def create_threads(self):
# Create the external threads.
self.tasks = ThreadedTasks()
self.tasks.selection_task.set_loader(self.loader)
self.tasks.selection_task.selectionDone.connect(
self.selection_done_callback)
self.tasks.correlograms_task.correlogramsComputed.connect(
self.correlograms_computed_callback)
self.tasks.similarity_matrix_task.correlationMatrixComputed.connect(
self.similarity_matrix_computed_callback)
def create_threads(self):
# Create the external threads.
self.tasks = ThreadedTasks()
self.tasks.selection_task.set_loader(self.loader)
self.tasks.selection_task.selectionDone.connect(self.selection_done_callback)
self.tasks.recluster_task.reclusterDone.connect(self.recluster_done_callback)
self.tasks.correlograms_task.correlogramsComputed.connect(self.correlograms_computed_callback)
self.tasks.similarity_matrix_task.correlationMatrixComputed.connect(self.similarity_matrix_computed_callback)
class TaskGraph(AbstractTaskGraph):
def __init__(self, mainwindow):
# Shortcuts for the main window.
self.set(mainwindow)
# Create external threads/processes for long-lasting tasks.
self.create_threads()
def set(self, mainwindow):
# Shortcuts for the main window.
self.mainwindow = mainwindow
self.get_view = self.mainwindow.get_view
self.get_views = self.mainwindow.get_views
self.loader = self.mainwindow.loader
self.wizard = self.mainwindow.wizard
self.controller = self.mainwindow.controller
self.statscache = self.mainwindow.statscache
def create_threads(self):
# Create the external threads.
self.tasks = ThreadedTasks()
self.tasks.selection_task.set_loader(self.loader)
self.tasks.selection_task.selectionDone.connect(
self.selection_done_callback)
self.tasks.correlograms_task.correlogramsComputed.connect(
self.correlograms_computed_callback)
self.tasks.similarity_matrix_task.correlationMatrixComputed.connect(
self.similarity_matrix_computed_callback)
def join(self):
self.tasks.join()
# Selection.
# ----------
def _select(self, clusters, wizard=False):
self.tasks.selection_task.select(clusters, wizard)
def _select_done(self, clusters, wizard=False):
if wizard:
target = (self.wizard.current_target(),)
else:
target = ()
# self.loader.select(clusters=clusters)
log.debug("Selected clusters {0:s}.".format(str(clusters)))
return [
('_update_feature_view', target),
('_update_waveform_view', (), dict(wizard=wizard)),
('_show_selection_in_matrix', (clusters,)),
('_compute_correlograms', (clusters,),),
]
def _select_in_cluster_view(self, clusters, groups=[], wizard=False):
self.get_view('ClusterView').select(clusters, groups=groups,
wizard=wizard)
# Callbacks.
# ----------
def selection_done_callback(self, clusters, wizard):
self.select_done(clusters, wizard=wizard)
def correlograms_computed_callback(self, clusters, correlograms, ncorrbins,
corrbin):
# Execute the callback function under the control of the task manager
# (which handles the graph dependency).
self.correlograms_computed(clusters, correlograms, ncorrbins, corrbin)
def similarity_matrix_computed_callback(self, clusters_selected, matrix,
clusters, cluster_groups, target_next=None):
# Execute the callback function under the control of the task manager
# (which handles the graph dependency).
self.similarity_matrix_computed(clusters_selected, matrix, clusters,
cluster_groups, target_next=target_next)
# Computations.
# -------------
def _compute_correlograms(self, clusters_selected):
# Get the correlograms parameters.
spiketimes = get_array(self.loader.get_spiketimes('all'))
# Make a copy of the array so that it does not change before the
# computation of the correlograms begins.
clusters = np.array(get_array(self.loader.get_clusters('all')))
# corrbin = self.loader.corrbin
# ncorrbins = self.loader.ncorrbins
corrbin = SETTINGS.get('correlograms.corrbin', .001)
ncorrbins = SETTINGS.get('correlograms.ncorrbins', 100)
# Get cluster indices that need to be updated.
clusters_to_update = (self.statscache.correlograms.
not_in_key_indices(clusters_selected))
# If there are pairs that need to be updated, launch the task.
if len(clusters_to_update) > 0:
# Set wait cursor.
self.mainwindow.set_busy(computing_correlograms=True)
# Launch the task.
self.tasks.correlograms_task.compute(spiketimes, clusters,
clusters_to_update=clusters_to_update,
clusters_selected=clusters_selected,
ncorrbins=ncorrbins, corrbin=corrbin)
# Otherwise, update directly the correlograms view without launching
# the task in the external process.
else:
# self.update_correlograms_view()
return '_update_correlograms_view'
def _compute_similarity_matrix(self, target_next=None):
similarity_measure = self.loader.similarity_measure
features = self.loader.background_features
masks = self.loader.background_masks
clusters = get_array(self.loader.get_clusters(
spikes=self.loader.background_spikes))
cluster_groups = get_array(self.loader.get_cluster_groups('all'))
clusters_all = self.loader.get_clusters_unique()
# Get cluster indices that need to be updated.
# if clusters_to_update is None:
# NOTE: not specifying explicitely clusters_to_update ensures that
# all clusters that need to be updated are updated.
# Allows to fix a bug where the matrix is not updated correctly
# when multiple calls to this functions are called quickly.
clusters_to_update = (self.statscache.similarity_matrix.
