def extract_master_keyword(self, title, article):
assert article != None, 'article is invalid'
assert len(article) > 0, 'article is invalid'
title_morpheme_list = self._morpheme_analyzer.analyze_morpheme(title)
title_keywords = [
morpheme.morpheme for morpheme in title_morpheme_list
]
article_morpheme_list = self._morpheme_analyzer.analyze_morpheme(
article)
#article_morpheme_list = filter(lambda x:x.frequency > 1, article_morpheme_list)
article_matrix = self._create_article_matrix(article,
article_morpheme_list)
if not isinstance(article_matrix, ndarray):
return None
pca = PCANode(svd=True)
try:
pca.execute(article_matrix)
except Exception, e:
print str(e), title
logging.exception(
'Exception raised in method extract_master_keyword with article title='
+ title)
def bench_mdp():
#
# .. MDP ..
#
from mdp.nodes import PCANode
start = datetime.now()
mdp_clf = PCANode(output_dim=n_components)
mdp_clf.train(X)
return datetime.now() - start
def _train(self, data, label=None):
""" Updates the estimated covariance matrix based on *data*. """
# We simply ignore the class label since we
# are doing unsupervised learning
if self.wrapped_node is None:
self.wrapped_node = PCANode(svd=self.svd, reduce=self.reduce)
x = 1.0 * data.view(type=numpy.ndarray)
self.wrapped_node.train(x)
if self.channel_names is None:
self.channel_names = data.channel_names
def pca(multidim_data, output_dim):
"""Principal Component Analysis"""
pcanode = PCANode(input_dim=multidim_data.shape[1],
output_dim=output_dim,
dtype=np.float32,
svd=True,
reduce=True,
var_rel=1E-15,
var_abs=1E-15)
pcanode.train(data)
return np.dot(multidim_data, pcanode.get_projmatrix())
def applyPCA(features, components):
"""
Dimension reduction of a matrix using pca
"""
features = np.array(features)
pcanode = PCANode(output_dim=components)
pcanode.train(features)
pcanode.stop_training()
# print "Explained variance: %s" % pcanode.explained_variance
return pcanode(features)
def bench_mdp(X, y, T, valid):
#
# .. MDP ..
#
from mdp.nodes import PCANode
start = datetime.now()
clf = PCANode(output_dim=n_components)
clf.train(X)
clf.stop_training()
delta = datetime.now() - start
ev = explained_variance(X, clf.v.T).sum()
return ev, delta
def PCA(): col1 = np.array([[1.0, 2, 3, 4, 5, 6, 7, 8, 9]]).T col2 = np.array([[2.0, 4, 6, 8, 10, 12, 14, 16, 18]]).T matr12 = np.hstack((col1, col2)) matr_arr = [matr12] pca_node = PCANode() d_arr = [] for arr in matr_arr: result = pca_node.execute(arr) d_arr.append(pca_node.d) print pca_node.d
def pca(data, output_dim=2, standardize=True):
""" This package is using the PCA module form
url : http://folk.uio.no/henninri/pca_module/
"""
"""
T, P, explained_var = PCA_svd(data, standardize=True)
return T,P,explained_var
"""
""" ALTERNATE MPLEMENTATION FROM MDP """
pca_node = PCANode(output_dim=output_dim)
pca_node.train(data)
pca_node.stop_training()
return pca_node
def test_mcanode_v1():
line_x = numx.zeros((1000, 2), "d")
line_y = numx.zeros((1000, 2), "d")
line_x[:, 0] = numx.linspace(-1, 1, num=1000, endpoint=1)
line_y[:, 1] = numx.linspace(-0.2, 0.2, num=1000, endpoint=1)
mat = numx.concatenate((line_x, line_y))
utils.rotate(mat, uniform() * 2 * numx.pi)
mat += uniform(2)
mat -= mat.mean(axis=0)
mca = MCANode()
for i in xrange(5):
mca.train(mat)
bpca = PCANode()
bpca.train(mat)
bpca.stop_training()
v = mca.get_projmatrix()
bv = bpca.get_projmatrix()[:, ::-1]
dcosines = numx.zeros(v.shape[1])
for dim in xrange(v.shape[1]):
dcosines[dim] = numx.fabs(numx.dot(v[:, dim], bv[:, dim].T)) / (
numx.linalg.norm(v[:, dim]) * numx.linalg.norm(bv[:, dim]))
assert_almost_equal(numx.ones(v.shape[1]), dcosines)
def _train(self, data, label = None):
""" Updates the estimated covariance matrix based on *data*. """
# We simply ignore the class label since we
# are doing unsupervised learning
if self.wrapped_node is None:
self.wrapped_node = PCANode()
x=data.view(numpy.ndarray)
self.wrapped_node.train(x)
if self.channel_names is None:
self.channel_names = data.channel_names
def pre_processing(spks, ndet, tf, pca_dim=4):
print 'starting to prewhiten w.r.t. noise..',
spks_prw = sp.dot(spks, ndet.get_whitening_op(tf=tf))
print 'done.'
print 'pca projection: %s' % pca_dim,
spks_pca = PCANode(output_dim=pca_dim)(spks_prw)
print 'done.'
return spks_pca
def test_ccipcanode_v1():
line_x = numx.zeros((1000, 2), "d")
line_y = numx.zeros((1000, 2), "d")
line_x[:, 0] = numx.linspace(-1, 1, num=1000, endpoint=1)
line_y[:, 1] = numx.linspace(-0.2, 0.2, num=1000, endpoint=1)
mat = numx.concatenate((line_x, line_y))
utils.rotate(mat, uniform() * 2 * numx.pi)
mat += uniform(2)
mat -= mat.mean(axis=0)
pca = CCIPCANode()
for i in range(5):
pca.train(mat)
bpca = PCANode()
bpca.train(mat)
bpca.stop_training()
v = pca.get_projmatrix()
bv = bpca.get_projmatrix()
dcosines = numx.zeros(v.shape[1])
for dim in range(v.shape[1]):
dcosines[dim] = numx.fabs(numx.dot(v[:, dim], bv[:, dim].T)) / (
numx.linalg.norm(v[:, dim]) * numx.linalg.norm(bv[:, dim]))
assert_almost_equal(numx.ones(v.shape[1]), dcosines)
def doPCA(in_pts, D, tol=.8):
pts=copy(in_pts)
pts.shape = (len(in_pts),D)
p=PCANode(output_dim=tol)
p.train(pts)
p.stop_training()
return p
def bench_mdp(X, y, T, valid):
