def get_traces(self, interval):
"""Load traces and spikes in an interval."""
tr = select_traces(
self.all_traces,
interval,
sample_rate=self.sample_rate,
)
tr = tr - np.mean(tr, axis=0)
a, b = self.spike_times.searchsorted(interval)
sc = self.spike_templates[a:b]
# Remove templates.
tr_sub = subtract_templates(
tr,
start=interval[0],
spike_times=self.spike_times[a:b],
spike_clusters=sc,
amplitudes=self.all_amplitudes[a:b],
spike_templates=self.templates_unw[sc],
sample_rate=self.sample_rate,
)
return [
Bunch(traces=tr),
Bunch(traces=tr_sub, color=(.5, .5, .5, .75)),
]
def test_gui_state_view(tempdir):
view = Bunch(name='MyView0')
state = GUIState(config_dir=tempdir)
state.update_view_state(view, dict(hello='world'))
assert not state.get_view_state(Bunch(name='MyView'))
assert not state.get_view_state(Bunch(name='MyView1'))
assert state.get_view_state(view) == Bunch(hello='world')
def get_features(self, cluster_id, load_all=False):
# Overriden to take into account the sparse structure.
# Only keep spikes belonging to the features spike ids.
if self.features_spike_ids is not None:
# All spikes
spike_ids = self._select_spikes(cluster_id)
spike_ids = np.intersect1d(spike_ids, self.features_spike_ids)
# Relative indices of the spikes in the self.features_spike_ids
# array, necessary to load features from all_features which only
# contains the subset of the spikes.
spike_ids_rel = _index_of(spike_ids, self.features_spike_ids)
else:
spike_ids = self._select_spikes(cluster_id,
self.n_spikes_features
if not load_all else None)
spike_ids_rel = spike_ids
st = self.spike_templates[spike_ids]
nc = self.n_channels
nfpc = self.n_features_per_channel
ns = len(spike_ids)
f = _densify(spike_ids_rel, self.all_features,
self.features_ind[st, :], self.n_channels)
f = np.transpose(f, (0, 2, 1))
assert f.shape == (ns, nc, nfpc)
b = Bunch()
# Normalize features.
m = self.get_feature_lim()
f = _normalize(f, -m, m)
b.data = f
b.spike_ids = spike_ids
b.spike_clusters = self.spike_clusters[spike_ids]
b.masks = self.all_masks[spike_ids]
return b
def _get_traces(self, interval):
"""Get traces and spike waveforms."""
k = self.model.n_samples_templates
m = self.model
traces_interval = select_traces(m.traces, interval,
sample_rate=m.sample_rate)
# Reorder vertically.
out = Bunch(data=traces_interval)
out.waveforms = []
def gbc(cluster_id):
return self.get_best_channels(cluster_id)
for b in _iter_spike_waveforms(interval=interval,
traces_interval=traces_interval,
model=self.model,
supervisor=self.supervisor,
color_selector=self.color_selector,
n_samples_waveforms=k,
get_best_channels=gbc,
show_all_spikes=self._show_all_spikes,
):
i = b.spike_id
# Compute the residual: waveform - amplitude * template.
residual = b.copy()
template_id = m.spike_templates[i]
template = m.get_template(template_id).template
amplitude = m.amplitudes[i]
residual.data = residual.data - amplitude * template
out.waveforms.extend([b, residual])
return out
def _get_traces(self, interval):
"""Get traces and spike waveforms."""
ns = self.model.n_samples_waveforms
m = self.model
c = self.channel_vertical_order
traces_interval = select_traces(m.traces,
interval,
sample_rate=m.sample_rate)
# Reorder vertically.
traces_interval = traces_interval[:, c]
def gbc(cluster_id):
ch = self.get_best_channels(cluster_id)
return ch
out = Bunch(data=traces_interval)
out.waveforms = []
for b in _iter_spike_waveforms(
interval=interval,
traces_interval=traces_interval,
model=self.model,
supervisor=self.supervisor,
color_selector=self.color_selector,
n_samples_waveforms=ns,
get_best_channels=gbc,
show_all_spikes=self._show_all_spikes,
):
b.channel_labels = m.channel_order[b.channel_ids]
out.waveforms.append(b)
return out
def state(tempdir):
# Save a test GUI state JSON file in the tempdir.
state = Bunch()
state.WaveformView0 = Bunch(overlap=False)
state.TraceView0 = Bunch(scaling=1.)