not_in_key_indices(clusters_all))
log.debug("Clusters to update: {0:s}".format(str(clusters_to_update)))
# If there are pairs that need to be updated, launch the task.
if len(clusters_to_update) > 0:
self.mainwindow.set_busy(computing_matrix=True)
# Launch the task.
self.tasks.similarity_matrix_task.compute(features,
clusters, cluster_groups, masks, clusters_to_update,
target_next=target_next, similarity_measure=similarity_measure)
# Otherwise, update directly the correlograms view without launching
# the task in the external process.
else:
return [('_wizard_update', (target_next,)),
('_update_similarity_matrix_view',),
]
def _correlograms_computed(self, clusters, correlograms, ncorrbins, corrbin):
clusters_selected = self.loader.get_clusters_selected()
# Abort if the selection has changed during the computation of the
# correlograms.
# Reset the cursor.
self.mainwindow.set_busy(computing_correlograms=False)
if not np.array_equal(clusters, clusters_selected):
log.debug("Skip update correlograms with clusters selected={0:s}"
" and clusters updated={1:s}.".format(clusters_selected, clusters))
return
if self.statscache.ncorrbins != ncorrbins:
log.debug(("Skip updating correlograms because ncorrbins has "
"changed (from {0:d} to {1:d})".format(
ncorrbins, self.statscache.ncorrbins)))
return
# Put the computed correlograms in the cache.
self.statscache.correlograms.update(clusters, correlograms)
# Update the view.
# self.update_correlograms_view()
return '_update_correlograms_view'
def _similarity_matrix_computed(self, clusters_selected, matrix, clusters,
cluster_groups, target_next=None):
self.mainwindow.set_busy(computing_matrix=False)
# spikes_slice = _get_similarity_matrix_slice(
# self.loader.nspikes,
# len(self.loader.get_clusters_unique()))
clusters_now = self.loader.get_clusters(
spikes=self.loader.background_spikes)
if not np.array_equal(clusters, clusters_now):
return False
self.statscache.similarity_matrix.update(clusters_selected, matrix)
self.statscache.similarity_matrix_normalized = normalize(
self.statscache.similarity_matrix.to_array(copy=True))
# Update the cluster view with cluster quality.
quality = np.diag(self.statscache.similarity_matrix_normalized)
self.statscache.cluster_quality = pd.Series(
quality,
index=self.statscache.similarity_matrix.indices,
)
self.get_view('ClusterView').set_quality(
self.statscache.cluster_quality)
return [('_wizard_update', (target_next,)),
('_update_similarity_matrix_view',),
]
def _invalidate(self, clusters):
self.statscache.invalidate(clusters)
# View updates.
# -------------
def _update_correlograms_view(self):
data = vd.get_correlogramsview_data(self.loader, self.statscache)
[view.set_data(**data) for view in self.get_views('CorrelogramsView')]
def _update_similarity_matrix_view(self):
data = vd.get_similaritymatrixview_data(self.loader, self.statscache)
[view.set_data(**data)
for view in self.get_views('SimilarityMatrixView')]
# Show selected clusters when the matrix has been updated.
clusters = self.loader.get_clusters_selected()
return ('_show_selection_in_matrix', (clusters,))
def _update_feature_view(self, autozoom=None):
data = vd.get_featureview_data(self.loader,
autozoom=autozoom)
[view.set_data(**data) for view in self.get_views('FeatureView')]
def _update_waveform_view(self, autozoom=None, wizard=None):
data = vd.get_waveformview_data(self.loader, autozoom=autozoom,
wizard=wizard)
[view.set_data(**data) for view in self.get_views('WaveformView')]
def _update_trace_view(self):
data = vd.get_traceview_data(self.loader)
[view.set_data(**data) for view in self.get_views('TraceView')]
def _update_cluster_view(self, clusters=None):
"""Update the cluster view using the data stored in the loader
object."""
data = vd.get_clusterview_data(self.loader, self.statscache,
clusters=clusters)
self.get_view('ClusterView').set_data(**data)
if clusters is not None:
return
def _show_selection_in_matrix(self, clusters):
if clusters is not None and 1 <= len(clusters) <= 2:
[view.show_selection(clusters[0], clusters[-1])
for view in self.get_views('SimilarityMatrixView')]
# Override colors.
# ----------------
def _override_color(self, override_color):
self.loader.set_override_color(override_color)
return ['_update_feature_view', '_update_waveform_view', '_update_correlograms_view']
# Change correlograms parameter.
# ------------------------------
def _change_correlograms_parameters(self, ncorrbins=None, corrbin=None):
# Update the correlograms parameters.
if ncorrbins is not None:
SETTINGS['correlograms.ncorrbins'] = ncorrbins
if corrbin is not None:
SETTINGS['correlograms.corrbin'] = corrbin
# Reset the cache.
self.statscache.reset(ncorrbins)
# Update the correlograms.
clusters = self.loader.get_clusters_selected()
return ('_compute_correlograms', (clusters,))
# Merge/split actions.
# --------------------
def _merge(self, clusters, wizard=False):
if len(clusters) >= 2:
action, output = self.controller.merge_clusters(clusters)
# Tell the next nodes whether the merge occurred after a wizard
# selection or not, so that the merged cluster background is
# highlighted or not.
output['wizard'] = wizard
return after_merge(output)
def _split(self, clusters, spikes_selected, wizard=False):
if len(spikes_selected) >= 1:
action, output = self.controller.split_clusters(clusters,
spikes_selected)
output['wizard'] = wizard
return after_split(output)
def _undo(self, wizard=False):
undo = self.controller.undo()
if undo is None:
return
action, output = undo
output['wizard'] = wizard
if action == 'merge_clusters_undo':
return after_merge_undo(output)
elif action == 'split_clusters_undo':
return after_split_undo(output)
elif action == 'change_cluster_color_undo':
return after_cluster_color_changed_undo(output)
elif action == 'change_group_color_undo':
return after_group_color_changed(output)
elif action == 'move_clusters_undo':
return after_clusters_moved_undo(output)
elif action == 'add_group_undo':
return after_group_added(output)
elif action == 'rename_group_undo':
return after_group_renamed(output)
elif action == 'remove_group_undo':
return after_group_removed(output)
def _redo(self, wizard=False):
redo = self.controller.redo()
if redo is None:
return
action, output = redo
output['wizard'] = wizard
if action == 'merge_clusters':
return after_merge(output)
elif action == 'split_clusters':
return after_split(output)
elif action == 'change_cluster_color':
return after_cluster_color_changed(output)
elif action == 'change_group_color':
return after_group_color_changed(output)
elif action == 'move_clusters':
return after_clusters_moved(output)
elif action == 'add_group':
return after_group_added(output)
elif action == 'rename_group':
return after_group_renamed(output)
elif action == 'remove_group':
return after_group_removed(output)