#
# .. MDP ..
#
from mdp.nodes import PCANode
start = datetime.now()
clf = PCANode(output_dim=n_components)
clf.train(X)
clf.stop_training()
delta = datetime.now() - start
ev = explained_variance(X, clf.v.T).sum()
return ev, delta
def test_mcanode_v2():
iterval = 30
t = numx.linspace(0, 4 * numx.pi, 500)
x = numx.zeros([t.shape[0], 2])
x[:, 0] = numx.real(numx.sin(t) + numx.power(numx.cos(11 * t), 2))
x[:, 1] = numx.cos(11 * t)
expnode = PolynomialExpansionNode(2)
input_data = expnode(x)
input_data = input_data - input_data.mean(axis=0)
wtnnode = WhiteningNode()
input_data = wtnnode(input_data)
input_data = mdp.utils.timediff(input_data)
##Setup node/trainer
output_dim = 4
node = MCANode(output_dim=output_dim, eps=0.05)
bpcanode = PCANode()
bpcanode(input_data)
# bv = bpcanode.v / numx.linalg.norm(bpcanode.v, axis=0)
bv = bpcanode.v / mdp.numx.sum(bpcanode.v**2, axis=0)**0.5
bv = bv[:, ::-1][:, :output_dim]
_tcnt = time.time()
v = []
for i in xrange(iterval * input_data.shape[0]):
node.train(input_data[i % input_data.shape[0]:i % input_data.shape[0] +
1])
if (node.get_current_train_iteration() % 100 == 0):
v.append(node.v)
dcosines = numx.zeros([len(v), output_dim])
for i in xrange(len(v)):
for dim in xrange(output_dim):
dcosines[i,
dim] = numx.fabs(numx.dot(v[i][:, dim], bv[:, dim].T)) / (
numx.linalg.norm(v[i][:, dim]) *
numx.linalg.norm(bv[:, dim]))
print('\nTotal Time for {} iterations: {}'.format(iterval,
time.time() - _tcnt))
assert_almost_equal(numx.ones(output_dim), dcosines[-1], decimal=2)
def test_ccipcanode_v2():
iterval = 30
t = numx.linspace(0, 4 * numx.pi, 500)
x = numx.zeros([t.shape[0], 2])
x[:, 0] = numx.real(numx.sin(t) + numx.power(numx.cos(11 * t), 2))
x[:, 1] = numx.cos(11 * t)
expnode = PolynomialExpansionNode(2)
input_data = expnode(x)
input_data = input_data - input_data.mean(axis=0)
##Setup node/trainer
output_dim = 4
node = CCIPCANode(output_dim=output_dim)
bpcanode = PCANode(output_dim=output_dim)
bpcanode(input_data)
# bv = bpcanode.v / numx.linalg.norm(bpcanode.v, axis=0)
bv = bpcanode.v / mdp.numx.sum(bpcanode.v**2, axis=0)**0.5
v = []
_tcnt = time.time()
for i in xrange(iterval * input_data.shape[0]):
node.train(input_data[i % input_data.shape[0]:i % input_data.shape[0] +
1])
if (node.get_current_train_iteration() % 100 == 0):
v.append(node.v)
dcosines = numx.zeros([len(v), output_dim])
for i in xrange(len(v)):
for dim in xrange(output_dim):
dcosines[i,
dim] = numx.fabs(numx.dot(v[i][:, dim], bv[:, dim].T)) / (
numx.linalg.norm(v[i][:, dim]) *
numx.linalg.norm(bv[:, dim]))
print
print 'Total Time for %d iterations: ' % (iterval), time.time() - _tcnt
assert_almost_equal(numx.ones(output_dim), dcosines[-1], decimal=3)
class PCAWrapperNode(SpatialFilteringNode): #, PCANode):
""" Reuse the implementation of the Principal Component Analysis of mdp
This node implements the unsupervised principal component
analysis algorithm for spatial filtering.
**Parameters**
:retained_channels: Determines how many of the PCA pseudo channels
are retained. Default is None which means "all channels".
:load_path: An absolute path from which the PCA eigenmatrix
is loaded.
If not specified, this matrix is learned from the training data.
(*optional, default: None*)
**Exemplary Call**
.. code-block:: yaml
-
node : PCA
parameters:
retained_channels : 42
"""
def __init__(self, retained_channels=None, load_path=None, **kwargs):
# Must be set before constructor of superclass is set
self.trainable = (load_path == None)
super(PCAWrapperNode, self).__init__(**kwargs)
self.output_dim = retained_channels
# Load patterns from file if requested
if load_path != None:
filter_file = open(load_path, 'r')
avg, v = cPickle.load(filter_file)
self.set_permanent_attributes(avg=avg,
v=v,
trainable=False,
filters=v)
self.set_permanent_attributes( # The number of channels that will be retained
retained_channels=retained_channels,
output_dim=retained_channels,
channel_names=None,
new_channels=None,
wrapped_node=None)
def is_trainable(self):
""" Returns whether this node is trainable. """
return self.trainable
def is_supervised(self):
""" Returns whether this node requires supervised training. """
return False
def _train(self, data, label=None):
""" Updates the estimated covariance matrix based on *data*. """
# We simply ignore the class label since we
# are doing unsupervised learning
if self.wrapped_node is None:
self.wrapped_node = PCANode()
x = data.view(numpy.ndarray)
self.wrapped_node.train(x)
if self.channel_names is None:
self.channel_names = data.channel_names
def _stop_training(self, debug=False):
""" Stops training by forwarding to super class. """
self.wrapped_node.stop_training()
#super(PCAWrapperNode, self)._stop_training(debug)
self.v = self.wrapped_node.v
self.avg = self.wrapped_node.avg
self.filters = self.v
def _execute(self, data, n=None):
""" Execute learned transformation on *data*.
Projects the given data to the axis of the most significant
eigenvectors and returns the data in this lower-dimensional subspace.
"""
# 'INITIALIZATION'
if self.retained_channels == None:
self.retained_channels = data.shape[1]
if n is None:
n = self.retained_channels
if self.channel_names is None:
self.channel_names = data.channel_names
if len(self.channel_names) < self.retained_channels:
self.retained_channels = len(self.channel_names)
self._log(
"To many channels chosen for the retained channels! Replaced by maximum number.",
level=logging.CRITICAL)
if not (self.output_dim == self.retained_channels):
# overwrite internal output_dim variable, since it is set wrong
self._output_dim = self.retained_channels
# 'Real' Processing
#projected_data = super(PCANodeWrapper, self)._execute(data, n)
x = data.view(numpy.ndarray)
projected_data = mult(x - self.avg, self.v[:, :self.retained_channels])
if self.new_channels is None:
self.new_channel_names = [
"pca%03d" % i for i in range(projected_data.shape[1])
]
return TimeSeries(projected_data, self.new_channel_names,
data.sampling_frequency, data.start_time,
data.end_time, data.name, data.marker_name)
def store_state(self, result_dir, index=None):
""" Stores this node in the given directory *result_dir*. """
if self.store:
node_dir = os.path.join(result_dir, self.__class__.__name__)
create_directory(node_dir)
# This node only stores the learned eigenvector and eigenvalues
name = "%s_sp%s.pickle" % ("eigenmatrix", self.current_split)
result_file = open(os.path.join(node_dir, name), "wb")
result_file.write(cPickle.dumps((self.avg, self.v), protocol=2))
result_file.close()
class spikesorter(graphics.diagnosticGUI):
def __init__(self, title):
#global plotter
plotter = graphics.Plotter()
graphics.diagnosticGUI.__init__(self, plotter)
#Recover data:
#data = importPointset('shortsimdata1.dat',t=0,sep=',')
data = importPointset('simdata1_100000.dat', t=0, sep=',')
vs = data['vararray'][0]
vs = self.bandpass_filter(vs, 300, 3000, 32000)
ts = data['t']
self.N = len(vs)
self.traj = numeric_to_traj([vs],
'test_traj', ['x'],
ts,
discrete=False)
#Threshold used in martinez et al.
self.mthresh = 4 * (np.median(np.abs(vs)) / 0.6475)
self.selected_pcs = []
self.fovea_setup()
#x = self.traj.sample(tlo= 0, thi = self.N)['x']
#r = x > 10
#above_thresh = np.where(r == True)[0]
#spike = []
#peaks = []
#crosses = []
#last_i = above_thresh[0] - 1
#for i in above_thresh:
#if i - 1 != last_i:
#crosses.append(i)
##Return x value of the highest y value.
#peaks.append(np.where(x == max(list(x[spike])))[0][0])
#spike = []
#spike.append(i)
#last_i = i
if tutorial_on:
print("STEP 1:")
print("Create a horizontal line of interest by pressing 'l'.")
print(
"Once created, this line can be forced to extent by pressing 'm'."
)
print(
"Enter 'ssort.selected_object.update(name = 'thresh')' to identify the line as a threshold for spike detection"
)
print(
"Once the line is renamed to 'thresh', the arrow keys can be used to move it up and down."