state.FeatureView0 = Bunch(feature_scaling=.5)
state.CorrelogramView0 = Bunch(uniform_normalization=True)
return state
def state(tempdir):
# Save a test GUI state JSON file in the tempdir.
state = Bunch()
state.WaveformView0 = Bunch(overlap=False)
state.TraceView0 = Bunch(scaling=1.)
state.FeatureView0 = Bunch(feature_scaling=.5)
state.CorrelogramView0 = Bunch(uniform_normalization=True)
return state
def _select_data(self, cluster_id, arr, n_max=None, batch_size=None):
spike_ids = self._select_spikes(cluster_id, n_max,
batch_size=batch_size)
b = Bunch()
b.data = arr[spike_ids]
b.spike_ids = spike_ids
b.masks = self.all_masks[spike_ids]
return b
def _select_data(self, cluster_id, arr, n_max=None):
spike_ids = self._select_spikes(cluster_id, n_max)
b = Bunch()
b.data = arr[spike_ids]
b.spike_ids = spike_ids
b.spike_clusters = self.spike_clusters[spike_ids]
b.masks = self.all_masks[spike_ids]
return b
def get_amplitudes(self, cluster_id):
spike_ids = self._select_spikes(cluster_id, self.n_spikes_features)
d = Bunch()
d.spike_ids = spike_ids
d.spike_clusters = cluster_id * np.ones(len(spike_ids), dtype=np.int32)
d.x = self.spike_times[spike_ids]
d.y = self.all_amplitudes[spike_ids]
return d
def get_background_features(self):
k = max(1, int(self.n_spikes // self.n_spikes_background_features))
spike_ids = slice(None, None, k)
b = Bunch()
b.data = self.all_features[spike_ids]
b.spike_ids = spike_ids
b.spike_clusters = self.spike_clusters[spike_ids]
b.masks = self.all_masks[spike_ids]
return b
def get_amplitudes(self, cluster_id):
spike_ids = self._select_spikes(cluster_id, self.n_spikes_features)
d = Bunch()
d.spike_ids = spike_ids
d.x = self.spike_times[spike_ids]
d.y = self.all_amplitudes[spike_ids]
M = d.y.max()
d.data_bounds = [0, 0, self.duration, M]
return d
def wrapped(cluster_ids, **kwargs):
# Single cluster.
if not hasattr(cluster_ids, '__len__'):
return f(cluster_ids, **kwargs)
# Concatenate the result of multiple clusters.
l = [f(c, **kwargs) for c in cluster_ids]
# Handle the case where every function returns a list of Bunch.
if l and isinstance(l[0], list):
# We assume that all items have the same length.
n = len(l[0])
return [
Bunch(_accumulate([item[i] for item in l])) for i in range(n)
]
else:
return Bunch(_accumulate(l))
def validate(pos=None, text=None, anchor=None,
data_bounds=None,
):
if text is None:
text = []
if isinstance(text, string_types):
text = [text]
if pos is None:
pos = np.zeros((len(text), 2))
assert pos is not None
pos = np.atleast_2d(pos)
assert pos.ndim == 2
assert pos.shape[1] == 2
n_text = pos.shape[0]
assert len(text) == n_text
anchor = anchor if anchor is not None else (0., 0.)
