def test_spectral_embedding_unknown_affinity(seed=36):
# Test that SpectralClustering fails with an unknown affinity type
se = SpectralEmbedding(n_components=1,
affinity="<unknown>",
random_state=np.random.RandomState(seed))
with pytest.raises(ValueError):
se.fit(S)
def test_spectral_embedding_unknown_eigensolver(seed=36):
# Test that SpectralClustering fails with an unknown eigensolver
se = SpectralEmbedding(n_components=1, affinity="precomputed",
random_state=np.random.RandomState(seed),
eigen_solver="<unknown>")
with pytest.raises(ValueError):
se.fit(S)
def swiss_roll_test():
import matplotlib.pyplot as plt
plt.style.use('ggplot')
from time import time
from sklearn import manifold, datasets
from sklearn.manifold import SpectralEmbedding
from lpproj import LocalityPreservingProjection
n_points = 1000
X, color = datasets.samples_generator.make_s_curve(n_points,
random_state=0)
n_neighbors = 20
n_components = 2
# original lE algorithm
t0 = time()
ml_model = SpectralEmbedding(n_neighbors=n_neighbors,
n_components=n_components)
Y = ml_model.fit_transform(X)
t1 = time()
# 2d projection
fig, ax = plt.subplots(nrows=3, ncols=1, figsize=(5, 10))
ax[0].scatter(Y[:, 0], Y[:, 1], c=color, label='scikit')
ax[0].set_title('Sklearn-LE: {t:.2g}'.format(t=t1 - t0))
# Jakes LPP Algorithm
t0 = time()
ml_model = LocalityPreservingProjection(n_components=n_components)
ml_model.fit(X)
Y = ml_model.transform(X)
t1 = time()
ax[1].scatter(Y[:, 0], Y[:, 1], c=color, label='Jakes Algorithm')
ax[1].set_title('Jakes LPP: {t:.2g}'.format(t=t1 - t0))
# my SSSE algorith,
t0 = time()
ml_model = LocalityPreservingProjections(weight='angle',
n_components=n_components,
n_neighbors=n_neighbors,
sparse=True,
eig_solver='dense')
ml_model.fit(X)
Y = ml_model.transform(X)
t1 = time()
ax[2].scatter(Y[:, 0], Y[:, 1], c=color, label='My LPP Algorithm')
ax[2].set_title('My LPP: {t:.2g}'.format(t=t1 - t0))
plt.show()
def swiss_roll_test():
n_points = 1000
X, color = datasets.samples_generator.make_s_curve(n_points,
random_state=0)
n_neighbors=20
n_components=2
# original lE algorithm
t0 = time()
ml_model = SpectralEmbedding(n_neighbors=n_neighbors,
n_components=n_components)
Y = ml_model.fit_transform(X)
t1 = time()
# 2d projection
fig, ax = plt.subplots(nrows=3, ncols=1, figsize=(5,10))
ax[0].scatter(Y[:,0], Y[:,1], c=color, label='scikit')
ax[0].set_title('Sklearn-LE: {t:.2g}'.format(t=t1-t0))
# Jakes LPP Algorithm
t0 = time()
ml_model = LocalityPreservingProjection(n_components=n_components)
ml_model.fit(X)
Y = ml_model.transform(X)
t1 = time()
ax[1].scatter(Y[:,0], Y[:,1], c=color, label='Jakes Algorithm')
ax[1].set_title('Jakes LPP: {t:.2g}'.format(t=t1-t0))
# my SSSE algorith,
t0 = time()
ml_model = LocalityPreservingProjections(weight='angle',
n_components=n_components,
n_neighbors=n_neighbors,
sparse=True,
eig_solver='dense')
ml_model.fit(X)
Y = ml_model.transform(X)
t1 = time()
ax[2].scatter(Y[:,0], Y[:,1], c=color, label='My LPP Algorithm')
ax[2].set_title('My LPP: {t:.2g}'.format(t=t1-t0))
plt.show()
class LaplacianEigenmaps(AbstractReducer):
def __init__(self, d: int = 2, random_state: int = 0, **kwargs):
super().__init__(d, random_state)
self._main = SpectralEmbedding(n_components=d,
random_state=random_state,
**kwargs)
def fit_transform(self, x: np.ndarray, **kwargs) -> np.ndarray:
return self._main.fit_transform(x)
def fit(self, x: np.ndarray, **kwargs):
return self._main.fit(x)
def transform(self, x: np.ndarray, **kwargs) -> np.ndarray:
raise NotImplementedError
def set_random_state(self, random_state: int = 0):
self.random_state = random_state
self._main.random_state = random_state
@property
def is_deterministic(self) -> bool:
return False
@property