# Other actions.
# --------------
def _cluster_color_changed(self, cluster, color, wizard=True):
action, output = self.controller.change_cluster_color(cluster, color)
# if cluster == self.wizard.current_target():
output['wizard'] = wizard
return after_cluster_color_changed(output)
def _group_color_changed(self, group, color):
action, output = self.controller.change_group_color(group, color)
return after_group_color_changed(output)
def _group_renamed(self, group, name):
action, output = self.controller.rename_group(group, name)
return after_group_renamed(output)
def _clusters_moved(self, clusters, group, wizard=False,):
action, output = self.controller.move_clusters(clusters, group)
output['wizard'] = wizard
return after_clusters_moved(output)
def _group_removed(self, group):
action, output = self.controller.remove_group(group)
return after_group_removed(output)
def _group_added(self, group, name, color):
action, output = self.controller.add_group(group, name, color)
return after_group_added(output)
# Wizard.
# -------
def _wizard_update(self, target=None, update_matrix=True):
if update_matrix:
self.wizard.set_data(
cluster_groups=self.loader.get_cluster_groups('all'),
similarity_matrix=self.statscache.similarity_matrix_normalized,
)
else:
self.wizard.set_data(
cluster_groups=self.loader.get_cluster_groups('all'),
)
self.wizard.update_candidates(target)
def _wizard_change_color(self, clusters):
if clusters is not None:
# Set the background color in the cluster view for the wizard
# target and candidate.
self.get_view('ClusterView').set_background(
{cluster: {0: 'target', 1: 'candidate'}.get(i, None)
for i, cluster in enumerate(clusters[:2])})
def _wizard_change_candidate_color(self):
candidate = self.wizard.current_candidate()
target = self.wizard.current_target()
# color = self.loader.get_cluster_color(candidate)
return ('_cluster_color_changed', (candidate, random_color(),))
def _wizard_show_pair(self, target=None, candidate=None):
if target is None:
target = (self.wizard.current_target(),
self.loader.get_cluster_color(self.wizard.current_target()))
if candidate is None:
candidate = (self.wizard.current_candidate(),
self.loader.get_cluster_color(self.wizard.current_candidate()))
[view.set_wizard_pair(target, candidate)
for view in self.get_views('FeatureView')]
# Navigation.
def _wizard_reset(self):
clusters = self.wizard.reset()
return ['_wizard_update', '_wizard_current_candidate']
def _wizard_previous_candidate(self):
clusters = self.wizard.previous_pair()
return after_wizard_selection(clusters)
def _wizard_current_candidate(self):
clusters = self.wizard.current_pair()
return after_wizard_selection(clusters)
def _wizard_next_candidate(self):
clusters = self.wizard.next_pair()
return after_wizard_selection(clusters)
def _wizard_skip_target(self):
# Skip the current target and go the next target.
self.wizard.skip_target()
return [('_wizard_update', ()),
('_wizard_next_candidate',),]
def _wizard_reset_skipped(self):
self.wizard.reset_skipped()
# Control.
def _wizard_move_and_next(self, what, group):
"""Move target, candidate, or both, to a given group, and go to
the next proposition."""
# Current proposition.
clusters = self.wizard.current_pair()
if clusters is None:
return
target, candidate = clusters
# Select appropriate clusters to move.
if what == 'candidate':
clusters = [candidate]
# Keep the current target.
target_next = target
reset_skipped = False
elif what == 'target':
clusters = [target]
# Go to the next best target cluster.
target_next = None
reset_skipped = True
elif what == 'both':
clusters = [candidate, target]
# Go to the next best target cluster.
target_next = None
reset_skipped = True
# Move clusters, and select next proposition.
r = [('_clusters_moved', (clusters, group, True)),
]
if reset_skipped:
r += [('_wizard_reset_skipped',),]
r += [('_wizard_update', (target_next,)),
('_wizard_next_candidate',),
]
return r
class TaskGraph(AbstractTaskGraph):
def __init__(self, mainwindow):
# Shortcuts for the main window.
self.set(mainwindow)
# Create external threads/processes for long-lasting tasks.
self.create_threads()
def set(self, mainwindow):
# Shortcuts for the main window.
self.mainwindow = mainwindow
self.get_view = self.mainwindow.get_view
self.get_views = self.mainwindow.get_views
self.loader = self.mainwindow.loader
self.experiment = self.loader.experiment
self.wizard = self.mainwindow.wizard
self.controller = self.mainwindow.controller
self.statscache = self.mainwindow.statscache
def create_threads(self):
# Create the external threads.
self.tasks = ThreadedTasks()
self.tasks.selection_task.set_loader(self.loader)
self.tasks.selection_task.selectionDone.connect(self.selection_done_callback)
self.tasks.recluster_task.reclusterDone.connect(self.recluster_done_callback)
self.tasks.correlograms_task.correlogramsComputed.connect(self.correlograms_computed_callback)
self.tasks.similarity_matrix_task.correlationMatrixComputed.connect(self.similarity_matrix_computed_callback)
def join(self):
self.tasks.join()
# Selection.
# ----------
def _select(self, clusters, wizard=False):
self.tasks.selection_task.select(clusters, wizard)
def _select_done(self, clusters, wizard=False):
if wizard:
target = (self.wizard.current_target(),)
else:
target = ()
# self.loader.select(clusters=clusters)
log.debug("Selected clusters {0:s}.".format(str(clusters)))
return [
("_update_feature_view", target, dict()),
("_update_waveform_view", (), dict(wizard=wizard)),
("_show_selection_in_matrix", (clusters,)),
("_compute_correlograms", (clusters,), dict(wizard=wizard)),
]
def _select_in_cluster_view(self, clusters, groups=[], wizard=False):
self.get_view("ClusterView").select(clusters, groups=groups, wizard=wizard)
# Callbacks.
# ----------
def selection_done_callback(self, clusters, wizard):
self.select_done(clusters, wizard=wizard)
def recluster_done_callback(self, channel_group, clusters, spikes, clu, wizard):
self.recluster_done(channel_group=channel_group, clusters=clusters, spikes=spikes, clu=clu, wizard=wizard)
def correlograms_computed_callback(self, clusters, correlograms, ncorrbins, corrbin, sample_rate, wizard):
# Execute the callback function under the control of the task manager
# (which handles the graph dependency).
self.correlograms_computed(clusters, correlograms, ncorrbins, corrbin, sample_rate, wizard)
def similarity_matrix_computed_callback(
self, clusters_selected, matrix, clusters, cluster_groups, target_next=None
):
# Execute the callback function under the control of the task manager
# (which handles the graph dependency).
self.similarity_matrix_computed(clusters_selected, matrix, clusters, cluster_groups, target_next=target_next)