anchor = np.atleast_2d(anchor)
if anchor.shape[0] == 1:
anchor = np.repeat(anchor, n_text, axis=0)
assert anchor.ndim == 2
assert anchor.shape == (n_text, 2)
if data_bounds is not None:
data_bounds = _get_data_bounds(data_bounds, pos)
assert data_bounds.shape[0] == n_text
data_bounds = data_bounds.astype(np.float64)
assert data_bounds.shape == (n_text, 4)
return Bunch(pos=pos, text=text, anchor=anchor,
data_bounds=data_bounds)
def _get_amplitudes(self, cluster_id):
n = self.n_spikes_amplitudes
m = self.model
spike_ids = self.selector.select_spikes([cluster_id], n)
x = m.spike_times[spike_ids]
y = m.amplitudes[spike_ids]
return Bunch(x=x, y=y, data_bounds=(0., 0., m.duration, y.max()))
def state(self):
return Bunch(box_scaling=tuple(self.box_scaling),
probe_scaling=tuple(self.probe_scaling),
overlap=self.overlap,
do_zoom_on_channels=self.do_zoom_on_channels,
do_show_labels=self.do_show_labels,
)
def get_traces(self, interval):
tr = select_traces(
self.all_traces,
interval,
sample_rate=self.sample_rate,
)
return [Bunch(traces=tr)]
def validate(
x=None,
y=None,
pos=None,
color=None,
size=None,
depth=None,
data_bounds='auto',
):
if pos is None:
x, y = _get_pos(x, y)
pos = np.c_[x, y]
pos = np.asarray(pos)
assert pos.ndim == 2
assert pos.shape[1] == 2
n = pos.shape[0]
# Validate the data.
color = _get_array(color, (n, 4),
ScatterVisual._default_color,
dtype=np.float32)
size = _get_array(size, (n, 1), ScatterVisual._default_marker_size)
depth = _get_array(depth, (n, 1), 0)
if data_bounds is not None:
data_bounds = _get_data_bounds(data_bounds, pos)
assert data_bounds.shape[0] == n
return Bunch(pos=pos,
color=color,
size=size,
depth=depth,
data_bounds=data_bounds)
def state(self):
return Bunch(speed_threshold=self.speed_threshold,
speed_threshold_mode=self.speed_threshold_mode,
time_range=self.time_ranges,
n_rate_map_contours=self.n_rate_map_contours,
rate_map_contour_mode=self.rate_map_contour_mode,
spike_pos_shift=self.spike_pos_shift)
def validate(
x=None,
y=None,
pos=None,
masks=None,
data_bounds='auto',
):
if pos is None:
x, y = _get_pos(x, y)
pos = np.c_[x, y]
pos = np.asarray(pos)
assert pos.ndim == 2
assert pos.shape[1] == 2
n = pos.shape[0]
masks = _get_array(masks, (n, 1), 1., np.float32)
assert masks.shape == (n, 1)
# Validate the data.
if data_bounds is not None:
data_bounds = _get_data_bounds(data_bounds, pos)
assert data_bounds.shape[0] == n
return Bunch(
pos=pos,
masks=masks,
data_bounds=data_bounds,
)
def validate(hist=None, color=None, ylim=None):
assert hist is not None
hist = np.asarray(hist, np.float64)
if hist.ndim == 1:
hist = hist[None, :]
assert hist.ndim == 2
n_hists, n_bins = hist.shape
# Validate the data.
color = _get_array(
color,
(n_hists, 4),
HistogramVisual._default_color,
dtype=np.float32,
)
# Validate ylim.
if ylim is None:
ylim = hist.max() if hist.size > 0 else 1.
ylim = np.atleast_1d(ylim)
if len(ylim) == 1:
ylim = np.tile(ylim, n_hists)
if ylim.ndim == 1:
ylim = ylim[:, np.newaxis]
assert ylim.shape == (n_hists, 1)
return Bunch(
hist=hist,
ylim=ylim,
color=color,
)
def _get_template_waveforms(self, cluster_id):
"""Return the waveforms of the templates corresponding to a cluster."""
pos = self.model.channel_positions
count = self.get_template_counts(cluster_id)
template_ids = np.nonzero(count)[0]
count = count[template_ids]
# Get local channels.
channel_ids = self.get_best_channels(cluster_id)
# Get masks.
masks = count / float(count.max())
masks = np.tile(masks.reshape((-1, 1)), (1, len(channel_ids)))
# Get the mean amplitude for the cluster.
mean_amp = self._get_amplitudes(cluster_id).y.mean()
# Get all templates from which this cluster stems from.
templates = [self.model.get_template(template_id)
for template_id in template_ids]
data = np.stack([b.template * mean_amp for b in templates], axis=0)
cols = np.stack([b.channel_ids for b in templates], axis=0)