)
self.tutorial = 'step2'
else:
self.tutorial = None
def fovea_setup(self):
#Setup code
DOI = [(0, self.N), (-30, 30)]
self.plotter.clean() # in case rerun in same session
self.plotter.add_fig('master',
title='spikesort',
xlabel='time',
ylabel='mV',
domain=DOI)
#Setup all layers
self.plotter.add_layer('spikes')
self.plotter.add_layer('thresh_crosses')
self.plotter.add_layer('detected')
self.plotter.add_layer('pcs')
self.plotter.add_layer('scores')
self.setup(
{
'11': {
'name': 'Waveform',
'scale': DOI,
'layers': ['spikes', 'thresh_crosses'],
'callbacks': '*',
'axes_vars': ['x', 'y']
},
'12': {
'name': 'Detected Spikes',
'scale': [(0, default_sw), (-80, 80)],
'layers': ['detected'],
#'callbacks':'*',
'axes_vars': ['x', 'y']
},
'21': {
'name': 'Principal Components',
'scale': [(0, default_sw), (-0.5, 0.5)],
'layers': ['pcs'],
#'callbacks':'*',
'axes_vars': ['x', 'y']
},
'22': {
'name': 'Projected Spikes',
#'scale': [(-100, 100), (-100, 100)],
'scale': [(-300, 300), (-300, 300)],
'layers': ['scores'],
'callbacks': '*',
'axes_vars': ['firstPC', 'secondPC']
}
},
size=(8, 8),
with_times=False,
basic_widgets=True)
#self.plotter.set_text('load_perc', Loading: %d\%'%n, 'loading')
#Bad code carried over from fovea_game:
fig_struct, figure = self.plotter._resolve_fig(None)
#self.ax = fig_struct.arrange['11']['axes_obj']
coorddict = {'x': {'x': 't', 'layer': 'spikes', 'style': 'b-'}}
self.add_data_points(self.traj.sample(), coorddict=coorddict)
evKeyOn = self.fig.canvas.mpl_connect('key_press_event',
self.ssort_key_on)
self.plotter.auto_scale_domain(subplot='11', xcushion=0)
self.plotter.show()
def bandpass_filter(self, data, lowcut, highcut, fs, order=5):
nyq = 0.5 * fs
low = lowcut / nyq
high = highcut / nyq
b, a = butter(order, [low, high], btype='band')
y = lfilter(b, a, data)
return y
def user_pick_func(self, ev):
if self.selected_object.layer == 'detected' or self.selected_object.layer == 'scores':
if hasattr(self, 'last_name'):
self.plotter.set_data_2(
self.last_name,
layer='scores',
markersize=6,
zorder=1,
style=self.default_colors[self.last_name])
self.plotter.set_data_2(
self.last_name,
layer='detected',
linewidth=1,
zorder=1,
style=self.default_colors[self.last_name])
self.plotter.set_data_2(self.selected_object.name,
layer='scores',
markersize=12,
zorder=10,
style='y*')
self.plotter.set_data_2(self.selected_object.name,
layer='detected',
linewidth=2.5,
zorder=10,
style='y-')
self.last_name = self.selected_object.name
elif self.selected_object.layer == 'pcs':
self.proj_PCs.insert(0, self.selected_object.name)
self.proj_PCs = self.proj_PCs[0:2]
for name in fig_struct['layers']['pcs']['data'].keys():
if name not in self.proj_PCs:
self.plotter.set_data_2(name,
layer='pcs',
style=fig_struct['layers']['pcs']
['data'][name]['style'][0] + '--')
for pc in self.proj_PCs:
self.plotter.set_data_2(
pc,
layer='pcs',
style=fig_struct['layers']['pcs']['data'][pc]['style'][0] +
'-')
self.proj_vec1 = fig_struct['layers']['pcs']['handles'][
self.proj_PCs[0]].get_ydata()
self.proj_vec2 = fig_struct['layers']['pcs']['handles'][
self.proj_PCs[1]].get_ydata()
fig_struct.arrange['22']['axes_vars'] = list(
reversed(self.proj_PCs))
self.project_to_PC()
self.plotter.show()
def user_update_func(self):
if self.selected_object.name is 'thresh':
if self.selected_object.m != 0:
print("Make 'thresh' a horizontal threshold by pressing 'm'.")
return
try:
self.search_width = self.context_objects['ref_box'].dx
self.context_objects['ref_box'].remove()
except KeyError:
self.search_width = default_sw
if self.tutorial == 'step2':
print("STEP 2: ")
print(
"When thresh is in place, press 'd' to capture each spike crossing the threshold in a bounding box."
)
print(
"Each detected spike will be placed in the top right subplot."
)
self.tutorial = 'step3'
cutoff = self.selected_object.y1
traj_samp = self.traj.sample()['x']
r = traj_samp > cutoff
above_thresh = np.where(r == True)[0]
spike = []
spikes = []
crosses = []
last_i = above_thresh[0] - 1
for i in above_thresh:
if i - 1 != last_i:
crosses.append(i)
#Return x value of the highest y value.
spikes.append(spike)
spike = []
spike.append(i)
last_i = i
self.traj_samp = traj_samp
self.crosses = crosses
self.spikes = spikes
self.plotter.add_data([self.crosses, [cutoff] * len(self.crosses)],
layer='thresh_crosses',
style='r*',
name='crossovers',
force=True)
self.show()
def compute_bbox(self):
fig_struct, figs = self.plotter._resolve_fig(None)
#Clear existing bounding boxes
rem_names = []
for con_name, con_obj in self.context_objects.items():
if isinstance(con_obj, box_GUI) and con_name is not 'ref_box':
rem_names.append(con_name)
for name in rem_names:
self.context_objects[name].remove(draw=False)
fig_struct['layers']['detected']['data'] = {}
self.plotter.show(rebuild=True)
#Create new bounding boxes
c = 0
for spike in self.spikes:
peak = np.where(
self.traj_samp == max(list(self.traj_samp[spike])))[0][0]
tlo = peak - 20
thi = tlo + self.search_width
valley = min(self.traj.sample()['x'][tlo:thi])
box_GUI(self,
pp.Point2D(tlo,
self.traj.sample()['x'][peak]),
pp.Point2D(thi, valley),
name='spike_box' + str(c),
select=False)
spike_seg = self.traj_samp[tlo:thi]
try:
X = np.row_stack((X, spike_seg))
except NameError:
X = spike_seg
c += 1
return X
def project_to_PC(self):
Y = np.dot(self.X, np.column_stack((self.proj_vec1, self.proj_vec2)))
#If moving to a smaller number of spikes, just forcing out data by reassigning names won't work. Must clear.
self.clear_data('scores')
self.show()
self.default_colors = {}
#Add spikes as individual lines, so they can be referenced individually.
c = 0
for spike in Y:
name = 'spike' + str(c)
self.default_colors[name] = 'k'
self.add_data_points([spike[0], spike[1]],
layer='scores',
style=self.default_colors[name] + '*',
name=name)
c += 1
self.plotter.auto_scale_domain(subplot='22')
self.show(rebuild=True)
def ssort_key_on(self, ev):
self._key = k = ev.key # keep record of last keypress
fig_struct, fig = self.plotter._resolve_fig(None)
class_keys = ['1', '2', '3', '0']
if k in class_keys:
if isinstance(self.selected_object, box_GUI):
for dname, dstruct in fig_struct['layers']['scores'][
'data'].items():
if self.selected_object.x1 < dstruct['data'][0] < self.selected_object.x2 and \
self.selected_object.y1 < dstruct['data'][1] < self.selected_object.y2:
if k == '1':
self.default_colors[dname] = 'r'
self.plotter.set_data_2(dname,
layer='detected',
style='r-')
self.plotter.set_data_2(dname,
layer='scores',
style='r*')
if k == '2':
self.default_colors[dname] = 'g'
self.plotter.set_data_2(dname,
layer='detected',
style='g-')
self.plotter.set_data_2(dname,
layer='scores',
style='g*')
if k == '3':
self.default_colors[dname] = 'b'
self.plotter.set_data_2(dname,
layer='detected',
style='b-')
self.plotter.set_data_2(dname,
layer='scores',
style='b*')
if k == '0':
self.default_colors[dname] = 'k'
self.plotter.set_data_2(dname,
layer='detected',
style='k-')
self.plotter.set_data_2(dname,
layer='scores',
style='k*')
self.plotter.show()
if k == 'd':
try:
self.crosses
except AttributeError:
print(
"Can't detect spikes until threshold crossings have been found."
