def compress(self, X):
n = X.shape[0]
# nearest_neighbours = np.zeros((n, self.nn))
# Compute Euclidean distances
D = utils.euclidean_dist_squared(X, X)
D = np.sqrt(D)
# If two points are disconnected (distance is Inf)
# then set their distance to the maximum
# distance in the graph, to encourage them to be far apart.
adjacency_matrix = np.zeros((n, n))
nearest_neighbours = self.knn(X)
for i, j in enumerate(nearest_neighbours):
for neighbour in j:
adjacency_matrix[i, neighbour] = D[i, neighbour]
adjacency_matrix[neighbour, i] = D[neighbour, i]
dijkstra = utils.dijkstra(adjacency_matrix)
dijkstra[np.isinf(dijkstra)] = dijkstra[~np.isinf(dijkstra)].max()
# Initialize low-dimensional representation with PCA
Z = PCA(self.k).fit(X).compress(X)
# Solve for the minimizer
z = find_min(self._fun_obj_z, Z.flatten(), 500, False, dijkstra)
Z = z.reshape(n, self.k)
return Z
def hullselect(self):
def selectHullPoints(data, n=20):
""" select data points for pairwise projections of the first n
dimensions """
# iterate over all projections and select data points
idx = np.array([])
# iterate over some pairwise combinations of dimensions
for i in combinations(range(n), 2):
# sample convex hull points in 2D projection
convex_hull_d = quickhull(data[i, :].T)
# get indices for convex hull data points
idx = np.append(idx, dist.vq(data[i, :], convex_hull_d.T))
idx = np.unique(idx)
return np.int32(idx)
# determine convex hull data points only if the total
# amount of available data is >50
#if self.data.shape[1] > 50:
pcamodel = PCA(self.data, show_progress=self._show_progress)
pcamodel.factorize()
idx = selectHullPoints(pcamodel.H, n=self._base_sel)
# set the number of subsampled data
self.nsub = len(idx)
return idx
def __init__(self, img_dataset: np.ndarray):
"""
:param img_dataset: An image dataset, which is a matrix with the shape of (N x H x W), where:
- N: number of images
- H: height of images
- W: width of images
- each item of the matrix is an real value between 0 and 1
Notes: All images should have same width and height
"""
# Get the shape of the input data
super().__init__()
assert len(img_dataset.shape) == 3
self._n_samples, self._height, self._width = img_dataset.shape
self.logger.info({
'msg': 'Image dataset shape',
'shape': img_dataset.shape
})
self._n_features = self._height * self._width
# Flatten the images of shape (height, width) to vectors of length height x width
self._flatten_dataset = img_dataset.reshape(
(self._n_samples, self._n_features))
# Build the PCA transformer
self._pca_transformer = PCA(self._flatten_dataset)
def compress(self, X):
n = X.shape[0]
# Compute Euclidean distances
D = utils.euclidean_dist_squared(X, X)
D = np.sqrt(D)
# Construct nearest neighbour graph
G = np.zeros([n, n])
for i in range(n):
neighbours = np.argsort(D[i])[:self.nn + 1]
for j in neighbours:
G[i, j] = D[i, j]
G[j, i] = D[j, i]
# Compute ISOMAP distances
D = utils.dijkstra(G)
# If two points are disconnected (distance is Inf)
# then set their distance to the maximum
# distance in the graph, to encourage them to be far apart.
D[np.isinf(D)] = D[~np.isinf(D)].max()
# Initialize low-dimensional representation with PCA
pca = PCA(self.k)
pca.fit(X)
Z = pca.compress(X)
# Solve for the minimizer
z, f = findMin(self._fun_obj_z, Z.flatten(), 500, D)
Z = z.reshape(n, self.k)
return Z
def compress(self, X):
n = X.shape[0]
# Compute Euclidean distances
D = utils.euclidean_dist_squared(X,X)
D = np.sqrt(D)
#TODO:
D = self.construct_dist_graph(X , D)
# If two points are disconnected (distance is Inf)
# then set their distance to the maximum
# distance in the graph, to encourage them to be far apart.
D[np.isinf(D)] = D[~np.isinf(D)].max()
# Initialize low-dimensional representation with PCA
pca = PCA(self.k)
pca.fit(X)
Z = pca.compress(X)
# Solve for the minimizer
z,f = findMin(self._fun_obj_z, Z.flatten(), 500, D)
Z = z.reshape(n, self.k)
return Z
def toyExample():
mat = scipy.io.loadmat('../data/toy_data.mat')
data = mat['toy_data']
# TODO: Train PCA
pca = PCA(-1)
pca.train(data)
print("Variance of the data")
# TODO 1.2: Compute data variance to the S vector computed by the PCA
data_variance = np.var(data, axis=1)
print(data_variance)
print(np.power(pca.S, 2) / data.shape[1])
# TODO 1.3: Compute data variance for the projected data (into 1D) to the S vector computed by the PCA
Xout = pca.project(data, 1)
print("Variance of the projected data")
data_variance = np.var(Xout, axis=1)
print(data_variance)
print(np.power(pca.S[0], 2) / data.shape[1])
plt.figure()
plt.title('PCA plot')
plt.subplot(1, 2, 1) # Visualize given data and principal components
# TODO 1.1: Plot original data (hint, use the plot_pca function
pca.plot_pca(data)
plt.subplot(1, 2, 2)
# TODO 1.3: Plot data projected into 1 dimension
pca.S[1] = 0
pca.plot_pca(Xout)
plt.show()
def faceLoader() -> None:
'''
Face loader and visualizer example code
'''
gall = importGallery()
print(gall.shape)
gall = gall[:, :10]
print(gall.shape)
# Show first image
plt.figure(0)
plt.title('First face')
n = 0
nComponents = 10
pca = PCA(nComponents)
face = gall[:, :1]
print(face.shape)
# face = face.reshape(24576,1)
# print(face.shape)
mu, U, C, data = pca.train(gall)
alpha = pca.to_pca(data)
# print(alpha.shape)
# faceId = gall.item(n)[0][0]
# print('Face got face id: {}'.format(faceId))
face = alpha[:, :1]
print(face.shape)
face = face.reshape(192,128)
plt.imshow(face, cmap='gray')
plt.show()
def compress(self, X):
n = X.shape[0]
# Compute Euclidean distances
D = utils.euclidean_dist_squared(X, X)
D = np.sqrt(D)
# D is symmetric matrix
geoD = np.zeros((n, n))
