def test_timeseries_1D_definition(self):
with pytest.raises(ValueError):
Timeseries(self.data_1D,
labels_dimensions={},
start_time=0,
sample_period=0.01,
sample_period_unit="ms")
def plot_timeseries_interactive(self, timeseries, first_n=-1, **kwargs):
if isinstance(timeseries, TimeSeries):
self.plot_tvb_timeseries_interactive(timeseries, first_n, **kwargs)
elif isinstance(timeseries, Timeseries):
self.plot_tvb_timeseries_interactive(timeseries._tvb, first_n, **kwargs)
elif isinstance(timeseries, numpy.ndarray):
self.plot_tvb_timeseries_interactive(Timeseries(data=timeseries), first_n, **kwargs)
elif isinstance(timeseries, (list, tuple)):
self.plot_tvb_timeseries_interactive(Timeseries(data=TimeSeries(data=numpy.stack(timeseries, axis=1))),
first_n, **kwargs)
elif isinstance(timeseries, dict):
ts = numpy.stack(timeseries.values(), axis=1)
timeseries = TimeSeries(data=ts, labels_dimensions={"State Variable": timeseries.keys()})
self.plot_tvb_timeseries_interactive(timeseries, first_n, **kwargs)
else:
raise_value_error("Input timeseries: %s \n" "is not on of one of the following types: "
"[Timeseries (tvb-scripts), TimeSeries (TVB), numpy.ndarray, dict, list, tuple]" %
str(timeseries))
def test_timeseries_3D(self):
ts_3D = Timeseries(self.data_3D,
labels_dimensions={
TimeseriesDimensions.SPACE.value:
numpy.array([]),
TimeseriesDimensions.VARIABLES.value:
numpy.array([])
},
start_time=self.start_time,
sample_period=self.sample_period,
sample_period_unit=self.sample_period_unit)
assert ts_3D.data.ndim == 4
assert ts_3D.data.shape[3] == 1
def test_timeseries_4D(self):
ts_4D = Timeseries(self.data_4D,
labels_dimensions={
TimeseriesDimensions.SPACE.value:
numpy.array(["r1", "r2", "r3", "r4"]),
TimeseriesDimensions.VARIABLES.value:
numpy.array([
PossibleVariables.X1.value,
PossibleVariables.X2.value, "sv3"
])
},
start_time=self.start_time,
sample_period=self.sample_period,
sample_period_unit=self.sample_period_unit)
assert ts_4D.data.shape == (3, 4, 3, 4)
assert ts_4D.x1.data.shape == (3, 4, 1, 4)
def concatenate(self, timeseries_list, dim, **kwargs):
timeseries_list = ensure_list(timeseries_list)
n_ts = len(timeseries_list)
if n_ts > 0:
out_timeseries, select_fun = self.select(timeseries_list[0],
**kwargs)
if n_ts > 1:
for id, timeseries in enumerate(timeseries_list[1:]):
if np.float32(out_timeseries.sample_period) != np.float32(
timeseries.sample_period):
raise_value_error(
"Timeseries concatenation failed!\n"
"Timeseries %d have a different time step (%s) \n "
"than the concatenated ones (%s)!" %
(id, str(np.float32(timeseries.sample_period)),
str(np.float32(out_timeseries.sample_period))))
else:
timeseries = self.select(timeseries, select_fun)[0]
try:
out_timeseries.set_data(
np.concatenate(
[out_timeseries.data, timeseries.data],
axis=dim))
if len(out_timeseries.labels_dimensions[
out_timeseries.labels_ordering[dim]]) > 0:
dim_label = out_timeseries.labels_ordering[dim]
out_timeseries.labels_dimensions[dim_label] = \
np.array(ensure_list(out_timeseries.labels_dimensions[dim_label]) +
ensure_list(timeseries.labels_dimensions[dim_label]))
except:
raise_value_error(
"Timeseries concatenation failed!\n"
"Timeseries %d have a shape (%s) and the concatenated ones (%s)!"