# NOTE: transposition because the channels should be in the second
# dimension for from_sparse.
data = data.transpose((0, 2, 1))
assert data.ndim == 3
assert data.shape[1] == cols.shape[1]
waveforms = from_sparse(data, cols, channel_ids)
# Transpose back.
waveforms = waveforms.transpose((0, 2, 1))
return Bunch(data=waveforms,
channel_ids=channel_ids,
channel_positions=pos[channel_ids],
masks=masks,
alpha=1.,
)
def _get_template_features(self, cluster_ids):
assert len(cluster_ids) == 2
clu0, clu1 = cluster_ids
s0 = self._get_spike_ids(clu0)
s1 = self._get_spike_ids(clu1)
n0 = self.get_template_counts(clu0)
n1 = self.get_template_counts(clu1)
t0 = self.model.get_template_features(s0)
t1 = self.model.get_template_features(s1)
x0 = np.average(t0, weights=n0, axis=1)
y0 = np.average(t0, weights=n1, axis=1)
x1 = np.average(t1, weights=n0, axis=1)
y1 = np.average(t1, weights=n1, axis=1)
return Bunch(x0=x0, y0=y0, x1=x1, y1=y1,
data_bounds=(min(x0.min(), x1.min()),
min(y0.min(), y1.min()),
max(y0.max(), y1.max()),
max(y0.max(), y1.max()),
),
)
def _get_axis_data(self, bunch, dim, cluster_id=None, load_all=None):
"""Extract the points from the data on a given dimension.
bunch is returned by the features() function.
dim is the string specifying the dimensions to extract for the data.
"""
if dim in self.attributes:
return self.attributes[dim](cluster_id, load_all=load_all)
masks = bunch.get('masks', None)
assert dim not in self.attributes # This is called only on PC data.
s = 'ABCDEFGHIJ'
# Channel relative index.
c_rel = int(dim[:-1])
# Get the channel_id from the currently-selected channels.
channel_id = self.channel_ids[c_rel % len(self.channel_ids)]
# Skup the plot if the channel id is not displayed.
if channel_id not in bunch.channel_ids: # pragma: no cover
return None
# Get the column index of the current channel in data.
c = list(bunch.channel_ids).index(channel_id)
# Principal component: A=0, B=1, etc.
d = s.index(dim[-1])
if masks is not None:
masks = masks[:, c]
return Bunch(
data=bunch.data[:, c, d],
masks=masks,
)
def state(self):
return Bunch(
scaling=self.scaling,
origin=self.origin,
interval=self._interval,
do_show_labels=self.do_show_labels,
)
def waveform_loader(request):
scale_factor, dc_offset = request.param
n_samples_trace, n_channels = 1000, 5
h = 10
n_samples_waveforms = 2 * h
n_spikes = n_samples_trace // (2 * n_samples_waveforms)
traces = artificial_traces(n_samples_trace, n_channels)
spike_samples = artificial_spike_samples(n_spikes,
max_isi=2 * n_samples_waveforms)
with raises(ValueError):
WaveformLoader(traces)
loader = WaveformLoader(
traces=traces,
n_samples_waveforms=n_samples_waveforms,
scale_factor=scale_factor,
dc_offset=dc_offset,
)
b = Bunch(
loader=loader,
n_samples_waveforms=n_samples_waveforms,
n_spikes=n_spikes,
spike_samples=spike_samples,
)
yield b
def state(self):
return Bunch(
bin_size=self.bin_size,
window_size=self.window_size,
excerpt_size=self.excerpt_size,
n_excerpts=self.n_excerpts,
uniform_normalization=self.uniform_normalization,
)
def _get_amplitudes(self, cluster_id):
n = self.n_spikes_amplitudes
m = self.model
spike_ids = self.selector.select_spikes([cluster_id], n)
channel_id = self.get_best_channel(cluster_id)
x = m.spike_times[spike_ids]
y = m.amplitudes[spike_ids, channel_id]
return Bunch(x=x, y=y, data_bounds=(0., y.min(), m.duration, y.max()))
def state(self):
"""View state.
This Bunch will be automatically persisted in the GUI state when the
GUI is closed.
To be overriden.