# find nn-neighbours
for i in range(n):
sort = np.argsort(D[:, i])
neigh = np.setdiff1d(sort[0:self.nn + 1], i)
# find the nn+1 smallest indexes that are not i
for j in range(len(neigh)):
t = neigh[j]
geoD[i, t] = D[i, t]
geoD[t, i] = D[t, i]
D = utils.dijkstra(geoD)
# for disconnected vertices (distance is Inf)
# set their dist = max_dist(graph)
# to encourage they are far away from each other
D[np.isinf(D)] = D[~np.isinf(D)].max()
# Initialize low-dimensional representation with PCA
pca = PCA(self.k)
pca.fit(X)
Z = pca.compress(X)
# Solve for the minimizer
z, f = findMin(self._fun_obj_z, Z.flatten(), 500, D)
Z = z.reshape(n, self.k)
return Z
def compress(self, X):
n = X.shape[0]
# Compute Euclidean distances
D = utils.euclidean_dist_squared(X, X)
D = np.sqrt(D)
sorted_indices = np.argsort(D)
G = np.zeros((n, n))
for i in range(D.shape[0]):
for j in range(self.nn + 1):
G[i, sorted_indices[i, j]] = D[i, sorted_indices[i, j]]
G[sorted_indices[i, j], i] = D[sorted_indices[i, j], i]
dist = utils.dijkstra(G)
dist[np.isinf(dist)] = dist[~np.isinf(dist)].max()
# Initialize low-dimensional representation with PCA
pca = PCA(self.k)
pca.fit(X)
Z = pca.compress(X)
# Solve for the minimizer
z, f = findMin(self._fun_obj_z, Z.flatten(), 500, dist)
Z = z.reshape(n, self.k)
return Z
def hullselect(self):
def selectHullPoints(data, n=20):
""" select data points for pairwise projections of the first n
dimensions """
# iterate over all projections and select data points
idx = np.array([])
# iterate over some pairwise combinations of dimensions
for i in combinations(range(n), 2):
# sample convex hull points in 2D projection
convex_hull_d = quickhull(data[i, :].T)
# get indices for convex hull data points
idx = np.append(idx, dist.vq(data[i, :], convex_hull_d.T))
idx = np.unique(idx)
return np.int32(idx)
# determine convex hull data points only if the total
# amount of available data is >50
#if self.data.shape[1] > 50:
pcamodel = PCA(self.data, show_progress=self._show_progress)
pcamodel.factorize()
idx = selectHullPoints(pcamodel.H, n=self._base_sel)
# set the number of subsampled data
self.nsub = len(idx)
return idx
def pic_handle(img_data, sd, ori_size):
"""
做PCA处理之后，在还原到原来的维度，然后显示，之后输出信噪比
"""
Pca = PCA(sd, img_data)
c_data, w_star = Pca.pca() # 进行pca降维,获取投影矩阵
w_star = np.real(w_star)
print(w_star)
new_data = w_star * w_star.T * c_data + Pca.mean # 还原到原来的维度
total_img = []
# 图片混合
for i in range(Pca.data_size):
if len(total_img) == 0:
total_img = new_data[:, i].T.reshape(ori_size)
else:
total_img = np.hstack(
[total_img, new_data[:, i].T.reshape(ori_size)])
# 计算信噪比
print('信噪比:')
for i in range(Pca.data_size):
a = psnr(np.array(data[:, i].T), np.array(new_data[:, i].T))
print('图', i, '的信噪比为:', a, 'dB')
# 处理图片
total_img = np.array(total_img).astype(np.uint8)
cv2.imwrite('pca image.jpg', total_img) # 图片显示
cv2.imshow('pca image', total_img)
cv2.waitKey(0)
def plot_iris(y, y_classes, maxit=25, *args, **kwargs):
# np.random.seed(0)
fig, ax = plot_grid(5)
#Variational bayes
vbpca = VBPCA(y, *args, **kwargs)
for i in range(maxit):
vbpca.update()
plot_scatter(vbpca.transform(), y_classes, ax[0])
ax[0].set_title('VBPCA')
#Laplace approximation
lbpca = LBPCA(y.T)
lbpca.fit(maxit)
plot_scatter(lbpca.transform(2).T, y_classes, ax[1])
ax[1].set_title('LBPCA')
#Streaming LBPCA
stream = create_distributed(np.copy(y.T), 10)
stream.randomized_fit(1)
plot_scatter(stream.transform(y.T, 2).T, y_classes, ax[2])
ax[2].set_title('Batch BPCA')
#Distributed LBPCA
stream = create_distributed(np.copy(y.T), 10)
stream.averaged_fit(maxit)
plot_scatter(stream.transform(y.T, 2).T, y_classes, ax[3])
ax[3].set_title('Parallel BPCA')
#PCA
pca = PCA(y.T)
plot_scatter(pca.fit_transform().T, y_classes, ax[4])
ax[4].set_title('PCA')
plt.show()
def compress(self, X):
n = X.shape[0]
k = self.k
K = self.K
# Compute Euclidean distances
D = utils.euclidean_dist_squared(X, X)
D = np.sqrt(D)
nbrs = np.argsort(D, axis=1)[:, 1:K + 1]
G = np.zeros((n, n))
for i in range(n):
for j in nbrs[i]:
G[i, j] = D[i, j]
G[j, i] = D[j, i]
D = utils.dijkstra(G)
D[D == np.inf] = -np.inf
max = np.max(D)
D[D == -np.inf] = max
# Initialize low-dimensional representation with PCA
Z = PCA(k).fit(X).compress(X)
# Solve for the minimizer
z = find_min(self._fun_obj_z, Z.flatten(), 500, False, D)
Z = z.reshape(n, k)
return Z
def getCentVec(self, contextVecs):
sample, rank, dim = contextVecs.shape
contexts = np.reshape(contextVecs, (sample * rank, dim))
pca = PCA(n_components=1)
pca.fit(contexts)
return pca.components_[0]
def main():
# Load dataset
data = datasets.load_iris()
X = normalize(data.data[data.target != 0])
y = data.target[data.target != 0]
y[y == 1] = 0
y[y == 2] = 1
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.33,
seed=1)
clf = LogisticRegression(gradient_descent=True)
clf.fit(X_train, y_train)
y_pred = clf.predict(X_test)
accuracy = accuracy_score(y_test, y_pred)
print("Accuracy:", accuracy)
# Reduce dimension to two using PCA and plot the results
pca = PCA()
pca.plot_in_2d(X_test,
y_pred,
title="Logistic Regression",
accuracy=accuracy)
def testePCA(self):
pca = PCA()
matrizX = Matrizx()
idModelo = self.txtIdModelo.get()
matrizPrincipal = matrizX.selectMatrizXModeloMMM(idModelo)
pca.testePCA(matrizPrincipal)
def test_pca(self):
X = np.array([[-1, -1], [-2, -1], [-3, -2], [1, 1], [2, 1], [3, 2]])
pca = PCA(n_comp=2)
pca.fit(X)
self.assertEqual(
np.allclose(pca.explained_variance,
np.array([0.9924, 0.0075]),
atol=1e-3), True)
def train_classifiers(all_training_idx, samples_training, color_channels):
logging.debug("Training classifiers..")