)
return
self.X = self.compute_bbox()
self.default_colors = {}
if len(self.X.shape) == 1:
self.default_colors['spike0'] = 'k'
self.add_data_points([list(range(0, len(self.X))), self.X],
layer='detected',
style=self.default_colors['spike0'] + '-',
name='spike0',
force=True)
else:
c = 0
for spike in self.X:
name = 'spike' + str(c)
self.default_colors[name] = 'k'
self.add_data_points([list(range(0, len(spike))), spike],
layer='detected',
style=self.default_colors[name] + '-',
name=name,
force=True)
c += 1
self.plotter.auto_scale_domain(xcushion=0, subplot='12')
self.show()
if self.tutorial == 'step3':
print("STEP 3: ")
print(
"You can now press 'p' to perform PCA on the detected spikes."
)
print(
"The bottom right subplot will display the first 3 principal components (in red, green, and yellow respectively.)"
)
print(
"The bottom left subplot will show the detected spikes projected onto the first two PCs"
)
self.tutorial = 'step4'
if k == 'p':
try:
X = self.X
except AttributeError:
print('Must detect spikes before performing PCA.')
return
print('doing PCA...')
self.p = PCANode(output_dim=0.99, reduce=True, svd=True)
self.p.train(X)
self.proj_vec1 = self.p.get_projmatrix()[:, 0]
self.proj_vec2 = self.p.get_projmatrix()[:, 1]
self.add_data_points(
[list(range(0, len(self.proj_vec1))), self.proj_vec1],
style='r-',
layer='pcs',
name='firstPC',
force=True)
self.add_data_points(
[list(range(0, len(self.proj_vec2))), self.proj_vec2],
style='g-',
layer='pcs',
name='secondPC',
force=True)
self.plotter.show()
self.proj_PCs = ['firstPC', 'secondPC']
try:
self.add_data_points([
list(range(0, len(self.p.get_projmatrix()))),
self.p.get_projmatrix()[:, 2]
],
style='y--',
layer='pcs',
name='thirdPC',
force=True)
except IndexError:
pass
self.add_legend(['r', 'g', 'y'], ['1st PC', '2nd PC', '3rd PC'],
'21')
self.plotter.auto_scale_domain(xcushion=0, subplot='21')
self.show()
self.project_to_PC()
if self.tutorial == 'step4':
print("STEP 4: ")
print("Use mouse clicks to explore the data.")
print(
"Clicking on detected spikes in the top-right will highlight the corresponding projection in the bottom right (and vice versa)."
)
print(
"You can also change the set of PCs onto which the data are projected by clicking the desired projection PCs in the bottom left"
)
print("NOTE ALSO: ")
print(
"Creating a bounding box in the upper-left plot and renaming it to 'ref_box', will change the search width of the detected spike."
)
print(
"e.g., if you want detected spikes to be 30 msec long, the box's .dx value must be 30."
)
print(
"After creating the box, it will be set to the current selected object. You can select the thresh line again by clicking on it."
)
def ssort_key_on(self, ev):
self._key = k = ev.key # keep record of last keypress
fig_struct, fig = self.plotter._resolve_fig(None)
class_keys = ['1', '2', '3', '0']
if k in class_keys:
if isinstance(self.selected_object, box_GUI):
for dname, dstruct in fig_struct['layers']['scores'][
'data'].items():
if self.selected_object.x1 < dstruct['data'][0] < self.selected_object.x2 and \
self.selected_object.y1 < dstruct['data'][1] < self.selected_object.y2:
if k == '1':
self.default_colors[dname] = 'r'
self.plotter.set_data_2(dname,
layer='detected',
style='r-')
self.plotter.set_data_2(dname,
layer='scores',
style='r*')
if k == '2':
self.default_colors[dname] = 'g'
self.plotter.set_data_2(dname,
layer='detected',
style='g-')
self.plotter.set_data_2(dname,
layer='scores',
style='g*')
if k == '3':
self.default_colors[dname] = 'b'
self.plotter.set_data_2(dname,
layer='detected',
style='b-')
self.plotter.set_data_2(dname,
layer='scores',
style='b*')
if k == '0':
self.default_colors[dname] = 'k'
self.plotter.set_data_2(dname,
layer='detected',
style='k-')
self.plotter.set_data_2(dname,
layer='scores',
style='k*')
self.plotter.show()
if k == 'd':
try:
self.crosses
except AttributeError:
print(
"Can't detect spikes until threshold crossings have been found."
)
return
self.X = self.compute_bbox()
self.default_colors = {}
if len(self.X.shape) == 1:
self.default_colors['spike0'] = 'k'
self.add_data_points([list(range(0, len(self.X))), self.X],
layer='detected',
style=self.default_colors['spike0'] + '-',
name='spike0',
force=True)
else:
c = 0
for spike in self.X:
name = 'spike' + str(c)
self.default_colors[name] = 'k'
self.add_data_points([list(range(0, len(spike))), spike],
layer='detected',
style=self.default_colors[name] + '-',
name=name,
force=True)
c += 1
self.plotter.auto_scale_domain(xcushion=0, subplot='12')
self.show()
if self.tutorial == 'step3':
print("STEP 3: ")
print(
"You can now press 'p' to perform PCA on the detected spikes."
)
print(
"The bottom right subplot will display the first 3 principal components (in red, green, and yellow respectively.)"
)
print(
"The bottom left subplot will show the detected spikes projected onto the first two PCs"
)
self.tutorial = 'step4'
if k == 'p':
try:
X = self.X
except AttributeError:
print('Must detect spikes before performing PCA.')
return
print('doing PCA...')
self.p = PCANode(output_dim=0.99, reduce=True, svd=True)
self.p.train(X)
self.proj_vec1 = self.p.get_projmatrix()[:, 0]
self.proj_vec2 = self.p.get_projmatrix()[:, 1]
self.add_data_points(
[list(range(0, len(self.proj_vec1))), self.proj_vec1],
style='r-',
layer='pcs',
name='firstPC',
force=True)
self.add_data_points(
[list(range(0, len(self.proj_vec2))), self.proj_vec2],
style='g-',
layer='pcs',
name='secondPC',
force=True)
self.plotter.show()
self.proj_PCs = ['firstPC', 'secondPC']
try:
self.add_data_points([
list(range(0, len(self.p.get_projmatrix()))),
self.p.get_projmatrix()[:, 2]
],
style='y--',
layer='pcs',
name='thirdPC',
force=True)
except IndexError:
pass
self.add_legend(['r', 'g', 'y'], ['1st PC', '2nd PC', '3rd PC'],
'21')
self.plotter.auto_scale_domain(xcushion=0, subplot='21')
self.show()
self.project_to_PC()
if self.tutorial == 'step4':
print("STEP 4: ")
print("Use mouse clicks to explore the data.")
print(
"Clicking on detected spikes in the top-right will highlight the corresponding projection in the bottom right (and vice versa)."
)
print(
"You can also change the set of PCs onto which the data are projected by clicking the desired projection PCs in the bottom left"
)
print("NOTE ALSO: ")
print(
"Creating a bounding box in the upper-left plot and renaming it to 'ref_box', will change the search width of the detected spike."
)
print(
"e.g., if you want detected spikes to be 30 msec long, the box's .dx value must be 30."
)
print(
"After creating the box, it will be set to the current selected object. You can select the thresh line again by clicking on it."
)
class PCAWrapperNode(SpatialFilteringNode): #, PCANode):
""" Reuse the implementation of the Principal Component Analysis of mdp
For a theoretical description of how PCA works, the following tutorial
is extremely useful.
