = np.array([[1, 2, 1], [1, 3, 2], [1, 3, 4], [1, 5, 5],
+                   [1, 7, 5], [1, 2, 5], [1, 6, 6], [1, 7, 7]], dtype=np.float32)
+y_data = np.array([[0, 0, 1], [0, 0, 1], [0, 0, 1], [0, 1, 0],
+                   [0, 1, 0], [0, 1, 0], [1, 0, 0], [1, 0, 0]], dtype=np.float32)
+
+X = tf.placeholder("float", [None, 3])
+Y = tf.placeholder("float", [None, 3])
+
+W = tf.Variable(tf.zeros([3, 3]))
+
+# Softmax
+hypothesis = tf.nn.softmax(tf.matmul(X, W))
+# tf.nn.softmax computes softmax activations
+# softmax = exp(logits) / reduce_sum(exp(logits), dim)
+
+# Cross entropy cost
+cost = tf.reduce_mean(-tf.reduce_sum(Y *
+                                     tf.log(hypothesis), axis=1))
+
+optimizer = tf.train.GradientDescentOptimizer(
+    learning_rate=0.1).minimize(cost)
+
+init = tf.global_variables_initializer()
+
+# Launch graph
+with tf.Session() as sess:
+    sess.run(init)
+
+    for step in range(2001):
+        sess.run(optimizer, feed_dict={X: x_data, Y: y_data})
+        if step % 200 == 0:
+            print(step, sess.run(cost, feed_dict={
+                  X: x_data, Y: y_data}), sess.run(W))
+
+    print('--------------')
+
+    # Testing & One-hot encoding
+    a = sess.run(hypothesis, feed_dict={X: [[1, 11, 7]]})
+    print(a, sess.run(tf.arg_max(a, 1)))
+
+    print('--------------')
+
+    b = sess.run(hypothesis, feed_dict={X: [[1, 3, 4]]})
+    print(a, sess.run(tf.arg_max(b, 1)))
+
+    print('--------------')
+
+    c = sess.run(hypothesis, feed_dict={X: [[1, 1, 0]]})
+    print(a, sess.run(tf.arg_max(c, 1)))
+
+    print('--------------')
+
+    all = sess.run(hypothesis, feed_dict={
+                   X: [[1, 11, 7], [1, 3, 4], [1, 1, 0]]})
+    print(all, sess.run(tf.arg_max(all, 1)))

lab-09-1-xor.py

@@ -9,38 +9,35 @@
 X = tf.placeholder(tf.float32)
 Y = tf.placeholder(tf.float32)
 
-
 W = tf.Variable(tf.random_uniform(
     shape=[2, 1], minval=-1.0, maxval=1.0, dtype=tf.float32))
 
-# Hypothesis
-h = tf.matmul(X, W)
-hypothesis = tf.div(1., 1. + tf.exp(-h))
+# Hypothesis using sigmoid: tf.div(1., 1. + tf.exp(tf.matmul(X, W)))
+hypothesis = tf.sigmoid(tf.matmul(X, W))
 
 # Cost function
 cost = -tf.reduce_mean(Y * tf.log(hypothesis) + (1 - Y)
                        * tf.log(1 - hypothesis))
 
 train = tf.train.GradientDescentOptimizer(learning_rate=0.01).minimize(cost)
 
+# Accuracy computation
+predicted = tf.cast(hypothesis > 0.5, dtype=tf.float32) # True if hypothesis>0.5 else False
+accuracy = tf.reduce_mean(tf.cast(tf.equal(predicted, Y), dtype=tf.float32))
+
 # Initialize variables
 init = tf.global_variables_initializer()
 
 # Launch graph
 with tf.Session() as sess:
     sess.run(init)
 
-    for step in range(1001):
+    for step in range(5001):
         sess.run(train, feed_dict={X: x_data, Y: y_data})
         if step % 200 == 0:
             print(step, sess.run(cost, feed_dict={
                   X: x_data, Y: y_data}), sess.run(W))
 
-    # Test model
-    correct_prediction = tf.equal(tf.floor(hypothesis + 0.5), Y)
-
-    # Accuracy
-    accuracy = tf.reduce_mean(tf.cast(correct_prediction, dtype=tf.float32))
-    print(sess.run([hypothesis, tf.floor(hypothesis + 0.5),
-                    correct_prediction, accuracy], feed_dict={X: x_data, Y: y_data}))
-    print("Accuracy: ", accuracy.eval({X: x_data, Y: y_data}))
+    # 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)

lab-09-2-xor-nn.py

@@ -0,0 +1,46 @@
+# Lab 9 XOR
+# This example does not work
+import tensorflow as tf
+import numpy as np
+
+x_data = np.array([[0, 0], [0, 1], [1, 0], [1, 1]], dtype=np.float32)
+y_data = np.array([[0], [1], [1], [0]], dtype=np.float32)
+
+X = tf.placeholder(tf.float32)
+Y = tf.placeholder(tf.float32)
+
+
+W = tf.Variable(tf.random_uniform(
+    shape=[2, 1], minval=-1.0, maxval=1.0, dtype=tf.float32))
+
+# Hypothesis
+h = tf.matmul(X, W)
+hypothesis = tf.div(1., 1. + tf.exp(-h))
+
+# Cost function
+cost = -tf.reduce_mean(Y * tf.log(hypothesis) + (1 - Y)
+                       * tf.log(1 - hypothesis))
+
+train = tf.train.GradientDescentOptimizer(learning_rate=0.01).minimize(cost)
+
+# Initialize variables
+init = tf.global_variables_initializer()
+
+# Launch graph
+with tf.Session() as sess:
+    sess.run(init)
+
+    for step in range(1001):
+        sess.run(train, feed_dict={X: x_data, Y: y_data})
+        if step % 200 == 0:
+            print(step, sess.run(cost, feed_dict={
+                  X: x_data, Y: y_data}), sess.run(W))
+
+    # Test model
+    correct_prediction = tf.equal(tf.floor(hypothesis + 0.5), Y)
+
+    # Accuracy
+    accuracy = tf.reduce_mean(tf.cast(correct_prediction, dtype=tf.float32))
+    print(sess.run([hypothesis, tf.floor(hypothesis + 0.5),
+                    correct_prediction, accuracy], feed_dict={X: x_data, Y: y_data}))
+    print("Accuracy: ", accuracy.eval({X: x_data, Y: y_data}))