"""
return Bunch()
def get_traces(interval):
out = Bunch(data=select_traces(traces, interval, sample_rate=sr),
color=(.75,) * 4,
)
a, b = st.searchsorted(interval)
out.waveforms = []
k = 20
for i in range(a, b):
t = st[i]
c = sc[i]
s = int(round(t * sr))
d = Bunch(data=traces[s - k:s + k, :],
start_time=t - k / sr,
color=cs.get(c),
channel_ids=np.arange(5),
cluster_id=c,
)
out.waveforms.append(d)
return out
def get_cluster_pair_features(self, ci, cj):
si = self._select_spikes(ci, self.n_spikes_features)
sj = self._select_spikes(cj, self.n_spikes_features)
ni = self.get_cluster_templates(ci)
nj = self.get_cluster_templates(cj)
ti = self._get_template_features(si)
x0 = np.average(ti, weights=ni, axis=1)
y0 = np.average(ti, weights=nj, axis=1)
tj = self._get_template_features(sj)
x1 = np.average(tj, weights=ni, axis=1)
y1 = np.average(tj, weights=nj, axis=1)
return [
Bunch(x=x0, y=y0, spike_ids=si),
Bunch(x=x1, y=y1, spike_ids=sj)
]
def get_traces(interval):
out = Bunch(data=select_traces(traces, interval, sample_rate=sr),
color=(.75,) * 4,
)
a, b = st.searchsorted(interval)
out.waveforms = []
k = 20
for i in range(a, b):
t = st[i]
c = sc[i]
s = int(round(t * sr))
d = Bunch(data=traces[s - k:s + k, :],
start_time=t - k / sr,
color=cs.get(c),
channel_ids=np.arange(5),
spike_id=i,
spike_cluster=c,
)
out.waveforms.append(d)
return out
def get_waveforms(self, cluster_id):
m, M = self.get_waveform_lims()
if self.all_waveforms is not None:
# Waveforms.
waveforms_b = self._select_data(
cluster_id,
self.all_waveforms,
self.n_spikes_waveforms,
)
w = waveforms_b.data
# Sparsify.
channels = np.nonzero(w.mean(axis=1).mean(axis=0))[0]
w = w[:, :, channels]
waveforms_b.channels = channels
# Normalize.
mean = w.mean(axis=1).mean(axis=1)
w = w.astype(np.float64)
w -= mean[:, np.newaxis, np.newaxis]
w = _normalize(w, m, M)
waveforms_b.data = w
waveforms_b.cluster_id = cluster_id
waveforms_b.tag = 'waveforms'
else:
waveforms_b = None
# Find the templates corresponding to the cluster.
template_ids = np.nonzero(self.get_cluster_templates(cluster_id))[0]
# Templates.
templates = self.templates_unw[template_ids]
assert templates.ndim == 3
# Masks.
masks = self.template_masks[template_ids]
assert masks.ndim == 2
assert templates.shape[0] == masks.shape[0]
# Find mean amplitude.
spike_ids = self._select_spikes(cluster_id,
self.n_spikes_waveforms_lim)
mean_amp = self.all_amplitudes[spike_ids].mean()
# Normalize.
# mean = templates.mean(axis=1).mean(axis=1)
templates = templates.astype(np.float64).copy()
# templates -= mean[:, np.newaxis, np.newaxis]
templates *= mean_amp
templates *= 2. / (M - m)
template_b = Bunch(
data=templates,
masks=masks,
alpha=1.,
cluster_id=cluster_id,
tag='templates',
)
if waveforms_b is not None:
return [waveforms_b, template_b]
else:
return [template_b]
def get_cluster_pair_features(self, ci, cj):
si = self._select_spikes(ci, self.n_spikes_features)
sj = self._select_spikes(cj, self.n_spikes_features)
ni = self.get_cluster_templates(ci)
nj = self.get_cluster_templates(cj)
ti = self._get_template_features(si)
x0 = np.sum(ti * ni[np.newaxis, :], axis=1) / ni.sum()
y0 = np.sum(ti * nj[np.newaxis, :], axis=1) / nj.sum()
tj = self._get_template_features(sj)
x1 = np.sum(tj * ni[np.newaxis, :], axis=1) / ni.sum()
y1 = np.sum(tj * nj[np.newaxis, :], axis=1) / nj.sum()
d = Bunch()
d.x = np.hstack((x0, x1))
d.y = np.hstack((y0, y1))
d.spike_ids = np.hstack((si, sj))
d.spike_clusters = self.spike_clusters[d.spike_ids]
return d
def get_features(self, cluster_id, load_all=False):