classifiers = []
for channel in color_channels:
classifier = PCA(samples_training)
classifier.train(channel[all_training_idx])
classifiers.append(classifier)
return classifiers
def test_transform_inverse_transform(self):
X = self.data(500)
pca = PCA(X)
x = pca.transform(X, ndims=3)
self.assertTrue(
np.allclose(np.cov(x.T), np.eye(3), atol=1e-4, rtol=1e-2))
X_ = pca.inverse_transform(x)
self.assertTrue(np.allclose(X, X_))
def train_models():
images, labels, labels_dic = collect_dat_set()
rec_eig = PCA(500, 5)
if images:
rec_eig.train(images, labels)
return rec_eig, labels_dic
def generate_pca_embedding_files():
'''
Generate PCA embedding csv files for the experiments.
'''
raw = genfromtxt('digits-raw.csv', delimiter=',')
X = raw[:, 2:]
pca = PCA(10)
X_new = pca.fit_transform(X)
raw_new = hstack((raw[:, :2], X_new))
savetxt('digits-pca-embedding.csv', raw_new, delimiter=',')
def main():
random.seed(1)
img_dim = 15 # 10, 15, ...
datasets = loadUpsonData('../data/upson_rovio_1/train_%d_50000.pkl.gz' % img_dim,
'../data/upson_rovio_1/test_%d_50000.pkl.gz' % img_dim)
print 'done loading.'
train_set_x_data, train_set_y = datasets[0]
pca = PCA(train_set_x_data)
print 'done PCA.'
image = Image.fromarray(tile_raster_images(
X = train_set_x_data,
img_shape = (img_dim,img_dim),tile_shape = (10,10),
tile_spacing=(1,1)))
image.save(os.path.join(resman.rundir, 'samplesData.png'))
pyplot.figure()
pyplot.subplot(221); pyplot.semilogy(pca.var); pyplot.title('pca.var')
pyplot.subplot(222); pyplot.semilogy(pca.std); pyplot.title('pca.std')
pyplot.subplot(223); pyplot.semilogy(pca.fracVar); pyplot.title('pca.fracVar')
pyplot.subplot(224); pyplot.semilogy(pca.fracStd); pyplot.title('pca.fracStd')
pyplot.savefig(os.path.join(resman.rundir, 'varstd.png'))
pyplot.close()
#font = ImageFont.truetype('/usr/share/fonts/truetype/ttf-lyx/cmr10.ttf', 10)
font = ImageFont.truetype('/usr/share/texmf/fonts/opentype/public/lm/lmmono12-regular.otf', 14)
def plotImage(xx, filename, str = None):
arr = tile_raster_images(X = xx,
img_shape = (img_dim,img_dim),tile_shape = (10,10),
tile_spacing=(1,1))
arrHeight = arr.shape[0]
if str is not None:
arr = vstack((arr, zeros((20, arr.shape[1]), dtype = arr.dtype)))
image = Image.fromarray(arr)
if str is not None:
draw = ImageDraw.Draw(image)
draw.text((2, arrHeight+2), str, 255, font = font)
draw = ImageDraw.Draw(image)
image.save(os.path.join(resman.rundir, filename))
plotImage(pca.pc().T, 'pc.png')
for dims in [1, 2, 5, 10, 20, 50, 100, 200, 225]:
plotImage(pca.pcaAndBack(train_set_x_data, dims),
'samplesPCA_%03d.png' % dims,
'dims=%d' % dims)
for ee, epsilon in enumerate([0, 1e-4, 1e-3, 1e-2, 1e-1, 1, 2, 5]):
plotImage(pca.toZca(train_set_x_data, dims, epsilon = epsilon),
'samplesZCA_%03d_%02d.png' % (dims, ee),
'dims=%d, eps=%s' % (dims, repr(epsilon)))
def main():
whiten = False
if len(sys.argv) > 1 and sys.argv[1] == '--whiten':
whiten = True
del sys.argv[1]
if len(sys.argv) <= 3:
print 'Usage: %s pcaDims n_hidden learningRate' % sys.argv[0]
sys.exit(1)
# loads data like datasets = ((train_x, train_y), ([], None), (test_x, None))
datasets = loadUpsonData('../data/upson_rovio_1/train_15_50000.pkl.gz',
'../data/upson_rovio_1/test_15_50000.pkl.gz')
img_dim = 15 # must match actual size of training data
print 'done loading.'