def is_stateful(self) -> bool:
return True
@staticmethod
def get_parameter_ranges() -> dict:
return {'n_neighbors': (int, 1, 20)}
def swiss_roll_test():
import matplotlib.pyplot as plt
plt.style.use('ggplot')
from time import time
from sklearn import manifold, datasets
from sklearn.manifold import SpectralEmbedding
n_points = 1000
X, color = datasets.samples_generator.make_s_curve(n_points,
random_state=0)
n_neighbors=10
n_components=2
# original scikit-learn lE algorithm
t0 = time()
ml_model = SpectralEmbedding(affinity='nearest_neighbors',
n_neighbors=n_neighbors,
n_components=n_components)
Y = ml_model.fit_transform(X)
t1 = time()
# 2d projection
fig, ax = plt.subplots(nrows=2, ncols=1, figsize=(5,10))
ax[0].scatter(Y[:,0], Y[:,1], c=color, label='scikit')
ax[0].set_title('Sklearn LE: {t:.2g}'.format(t=t1-t0))
# my Laplacian Eigenmaps algorithm
t0 = time()
ml_model = LaplacianEigenmaps(n_components=n_components,
n_neighbors=n_neighbors)
ml_model.fit(X)
Y = ml_model.fit_transform(X)
t1 = time()
ax[1].scatter(Y[:,0], Y[:,1], c=color, label='My LE Algorithm')
ax[1].set_title('My LE: {t:.2g}'.format(t=t1-t0))
plt.show()
def plot_data(r1_nc, r2_nc, rx_nc, r1_dur, r2_dur, knn):
ncon = r1_nc + r2_nc + rx_nc
#Create simulated data
data = create_sym_data(r1_nc, r2_nc, rx_nc, r1_dur, r2_dur)
#Carpet plot of connectivity
swc_plot = data.hvplot.heatmap(cmap='jet',
clim=(0, 1),
xlabel='Time (Windows)',
ylabel='Connections',
title='Simulated SWC Matrix',
shared_axes=False).opts(toolbar=None)
#Correlation matrix for SCC
cm = data.corr()
cm.columns = [
'WIN' + str(i + 1).zfill(4) for i in np.arange(r1_dur + r2_dur)
]
cm.index = [
'WIN' + str(i + 1).zfill(4) for i in np.arange(r1_dur + r2_dur)
]
cm_plot = cm.hvplot.heatmap(
cmap='RdBu_r',
clim=(-1, 1),
shared_axes=False,
xlabel='Time (Windows)',
ylabel='Time (Windows)',
title='Win-2-Win Correlation (Similarity across samples)')
# Create Embedding
se = SpectralEmbedding(n_components=2, n_neighbors=knn)
se.fit(data.T)
se_df = pd.DataFrame(se.embedding_)
se_df.columns = ['x', 'y']
se_df['class'] = 'Run 2'
se_df.loc[0:r1_dur, 'class'] = 'Run 1'
se_plot = se_df.hvplot.scatter(x='x',
y='y',
color='class',
title='Spectral Embedding',
xlabel='SE Dim 1',
ylabel='SE Dim 2')
return (swc_plot + cm_plot + se_plot).cols(1)
def mds_dyn_mat(self, ax=None, type="T", learned=True):
"""
dimensionality reduction plot for either est_SR or est_T depending on 'type' and 'learned'
"""
import networkx as nx
from sklearn.manifold import MDS, SpectralEmbedding
dr = SpectralEmbedding(
n_components=2,
affinity="precomputed",
gamma=None,
random_state=None,
eigen_solver=None,
n_neighbors=3,
n_jobs=None,
)
self.set_target_axis(ax)
if type is "T":
if learned:
S = (self.est_T + self.est_T.T) / 2.0
else:
S = (self.ENV.T + self.ENV.T.T) / 2.0
elif type is "SR":
if learned:
S = (self.est_SR + self.est_SR.T) / 2.0
else:
S = (self.SR + self.SR.T) / 2.0
elif type is "Q":
if learned:
S = (self.est_Q + self.est_Q.T) / 2.0
else:
S = (self.GEN.Q + self.GEN.Q.T) / 2.0
else:
raise ValueError("unrecognized learned dynamics matrix request")
pos = dr.fit(S).embedding_[:, :2]
G = nx.from_numpy_array(self.ENV.T)
pos_dict = dict([(i, posxy) for i, posxy in enumerate(pos)])
nx.draw(
G,
pos=pos_dict,
ax=self.ax,
node_size=10,
node_color="black",
edge_color="grey",
width=0.5,
)
for i in range(0, n):
rowTopN[i] = rowSorted[i, :][n - topN]
columnTopN[i] = columnSorted[:, i][n - topN]
for i in range(0, n):
for j in range(0, m):
if (sM[i][j] >= rowTopN[i] or sM[i][j] >= columnTopN[j]):
result[i][j] = sM[i][j]
return result
embedding = SpectralEmbedding(n_components=2,
affinity="precomputed",
n_neighbors=0)
coords = embedding.fit(thresholdMatrix(1 / (dM + 0.1), 20)).embedding_