# Computations.
# -------------
def _compute_correlograms(self, clusters_selected, wizard=None):
# Get the correlograms parameters.
spiketimes = get_array(self.loader.get_spiketimes("all"))
sample_rate = self.loader.freq
# print spiketimes.dtype
# Make a copy of the array so that it does not change before the
# computation of the correlograms begins.
clusters = np.array(get_array(self.loader.get_clusters("all")))
# Get excerpts
nexcerpts = USERPREF.get("correlograms_nexcerpts", 50)
excerpt_size = USERPREF.get("correlograms_excerpt_size", 10000)
spiketimes_excerpts = get_excerpts(spiketimes, nexcerpts=nexcerpts, excerpt_size=excerpt_size)
clusters_excerpts = get_excerpts(clusters, nexcerpts=nexcerpts, excerpt_size=excerpt_size)
# corrbin = self.loader.corrbin
# ncorrbins = self.loader.ncorrbins
corrbin = SETTINGS.get("correlograms.corrbin", 0.001)
ncorrbins = SETTINGS.get("correlograms.ncorrbins", 101)
# Ensure ncorrbins is odd.
if ncorrbins % 2 == 0:
ncorrbins += 1
# Get cluster indices that need to be updated.
# clusters_to_update = (self.statscache.correlograms.
# not_in_key_indices(clusters_selected))
clusters_to_update = clusters_selected
# If there are pairs that need to be updated, launch the task.
if len(clusters_to_update) > 0:
# Set wait cursor.
self.mainwindow.set_busy(computing_correlograms=True)
# Launch the task.
self.tasks.correlograms_task.compute(
spiketimes_excerpts,
clusters_excerpts,
clusters_to_update=clusters_to_update,
clusters_selected=clusters_selected,
ncorrbins=ncorrbins,
corrbin=corrbin,
sample_rate=sample_rate,
wizard=wizard,
)
# Otherwise, update directly the correlograms view without launching
# the task in the external process.
else:
# self.update_correlograms_view()
return ("_update_correlograms_view", (wizard,), {})
def _recluster(self):
exp = self.loader.experiment
channel_group = self.loader.shank
clusters_selected = self.loader.get_clusters_selected()
self.tasks.recluster_task.recluster(exp, channel_group=channel_group, clusters=clusters_selected)
def _recluster_done(self, channel_group=0, clusters=None, spikes=None, clu=None, wizard=False):
return [("_split2", (spikes, clu, wizard))]
def _compute_similarity_matrix(self, target_next=None):
exp = self.experiment
channel_group = self.loader.shank
clustering = "main" # TODO
fetdim = exp.application_data.spikedetekt.n_features_per_channel
clusters_data = getattr(exp.channel_groups[channel_group].clusters, clustering)
spikes_data = exp.channel_groups[channel_group].spikes
cluster_groups_data = getattr(exp.channel_groups[channel_group].cluster_groups, clustering)
clusters_all = sorted(clusters_data.keys())
cluster_groups = pd.Series([clusters_data[cl].cluster_group or 0 for cl in clusters_all], index=clusters_all)
spikes_selected, fm = spikes_data.load_features_masks(fraction=0.1)
clusters = getattr(spikes_data.clusters, clustering)[:][spikes_selected]
fm = np.atleast_3d(fm)
features = fm[:, :, 0]
if features.shape[1] <= 1:
return []
# masks = fm[:, ::fetdim, 1]
if fm.shape[2] > 1:
masks = fm[:, :, 1]
else:
masks = None
# features = pandaize(features, spikes_selected)
# masks = pandaize(masks, spikes_selected)
# Get cluster indices that need to be updated.
# if clusters_to_update is None:
# NOTE: not specifying explicitely clusters_to_update ensures that
# all clusters that need to be updated are updated.
# Allows to fix a bug where the matrix is not updated correctly
# when multiple calls to this functions are called quickly.
clusters_to_update = self.statscache.similarity_matrix.not_in_key_indices(clusters_all)
log.debug("Clusters to update: {0:s}".format(str(clusters_to_update)))
# If there are pairs that need to be updated, launch the task.
if len(clusters_to_update) > 0:
self.mainwindow.set_busy(computing_matrix=True)
# Launch the task.
self.tasks.similarity_matrix_task.compute(
features,
clusters,
cluster_groups,
masks,
clusters_to_update,
target_next=target_next,
similarity_measure=None,
)
# Otherwise, update directly the correlograms view without launching
# the task in the external process.
else:
return [("_wizard_update", (target_next,)), ("_update_similarity_matrix_view",)]
def _correlograms_computed(self, clusters, correlograms, ncorrbins, corrbin, sample_rate, wizard):
clusters_selected = self.loader.get_clusters_selected()
# Abort if the selection has changed during the computation of the
# correlograms.
# Reset the cursor.
self.mainwindow.set_busy(computing_correlograms=False)
if not np.array_equal(clusters, clusters_selected):
log.debug(
"Skip update correlograms with clusters selected={0:s}"
" and clusters updated={1:s}.".format(clusters_selected, clusters)
)
return
if self.statscache.ncorrbins != ncorrbins:
log.debug(
(
"Skip updating correlograms because ncorrbins has "
"changed (from {0:d} to {1:d})".format(ncorrbins, self.statscache.ncorrbins)
)
)
return
# Put the computed correlograms in the cache.
self.statscache.correlograms.update(clusters, correlograms)
# Update the view.
# self.update_correlograms_view()
return ("_update_correlograms_view", (), dict(wizard=wizard))
def _similarity_matrix_computed(self, clusters_selected, matrix, clusters, cluster_groups, target_next=None):
self.mainwindow.set_busy(computing_matrix=False)
# spikes_slice = _get_similarity_matrix_slice(
# self.loader.nspikes,
# len(self.loader.get_clusters_unique()))
# clusters_now = self.loader.get_clusters(
# spikes=self.loader.background_spikes)
# if not np.array_equal(clusters, clusters_now):
# return False
if len(matrix) == 0:
return []
self.statscache.similarity_matrix.update(clusters_selected, matrix)
self.statscache.similarity_matrix_normalized = normalize(self.statscache.similarity_matrix.to_array(copy=True))
# Update the cluster view with cluster quality.
quality = np.diag(self.statscache.similarity_matrix_normalized).copy()
self.statscache.cluster_quality = pd.Series(quality, index=self.statscache.similarity_matrix.indices)
self.get_view("ClusterView").set_quality(self.statscache.cluster_quality)
return [("_wizard_update", (target_next,)), ("_update_similarity_matrix_view",)]
def _invalidate(self, clusters):
self.statscache.invalidate(clusters)