pcaDims = int(sys.argv[1])
pca = PCA(datasets[0][0]) # train
datasets[0][0] = pca.toPC(datasets[0][0], pcaDims, whiten = whiten) # train
datasets[1][0] = pca.toPC(datasets[1][0], pcaDims, whiten = whiten) if len(datasets[1][0]) > 0 else array([]) # valid
datasets[2][0] = pca.toPC(datasets[2][0], pcaDims, whiten = whiten) # test
print 'reduced by PCA to'
print ('(%d, %d, %d) %d dimensional examples in (train, valid, test)' %
(datasets[0][0].shape[0], datasets[1][0].shape[0], datasets[2][0].shape[0], datasets[0][0].shape[1]))
# plot mean and principle components
image = Image.fromarray(tile_raster_images(
X = pca.meanAndPc(pcaDims).T,
img_shape = (img_dim,img_dim),tile_shape = (10,10),
tile_spacing=(1,1)))
image.save(os.path.join(resman.rundir, 'meanAndPc.png'))
# plot fractional stddev in PCA dimensions
pyplot.semilogy(pca.fracStd, 'bo-')
if pcaDims is not None:
pyplot.axvline(pcaDims)
pyplot.savefig(os.path.join(resman.rundir, 'fracStd.png'))
pyplot.clf()
test_rbm(datasets = datasets,
training_epochs = 45,
img_dim = img_dim,
n_input = pcaDims if pcaDims else img_dim * img_dim,
n_hidden = int(sys.argv[2]),
learning_rate = float(sys.argv[3]),
output_dir = resman.rundir,
quickHack = False,
visibleModel = 'real',
initWfactor = .01,
imgPlotFunction = lambda xx: pca.fromPC(xx, unwhiten = whiten))
def toyExample() -> None:
## Toy Data Set
mat = scipy.io.loadmat('../data/toy_data.mat')
data = mat['toy_data']
data = importGallery()
## limit datafor testing purposes
data = data[:, :144].T
print(data.shape)
## Iris dataset. Just for testing purposes
#iris = datasets.load_iris()
#data = iris['data'].astype(np.float32) # a 150x4 matrix with features
#data = data.T
# TODO: Train PCA
nComponents = 25
pca = PCA(nComponents)
## 1.1 Calculate PCA manuelly. SVD is following
#pca.pca_manuel(data)
## 1.2 Calculate PCA via SVD
mu, U, C, dataCenter = pca.train(data)
## 2. Transform RAW data using first n principal components
alpha = pca.to_pca(dataCenter)
## 3. Backtransform alpha to Raw data
Xout = pca.from_pca(alpha)
print("Variance")
# TODO 1.2: Compute data variance to the eigenvalue vector computed by the PCA
print(f'Total Variance: {np.var(data)}')
print(f'Eigenvalues: {C} \n')
# TODO 1.3: Compute data variance for the projected data (into 1D) to the S vector computed by the PCA
print(f'Total Variance Transform: {np.var(alpha)}')
print(f'Mean Eigenvalues: {np.mean(C)}')
## Plot only if fewer than 2 components
if nComponents == 2:
plt.figure()
plt.title('PCA plot')
plt.subplot(1, 2, 1) # Visualize given data and principal components
# TODO 1.1: Plot original data (hint, use the plot_pca function
pca.plot_pca(data)
plt.subplot(1, 2, 2)
# TODO 1.3: Plot data projected into 1 dimension
pca.plot_pca(Xout)
plt.show()
## Plot variances
else:
x = np.arange(1, len(C) + 1)
plt.bar(x, C)
plt.show()
def main():
data = datasets.load_iris()
X = normalize(data.data)
y = data.target
X_train, X_test, y_train, y_test = train_test_split(X, y, 0.3)
knn = KNN(3)
y_pred = knn.predict(X_test, X_train, y_train)
accuracy = accuracy_score(y_pred, y_test)
print("accuracy is ", accuracy)
pca = PCA()
pca.plot_in_2d(X_test, y_pred, title="knn", accuracy=accuracy)
def main(matrix, pcomps):
pca = PCA(pcomps, matrix)
# get covariance matrix - saved as instance var
pca.covariance_matrix()
# find evecs and evals of covariance matrix
pca.evecs_and_evals()
# get top x principal components aka evecs corresponding to top evals
pca.principal_components()
# reduce image on those components
return pca.get_evec_matrix()
def getCxtSubspace(wl, dim, var_threshold=0.45):
emb = []
for word in wl:
if (word not in vecDict):
print "non-exist:", word
continue
wordEmbed = dict[word]
emb.append(wordEmbed)
emb = np.array(emb)
pca = PCA()
pca.fit(emb)
varList = pca.explained_variance_ratio_
cand = 0
varSum = 0
for var in varList:
varSum += var
cand += 1
if (varSum >= var_threshold):
break
pca = PCA(n_components=cand)
pca.fit(emb)
top_embed = pca.components_
print "dim:", len(top_embed.tolist()), cand
return top_embed.tolist()
def Bonus3():
'''
Scatter plot of samples projected onto the first
two eigenvectors.
'''
raw = genfromtxt('digits-raw.csv', delimiter=',')
X = raw[:, 2:]
pca = PCA(2)
X_new = pca.fit_transform(X)
perm = permutation(X.shape[0])[:1000]
labels = array(raw[perm, 1], dtype=int)
colors = rand(10, 3)[labels, :]
scatter(X_new[perm, 0], X_new[perm, 1], c=colors, alpha=0.9, s=10)
show()
def compress(self, X):
n = X.shape[0]
k = self.k
# Compute Euclidean distances
D = utils.euclidean_dist_squared(X, X)
D = np.sqrt(D)
# Initialize low-dimensional representation with PCA
Z = PCA(k).fit(X).compress(X)
# Solve for the minimizer
z = find_min(self._fun_obj_z, Z.flatten(), 500, False, D)
Z = z.reshape(n, k)
return Z
def hack_pca(filename, threshold=0.6):
'''
Input: filename -- input image file name/path
Output: img -- image without rotation
'''
img_r = (plt.imread(filename)).astype(np.float64) / 255
# YOUR CODE HERE
img = img_r[:, :, 0] * 0.299 + img_r[:, :, 1] * 0.587 + img_r[:, :,
2] * 0.114
H, W = img.shape
data = []
for i in range(H):
# x axis
for j in range(W):
# y axis
if img[i, j] >= threshold:
data.append([i, j])
data = np.array(data)
N = data.shape[0]
eigvectors, eigvalues = PCA(data)
(vx, vy) = eigvectors[:, 0]
(vx, vy) = (vx, vy) if vy >= 0 else (-vx, -vy)
theta = -math.asin(-vx) * 180 / math.pi
R = np.array([[vy, vx], [-vx, vy]]) # rotate matrix
odata = np.matmul(data, R)
odata -= np.min(odata, axis=0)
odata = odata.astype(int)
nH, nW = np.max(odata, axis=0)
oimg = np.zeros((nH + 1, nW + 1))
for i in range(N):
oimg[odata[i, 0], odata[i, 1]] = 1.
return img, oimg, theta
def hack_pca(filename):
'''
Input: filename -- input image file name/path
Output: img -- image without rotation
'''
img_r = (plt.imread(filename)).astype(np.float64) # 4 channels: R,G,B,A
img_gray = img_r[:, :, 0] * 0.3 + img_r[:, :, 1] * 0.59 + img_r[:, :,
2] * 0.11
X_int = np.array(np.where(img_gray > 0))
X = X_int.astype(np.float64)
D, N = X.shape
eigen_vec, eigen_val = PCA(X)
print(eigen_vec, eigen_val)
Y = np.matmul(X.T, eigen_vec).T
Y_int = Y.astype(np.int32)
dmin = np.min(Y_int, axis=1).reshape(D, 1)
Y_int = Y_int - dmin
bound = np.max(Y_int, axis=1) + 1
new_img = np.zeros(bound)
for t in range(Y_int.shape[1]):
new_img[tuple(Y_int[:, t])] = img_gray[tuple(X_int[:, t])]
new_img = new_img.T[::-1, ::-1]
return new_img
def hack_pca(filename):
'''
Input: filename -- input image file name/path
Output: img -- image without rotation
'''
img_r = (plt.imread(filename)).astype(np.float64) / 255
img_r = rgb2gray(img_r)
plt.imshow(img_r, cmap='gray')
plt.show()
m, n = img_r.shape
xy = []
xyv = []
for i in range(m):
for j in range(n):
if img_r[i, j] > 0:
xy.append((i, j))
xyv.append((i, j, img_r[i, j]))
xy = np.array(xy)
vector, value = PCA(xy)
d = np.array(np.round(np.dot(xy, vector))).astype(np.int)
min_xy = np.min(d, axis=0)
d -= min_xy
max_xy = np.max(d, axis=0)
img = np.zeros((max_xy[1] + 1, max_xy[0] + 1))
for i in range(xy.shape[0]):
img[max_xy[1] - d[i, 1], max_xy[0] - d[i, 0]] = xyv[i][2]
plt.imshow(img, cmap='gray')
plt.show()
return img
def main():