% (id, str(out_timeseries.shape),
str(timeseries.shape)))
return out_timeseries
else:
return out_timeseries
else:
return Timeseries(**kwargs)
def read_edf_to_Timeseries(path, sensors, rois_selection=None, label_strip_fun=None, time_unit="ms", **kwargs):
data, times, rois, rois_inds, rois_lbls = \
read_edf(path, sensors, rois_selection, label_strip_fun, time_unit)
return Timeseries(data, time=times, labels_dimensions={TimeseriesDimensions.SPACE.value: rois_lbls},
sample_period=np.mean(np.diff(times)), sample_period_unit=time_unit, **kwargs)
def test_timeseries_2D(self):
ts_from_2D = Timeseries(self.data_2D,
labels_dimensions={
TimeseriesDimensions.SPACE.value:
numpy.array(["r1", "r2", "r3"])
},
start_time=self.start_time,
sample_period=self.sample_period,
sample_period_unit=self.sample_period_unit)
assert ts_from_2D.data.ndim == 4
assert ts_from_2D.data.shape[2] == ts_from_2D.data.shape[3] == 1
assert ts_from_2D.data.shape == (3, 3, 1, 1)
assert ts_from_2D.end_time == 0.02
assert all(ts_from_2D.time == numpy.array([0, 0.01, 0.02]))
with pytest.raises(KeyError):
ts_from_2D.get_variables_by_label("")
ts_r2r3 = ts_from_2D.get_subspace_by_labels(["r2", "r3"])
assert ts_r2r3.data.ndim == 4
assert ts_r2r3.data.shape == (3, 2, 1, 1)
ts_r2 = ts_from_2D.get_subspace_by_labels(["r2"])
assert ts_r2.data.ndim == 4
assert ts_r2.data.shape == (3, 1, 1, 1)
assert ts_r2.labels_dimensions[
TimeseriesDimensions.SPACE.value] == numpy.array(["r2"])
assert ts_r2.get_subspace_by_labels(["r2"]).labels_dimensions[TimeseriesDimensions.SPACE.value] == \
numpy.array(["r2"])
with pytest.raises(IndexError):
ts_r2.get_subspace_by_labels(["r1"])
ts_r2r3_idx = ts_from_2D.get_subspace_by_index([1, 2])
assert ts_r2r3_idx.data.ndim == 4
assert ts_r2r3_idx.data.shape == (3, 2, 1, 1)
ts_r2_idx = ts_r2r3_idx.get_subspace_by_index([0])
assert ts_r2_idx.data.ndim == 4
assert ts_r2_idx.data.shape == (3, 1, 1, 1)
assert ts_r2_idx.labels_dimensions[
TimeseriesDimensions.SPACE.value] == numpy.array(["r2"])
assert ts_r2_idx.get_subspace_by_index([0]).labels_dimensions[TimeseriesDimensions.SPACE.value] == \
numpy.array(["r2"])
with pytest.raises(IndexError):
ts_r2_idx.get_subspace_by_index([2])
assert ts_r2r3_idx.data.all() == ts_r2r3.data.all()
assert all(ts_r2_idx.data == ts_r2.data)
ts_time_window = ts_from_2D.get_time_window(1, 2)
assert ts_time_window.data.ndim == 4
assert ts_time_window.data.shape == (1, 3, 1, 1)
assert numpy.array_equal(
ts_time_window.labels_dimensions[TimeseriesDimensions.SPACE.value],
numpy.array(["r1", "r2", "r3"]))
assert ts_time_window.start_time == 0.01
ts_time_window_units = ts_from_2D.get_time_window_by_units(0.01, 0.02)
assert ts_time_window_units.data.ndim == 4
assert ts_time_window_units.data.shape == (1, 3, 1, 1)
assert numpy.array_equal(
ts_time_window_units.labels_dimensions[
TimeseriesDimensions.SPACE.value],
numpy.array(["r1", "r2", "r3"]))
assert ts_time_window_units.start_time == 0.01
with pytest.raises(IndexError):
ts_from_2D.get_time_window(2, 4)
with pytest.raises(ValueError):
ts_from_2D.get_time_window_by_units(0, 0.025)
with pytest.raises(AttributeError):
ts_from_2D.lfp
def test_timeseries_data_access(self):
ts = Timeseries(self.data_3D,
labels_dimensions={
TimeseriesDimensions.SPACE.value:
numpy.array(["r1", "r2", "r3", "r4"]),
TimeseriesDimensions.VARIABLES.value:
numpy.array(["sv1", "sv2", "sv3"])
},
start_time=self.start_time,
sample_period=self.sample_period,
sample_period_unit=self.sample_period_unit)
assert isinstance(ts.r1, Timeseries)
assert ts.r1.data.shape == (3, 1, 3, 1)
assert isinstance(ts.sv1, Timeseries)
assert ts.sv1.data.shape == (3, 4, 1, 1)
with pytest.raises(AttributeError):
ts.r9
with pytest.raises(AttributeError):
ts.sv0
assert ts[:, :, :, :].shape == ts.data.shape
assert ts[1:, :, :, :].shape == ts.data[1:, :, :, :].shape
assert ts[1:2, :, :, :].shape == ts.data[1:2, :, :, :].shape
assert ts[1, :, :, :].shape == ts.data[1, :, :, :].shape
assert ts[:, 1:, :, :].shape == ts.data[:, 1:, :, :].shape
assert ts[:, :1, :, :].shape == ts.data[:, :1, :, :].shape
assert ts[:, 1:3, :, :].shape == ts.data[:, 1:3, :, :].shape
assert ts[:, 1, :, :].shape == ts.data[:, 1, :, :].shape
assert ts[:, "r2":, :, :].shape == ts.data[:, 1:, :, :].shape
assert ts[:, :"r2", :, :].shape == ts.data[:, :1, :, :].shape
assert ts[:, "r2", :, :].shape == ts.data[:, 1, :, :].shape
assert ts[:, "r2":"r4", :, :].shape == ts.data[:, 1:3, :, :].shape
assert ts[1:2, "r2":"r4", :, :].shape == ts.data[1:2, 1:3, :, :].shape
assert ts[1, "r2":"r4", :, :].shape == ts.data[1, 1:3, :, :].shape
assert ts[:, :, 1:, :].shape == ts.data[:, :, 1:, :].shape
assert ts[:, :, :1, :].shape == ts.data[:, :, :1, :].shape
assert ts[:, :, 0:2, :].shape == ts.data[:, :, 0:2, :].shape
assert ts[:, :, 2, :].shape == ts.data[:, :, 2, :].shape
assert ts[:, :, "sv2":, :].shape == ts.data[:, :, 1:, :].shape
assert ts[:, :, :"sv2", :].shape == ts.data[:, :, :1, :].shape
assert ts[:, :, "sv1":"sv3", :].shape == ts.data[:, :, 0:2, :].shape
assert ts[:, :, "sv3", :].shape == ts.data[:, :, 2, :].shape
assert ts[1:2, :, "sv2":, :].shape == ts.data[1:2, :, 1:, :].shape
assert ts[1:2, :, :"sv2", :].shape == ts.data[1:2, :, :1, :].shape
assert ts[1:2, :, "sv1":"sv3", :].shape == ts.data[1:2, :,
0:2, :].shape
assert ts[1:2, :, "sv3", :].shape == ts.data[1:2, :, 2, :].shape
assert ts[2, :, "sv3", :].shape == ts.data[2, :, 2, :].shape
assert ts[2, 0:3, "sv3", :].shape == ts.data[2, 0:3, 2, :].shape
assert ts[2, "r1":"r4", "sv3", :].shape == ts.data[2, 0:3, 2, :].shape
assert ts[0:2, "r1":"r4", "sv3", :].shape == ts.data[0:2, 0:3,
2, :].shape
assert ts[0:2, :"r2", "sv3", :].shape == ts.data[0:2, :1, 2, :].shape
assert ts[0:2, "r2":, "sv3", :].shape == ts.data[0:2, 1:, 2, :].shape
assert ts[0:2, "r1", "sv3", :].shape == ts.data[0:2, 0, 2, :].shape
assert all(ts[0:2, "r1", "sv3", :] == ts.data[0:2, 0, 2, :])
assert ts[0:2, "r1":"r3", "sv3", :].all() == ts.data[0:2, 0:2,
2, :].all()
assert ts[0:2,
"r1":"r3", :"sv2", :].all() == ts.data[0:2,
0:2, :1, :].all()
assert ts[2, "r1":"r3", :"sv2", :].all() == ts.data[2,
0:2, :1, :].all()
assert ts[2, "r3", "sv2", :].all() == ts.data[2, 2, 1, :].all()
assert ts[2, "r3", "sv2", 0] == ts.data[2, 2, 1, 0]
# IndexError because of [0][0] on numpy array in TS._get_index_of_state_variable
with pytest.raises(IndexError):
ts[:, :, "sv0", :]
with pytest.raises(IndexError):
ts[0, "r1":"r5", :, :]
with pytest.raises(IndexError):
ts[0, :, 10, :]
with pytest.raises(AttributeError):
ts.lfp