# View updates.
# -------------
def _update_correlograms_view(self, wizard=None):
clu = self.loader.get_clusters_selected()
# HACK: work around a bug with some GPU drivers and empty selections
if len(clu) == 0:
return
data = vd.get_correlogramsview_data(
self.experiment, self.statscache.correlograms, clusters=clu, channel_group=self.loader.shank, wizard=wizard
)
[view.set_data(**data) for view in self.get_views("CorrelogramsView")]
def _update_similarity_matrix_view(self):
data = vd.get_similaritymatrixview_data(
self.experiment, self.statscache.similarity_matrix_normalized, channel_group=self.loader.shank
)
[view.set_data(**data) for view in self.get_views("SimilarityMatrixView")]
# Show selected clusters when the matrix has been updated.
clusters = self.loader.get_clusters_selected()
return ("_show_selection_in_matrix", (clusters,))
def _update_feature_view(self, autozoom=None):
clu = self.loader.clusters_selected
# HACK: work around a bug with some GPU drivers and empty selections
if len(clu) == 0:
return
data = vd.get_featureview_data(
self.experiment, clusters=clu, autozoom=autozoom, channel_group=self.loader.shank
)
[view.set_data(**data) for view in self.get_views("FeatureView")]
def _update_waveform_view(self, autozoom=None, wizard=None):
clu = self.loader.clusters_selected
# HACK: work around a bug with some GPU drivers and empty selections
if len(clu) == 0:
return
data = vd.get_waveformview_data(
self.experiment, clusters=clu, autozoom=autozoom, wizard=wizard, channel_group=self.loader.shank
)
[view.set_data(**data) for view in self.get_views("WaveformView")]
def _update_trace_view(self):
data = vd.get_traceview_data(self.experiment, channel_group=self.loader.shank)
[view.set_data(**data) for view in self.get_views("TraceView")]
def _update_cluster_view(self, clusters=None):
"""Update the cluster view using the data stored in the loader
object."""
data = vd.get_clusterview_data(self.experiment, self.statscache, channel_group=self.loader.shank)
self.get_view("ClusterView").set_data(**data)
if clusters is not None:
return
def _show_selection_in_matrix(self, clusters):
if clusters is not None and 1 <= len(clusters) <= 2:
[view.show_selection(clusters[0], clusters[-1]) for view in self.get_views("SimilarityMatrixView")]
# Override colors.
# ----------------
def _override_color(self, override_color):
self.loader.set_override_color(override_color)
return ["_update_feature_view", "_update_waveform_view", "_update_correlograms_view"]
# Change correlograms parameter.
# ------------------------------
def _change_correlograms_parameters(self, ncorrbins=None, corrbin=None):
if ncorrbins % 2 == 0:
ncorrbins += 1
# Update the correlograms parameters.
if ncorrbins is not None:
SETTINGS["correlograms.ncorrbins"] = ncorrbins
if corrbin is not None:
SETTINGS["correlograms.corrbin"] = corrbin
# Reset the cache.
self.statscache.reset(ncorrbins)
# Update the correlograms.
clusters = self.loader.get_clusters_selected()
return ("_compute_correlograms", (clusters,))
# Merge/split actions.
# --------------------
def _merge(self, clusters, wizard=False):
if len(clusters) >= 2:
action, output = self.controller.merge_clusters(clusters)
# Tell the next nodes whether the merge occurred after a wizard
# selection or not, so that the merged cluster background is
# highlighted or not.
output["wizard"] = wizard
return after_merge(output)
def _split(self, clusters, spikes_selected, wizard=False):
if len(spikes_selected) >= 1:
action, output = self.controller.split_clusters(clusters, spikes_selected)
output["wizard"] = wizard
return after_split(output)
def _split2(self, spikes, clusters, wizard=False):
if len(spikes) >= 1:
action, output = self.controller.split2_clusters(spikes, clusters)
output["wizard"] = wizard
return after_split(output)
def _undo(self, wizard=False):
undo = self.controller.undo()
if undo is None:
return
action, output = undo
output["wizard"] = wizard
if action == "merge_clusters_undo":
return after_merge_undo(output)
elif action == "split_clusters_undo":
return after_split_undo(output)
elif action == "split2_clusters_undo":
return after_split_undo(output)
elif action == "change_cluster_color_undo":
return after_cluster_color_changed_undo(output)
elif action == "change_group_color_undo":
return after_group_color_changed(output)
elif action == "move_clusters_undo":
return after_clusters_moved_undo(output)
elif action == "add_group_undo":
return after_group_added(output)
elif action == "rename_group_undo":
return after_group_renamed(output)
elif action == "remove_group_undo":
return after_group_removed(output)
def _redo(self, wizard=False):
redo = self.controller.redo()
if redo is None:
return
action, output = redo
output["wizard"] = wizard
if action == "merge_clusters":
return after_merge(output)
elif action == "split_clusters":
return after_split(output)
elif action == "split2_clusters":
return after_split(output)
elif action == "change_cluster_color":
return after_cluster_color_changed(output)
elif action == "change_group_color":
return after_group_color_changed(output)
elif action == "move_clusters":
return after_clusters_moved(output)
elif action == "add_group":
return after_group_added(output)
elif action == "rename_group":
return after_group_renamed(output)
elif action == "remove_group":
return after_group_removed(output)
# Other actions.
# --------------
def _cluster_color_changed(self, cluster, color, wizard=True):
action, output = self.controller.change_cluster_color(cluster, color)
# if cluster == self.wizard.current_target():
output["wizard"] = wizard
return after_cluster_color_changed(output)
def _group_color_changed(self, group, color):
action, output = self.controller.change_group_color(group, color)
return after_group_color_changed(output)
def _group_renamed(self, group, name):
action, output = self.controller.rename_group(group, name)
return after_group_renamed(output)
def _clusters_moved(self, clusters, group, wizard=False):
action, output = self.controller.move_clusters(clusters, group)
output["wizard"] = wizard
return after_clusters_moved(output)
def _group_removed(self, group):
action, output = self.controller.remove_group(group)
return after_group_removed(output)
def _group_added(self, group, name, color):
action, output = self.controller.add_group(group, name, color)
return after_group_added(output)