# Overriden to take into account the sparse structure.
# Only keep spikes belonging to the features spike ids.
if self.features_spike_ids is not None:
# All spikes
spike_ids = self._select_spikes(cluster_id)
spike_ids = np.intersect1d(spike_ids, self.features_spike_ids)
# Relative indices of the spikes in the self.features_spike_ids
# array, necessary to load features from all_features which only
# contains the subset of the spikes.
spike_ids_rel = _index_of(spike_ids, self.features_spike_ids)
else:
spike_ids = self._select_spikes(
cluster_id, self.n_spikes_features if not load_all else None)
spike_ids_rel = spike_ids
st = self.spike_templates[spike_ids]
nc = self.n_channels
nfpc = self.n_features_per_channel
ns = len(spike_ids)
f = _densify(spike_ids_rel, self.all_features,
self.features_ind[st, :], self.n_channels)
f = np.transpose(f, (0, 2, 1))
assert f.shape == (ns, nc, nfpc)
b = Bunch()
# Normalize features.
m = self.get_feature_lim()
f = _normalize(f, -m, m)
b.data = f
b.spike_ids = spike_ids
b.spike_clusters = self.spike_clusters[spike_ids]
b.masks = self.all_masks[spike_ids]
return b
def get_background_features(self):
k = max(1, int(self.n_spikes // self.n_spikes_background_features))
spike_ids = slice(None, None, k)
b = Bunch()
b.data = self.all_features[spike_ids]
m = self.get_feature_lim()
b.data = _normalize(b.data.copy(), -m, +m)
b.spike_ids = spike_ids
b.spike_clusters = self.spike_clusters[spike_ids]
b.masks = self.all_masks[spike_ids]
return b
def get_amplitudes(self, cluster_id):
spike_ids = self._select_spikes(cluster_id, self.n_spikes_features)
d = Bunch()
d.spike_ids = spike_ids
d.spike_clusters = cluster_id * np.ones(len(spike_ids), dtype=np.int32)
d.x = self.spike_times[spike_ids]
d.y = self.all_amplitudes[spike_ids]
return d
def _get_waveforms(self, cluster_id):
"""Return a selection of waveforms for a cluster."""
pos = self.model.channel_positions
spike_ids = self.selector.select_spikes([cluster_id],
self.n_spikes_waveforms,
self.batch_size_waveforms,
)
channel_ids = self.get_best_channels(cluster_id)
data = self.model.get_waveforms(spike_ids, channel_ids)
data = data - data.mean()
return Bunch(data=data,
channel_ids=channel_ids,
channel_positions=pos[channel_ids],
)
def extract_spikes(traces, interval, sample_rate=None,
spike_times=None, spike_clusters=None,
all_masks=None,
n_samples_waveforms=None):
sr = sample_rate
ns = n_samples_waveforms
if not isinstance(ns, tuple):
ns = (ns // 2, ns // 2)
offset_samples = ns[0]
wave_len = ns[0] + ns[1]
# Find spikes.
a, b = spike_times.searchsorted(interval)
st = spike_times[a:b]
sc = spike_clusters[a:b]
m = all_masks[a:b]
n = len(st)
assert len(sc) == n
assert m.shape[0] == n
# Extract waveforms.
spikes = []
for i in range(n):
b = Bunch()
# Find the start of the waveform in the extracted traces.
sample_start = int(round((st[i] - interval[0]) * sr))
sample_start -= offset_samples
o = _extract_wave(traces, sample_start, m[i], wave_len)
if o is None: # pragma: no cover
logger.debug("Unable to extract spike %d.", i)
continue
b.waveforms, b.channels = o
# Masks on unmasked channels.
b.masks = m[i, b.channels]
b.spike_time = st[i]
b.spike_cluster = sc[i]
b.offset_samples = offset_samples
spikes.append(b)
return spikes
