model.update(chain_d, chain_m, lr_W1, lr_b1, lr_o1)
# Print Norms of the Parameters
print numx.mean(numxExt.get_norms(wl1.weights)), '\t', numx.mean(
numxExt.get_norms(wl2.weights)), '\t',
print numx.mean(numxExt.get_norms(l1.bias)), '\t', numx.mean(
numxExt.get_norms(l2.bias)), '\t',
print numx.mean(numxExt.get_norms(l3.bias)), '\t', numx.mean(
l1.offset), '\t', numx.mean(l2.offset), '\t', numx.mean(l3.offset)
# Show weights
VIS.imshow_matrix(
VIS.tile_matrix_rows(wl1.weights,
v11,
v12,
v21,
v22,
border_size=1,
normalized=False), 'Weights 1')
VIS.imshow_matrix(
VIS.tile_matrix_rows(numx.dot(wl1.weights, wl2.weights),
v11,
v12,
v31,
v32,
border_size=1,
normalized=False), 'Weights 2')
# # Samplesome steps
chain_m = [
numx.float64(numx.random.rand(10 * batch_size, v11 * v12) < 0.5),
corruptor=None,
# Rather strong sparsity regularization
reg_sparseness=2.0,
desired_sparseness=0.001,
reg_contractive=0.0,
reg_slowness=0.0,
data_next=None,
# The gradient restriction is important for fast learning, see also GRBMs
restrict_gradient=0.1,
restriction_norm='Cols')
# Show filters/features
filters = vis.tile_matrix_rows(ae.w,
v1,
v2,
h1,
h2,
border_size=1,
normalized=True)
vis.imshow_matrix(filters, 'Filter')
# Show samples
samples = vis.tile_matrix_rows(train_data[0:100].T,
v1,
v2,
10,
10,
border_size=1,
normalized=True)
vis.imshow_matrix(samples, 'Data samples')
# Load data (download is not existing)
data = io.load_olivetti_faces(path='olivettifaces.mat')
# Specify image width and height for displaying
width = height = 64
# PCA
pca = PCA(input_dim=width * height)
pca.train(data=data)
# Show the first 100 eigenvectors of the covariance matrix
eigenvectors = vis.tile_matrix_rows(matrix=pca.projection_matrix,
tile_width=width,
tile_height=height,
num_tiles_x=10,
num_tiles_y=10,
border_size=1,
normalized=True)
vis.imshow_matrix(
matrix=eigenvectors,
windowtitle='First 100 Eigenvectors of the covariance matrix')
# Show the first 100 images
images = vis.tile_matrix_rows(matrix=data[0:100].T,
tile_width=width,
tile_height=height,
num_tiles_x=10,
num_tiles_y=10,
border_size=1,
normalized=True)
trainer.model, betas=betas)
train_LL = numx.mean(
ESTIMATOR.LL_lower_bound(trainer.model, train_set, logZ))
print("AIS LL: ", 2**(v11 + 1) * train_LL)
logZ = ESTIMATOR.partition_function_exact(trainer.model)
train_LL = numx.mean(ESTIMATOR.LL_exact(trainer.model, train_set,
logZ))
print("True LL: ", 2**(v11 + 1) * train_LL)
print()
# Show weights
VIS.imshow_matrix(
VIS.tile_matrix_rows(dbm.W1,
v11,
v12,
v21,
v22,
border_size=1,
normalized=False), 'Weights 1')
VIS.imshow_matrix(
VIS.tile_matrix_rows(numx.dot(dbm.W1, dbm.W2),
v11,
v12,
v31,
v32,
border_size=1,
normalized=False), 'Weights 2')
VIS.show()
# ZCA projects the whitened data back to the original space, thus does not
# perform a dimensionality reduction but a whitening in the original space
whitened_data = zca.project(data)
# Create a ZCA node and train it (you could also use PCA whitened=True)
ica = ICA(input_dim=width * height)
ica.train(data=whitened_data,
iterations=100,
convergence=1.0,
status=True)
# Show whitened images
images = vis.tile_matrix_rows(matrix=data[0:100].T,
tile_width=width,
tile_height=height,
num_tiles_x=10,
num_tiles_y=10,
border_size=1,
normalized=True)
vis.imshow_matrix(matrix=images,
windowtitle='First 100 image patches')
# Show some whitened images
images = vis.tile_matrix_rows(matrix=whitened_data[0:100].T,
tile_width=width,
tile_height=height,
num_tiles_x=10,
num_tiles_y=10,
border_size=1,
normalized=True)
vis.imshow_matrix(matrix=images,
def generate_samples(rbm,
data,
iterations,
stepsize,
v1,
v2,
sample_states = False,
whitening = None):
''' Generates samples from the given RBM model.
:Parameters:
rbm: RBM model.
-type: RBM model object.
data: Data to sample from.
-type: numpy array [num samples, dimensions]
iterations: Number of Gibbs sampling steps.
-type: int
stepsize: After how many steps a sample should be plotted.
-type: int
v1: X-Axis of the reorder image patch.
-type: int
v2: Y-Axis of the reorder image patch.
-type: int
sample_states: If true returns the sates , probabilities otherwise.
-type: bool
whitening: If the data has been preprocessed it needs to be
undone.
-type: preprocessing object or None
:Returns:
Matrix with image patches order along X-Axis and it's evolution in
Y-Axis.
-type: numpy array
'''
result = data
if whitening != None:
result = whitening.unproject(data)
vis_states = data
for i in xrange(1,iterations+1):
hid_probs = rbm.probability_h_given_v(vis_states)
hid_states = rbm.sample_h(hid_probs)
vis_probs = rbm.probability_v_given_h(hid_states)
vis_states = rbm.sample_v(vis_probs)
if i % stepsize == 0:
if whitening != None:
if sample_states:
result = numx.vstack((result,
whitening.unproject(vis_states)))
else:
result = numx.vstack((result,
whitening.unproject(vis_probs)))
else:
if sample_states:
result = numx.vstack((result,vis_states))
else:
result = numx.vstack((result,vis_probs))
import pydeep.misc.visualization as Vis
return Vis.tile_matrix_rows(result.T, v1,v2,iterations/stepsize+1
,data.shape[0], border_size = 1,
normalized = False)