# reduce the dimensionality of the data to two dimension and plot the results.
data = datasets.load_digits()
X = data.data
y = data.target
# Project the data onto the 2 primary principal components
X_trans = PCA().transform(X, 2)
x1 = X_trans[:, 0]
x2 = X_trans[:, 1]
cmap = plt.get_cmap('viridis')
colors = [cmap(i) for i in np.linspace(0, 1, len(np.unique(y)))]
class_distr = []
# Plot the different class distributions
for i, l in enumerate(np.unique(y)):
_x1 = x1[y == l]
_x2 = x2[y == l]
_y = y[y == l]
class_distr.append(plt.scatter(_x1, _x2, color=colors[i]))
plt.legend(class_distr, y, loc=1)
plt.suptitle("PCA Dimensionality Reduction")
plt.title("Digit Dataset")
plt.xlabel('Principal Component 1')
plt.ylabel('Principal Component 2')
plt.show()
def pcaSenEmb(sent_vecs, var_threshold=0.6):
"""
output: basis of context space
"""
pca = PCA()
pca.fit(sent_vecs)
var_list = pca.explained_variance_ratio_
cand = 0
var_sum = 0
for var in var_list:
var_sum += var
cand += 1
if (var_sum >= var_threshold):
break
basis = pca.components_
return basis
class PCANNClassifier(object):
def __init__(self, trials, trial_angle, **kwargs):
if "pca" in kwargs and not kwargs["pca"]:
self.pca = None;
else:
# Prepare PCA only on training data:
self.pca = PCA(np.vstack(trials), None, 0.95);
tr = [];
tr_a = [];
for t,a in zip(trials, trial_angle):
# Project it onto PCA space
p = t;
if self.pca:
p = self.pca.proj(t);
for i in range(len(p)):
tr.append(p[i,:]);
tr_a.append(a);
# Prepare NN classifier
self.nn = NearestNeighbors(n_neighbors=19)
self.nn.fit(np.vstack(tr));
self.class_labels = tr_a;
def classify(self, trials):
rv = [];
for tr in trials:
p = tr;
if self.pca:
p = self.pca.proj(tr);
p = np.array(p);
votes = defaultdict(lambda: 0);
for i in range(len(p)):
dist, idx = self.nn.kneighbors(p[i,:]);
for j in idx[0,:]:
votes[self.class_labels[j]] += 1;
max_a, max_c = max(votes.iteritems(), key = lambda x: x[1]);
rv.append(max_a);
return rv;
def __init__(self, trials, bin_width, trial_angles, trial_moves):
self.trials = trials
self.bin_width = bin_width
self.trial_angles = trial_angles
self.trial_moves = trial_moves
self.pca = PCA(np.vstack(trials), 3);
fig = plt.figure()
self.ax = fig.add_subplot(111, projection='3d')
self.ax.set_xlabel('PCA 1')
self.ax.set_ylabel('PCA 2')
self.ax.set_zlabel('PCA 3')
def update_w(self):
""" compute new W """
def select_hull_points(data, n=3):
""" select data points for pairwise projections of the first n
dimensions """
# iterate over all projections and select data points
idx = np.array([])
# iterate over some pairwise combinations of dimensions
for i in combinations(range(n), 2):
# sample convex hull points in 2D projection
convex_hull_d = quickhull(data[i, :].T)
# get indices for convex hull data points
idx = np.append(idx, vq(data[i, :], convex_hull_d.T))
idx = np.unique(idx)
return np.int32(idx)
# determine convex hull data points using either PCA or random
# projections
method = 'randomprojection'
if method == 'pca':
pcamodel = PCA(self.data)
pcamodel.factorize(show_progress=False)
proj = pcamodel.H
else:
R = np.random.randn(self._base_sel, self._data_dimension)
proj = np.dot(R, self.data)
self._hull_idx = select_hull_points(proj, n=self._base_sel)
aa_mdl = AA(self.data[:, self._hull_idx], num_bases=self._num_bases)
# determine W
aa_mdl.factorize(niter=50, compute_h=True, compute_w=True,
compute_err=True, show_progress=False)
self.W = aa_mdl.W
self._map_w_to_data()
def __analyze(self, img):
_data = PCA.load_image(img)
data = self.__featrue.extract(_data)
min = 65536
l = None
_l = 97
for _d in self.__crops:
_m = Analyzer.__compare(_d, data)
if _m < min:
min = _m
l = chr(_l)
_l+=1
return l
class View3D(object):
def __init__(self, trials, bin_width, trial_angles, trial_moves):
self.trials = trials
self.bin_width = bin_width
self.trial_angles = trial_angles
self.trial_moves = trial_moves
self.pca = PCA(np.vstack(trials), 3);
fig = plt.figure()
self.ax = fig.add_subplot(111, projection='3d')
self.ax.set_xlabel('PCA 1')
self.ax.set_ylabel('PCA 2')
self.ax.set_zlabel('PCA 3')
def scatter_plot(self):
for angle, trial in zip(self.trial_angles, self.trials):
projection = self.pca.proj_whiten(trial)
projection = np.array(projection)
xs = np.array(projection[0])
ys = np.array(projection[1])
zs = np.array(projection[2])
self.ax.scatter(xs, ys, zs, c = COLOR_MAP[angle])
plt.show()
def line_plot(self):
for angle, trial in zip(self.trial_angles, self.trials):
projection = self.pca.proj_whiten(trial)
projection = np.array(projection)
xs = np.array(projection[0])
ys = np.array(projection[1])
zs = np.array(projection[2])
self.ax.plot(xs, ys, zs, c = COLOR_MAP[angle])
plt.show()
def __init__(self, trials, trial_angle, **kwargs):
if "pca" in kwargs and not kwargs["pca"]:
self.pca = None;
else:
# Prepare PCA only on training data:
self.pca = PCA(np.vstack(trials), None, 0.95);
tr = [];
tr_a = [];
for t,a in zip(trials, trial_angle):
# Project it onto PCA space
p = t;
if self.pca:
p = self.pca.proj(t);
for i in range(len(p)):
tr.append(p[i,:]);
tr_a.append(a);
# Prepare NN classifier
self.nn = NearestNeighbors(n_neighbors=19)
self.nn.fit(np.vstack(tr));
self.class_labels = tr_a;
def _choose_rule_pca(self, data, show_progress):
"""Our projections are onto the PCA vectors"""
pca = PCA(data, num_bases=1)
pca.factorize(show_progress)
primary_vec = pca.W.reshape(self._data_dimension)
return self._choose_rule_vecproject(data, primary_vec)
imlist = glob.glob("data/a_thumbs/*.jpg")
#print imlist
im = array(Image.open(imlist[0]))
#imshow(im)
#show()
m, n = im.shape[0:2]
imageCount = len(imlist)
#create matrix to stor all flattened images
imageMatrix = array([array(Image.open(im)).flatten()
for im in imlist], 'f')
#perform pca
V, S, immean = PCA.pca(imageMatrix)
#show some images (mean and 7 first modes
figure()
gray()
subplot(2,4,1)
imshow(immean.reshape(m,n))
for i in range(7):
subplot(2,4,i+2)
#convert images back to one dimension
imshow(V[i].reshape(m,n))
#imshow(imageMatrix[i].reshape(m,n))
#show()
f = open('font_pca_modes.pkl', 'wb')
from ResultsManager import resman
from pca import PCA
if __name__ == '__main__':
resman.start('junk', diary = True)
datasets = loadUpsonData('../data/upson_rovio_1/train_15_50000.pkl.gz',
'../data/upson_rovio_1/test_15_50000.pkl.gz')
#meanTrain = mean(datasets[0][0])
#stdTrain = std(datasets[0][0])
#datasets[0][0] = (datasets[0][0] - meanTrain) / stdTrain
#datasets[2][0] = (datasets[2][0] - meanTrain) / stdTrain
pca = PCA(datasets[0][0])
datasets[0][0] = pca.toZca(datasets[0][0], None, epsilon = .1)
datasets[2][0] = pca.toZca(datasets[2][0], None, epsilon = .1)
print 'done loading.'