======= =========================================================
Title A TUTORIAL ON PRINCIPAL COMPONENT ANALYSIS Derivation, Discussion and Singular Value Decomposition
Author Jon Shlens
Link http://www.cs.princeton.edu/picasso/mats/PCA-Tutorial-Intuition_jp.pdf
======= =========================================================
This node implements the unsupervised principal component
analysis algorithm for spatial filtering.
.. note:: The original PCANode can execute the Principal Component Analysis
in 2 ways. The first method(which is also the default) involves the
computation of the eigenvalues of a symmetric matrix. This is obviously
a rather fast approach. Nonetheless, this approach sometimes fails and
negative eigenvalues are obtained from the computation. The problem can
be solved by using the Singular Value Decomposition method in the PCA.
This is easily done by setting ``svd=True`` when initializing the
:mod:`~pySPACE.missions.nodes.spatial_filtering.pca`. The SVD approach
is more robust but also less cost-effective when it comes to computation
time.
**Parameters**
:retained_channels: Determines how many of the PCA pseudo channels
are retained. Default is None which means "all channels".
:load_path: An absolute path from which the PCA eigenmatrix
is loaded.
If not specified, this matrix is learned from the training data.
(*optional, default: None*)
**mdp parameters**
:svd: if True use Singular Value Decomposition instead of the
standard eigenvalue problem solver. Use it when PCANode
complains about singular covariance matrices
(*optional, default: False*)
:reduce: Keep only those principal components which have a variance
larger than 'var_abs' and a variance relative to the
first principal component larger than 'var_rel' and a
variance relative to total variance larger than 'var_part'
(set var_part to None or 0 for no filtering).
Note: when the 'reduce' switch is enabled, the actual number
of principal components (self.output_dim) may be different
from that set when creating the instance.
(*optional, default: False*)
**Exemplary Call**
.. code-block:: yaml
-
node : PCA
parameters:
retained_channels : 42
"""
def __init__(self,
retained_channels=None,
load_path=None,
svd=False,
reduce=False,
**kwargs):
# Must be set before constructor of superclass is set
self.trainable = (load_path == None)
super(PCAWrapperNode, self).__init__(**kwargs)
self.output_dim = retained_channels
# Load patterns from file if requested
if load_path != None:
filter_file = open(load_path, 'r')
avg, v = cPickle.load(filter_file)
self.set_permanent_attributes(avg=avg,
v=v,
trainable=False,
filters=v)
self.set_permanent_attributes( # The number of channels that will be retained
retained_channels=retained_channels,
output_dim=retained_channels,
channel_names=None,
new_channels=None,
wrapped_node=None,
svd=svd, # choose the method to use when computing the PCA
reduce=reduce)
def is_trainable(self):
""" Returns whether this node is trainable. """
return self.trainable
def is_supervised(self):
""" Returns whether this node requires supervised training. """
return False
def _train(self, data, label=None):
""" Updates the estimated covariance matrix based on *data*. """
# We simply ignore the class label since we
# are doing unsupervised learning
if self.wrapped_node is None:
self.wrapped_node = PCANode(svd=self.svd, reduce=self.reduce)
x = 1.0 * data.view(type=numpy.ndarray)
self.wrapped_node.train(x)
if self.channel_names is None:
self.channel_names = data.channel_names
def _stop_training(self, debug=False):
""" Stops training by forwarding to super class. """
self.wrapped_node.stop_training()
#super(PCAWrapperNode, self)._stop_training(debug)
self.v = self.wrapped_node.v
self.avg = self.wrapped_node.avg
self.filters = self.v
def _execute(self, data, n=None):
""" Execute learned transformation on *data*.
Projects the given data to the axis of the most significant
eigenvectors and returns the data in this lower-dimensional subspace.
"""
# 'INITIALIZATION'
if self.retained_channels == None:
self.retained_channels = data.shape[1]
if n is None:
n = self.retained_channels
if self.channel_names is None:
self.channel_names = data.channel_names
if len(self.channel_names) < self.retained_channels:
self.retained_channels = len(self.channel_names)
self._log(
"To many channels chosen for the retained channels! Replaced by maximum number.",
level=logging.CRITICAL)
if not (self.output_dim == self.retained_channels):
# overwrite internal output_dim variable, since it is set wrong
self._output_dim = self.retained_channels
# 'Real' Processing
#projected_data = super(PCANodeWrapper, self)._execute(data, n)
x = data.view(numpy.ndarray)
projected_data = mult(x - self.avg, self.v[:, :self.retained_channels])
if self.new_channels is None:
self.new_channel_names = [
"pca%03d" % i for i in range(projected_data.shape[1])
]
return TimeSeries(projected_data, self.new_channel_names,
data.sampling_frequency, data.start_time,
data.end_time, data.name, data.marker_name)
def store_state(self, result_dir, index=None):
""" Stores this node in the given directory *result_dir*. """
if self.store:
node_dir = os.path.join(result_dir, self.__class__.__name__)
create_directory(node_dir)
# This node only stores the learned eigenvector and eigenvalues
name = "%s_sp%s.pickle" % ("eigenmatrix", self.current_split)
result_file = open(os.path.join(node_dir, name), "wb")
result_file.write(cPickle.dumps((self.avg, self.v), protocol=2))
result_file.close()
class spikesorter(graphics.diagnosticGUI):
def __init__(self, title):
#global plotter
plotter = graphics.Plotter()
graphics.diagnosticGUI.__init__(self, plotter)
#Recover data:
#data = importPointset('shortsimdata1.dat',t=0,sep=',')
data = importPointset('simdata1_100000.dat',t=0,sep=',')
vs = data['vararray'][0]
vs = self.bandpass_filter(vs, 300, 3000, 32000)
ts = data['t']
self.N = len(vs)
self.traj = numeric_to_traj([vs], 'test_traj', ['x'], ts, discrete=False)
#Threshold used in martinez et al.
self.mthresh = 4*(np.median(np.abs(vs))/0.6475)
self.selected_pcs = []
self.fovea_setup()
#x = self.traj.sample(tlo= 0, thi = self.N)['x']
#r = x > 10
#above_thresh = np.where(r == True)[0]
#spike = []
#peaks = []
#crosses = []
#last_i = above_thresh[0] - 1
#for i in above_thresh:
#if i - 1 != last_i:
#crosses.append(i)
##Return x value of the highest y value.
#peaks.append(np.where(x == max(list(x[spike])))[0][0])
#spike = []
#spike.append(i)
#last_i = i
if tutorial_on:
print("STEP 1:")
print("Create a horizontal line of interest by pressing 'l'.")
print("Once created, this line can be forced to extent by pressing 'm'.")
print("Enter 'ssort.selected_object.update(name = 'thresh')' to identify the line as a threshold for spike detection")
print("Once the line is renamed to 'thresh', the arrow keys can be used to move it up and down.")
