sdae_test.py

from sdae import StackedDenoisingAE

from keras.utils.np_utils import to_categorical
from keras.datasets import mnist

from matplotlib import pyplot as plt

n_classes = 10

# input image dimensions
img_rows, img_cols = 28, 28

# the data, shuffled and split between train and test sets
(X_train, y_train), (X_test, y_test) = mnist.load_data()


X_train = X_train.reshape(X_train.shape[0], 784)
X_test = X_test.reshape(X_test.shape[0], 784)
X_train = X_train.astype('float32')
X_test = X_test.astype('float32')
X_train /= 255
X_test /= 255
print(X_train.shape[0], 'train samples')
print(X_test.shape[0], 'test samples')

# convert class vectors to binary class matrices
Y_train = to_categorical(y_train, n_classes)
Y_test = to_categorical(y_test, n_classes)

X_train_50 = X_train[0:50]

n_in = n_out = X_train.shape[1];
n_hid = 576

cur_sdae = StackedDenoisingAE(n_layers = 3, n_hid = [400], dropout = [0.1], nb_epoch = 20)

#train a stacked denoising autoencoder and get the trained model, dense representations of the final hidden layer, and reconstruction error
model, (dense_train, dense_val, dense_test), recon_mse = cur_sdae.get_pretrained_sda(X_train, X_test, X_test, dir_out = '../output/')

X_train_50 *= 255
output_dense = dense_train[0:50,:]
output_dense *= 255

#plotting digits - original and reconstructed
plt.figure(figsize=(10, 10))
for i in range(50):
    ax1 = plt.subplot(10, 10, i * 2 + 1)
    ax1.imshow(X_train_50[i].reshape((28, 28)), interpolation='nearest')
    
    ax2 = plt.subplot(10, 10, i * 2 + 2)
    ax2.imshow(output_dense[i].reshape((20, 20)), interpolation='nearest')
plt.savefig('../output/sdae_test_mnist_dense.png')


# fit_classifier = cur_sdae.supervised_classification(model=model, x_train=X_train, x_val=X_test, y_train=Y_train, y_val=Y_test, n_classes=n_classes)
# pred = cur_sdae.predict(fit_classifier, X_test, 1337)
# score, conf_matrix, error_idx = model_utils.score(y_true, y_pred, y_pred_score, cfg, n_classes)