Refactoring

Author: jatinshah

load_MNIST_images.py renamed to load_MNIST.py

@@ -20,4 +20,23 @@ def load_MNIST_images(filename):
 
         f.close()
 
-        return images
+        return images
+
+
+def load_MNIST_labels(filename):
+    """
+    returns a [number of MNIST images]x1 matrix containing
+    the labels for the MNIST images
+
+    :param filename: input file with labels
+    """
+    with open(filename, 'r') as f:
+        magic = np.fromfile(f, dtype=np.dtype('>i4'), count=1)
+
+        num_labels = np.fromfile(f, dtype=np.dtype('>i4'), count=1)
+
+        labels = np.fromfile(f, dtype=np.ubyte)
+
+        f.close()
+
+        return labels

load_MNIST_labels.py

@@ -1,5 +1,4 @@
-import load_MNIST_images
-import load_MNIST_labels
+import load_MNIST
 import numpy as np
 import softmax
 import compute_gradient
@@ -33,8 +32,8 @@
 # On some platforms, the files might be saved as
 # train-images.idx3-ubyte / train-labels.idx1-ubyte
 
-images = load_MNIST_images.load_MNIST_images('data/mnist/train-images-idx3-ubyte')
-labels = load_MNIST_labels.load_MNIST_labels('data/mnist/train-labels-idx1-ubyte')
+images = load_MNIST.load_MNIST_images('data/mnist/train-images-idx3-ubyte')
+labels = load_MNIST.load_MNIST_labels('data/mnist/train-labels-idx1-ubyte')
 
 if debug:
     input_size = 8 * 8
@@ -93,7 +92,7 @@
 #  (in softmaxPredict.m), which should return predictions
 #  given a softmax model and the input data.
 
-test_images = load_MNIST_images.load_MNIST_images('data/mnist/t10k-images.idx3-ubyte')
-test_labels = load_MNIST_labels.load_MNIST_labels('data/mnist/t10k-labels.idx1-ubyte')
+test_images = load_MNIST.load_MNIST_images('data/mnist/t10k-images.idx3-ubyte')
+test_labels = load_MNIST_labels('data/mnist/t10k-labels.idx1-ubyte')
 predictions = softmax.softmax_predict((opt_theta, input_size, num_classes), test_images)
 print "Accuracy: {0:.2f}%".format(100 * np.sum(predictions == test_labels, dtype=np.float64) / test_labels.shape[0])

softmax_exercise.py

@@ -1,5 +1,4 @@
-import load_MNIST_labels
-import load_MNIST_images
+import load_MNIST
 import numpy as np
 import sparse_autoencoder
 import scipy.optimize
@@ -26,8 +25,8 @@
 #  We have sorted the data for you in this so that you will not have to
 #  change it.
 
-images = load_MNIST_images.load_MNIST_images('data/mnist/train-images-idx3-ubyte')
-labels = load_MNIST_labels.load_MNIST_labels('data/mnist/train-labels-idx1-ubyte')
+images = load_MNIST.load_MNIST_images('data/mnist/train-images-idx3-ubyte')
+labels = load_MNIST.load_MNIST_labels('data/mnist/train-labels-idx1-ubyte')
 
 unlabeled_index = np.argwhere(labels >= 5).flatten()
 labeled_index = np.argwhere(labels < 5).flatten()
@@ -60,7 +59,7 @@
                                                          lambda_, sparsity_param,
                                                          beta, unlabeled_data)
 
-options_ = {'maxiter': 400, 'disp': True}
+options_ = {'maxiter': 100, 'disp': True}
 result = scipy.optimize.minimize(J, theta, method='L-BFGS-B', jac=True, options=options_)
 opt_theta = result.x
 
@@ -86,7 +85,7 @@
 ## STEP 4: Train the softmax classifier
 
 lambda_ = 1e-4
-options_ = {'maxiter': 400, 'disp': True}
+options_ = {'maxiter': 100, 'disp': True}
 
 opt_theta, input_size, num_classes = softmax.softmax_train(hidden_size, num_labels,
                                                            lambda_, train_features,

stl_exercise.py

@@ -6,7 +6,7 @@
 import check_gradient
 import compute_gradient
 import display_network
-import load_MNIST_images
+import load_MNIST
 
 
 ##======================================================================
@@ -41,7 +41,7 @@
 # patches = sample_images.sample_images()
 
 # Loading 10K images from MNIST database
-images = load_MNIST_images.load_MNIST_images('data/mnist/train-images-idx3-ubyte')
+images = load_MNIST.load_MNIST_images('data/mnist/train-images-idx3-ubyte')
 patches = images[:, 0:10000]
 
 #  Obtain random parameters theta