self.tutorial = 'step2'
else:
self.tutorial = None
def fovea_setup(self):
#Setup code
DOI = [(0,self.N),(-30,30)]
self.plotter.clean() # in case rerun in same session
self.plotter.add_fig('master',
title='spikesort',
xlabel='time', ylabel='mV',
domain=DOI)
#Setup all layers
self.plotter.add_layer('spikes')
self.plotter.add_layer('thresh_crosses')
self.plotter.add_layer('detected')
self.plotter.add_layer('pcs')
self.plotter.add_layer('scores')
self.setup({'11':
{'name': 'Waveform',
'scale': DOI,
'layers':['spikes', 'thresh_crosses'],
'callbacks':'*',
'axes_vars': ['x', 'y']
},
'12':
{'name': 'Detected Spikes',
'scale': [(0, default_sw), (-80, 80)],
'layers':['detected'],
#'callbacks':'*',
'axes_vars': ['x', 'y']
},
'21':
{'name': 'Principal Components',
'scale': [(0, default_sw), (-0.5, 0.5)],
'layers':['pcs'],
#'callbacks':'*',
'axes_vars': ['x', 'y']
},
'22':
{'name': 'Projected Spikes',
#'scale': [(-100, 100), (-100, 100)],
'scale': [(-300, 300), (-300, 300)],
'layers':['scores'],
'callbacks':'*',
'axes_vars': ['firstPC', 'secondPC']
}
},
size=(8, 8), with_times=False, basic_widgets=True)
#self.plotter.set_text('load_perc', Loading: %d\%'%n, 'loading')
#Bad code carried over from fovea_game:
fig_struct, figure = self.plotter._resolve_fig(None)
#self.ax = fig_struct.arrange['11']['axes_obj']
coorddict = {'x':
{'x':'t', 'layer':'spikes', 'style':'b-'}
}
self.add_data_points(self.traj.sample(), coorddict = coorddict)
evKeyOn = self.fig.canvas.mpl_connect('key_press_event', self.ssort_key_on)
self.plotter.auto_scale_domain(subplot= '11', xcushion= 0)
self.plotter.show()
def bandpass_filter(self, data, lowcut, highcut, fs, order= 5):
nyq = 0.5 * fs
low = lowcut / nyq
high = highcut / nyq
b, a = butter(order, [low, high], btype= 'band')
y = lfilter(b, a, data)
return y
def user_pick_func(self, ev):
if self.selected_object.layer == 'detected' or self.selected_object.layer == 'scores':
if hasattr(self, 'last_name'):
self.plotter.set_data_2(self.last_name, layer='scores', markersize= 6, zorder= 1, style=self.default_colors[self.last_name])
self.plotter.set_data_2(self.last_name, layer='detected', linewidth= 1, zorder= 1, style=self.default_colors[self.last_name])
self.plotter.set_data_2(self.selected_object.name, layer='scores', markersize= 12, zorder= 10, style='y*')
self.plotter.set_data_2(self.selected_object.name, layer='detected', linewidth= 2.5, zorder= 10, style='y-')
self.last_name = self.selected_object.name
elif self.selected_object.layer == 'pcs':
self.proj_PCs.insert(0, self.selected_object.name)
self.proj_PCs = self.proj_PCs[0:2]
for name in fig_struct['layers']['pcs']['data'].keys():
if name not in self.proj_PCs:
self.plotter.set_data_2(name, layer='pcs', style= fig_struct['layers']['pcs']['data'][name]['style'][0]+'--')
for pc in self.proj_PCs:
self.plotter.set_data_2(pc, layer='pcs', style= fig_struct['layers']['pcs']['data'][pc]['style'][0]+'-')
self.proj_vec1 = fig_struct['layers']['pcs']['handles'][self.proj_PCs[0]].get_ydata()
self.proj_vec2 = fig_struct['layers']['pcs']['handles'][self.proj_PCs[1]].get_ydata()
fig_struct.arrange['22']['axes_vars'] = list(reversed(self.proj_PCs))
self.project_to_PC()
self.plotter.show()
def user_update_func(self):
if self.selected_object.name is 'thresh':
if self.selected_object.m != 0:
print("Make 'thresh' a horizontal threshold by pressing 'm'.")
return
try:
self.search_width = self.context_objects['ref_box'].dx
self.context_objects['ref_box'].remove()
except KeyError:
self.search_width = default_sw
if self.tutorial == 'step2':
print("STEP 2: ")
print("When thresh is in place, press 'd' to capture each spike crossing the threshold in a bounding box.")
print("Each detected spike will be placed in the top right subplot.")
self.tutorial = 'step3'
cutoff = self.selected_object.y1
traj_samp = self.traj.sample()['x']
r = traj_samp > cutoff
above_thresh = np.where(r == True)[0]
spike = []
spikes = []
crosses = []
last_i = above_thresh[0] - 1
for i in above_thresh:
if i - 1 != last_i:
crosses.append(i)
#Return x value of the highest y value.
spikes.append(spike)
spike = []
spike.append(i)
last_i = i
self.traj_samp = traj_samp
self.crosses = crosses
self.spikes = spikes
self.plotter.add_data([self.crosses, [cutoff]*len(self.crosses)], layer='thresh_crosses', style='r*', name='crossovers', force= True)
self.show()
def compute_bbox(self):
fig_struct, figs = self.plotter._resolve_fig(None)
#Clear existing bounding boxes
rem_names = []
for con_name, con_obj in self.context_objects.items():
if isinstance(con_obj, box_GUI) and con_name is not 'ref_box':
rem_names.append(con_name)
for name in rem_names:
self.context_objects[name].remove(draw= False)
fig_struct['layers']['detected']['data'] = {}
self.plotter.show(rebuild= True)
#Create new bounding boxes
c = 0
for spike in self.spikes:
peak = np.where(self.traj_samp == max(list(self.traj_samp[spike])))[0][0]
tlo = peak - 20
thi = tlo + self.search_width
valley = min(self.traj.sample()['x'][tlo:thi])
box_GUI(self, pp.Point2D(tlo, self.traj.sample()['x'][peak]),
pp.Point2D(thi, valley),name= 'spike_box'+str(c), select= False)
spike_seg = self.traj_samp[tlo:thi]
try:
X = np.row_stack((X, spike_seg))
except NameError:
X = spike_seg
c += 1
return X
def project_to_PC(self):
Y = np.dot(self.X, np.column_stack((self.proj_vec1, self.proj_vec2)))
#If moving to a smaller number of spikes, just forcing out data by reassigning names won't work. Must clear.
self.clear_data('scores')
self.show()
self.default_colors = {}
#Add spikes as individual lines, so they can be referenced individually.
c = 0
for spike in Y:
name = 'spike'+str(c)
self.default_colors[name] = 'k'
self.add_data_points([spike[0], spike[1]], layer='scores', style=self.default_colors[name]+'*', name= name)
c += 1
self.plotter.auto_scale_domain(subplot = '22')
self.show(rebuild = True)
def ssort_key_on(self, ev):
self._key = k = ev.key # keep record of last keypress
fig_struct, fig = self.plotter._resolve_fig(None)
class_keys = ['1','2','3','0']
if k in class_keys:
if isinstance(self.selected_object, box_GUI):
for dname, dstruct in fig_struct['layers']['scores']['data'].items():
if self.selected_object.x1 < dstruct['data'][0] < self.selected_object.x2 and \
self.selected_object.y1 < dstruct['data'][1] < self.selected_object.y2:
if k == '1':
self.default_colors[dname] = 'r'
self.plotter.set_data_2(dname, layer='detected', style= 'r-')
self.plotter.set_data_2(dname, layer='scores', style= 'r*')
if k == '2':
self.default_colors[dname] = 'g'
self.plotter.set_data_2(dname, layer='detected', style= 'g-')
self.plotter.set_data_2(dname, layer='scores', style= 'g*')
if k == '3':
self.default_colors[dname] = 'b'
self.plotter.set_data_2(dname, layer='detected', style= 'b-')
self.plotter.set_data_2(dname, layer='scores', style= 'b*')
if k == '0':
self.default_colors[dname] = 'k'
self.plotter.set_data_2(dname, layer='detected', style= 'k-')
self.plotter.set_data_2(dname, layer='scores', style= 'k*')
self.plotter.show()
if k== 'd':
try:
self.crosses
except AttributeError:
print("Can't detect spikes until threshold crossings have been found.")
return
self.X = self.compute_bbox()
self.default_colors = {}
if len(self.X.shape) == 1:
self.default_colors['spike0'] = 'k'
self.add_data_points([list(range(0, len(self.X))), self.X], layer= 'detected', style= self.default_colors['spike0']+'-', name= 'spike0', force= True)
else:
c= 0
for spike in self.X:
name = 'spike'+str(c)
self.default_colors[name] = 'k'
self.add_data_points([list(range(0, len(spike))), spike], layer= 'detected', style= self.default_colors[name]+'-', name= name, force= True)
c += 1
self.plotter.auto_scale_domain(xcushion = 0, subplot = '12')
self.show()
if self.tutorial == 'step3':
print("STEP 3: ")
print("You can now press 'p' to perform PCA on the detected spikes.")
print("The bottom right subplot will display the first 3 principal components (in red, green, and yellow respectively.)")