test_rbm(datasets = datasets,
training_epochs = 45,
img_dim = 15, # must match actual size of training data
n_hidden = int(sys.argv[1]),
learning_rate = float(sys.argv[2]),
output_dir = resman.rundir,
quickHack = False,
visibleModel = 'real',
initWfactor = .01,
pcaDims = None)
def reduceModel(model, atoms, selstr):
"""Return reduced NMA model.
Reduces a :class:`NMA` model to a subset of *atoms* matching a selection
*selstr*. This function behaves differently depending on the type of the
*model* argument. For ANM and GNM or other NMA models, this functions
derives the force constant matrix for system of interest (specified by the
*selstr*) from the force constant matrix for the *model* by assuming that
for any given displacement of the system of interest, the other atoms move
along in such a way as to minimize the potential energy. This is based on
the formulation in in [KH00]_. For PCA models, this function simply takes
the sub-covariance matrix for the selected atoms.
:arg model: dynamics model
:type model: :class:`ANM`, :class:`GNM`, or :class:`PCA`
:arg atoms: atoms that were used to build the model
:arg selstr: a selection string specifying subset of atoms"""
linalg = importLA()
if not isinstance(model, NMA):
raise TypeError("model must be an NMA instance, not {0:s}".format(type(model)))
if not isinstance(atoms, (AtomGroup, AtomSubset, AtomMap)):
raise TypeError("atoms type is not valid")
if len(atoms) <= 1:
raise TypeError("atoms must contain more than 1 atoms")
if isinstance(model, GNM):
matrix = model._kirchhoff
elif isinstance(model, ANM):
matrix = model._hessian
elif isinstance(model, PCA):
matrix = model._cov
else:
raise TypeError("model does not have a valid type derived from NMA")
if matrix is None:
raise ValueError("model matrix (Hessian/Kirchhoff/Covariance) is not " "built")
system = SELECT.getBoolArray(atoms, selstr)
other = np.invert(system)
n_sel = sum(system)
if n_sel == 0:
LOGGER.warning("selection has 0 atoms")
return None
if len(atoms) == n_sel:
LOGGER.warning("selection results in same number of atoms, " "model is not reduced")
return None
if model.is3d():
system = np.tile(system, (3, 1)).transpose().flatten()
other = np.tile(other, (3, 1)).transpose().flatten()
ss = matrix[system, :][:, system]
if isinstance(model, PCA):
eda = PCA(model.getTitle() + " reduced")
eda.setCovariance(ss)
return eda, system
so = matrix[system, :][:, other]
os = matrix[other, :][:, system]
oo = matrix[other, :][:, other]
matrix = ss - np.dot(so, np.dot(linalg.inv(oo), os))
if isinstance(model, GNM):
gnm = GNM(model.getTitle() + " reduced")
gnm.setKirchhoff(matrix)
return gnm, system
elif isinstance(model, ANM):
anm = ANM(model.getTitle() + " reduced")
anm.setHessian(matrix)
return anm, system
elif isinstance(model, PCA):
eda = PCA(model.getTitle() + " reduced")
eda.setCovariance(matrix)
return eda, system
def main(self, algo="KNN", textview=None):
# Remplace "print"
def print_output(text):
if textview != None:
buf = textview.get_buffer()
buf.insert_at_cursor(text + "\n")
textview.scroll_mark_onscreen(buf.get_insert())
else:
log.info(text)
# liste des types de set
if self.validation == 1:
listeTypesSet = ["train", "validation", "test"]
else:
listeTypesSet = ["train", "test"]
# liste des resultats utilises pour les courbes
listeRes=[]
# creation des trainFile et testFile
log.debug("Construction des fichiers d'entrainement")
tools.constructLfwNamesCurrent( self.nbExemples )
#TODO ca ne sert plus a rien finalement
( nbClassesLFW, nbClassesORL ) = tools.trainAndTestConstruction( self.pourcentageTrain, self.nbExemples )
# Chargement des données
dataTrain, dataTrainIndices, nClass = tools.loadImageData( "train", self.categorie)
# tranformation pca
print_output("Calcul des vecteurs propres...")
pca_model = PCA( dataTrain )
pca_model.transform() # on transforme les donné dans un le "eigen space"
##### Recherche pas KNN
if algo == "KNN":
print_output("Début de l'algorithme des K plus proches voisins...")