# Wizard.
# -------
def _wizard_update(self, target=None, update_matrix=True):
if update_matrix:
self.wizard.set_data(
cluster_groups=self.loader.get_cluster_groups("all"),
similarity_matrix=self.statscache.similarity_matrix_normalized,
)
else:
self.wizard.set_data(cluster_groups=self.loader.get_cluster_groups("all"))
self.wizard.update_candidates(target)
def _wizard_change_color(self, clusters):
if clusters is not None:
# Set the background color in the cluster view for the wizard
# target and candidate.
self.get_view("ClusterView").set_background(
{cluster: {0: "target", 1: "candidate"}.get(i, None) for i, cluster in enumerate(clusters[:2])}
)
def _wizard_change_candidate_color(self):
candidate = self.wizard.current_candidate()
target = self.wizard.current_target()
# color = self.loader.get_cluster_color(candidate)
return ("_cluster_color_changed", (candidate, random_color()))
def _wizard_show_pair(self, target=None, candidate=None):
if target is None:
target = (self.wizard.current_target(), self.loader.get_cluster_color(self.wizard.current_target()))
if candidate is None:
try:
candidate = (
self.wizard.current_candidate(),
get_array(self.loader.get_cluster_color(self.wizard.current_candidate()))[0],
)
# HACK: this can fail because when merging clusters, the merged
# cluster (candidate) is deleted, and its color does not exist
# anymore.
except:
candidate = (self.wizard.current_candidate(), 0)
[view.set_wizard_pair(target, candidate) for view in self.get_views("FeatureView")]
# Navigation.
def _wizard_reset(self):
clusters = self.wizard.reset()
return ["_wizard_update", "_wizard_current_candidate"]
def _wizard_previous_candidate(self):
clusters = self.wizard.previous_pair()
return after_wizard_selection(clusters)
def _wizard_current_candidate(self):
clusters = self.wizard.current_pair()
return after_wizard_selection(clusters)
def _wizard_next_candidate(self):
clusters = self.wizard.next_pair()
return after_wizard_selection(clusters)
def _wizard_skip_target(self):
# Skip the current target and go the next target.
self.wizard.skip_target()
return [("_wizard_update", ()), ("_wizard_next_candidate",)]
def _wizard_reset_skipped(self):
self.wizard.reset_skipped()
# Control.
def _wizard_move_and_next(self, what, group):
"""Move target, candidate, or both, to a given group, and go to
the next proposition."""
# Current proposition.
clusters = self.wizard.current_pair()
if clusters is None:
return
target, candidate = clusters
# Select appropriate clusters to move.
if what == "candidate":
clusters = [candidate]
# Keep the current target.
target_next = target
reset_skipped = False
elif what == "target":
clusters = [target]
# Go to the next best target cluster.
target_next = None
reset_skipped = True
elif what == "both":
clusters = [candidate, target]
# Go to the next best target cluster.
target_next = None
reset_skipped = True
# Move clusters, and select next proposition.
r = [("_clusters_moved", (clusters, group, True))]
if reset_skipped:
r += [("_wizard_reset_skipped",)]
r += [("_wizard_update", (target_next,)), ("_wizard_next_candidate",)]
return r
class TaskGraph(AbstractTaskGraph):
def __init__(self, mainwindow):
# Shortcuts for the main window.
self.set(mainwindow)
# Create external threads/processes for long-lasting tasks.
self.create_threads()
def set(self, mainwindow):
# Shortcuts for the main window.
self.mainwindow = mainwindow
self.get_view = self.mainwindow.get_view
self.get_views = self.mainwindow.get_views
self.loader = self.mainwindow.loader
self.wizard = self.mainwindow.wizard
self.controller = self.mainwindow.controller
self.statscache = self.mainwindow.statscache
def create_threads(self):
# Create the external threads.
self.tasks = ThreadedTasks()
self.tasks.selection_task.set_loader(self.loader)
self.tasks.selection_task.selectionDone.connect(
self.selection_done_callback)
self.tasks.correlograms_task.correlogramsComputed.connect(
self.correlograms_computed_callback)
self.tasks.similarity_matrix_task.correlationMatrixComputed.connect(
self.similarity_matrix_computed_callback)
def join(self):
self.tasks.join()
# Selection.
# ----------
def _select(self, clusters, wizard=False):
self.tasks.selection_task.select(clusters, wizard)
def _select_done(self, clusters, wizard=False):
if wizard:
target = (self.wizard.current_target(), )
else:
target = ()
# self.loader.select(clusters=clusters)
log.debug("Selected clusters {0:s}.".format(str(clusters)))
return [
('_update_feature_view', target),
('_update_waveform_view', (), dict(wizard=wizard)),
('_show_selection_in_matrix', (clusters, )),
(
'_compute_correlograms',
(clusters, ),
),
]
def _select_in_cluster_view(self, clusters, groups=[], wizard=False):
self.get_view('ClusterView').select(clusters,
groups=groups,
wizard=wizard)
# Callbacks.
# ----------
def selection_done_callback(self, clusters, wizard):
self.select_done(clusters, wizard=wizard)
def correlograms_computed_callback(self, clusters, correlograms, ncorrbins,
corrbin):
# Execute the callback function under the control of the task manager
# (which handles the graph dependency).
self.correlograms_computed(clusters, correlograms, ncorrbins, corrbin)
def similarity_matrix_computed_callback(self,
clusters_selected,
matrix,
clusters,
cluster_groups,
target_next=None):
# Execute the callback function under the control of the task manager
# (which handles the graph dependency).
self.similarity_matrix_computed(clusters_selected,
matrix,
clusters,
cluster_groups,
target_next=target_next)