print("The bottom left subplot will show the detected spikes projected onto the first two PCs")
self.tutorial = 'step4'
if k == 'p':
try:
X = self.X
except AttributeError:
print('Must detect spikes before performing PCA.')
return
print('doing PCA...')
self.p = PCANode(output_dim=0.99, reduce= True, svd= True)
self.p.train(X)
self.proj_vec1 = self.p.get_projmatrix()[:, 0]
self.proj_vec2 = self.p.get_projmatrix()[:, 1]
self.add_data_points([list(range(0, len(self.proj_vec1))) , self.proj_vec1], style= 'r-', layer= 'pcs', name= 'firstPC', force= True)
self.add_data_points([list(range(0, len(self.proj_vec2))) , self.proj_vec2], style= 'g-', layer= 'pcs', name= 'secondPC', force= True)
self.plotter.show()
self.proj_PCs = ['firstPC', 'secondPC']
try:
self.add_data_points([list(range(0, len(self.p.get_projmatrix()))) ,self.p.get_projmatrix()[:,2]],
style= 'y--', layer= 'pcs', name= 'thirdPC', force= True)
except IndexError:
pass
self.add_legend(['r', 'g', 'y'], ['1st PC', '2nd PC', '3rd PC'], '21')
self.plotter.auto_scale_domain(xcushion = 0, subplot = '21')
self.show()
self.project_to_PC()
if self.tutorial == 'step4':
print("STEP 4: ")
print("Use mouse clicks to explore the data.")
print("Clicking on detected spikes in the top-right will highlight the corresponding projection in the bottom right (and vice versa).")
print("You can also change the set of PCs onto which the data are projected by clicking the desired projection PCs in the bottom left")
print("NOTE ALSO: ")
print("Creating a bounding box in the upper-left plot and renaming it to 'ref_box', will change the search width of the detected spike.")
print("e.g., if you want detected spikes to be 30 msec long, the box's .dx value must be 30.")
print("After creating the box, it will be set to the current selected object. You can select the thresh line again by clicking on it.")
class PCAWrapperNode(SpatialFilteringNode): #, PCANode):
""" Reuse the implementation of the Principal Component Analysis of mdp
This node implements the unsupervised principal component
analysis algorithm for spatial filtering.
**Parameters**
:retained_channels: Determines how many of the PCA pseudo channels
are retained. Default is None which means "all channels".
:load_path: An absolute path from which the PCA eigenmatrix
is loaded.
If not specified, this matrix is learned from the training data.
(*optional, default: None*)
**Exemplary Call**
.. code-block:: yaml
-
node : PCA
parameters:
retained_channels : 42
"""
def __init__(self, retained_channels=None, load_path=None, **kwargs):
# Must be set before constructor of superclass is set
self.trainable = (load_path == None)
super(PCAWrapperNode, self).__init__(**kwargs)
self.output_dim = retained_channels
# Load patterns from file if requested
if load_path != None:
filter_file = open(load_path, 'r')
avg, v = cPickle.load(filter_file)
self.set_permanent_attributes(avg=avg, v=v, trainable=False, filters=v)
self.set_permanent_attributes(# The number of channels that will be retained
retained_channels = retained_channels,
output_dim = retained_channels,
channel_names = None,
new_channels = None,
wrapped_node=None
)
def is_trainable(self):
""" Returns whether this node is trainable. """
return self.trainable
def is_supervised(self):
""" Returns whether this node requires supervised training. """
return False
def _train(self, data, label = None):
""" Updates the estimated covariance matrix based on *data*. """
# We simply ignore the class label since we
# are doing unsupervised learning
if self.wrapped_node is None:
self.wrapped_node = PCANode()
x=data.view(numpy.ndarray)
self.wrapped_node.train(x)
if self.channel_names is None:
self.channel_names = data.channel_names
def _stop_training(self, debug=False):
""" Stops training by forwarding to super class. """
self.wrapped_node.stop_training()
#super(PCAWrapperNode, self)._stop_training(debug)
self.v = self.wrapped_node.v
self.avg = self.wrapped_node.avg
self.filters = self.v
def _execute(self, data, n = None):
""" Execute learned transformation on *data*.
Projects the given data to the axis of the most significant
eigenvectors and returns the data in this lower-dimensional subspace.
"""
# 'INITIALIZATION'
if self.retained_channels==None:
self.retained_channels = data.shape[1]
if n is None:
n = self.retained_channels
if self.channel_names is None:
self.channel_names = data.channel_names
if len(self.channel_names)<self.retained_channels:
self.retained_channels = len(self.channel_names)
self._log("To many channels chosen for the retained channels! Replaced by maximum number.",level=logging.CRITICAL)
if not(self.output_dim==self.retained_channels):
# overwrite internal output_dim variable, since it is set wrong
self._output_dim = self.retained_channels
# 'Real' Processing
#projected_data = super(PCANodeWrapper, self)._execute(data, n)
x = data.view(numpy.ndarray)
projected_data = mult(x-self.avg, self.v[:, :self.retained_channels])
if self.new_channels is None:
self.new_channel_names = ["pca%03d" % i
for i in range(projected_data.shape[1])]
return TimeSeries(projected_data, self.new_channel_names,
data.sampling_frequency, data.start_time,
data.end_time, data.name, data.marker_name)
def store_state(self, result_dir, index=None):
""" Stores this node in the given directory *result_dir*. """
if self.store:
node_dir = os.path.join(result_dir, self.__class__.__name__)
create_directory(node_dir)
# This node only stores the learned eigenvector and eigenvalues
name = "%s_sp%s.pickle" % ("eigenmatrix", self.current_split)
result_file = open(os.path.join(node_dir, name), "wb")
result_file.write(cPickle.dumps((self.avg, self.v), protocol=2))
result_file.close()
def pca(multidim_data, output_dim):
"""Principal Component Analysis"""
pcanode = PCANode(input_dim=multidim_data.shape[1], output_dim=output_dim, dtype=np.float32,
svd=True, reduce=True, var_rel=1E-15, var_abs=1E-15)
pcanode.train(data)
return np.dot(multidim_data, pcanode.get_projmatrix())
def ssort_key_on(self, ev):
self._key = k = ev.key # keep record of last keypress
fig_struct, fig = self.plotter._resolve_fig(None)
class_keys = ['1','2','3','0']
if k in class_keys:
if isinstance(self.selected_object, box_GUI):
for dname, dstruct in fig_struct['layers']['scores']['data'].items():
if self.selected_object.x1 < dstruct['data'][0] < self.selected_object.x2 and \
self.selected_object.y1 < dstruct['data'][1] < self.selected_object.y2:
if k == '1':
self.default_colors[dname] = 'r'
self.plotter.set_data_2(dname, layer='detected', style= 'r-')
self.plotter.set_data_2(dname, layer='scores', style= 'r*')
if k == '2':
self.default_colors[dname] = 'g'
self.plotter.set_data_2(dname, layer='detected', style= 'g-')
self.plotter.set_data_2(dname, layer='scores', style= 'g*')
if k == '3':
self.default_colors[dname] = 'b'
self.plotter.set_data_2(dname, layer='detected', style= 'b-')
self.plotter.set_data_2(dname, layer='scores', style= 'b*')
if k == '0':
self.default_colors[dname] = 'k'
self.plotter.set_data_2(dname, layer='detected', style= 'k-')
self.plotter.set_data_2(dname, layer='scores', style= 'k*')
self.plotter.show()
if k== 'd':
try:
self.crosses
except AttributeError:
print("Can't detect spikes until threshold crossings have been found.")