# On build le model pour recherche par KNN
knn_model = KNN( pca_model.getWeightsVectors(), dataTrainIndices, nClass, self.K )
# On build le model pour Parzen
parzen_model = ParzenWindows( pca_model.getWeightsVectors(), dataTrainIndices, nClass, self.Theta )
## TEST ###########################
#TODO Toute cette partie est a revoir pour sortir des graphes
# de train, validation, test
for trainTest in listeTypesSet:
if trainTest == "train":
dataTest, dataTestIndices = dataTrain, dataTrainIndices
else :
### si l'on n'effectue pas de validation on concatene les entrees de test et de validation initiales pour obtenir le test
#if "validation" not in listeTypesSet:
#dataTestInitial, dataTestInitialIndices, nClass = tools.loadImageData( "test", self.categorie )
#dataValidation, dataValidationIndices, nClass = tools.loadImageData( "validation", self.categorie )
#dataTest = np.zeros(dataTestInitial.size + dataValidation.size)
#dataTestIndices = np.zeros( dataTest.size )
#dataTest[ : dataTestInitial.size], dataTestIndices[ : dataTestInitial.size] = dataTestInitial, dataTestInitialIndices
#dataTest[dataTestInitial.size : ], dataTestIndices[dataTestInitial.size : ] = dataValidation, dataValidationIndices
#else:
dataTest, dataTestIndices, nClass = tools.loadImageData( trainTest, self.categorie )
print_output("Projection des données de test...")
dataTest_proj = pca_model.getProjection( dataTest )
# compteurs de bons résultats
nbGoodResult = 0
nbGoodResult2 = 0
nbGoodResult3 = 0
t_start = time.clock()
for i in range(0, int( dataTest.shape[1] )):
# k = 1, pour réference
# on force k
knn_model.setK( 1 )
result1NN = knn_model.compute_predictions( dataTest_proj[:,i] )
if(result1NN == dataTestIndices[i]):
nbGoodResult += 1
# k = n
# replace k a ca position initial
knn_model.setK( self.K )
resultKNN = knn_model.compute_predictions( dataTest_proj[:,i] )
if(resultKNN == dataTestIndices[i]):
nbGoodResult2 += 1
resultParzen = parzen_model.compute_predictions( dataTest_proj[:,i] )
if(resultParzen == dataTestIndices[i]):
nbGoodResult3 += 1
out_str = "Classic method: "+ str( result1NN ) +" | KNN method: "+ str( resultKNN ) +" | KNN+Parzen method: "+ str( resultParzen ) +" | Expected: "+ str( dataTestIndices[i] ) +"\n" # +1 car l'index de la matrice commence a 0
print_output(out_str)
resClassic = (float(nbGoodResult) / float(dataTest.shape[1])) * 100.
out_str = "\nAccuracy with classic method: %.3f" % resClassic + "%\n"
resKNN = (nbGoodResult2 / float(dataTest.shape[1])) * 100.
out_str += "Accuracy with KNN method (k="+ str( self.K ) +"): %.3f" % resKNN + "%\n"
res = (nbGoodResult3 / float(dataTest.shape[1])) * 100.
out_str += "Accuracy with KNN + Parzen window method (theta="+ str( self.Theta ) +"): %.3f" % res + "%\n"
print_output(out_str)
t_stop = time.clock()
log.info("Temps total: %.4fs\n" % float(t_stop-t_start))
#### recupere les valeurs finale de l'erreur
listeRes.append( 100 - resClassic )
listeRes.append( 100 - resKNN )
listeRes.append( 100 - res )
#### Recherche pas NNET
elif algo == "NNET":
print_output("Début de l'algorithme du Perceptron multicouche...")
# parametre, donnees, etc...
dataTrain = pca_model.getWeightsVectors()
dataTrainTargets = (dataTrainIndices - 1).reshape(dataTrainIndices.shape[0], -1)
#! contrairement au KNN le NNET prends les vecteurs de features en ligne et non pas en colonne
train_set = np.concatenate((dataTrain.T, dataTrainTargets), axis=1)
# recuperation des données de validation
dataValidation, dataValidationIndices, nClass = tools.loadImageData( "validation", self.categorie )
print_output("Projection des données de validation...")
dataValidation_proj = pca_model.getProjection( dataValidation )
dataValidationTargets = (dataValidationIndices - 1).reshape(dataValidationIndices.shape[0], -1)
validation_set = np.concatenate((dataValidation_proj.T, dataValidationTargets), axis=1)
# recuperation des données de test
dataTest, dataTestIndices, nClass = tools.loadImageData( "test", self.categorie )
print_output("Projection des données de test...")
dataTest_proj = pca_model.getProjection( dataTest )
dataTestTargets = (dataTestIndices - 1).reshape(dataTestIndices.shape[0], -1)
test_set = np.concatenate((dataTest_proj.T, dataTestTargets), axis=1)
# On build et on entraine le model pour recherche par KNN
nnet_model = NeuralNetwork( dataTrain.shape[0], self.n_hidden, nClass, self.lr, self.wd )
if self.validation == 1:
train_out, valid_out, test_out = nnet_model.train( train_set, self.n_epoch, self.batch_size, valid_set=validation_set, test_set=test_set)
else :
train_out, test_out = nnet_model.train( train_set, self.n_epoch, self.batch_size, test_set=test_set)
# affichage des courbes d'entrainement
x = []
y = []
y_err = []
color = []
legend = []
legend_err = []
filename = IMG_DIR + "Risque__Epoch_"+ str(self.n_epoch) +"_Hidden_"+ str(self.n_hidden) +"_Lr_"+ str(self.lr) +"_L2_"+ str(self.wd) + "_Categorie_" + str(self.categorie) + "_Batch_" + str(self.batch_size) + "_"
filename_err = IMG_DIR + "Erreur_classification__Epoch_"+ str(self.n_epoch) +"_Hidden_"+ str(self.n_hidden) +"_Lr_"+ str(self.lr) +"_L2_"+ str(self.wd) + "_Categorie_" + str(self.categorie) + "_Batch_" + str(self.batch_size) + "_"
train_out = np.array(train_out)
x.append(np.array(xrange(train_out.shape[0])))
# parametres courbes train
color.append('g-')
legend.append("R Train")
filename += "_Train"
y.append(train_out[:,0])
y_err.append(train_out[:,1])
legend_err.append("Err Train")
filename_err += "_Train"
# parametre courbes validation
if self.validation == 1:
valid_out = np.array(valid_out)
x.append(np.array(xrange(valid_out.shape[0])))
y.append(valid_out[:,0])
y_err.append(valid_out[:,1])
color.append('b-')
legend.append("R Validation")
legend_err.append("Err Validation")
filename += "_Validation"
filename_err += "_Validation"
# parametre courbes test
test_out = np.array(test_out)
x.append(np.array(xrange(test_out.shape[0])))
y.append(test_out[:,0])
y_err.append(test_out[:,1])
color.append('r-')
legend.append("R Test")
legend_err.append("Err Test")
filename += "_Test"
filename_err += "_Test"
# affichage
title = u"\nEpoque: " + str(self.n_epoch) + " - Taille du batch: " + str(self.batch_size) + u" - Neurones cachés: " + str(self.n_hidden) + "\nL2: " + str(self.wd) + " - Taux d'apprentissage: " + str(self.lr) + u" - Catégorie: " + str(self.categorie)
tools.drawCurves(x, y, color, legend, bDisplay=True, filename=filename, title=title, xlabel="Epoque", ylabel=u"Risque régularisé")
tools.drawCurves(x, y_err, color, legend_err, bDisplay=True, filename=filename_err, title=title, xlabel="Epoque", ylabel="Erreur classification")
#### construction fichier pour courbes ameliorees
if self.stock == 1 :
fichier = open("curvErrorNNet"+''.join( ''.join( title.split(' ') ).split('\n') ),"w")
fichier.write("#epoch errorTrain errorValidation errorTest\n")
if len(x) == 3:
for j in range(len( x[0] )):
fichier.write(str( x[0][j] )+" "+str( y[0][j] )+" "+str( y[1][j] )+" "+str( y[2][j] )+"\n")
fichier.close()
"""
/!\ Cette partie n'est plus utile car effectué dans le nnet durant le train
## TEST ###########################
#TODO Toute cette partie est a revoir pour sortir des graphes
# de train, validation, test
# compteurs de bons résultats
nbGoodResult = 0
for i in range(0, int( dataTest.shape[1] )):
#
resultNNET = np.argmax(nnet_model.compute_predictions( dataTest_proj[:,i] ), axis=1)[0]
if(resultNNET == dataTestTargets[i]):
nbGoodResult += 1
out_str = "Result: "+ str( resultNNET ) + " | Expected: "+ str( dataTestTargets[i] ) +"\n" # +1 car l'index de la matrice commence a 0
print_output(out_str)
res = (float(nbGoodResult) / float(dataTest.shape[1])) * 100.