# Computations.
# -------------
def _compute_correlograms(self, clusters_selected):
# Get the correlograms parameters.
spiketimes = get_array(self.loader.get_spiketimes('all'))
# Make a copy of the array so that it does not change before the
# computation of the correlograms begins.
clusters = np.array(get_array(self.loader.get_clusters('all')))
# corrbin = self.loader.corrbin
# ncorrbins = self.loader.ncorrbins
corrbin = SETTINGS.get('correlograms.corrbin', .001)
ncorrbins = SETTINGS.get('correlograms.ncorrbins', 100)
# Get cluster indices that need to be updated.
clusters_to_update = (
self.statscache.correlograms.not_in_key_indices(clusters_selected))
# If there are pairs that need to be updated, launch the task.
if len(clusters_to_update) > 0:
# Set wait cursor.
self.mainwindow.set_busy(computing_correlograms=True)
# Launch the task.
self.tasks.correlograms_task.compute(
spiketimes,
clusters,
clusters_to_update=clusters_to_update,
clusters_selected=clusters_selected,
ncorrbins=ncorrbins,
corrbin=corrbin)
# Otherwise, update directly the correlograms view without launching
# the task in the external process.
else:
# self.update_correlograms_view()
return '_update_correlograms_view'
def _compute_similarity_matrix(self, target_next=None):
similarity_measure = self.loader.similarity_measure
features = self.loader.background_features
masks = self.loader.background_masks
clusters = get_array(
self.loader.get_clusters(spikes=self.loader.background_spikes))
cluster_groups = get_array(self.loader.get_cluster_groups('all'))
clusters_all = self.loader.get_clusters_unique()
# Get cluster indices that need to be updated.
# if clusters_to_update is None:
# NOTE: not specifying explicitely clusters_to_update ensures that
# all clusters that need to be updated are updated.
# Allows to fix a bug where the matrix is not updated correctly
# when multiple calls to this functions are called quickly.
clusters_to_update = (
self.statscache.similarity_matrix.not_in_key_indices(clusters_all))
log.debug("Clusters to update: {0:s}".format(str(clusters_to_update)))
# If there are pairs that need to be updated, launch the task.
if len(clusters_to_update) > 0:
self.mainwindow.set_busy(computing_matrix=True)
# Launch the task.
self.tasks.similarity_matrix_task.compute(
features,
clusters,
cluster_groups,
masks,
clusters_to_update,
target_next=target_next,
similarity_measure=similarity_measure)
# Otherwise, update directly the correlograms view without launching
# the task in the external process.
else:
return [
('_wizard_update', (target_next, )),
('_update_similarity_matrix_view', ),
]
def _correlograms_computed(self, clusters, correlograms, ncorrbins,
corrbin):
clusters_selected = self.loader.get_clusters_selected()
# Abort if the selection has changed during the computation of the
# correlograms.
# Reset the cursor.
self.mainwindow.set_busy(computing_correlograms=False)
if not np.array_equal(clusters, clusters_selected):
log.debug("Skip update correlograms with clusters selected={0:s}"
" and clusters updated={1:s}.".format(
clusters_selected, clusters))
return
if self.statscache.ncorrbins != ncorrbins:
log.debug(("Skip updating correlograms because ncorrbins has "
"changed (from {0:d} to {1:d})".format(
ncorrbins, self.statscache.ncorrbins)))
return
# Put the computed correlograms in the cache.
self.statscache.correlograms.update(clusters, correlograms)
# Update the view.
# self.update_correlograms_view()
return '_update_correlograms_view'
def _similarity_matrix_computed(self,
clusters_selected,
matrix,
clusters,
cluster_groups,
target_next=None):
self.mainwindow.set_busy(computing_matrix=False)
# spikes_slice = _get_similarity_matrix_slice(
# self.loader.nspikes,
# len(self.loader.get_clusters_unique()))
clusters_now = self.loader.get_clusters(
spikes=self.loader.background_spikes)
if not np.array_equal(clusters, clusters_now):
return False
self.statscache.similarity_matrix.update(clusters_selected, matrix)
self.statscache.similarity_matrix_normalized = normalize(
self.statscache.similarity_matrix.to_array(copy=True))
# Update the cluster view with cluster quality.
quality = np.diag(self.statscache.similarity_matrix_normalized)
self.statscache.cluster_quality = pd.Series(
quality,
index=self.statscache.similarity_matrix.indices,
)
self.get_view('ClusterView').set_quality(
self.statscache.cluster_quality)
return [
('_wizard_update', (target_next, )),
('_update_similarity_matrix_view', ),
]
def _invalidate(self, clusters):
self.statscache.invalidate(clusters)
# View updates.
# -------------
def _update_correlograms_view(self):
data = vd.get_correlogramsview_data(self.loader, self.statscache)
[view.set_data(**data) for view in self.get_views('CorrelogramsView')]
def _update_similarity_matrix_view(self):
data = vd.get_similaritymatrixview_data(self.loader, self.statscache)
[
view.set_data(**data)
for view in self.get_views('SimilarityMatrixView')
]
# Show selected clusters when the matrix has been updated.
clusters = self.loader.get_clusters_selected()
return ('_show_selection_in_matrix', (clusters, ))
def _update_feature_view(self, autozoom=None):
data = vd.get_featureview_data(self.loader, autozoom=autozoom)
[view.set_data(**data) for view in self.get_views('FeatureView')]
def _update_waveform_view(self, autozoom=None, wizard=None):
data = vd.get_waveformview_data(self.loader,
autozoom=autozoom,
wizard=wizard)
[view.set_data(**data) for view in self.get_views('WaveformView')]
def _update_trace_view(self):
data = vd.get_traceview_data(self.loader)
[view.set_data(**data) for view in self.get_views('TraceView')]
def _update_cluster_view(self, clusters=None):
"""Update the cluster view using the data stored in the loader
object."""