return
self.X = self.compute_bbox()
self.default_colors = {}
if len(self.X.shape) == 1:
self.default_colors['spike0'] = 'k'
self.add_data_points([list(range(0, len(self.X))), self.X], layer= 'detected', style= self.default_colors['spike0']+'-', name= 'spike0', force= True)
else:
c= 0
for spike in self.X:
name = 'spike'+str(c)
self.default_colors[name] = 'k'
self.add_data_points([list(range(0, len(spike))), spike], layer= 'detected', style= self.default_colors[name]+'-', name= name, force= True)
c += 1
self.plotter.auto_scale_domain(xcushion = 0, subplot = '12')
self.show()
if self.tutorial == 'step3':
print("STEP 3: ")
print("You can now press 'p' to perform PCA on the detected spikes.")
print("The bottom right subplot will display the first 3 principal components (in red, green, and yellow respectively.)")
print("The bottom left subplot will show the detected spikes projected onto the first two PCs")
self.tutorial = 'step4'
if k == 'p':
try:
X = self.X
except AttributeError:
print('Must detect spikes before performing PCA.')
return
print('doing PCA...')
self.p = PCANode(output_dim=0.99, reduce= True, svd= True)
self.p.train(X)
self.proj_vec1 = self.p.get_projmatrix()[:, 0]
self.proj_vec2 = self.p.get_projmatrix()[:, 1]
self.add_data_points([list(range(0, len(self.proj_vec1))) , self.proj_vec1], style= 'r-', layer= 'pcs', name= 'firstPC', force= True)
self.add_data_points([list(range(0, len(self.proj_vec2))) , self.proj_vec2], style= 'g-', layer= 'pcs', name= 'secondPC', force= True)
self.plotter.show()
self.proj_PCs = ['firstPC', 'secondPC']
try:
self.add_data_points([list(range(0, len(self.p.get_projmatrix()))) ,self.p.get_projmatrix()[:,2]],
style= 'y--', layer= 'pcs', name= 'thirdPC', force= True)
except IndexError:
pass
self.add_legend(['r', 'g', 'y'], ['1st PC', '2nd PC', '3rd PC'], '21')
self.plotter.auto_scale_domain(xcushion = 0, subplot = '21')
self.show()
self.project_to_PC()
if self.tutorial == 'step4':
print("STEP 4: ")
print("Use mouse clicks to explore the data.")
print("Clicking on detected spikes in the top-right will highlight the corresponding projection in the bottom right (and vice versa).")
print("You can also change the set of PCs onto which the data are projected by clicking the desired projection PCs in the bottom left")
print("NOTE ALSO: ")
print("Creating a bounding box in the upper-left plot and renaming it to 'ref_box', will change the search width of the detected spike.")
print("e.g., if you want detected spikes to be 30 msec long, the box's .dx value must be 30.")
print("After creating the box, it will be set to the current selected object. You can select the thresh line again by clicking on it.")
class PCAWrapperNode(SpatialFilteringNode): #, PCANode):
""" Reuse the implementation of the Principal Component Analysis of mdp
For a theoretical description of how PCA works, the following tutorial
is extremely useful.
======= =========================================================
Title A TUTORIAL ON PRINCIPAL COMPONENT ANALYSIS Derivation, Discussion and Singular Value Decomposition
Author Jon Shlens
Link http://www.cs.princeton.edu/picasso/mats/PCA-Tutorial-Intuition_jp.pdf
======= =========================================================
This node implements the unsupervised principal component
analysis algorithm for spatial filtering.
.. note:: The original PCANode can execute the Principal Component Analysis
in 2 ways. The first method(which is also the default) involves the
computation of the eigenvalues of a symmetric matrix. This is obviously
a rather fast approach. Nonetheless, this approach sometimes fails and
negative eigenvalues are obtained from the computation. The problem can
be solved by using the Singular Value Decomposition method in the PCA.
This is easily done by setting ``svd=True`` when initializing the
:mod:`~pySPACE.missions.nodes.spatial_filtering.pca`. The SVD approach
is more robust but also less cost-effective when it comes to computation
time.
**Parameters**
:retained_channels: Determines how many of the PCA pseudo channels
are retained. Default is None which means "all channels".
:load_path: An absolute path from which the PCA eigenmatrix
is loaded.
If not specified, this matrix is learned from the training data.
(*optional, default: None*)
**mdp parameters**
:svd: if True use Singular Value Decomposition instead of the
standard eigenvalue problem solver. Use it when PCANode
complains about singular covariance matrices
(*optional, default: False*)
:reduce: Keep only those principal components which have a variance
larger than 'var_abs' and a variance relative to the
first principal component larger than 'var_rel' and a
variance relative to total variance larger than 'var_part'
(set var_part to None or 0 for no filtering).
Note: when the 'reduce' switch is enabled, the actual number
of principal components (self.output_dim) may be different
from that set when creating the instance.
(*optional, default: False*)
**Exemplary Call**
.. code-block:: yaml
-
node : PCA
parameters:
retained_channels : 42
"""
def __init__(self, retained_channels=None, load_path=None,
svd=False, reduce=False, **kwargs):
# Must be set before constructor of superclass is set
self.trainable = (load_path == None)
super(PCAWrapperNode, self).__init__(**kwargs)
self.output_dim = retained_channels
# Load patterns from file if requested
if load_path != None:
filter_file = open(load_path, 'r')
avg, v = cPickle.load(filter_file)
self.set_permanent_attributes(avg=avg, v=v, trainable=False, filters=v)
self.set_permanent_attributes( # The number of channels that will be retained
retained_channels=retained_channels,
output_dim=retained_channels,
channel_names=None,
new_channels=None,
wrapped_node=None,
svd=svd, # choose the method to use when computing the PCA
reduce=reduce
)
def is_trainable(self):
""" Returns whether this node is trainable. """
return self.trainable
def is_supervised(self):
""" Returns whether this node requires supervised training. """
return False
def _train(self, data, label = None):
""" Updates the estimated covariance matrix based on *data*. """
# We simply ignore the class label since we
# are doing unsupervised learning
if self.wrapped_node is None:
self.wrapped_node = PCANode(svd=self.svd, reduce=self.reduce)
x = 1.0 * data.view(type=numpy.ndarray)
self.wrapped_node.train(x)
if self.channel_names is None:
self.channel_names = data.channel_names
def _stop_training(self, debug=False):
""" Stops training by forwarding to super class. """
self.wrapped_node.stop_training()
#super(PCAWrapperNode, self)._stop_training(debug)
self.v = self.wrapped_node.v
self.avg = self.wrapped_node.avg
self.filters = self.v
def _execute(self, data, n = None):
""" Execute learned transformation on *data*.
Projects the given data to the axis of the most significant
eigenvectors and returns the data in this lower-dimensional subspace.
"""
# 'INITIALIZATION'
if self.retained_channels==None:
self.retained_channels = data.shape[1]
if n is None:
n = self.retained_channels
if self.channel_names is None:
self.channel_names = data.channel_names
if len(self.channel_names)<self.retained_channels:
self.retained_channels = len(self.channel_names)
self._log("To many channels chosen for the retained channels! Replaced by maximum number.",level=logging.CRITICAL)
if not(self.output_dim==self.retained_channels):
# overwrite internal output_dim variable, since it is set wrong
self._output_dim = self.retained_channels
# 'Real' Processing
#projected_data = super(PCANodeWrapper, self)._execute(data, n)
x = data.view(numpy.ndarray)
projected_data = mult(x-self.avg, self.v[:, :self.retained_channels])
if self.new_channels is None:
self.new_channel_names = ["pca%03d" % i
for i in range(projected_data.shape[1])]
return TimeSeries(projected_data, self.new_channel_names,
data.sampling_frequency, data.start_time,
data.end_time, data.name, data.marker_name)
def store_state(self, result_dir, index=None):
""" Stores this node in the given directory *result_dir*. """
if self.store:
node_dir = os.path.join(result_dir, self.__class__.__name__)
create_directory(node_dir)
# This node only stores the learned eigenvector and eigenvalues
name = "%s_sp%s.pickle" % ("eigenmatrix", self.current_split)
result_file = open(os.path.join(node_dir, name), "wb")
result_file.write(cPickle.dumps((self.avg, self.v), protocol=2))
result_file.close()