print(coords.shape)
total_count = 0
similar_count1 = 0
similar_count2 = 0
sM = thresholdMatrix(1 / (dM + 0.1), 10)
for i in range(0, sM.shape[0]):
r = sM[i, :]
r = np.nonzero(r)
r = r[0]
l = r.size
for a in range(0, l):
for b in range(a + 1, l):
total_count += 1
if (sM[i, r[a]] >= sM[i, r[b]]
simMatAllTimes = setDiagToOne(forcePosDef(simMatAllTimes))
# tsneModel = TSNE(n_components=2, metric='precomputed',
# method='barnes_hut', perplexity=30.0)
# #method='barnes_hut'
# #method='exact'
# distMat = 1.0 / (simMatAllTimes + 1.01)
tsneModel = SpectralEmbedding(n_components=2, affinity='precomputed',
gamma=1.0, n_neighbors=6)
distMat = simMatAllTimes + 1.0
tsneModel = tsneModel.fit(distMat)
sizeScale = np.abs(tsneModel.embedding_.ravel()).max()
tsneModel.embedding_ /= sizeScale
#############################
#
# Plotting
#
# set plot parameters
comm1_str = np.vstack([np.sum(a[:, :nNodes_comm1], axis=1, keepdims=True)
for a in adjMat_timeseries])
comm2_str = np.vstack([np.sum(a[:, -nNodes_comm2:], axis=1, keepdims=True)
for a in adjMat_timeseries])
def spectral_emb(self, n):
semb = SpectralEmbedding(n_components=n)
semb.fit(self.sample)
semb_res = semb.fit_transform(self.sample)
import numpy as np from sklearn.utils.arpack import eigsh app = service.prodbox.CinemaService() X = app.getWeightedSearchFeatures(15) graph = kneighbors_graph(X, 10) lap = graph_laplacian(graph, True) from sklearn.decomposition import TruncatedSVD svd = TruncatedSVD(n_components = 30, algorithm="arpack") lap = spectral_embedding_._set_diag(lap, 1) svd.fit(-lap) eigenvalues = np.diag(svd.components_ * (-lap).todense() * svd.components_.T) eigenvalues2, _ = eigsh(-lap, k=30, which='LM', sigma=1) print(eigenvalues) print(eigenvalues2) se = SpectralEmbedding(n_components = 30, eigen_solver='arpack', affinity="nearest_neighbors") se.fit(X) app.quit() # TODO : check budget distribution, draw budget conditionnaly out = connected_components(graph)
def spectral_embed(affinity_mat, n_comps):
spec_emb = SpectralEmbedding(n_components=n_comps, affinity='precomputed')
spec_emb.fit(affinity_mat)
embed = spec_emb.embedding_
return embed.T
from sklearn.utils.arpack import eigsh
app = service.prodbox.CinemaService()
X = app.getWeightedSearchFeatures(15)
graph = kneighbors_graph(X, 10)
lap = graph_laplacian(graph, True)
from sklearn.decomposition import TruncatedSVD
svd = TruncatedSVD(n_components=30, algorithm="arpack")
lap = spectral_embedding_._set_diag(lap, 1)
svd.fit(-lap)
eigenvalues = np.diag(svd.components_ * (-lap).todense() * svd.components_.T)
eigenvalues2, _ = eigsh(-lap, k=30, which='LM', sigma=1)
print(eigenvalues)
print(eigenvalues2)
se = SpectralEmbedding(n_components=30,
eigen_solver='arpack',
affinity="nearest_neighbors")
se.fit(X)
app.quit()
# TODO : check budget distribution, draw budget conditionnaly
out = connected_components(graph)
norm_local_diss=False,
norm_sim=False,
merge_if_ccs=True,
min_cc_len=min_cc_len,
do_eps_graph=True,
preprocess_only=True)
# Run the spectral ordering method on the DNA reads similarity matrix
t0 = time()
reord_method.fit(new_mat)
my_ebd = reord_method.embedding
tme = time() - t0
print("my embedding in {}s".format(tme))
skl_method = SpectralEmbedding(n_components=dim, affinity='precomputed')
skl_method.fit(new_mat.toarray())
skl_ebd = skl_method.embedding_
tskl = time() - tme
print("sklearn embedding in {}s".format(tskl))
skl_amg = SpectralEmbedding(n_components=dim,
affinity='precomputed',
eigen_solver='amg')
skl_amg.fit(new_mat)
amg_ebd = skl_amg.embedding_
tamg = time() - tskl
print("sklearn amg embedding in {}s".format(tamg))
fig = plt.figure()
ax = Axes3D(fig)
ax.scatter(my_ebd[:, 0], my_ebd[:, 1], my_ebd[:, 2])