data = vd.get_clusterview_data(self.loader,
self.statscache,
clusters=clusters)
self.get_view('ClusterView').set_data(**data)
if clusters is not None:
return
def _show_selection_in_matrix(self, clusters):
if clusters is not None and 1 <= len(clusters) <= 2:
[
view.show_selection(clusters[0], clusters[-1])
for view in self.get_views('SimilarityMatrixView')
]
# Override colors.
# ----------------
def _override_color(self, override_color):
self.loader.set_override_color(override_color)
return [
'_update_feature_view', '_update_waveform_view',
'_update_correlograms_view'
]
# Change correlograms parameter.
# ------------------------------
def _change_correlograms_parameters(self, ncorrbins=None, corrbin=None):
# Update the correlograms parameters.
if ncorrbins is not None:
SETTINGS['correlograms.ncorrbins'] = ncorrbins
if corrbin is not None:
SETTINGS['correlograms.corrbin'] = corrbin
# Reset the cache.
self.statscache.reset(ncorrbins)
# Update the correlograms.
clusters = self.loader.get_clusters_selected()
return ('_compute_correlograms', (clusters, ))
# Merge/split actions.
# --------------------
def _merge(self, clusters, wizard=False):
if len(clusters) >= 2:
action, output = self.controller.merge_clusters(clusters)
# Tell the next nodes whether the merge occurred after a wizard
# selection or not, so that the merged cluster background is
# highlighted or not.
output['wizard'] = wizard
return after_merge(output)
def _split(self, clusters, spikes_selected, wizard=False):
if len(spikes_selected) >= 1:
action, output = self.controller.split_clusters(
clusters, spikes_selected)
output['wizard'] = wizard
return after_split(output)
def _undo(self, wizard=False):
undo = self.controller.undo()
if undo is None:
return
action, output = undo
output['wizard'] = wizard
if action == 'merge_clusters_undo':
return after_merge_undo(output)
elif action == 'split_clusters_undo':
return after_split_undo(output)
elif action == 'change_cluster_color_undo':
return after_cluster_color_changed_undo(output)
elif action == 'change_group_color_undo':
return after_group_color_changed(output)
elif action == 'move_clusters_undo':
return after_clusters_moved_undo(output)
elif action == 'add_group_undo':
return after_group_added(output)
elif action == 'rename_group_undo':
return after_group_renamed(output)
elif action == 'remove_group_undo':
return after_group_removed(output)
def _redo(self, wizard=False):
redo = self.controller.redo()
if redo is None:
return
action, output = redo
output['wizard'] = wizard
if action == 'merge_clusters':
return after_merge(output)
elif action == 'split_clusters':
return after_split(output)
elif action == 'change_cluster_color':
return after_cluster_color_changed(output)
elif action == 'change_group_color':
return after_group_color_changed(output)
elif action == 'move_clusters':
return after_clusters_moved(output)
elif action == 'add_group':
return after_group_added(output)
elif action == 'rename_group':
return after_group_renamed(output)
elif action == 'remove_group':
return after_group_removed(output)
# Other actions.
# --------------
def _cluster_color_changed(self, cluster, color, wizard=True):
action, output = self.controller.change_cluster_color(cluster, color)
# if cluster == self.wizard.current_target():
output['wizard'] = wizard
return after_cluster_color_changed(output)
def _group_color_changed(self, group, color):
action, output = self.controller.change_group_color(group, color)
return after_group_color_changed(output)
def _group_renamed(self, group, name):
action, output = self.controller.rename_group(group, name)
return after_group_renamed(output)
def _clusters_moved(
self,
clusters,
group,
wizard=False,
):
action, output = self.controller.move_clusters(clusters, group)
output['wizard'] = wizard
return after_clusters_moved(output)
def _group_removed(self, group):
action, output = self.controller.remove_group(group)
return after_group_removed(output)
def _group_added(self, group, name, color):
action, output = self.controller.add_group(group, name, color)
return after_group_added(output)
# Wizard.
# -------
def _wizard_update(self, target=None, update_matrix=True):
if update_matrix:
self.wizard.set_data(
cluster_groups=self.loader.get_cluster_groups('all'),
similarity_matrix=self.statscache.similarity_matrix_normalized,
)
else:
self.wizard.set_data(
cluster_groups=self.loader.get_cluster_groups('all'), )
self.wizard.update_candidates(target)
def _wizard_change_color(self, clusters):
if clusters is not None:
# Set the background color in the cluster view for the wizard
# target and candidate.
self.get_view('ClusterView').set_background({
cluster: {
0: 'target',
1: 'candidate'
}.get(i, None)
for i, cluster in enumerate(clusters[:2])
})
def _wizard_change_candidate_color(self):
candidate = self.wizard.current_candidate()
target = self.wizard.current_target()
# color = self.loader.get_cluster_color(candidate)
return ('_cluster_color_changed', (
candidate,
random_color(),
))
def _wizard_show_pair(self, target=None, candidate=None):
if target is None:
target = (self.wizard.current_target(),
self.loader.get_cluster_color(
self.wizard.current_target()))
if candidate is None:
candidate = (self.wizard.current_candidate(),
self.loader.get_cluster_color(
self.wizard.current_candidate()))
[
view.set_wizard_pair(target, candidate)
for view in self.get_views('FeatureView')
]
# Navigation.
def _wizard_reset(self):
clusters = self.wizard.reset()
return ['_wizard_update', '_wizard_current_candidate']
def _wizard_previous_candidate(self):
clusters = self.wizard.previous_pair()
return after_wizard_selection(clusters)
def _wizard_current_candidate(self):
clusters = self.wizard.current_pair()
return after_wizard_selection(clusters)
def _wizard_next_candidate(self):
clusters = self.wizard.next_pair()
return after_wizard_selection(clusters)
def _wizard_skip_target(self):
# Skip the current target and go the next target.
self.wizard.skip_target()
return [
('_wizard_update', ()),
('_wizard_next_candidate', ),
]
def _wizard_reset_skipped(self):
self.wizard.reset_skipped()
# Control.
def _wizard_move_and_next(self, what, group):
"""Move target, candidate, or both, to a given group, and go to
the next proposition."""
# Current proposition.
clusters = self.wizard.current_pair()
if clusters is None:
return
target, candidate = clusters
# Select appropriate clusters to move.
if what == 'candidate':
clusters = [candidate]
# Keep the current target.
target_next = target
reset_skipped = False
elif what == 'target':
clusters = [target]
# Go to the next best target cluster.
target_next = None
reset_skipped = True
elif what == 'both':
clusters = [candidate, target]
# Go to the next best target cluster.
target_next = None
reset_skipped = True
# Move clusters, and select next proposition.
r = [
('_clusters_moved', (clusters, group, True)),
]
if reset_skipped:
r += [
('_wizard_reset_skipped', ),
]
r += [
('_wizard_update', (target_next, )),
('_wizard_next_candidate', ),
]
return r