out_str = "\nAccuracy : %.3f" % res + "%\n"
print_output(out_str)
"""
return listeRes
def reduceModel(model, atoms, select):
"""Return reduced NMA model. Reduces a :class:`~.NMA` model to a subset of
*atoms* matching *select*. This function behaves differently depending on
the type of the *model* argument. For :class:`.ANM` and :class:`.GNM` or
other :class:`.NMA` models, force constant matrix for system of interest
(specified by the *select*) is derived from the force constant matrix for
the *model* by assuming that for any given displacement of the system of
interest, other atoms move along in such a way as to minimize the potential
energy. This is based on the formulation in [KH00]_. For :class:`.PCA`
models, this function simply takes the sub-covariance matrix for selection.
:arg model: dynamics model
:type model: :class:`.ANM`, :class:`.GNM`, or :class:`.PCA`
:arg atoms: atoms that were used to build the model
:type atoms: :class:`.Atomic`
:arg select: an atom selection or a selection string
:type select: :class:`.Selection`, str
:returns: (:class:`.NMA`, :class:`.Selection`)"""
linalg = importLA()
if not isinstance(model, NMA):
raise TypeError('model must be an NMA instance, not {0:s}'
.format(type(model)))
if not isinstance(atoms, (AtomGroup, AtomSubset, AtomMap)):
raise TypeError('atoms type is not valid')
if len(atoms) <= 1:
raise TypeError('atoms must contain more than 1 atoms')
if isinstance(model, GNM):
matrix = model._kirchhoff
elif isinstance(model, ANM):
matrix = model._hessian
elif isinstance(model, PCA):
matrix = model._cov
else:
raise TypeError('model does not have a valid type derived from NMA')
if matrix is None:
raise ValueError('model matrix (Hessian/Kirchhoff/Covariance) is not '
'built')
if isinstance(select, str):
system = SELECT.getBoolArray(atoms, select)
n_sel = sum(system)
if n_sel == 0:
raise ValueError('select matches 0 atoms')
if len(atoms) == n_sel:
raise ValueError('select matches all atoms')
if isinstance(atoms, AtomGroup):
ag = atoms
which = np.arange(len(atoms))[system]
else:
ag = atoms.getAtomGroup()
which = atoms._getIndices()[system]
sel = Selection(ag, which, select, atoms.getACSIndex())
elif isinstance(select, AtomSubset):
sel = select
if isinstance(atoms, AtomGroup):
if sel.getAtomGroup() != atoms:
raise ValueError('select and atoms do not match')
system = np.zeros(len(atoms), bool)
system[sel._getIndices()] = True
else:
if atoms.getAtomGroup() != sel.getAtomGroup():
raise ValueError('select and atoms do not match')
elif not sel in atoms:
raise ValueError('select is not a subset of atoms')
idxset = set(atoms._getIndices())
system = np.array([idx in idxset for idx in sel._getIndices()])
else:
raise TypeError('select must be a string or a Selection instance')
other = np.invert(system)
if model.is3d():
system = np.tile(system, (3,1)).transpose().flatten()
other = np.tile(other, (3,1)).transpose().flatten()
ss = matrix[system,:][:,system]
if isinstance(model, PCA):
eda = PCA(model.getTitle() + ' reduced')
eda.setCovariance(ss)
return eda, system
so = matrix[system,:][:,other]
os = matrix[other,:][:,system]
oo = matrix[other,:][:,other]
matrix = ss - np.dot(so, np.dot(linalg.inv(oo), os))
if isinstance(model, GNM):
gnm = GNM(model.getTitle() + ' reduced')
gnm.setKirchhoff(matrix)
return gnm, sel
elif isinstance(model, ANM):
anm = ANM(model.getTitle() + ' reduced')
anm.setHessian(matrix)
return anm, sel
elif isinstance(model, PCA):
eda = PCA(model.getTitle() + ' reduced')
eda.setCovariance(matrix)
return eda, sel
kernelpot = KernelPotential(options)
# FILL KERNEL
generate = False
if generate:
for struct in structures:
print struct.label
kernelpot.acquire(struct, 1., label=struct.label)
print kernelpot.IX.shape
np.savetxt('out.kernelpot.ix.txt', kernelpot.IX)
else:
IX = np.loadtxt('out.kernelpot.ix.txt')
kernelpot.importAcquire(IX, 1.)
# KERNEL PCA
pca = PCA()
pca.compute(IX, normalize_mean=False, normalize_std=False)
#pca = IPCA()
#pca.compute(IX, normalize_mean=False, normalize_std=False)
# =============================
# CHECK COMPONENT NORMALIZATION
# =============================
ones_vec = np.zeros(567)
ones_vec.fill(1.)
np.savetxt('out.pca.unnorm.txt', pca.unnormBlock(ones_vec))
# =================
# INDEX CUTOFF SCAN