def test_returns_correct_distance_for_multiunits(self):
a0 = neo.SpikeTrain(sp.array([1.0, 5.0, 7.0]) * pq.s, t_stop=8.0 * pq.s)
a1 = neo.SpikeTrain(sp.array([1.0, 2.0, 5.0]) * pq.s, t_stop=8.0 * pq.s)
b0 = neo.SpikeTrain(sp.array([2.0, 4.0, 5.0]) * pq.s, t_stop=8.0 * pq.s)
b1 = neo.SpikeTrain(sp.array([3.0, 8.0]) * pq.s, t_stop=9.0 * pq.s)
units = {0: [a0, a1], 1: [b0, b1]}
reassignment_cost = 0.7
expected = sp.array([[0.0, 4.4], [4.4, 0.0]])
actual = stm.victor_purpura_multiunit_dist(units, reassignment_cost)
assert_array_almost_equal(expected, actual)
import timeit
tau = 5.0 * pq.ms
def trains_as_multiunits(trains, trains_per_unit, num_units):
units = {}
for i in xrange(num_units):
units[i] = trains[i * trains_per_unit:(i + 1) * trains_per_unit]
return units
metrics = {
'vp': (r'$D_{\mathrm{V}}$ (multi-unit)',
lambda units: stm.victor_purpura_multiunit_dist(
units, 0.7, 2.0 / tau)),
'vr': (r'$D_{\mathrm{R}}$ (multi-unit)',
lambda units: stm.van_rossum_multiunit_dist(units, 0.5, tau))}
def print_available_metrics():
for key in metrics:
print "%s (%s)" % (key, metrics[key][0])
class BenchmarkData(object):
def __init__(
self, spike_count_range, train_count_range, num_units_range,
firing_rate=50 * pq.Hz):
self.spike_count_range = spike_count_range
self.train_count_range = train_count_range
def test_raises_exception_if_number_of_trials_differs(self):
st = create_empty_spike_train()
with self.assertRaises(ValueError):
stm.victor_purpura_multiunit_dist({0: [st], 1: [st, st]}, 1.0)
def test_returns_empty_array_if_trials_are_empty(self):
expected = sp.zeros((0, 0))
actual = stm.victor_purpura_multiunit_dist({0: [], 1: []}, 1.0)
assert_array_equal(expected, actual)
def test_returns_empty_array_if_empty_dict_is_passed(self):
expected = sp.zeros((0, 0))
actual = stm.victor_purpura_multiunit_dist({}, 1.0)
assert_array_equal(expected, actual)
def calc_metric(self, trains):
return stm.victor_purpura_multiunit_dist({0: trains}, 1)
lambda trains: stm.hunter_milton_similarity(trains, tau)),
'norm': ('norm distance',
lambda trains: stm.norm_dist(
trains, sigproc.CausalDecayingExpKernel(tau), sampling_rate)),
'ss': ('Schreiber et al. similarity measure',
lambda trains: stm.schreiber_similarity(
trains, sigproc.GaussianKernel(tau), sort=False)),
'vr': ('van Rossum distance',
lambda trains: stm.van_rossum_dist(trains, tau, sort=False)),
'vp': ('Victor-Purpura\'s distance',
lambda trains: stm.victor_purpura_dist(trains, 2.0 / tau)),
'vr_mu': ('van Rossum multi-unit distance',
lambda trains: stm.van_rossum_multiunit_dist(
trains_as_multiunits(trains), 0.5, tau)),
'vp_mu': ('Victor-Purpura\'s multi-unit distance',
lambda trains: stm.victor_purpura_multiunit_dist(
trains_as_multiunits(trains), 0.3, 2.0 / tau))}
def print_available_metrics():
for key in metrics:
print "%s (%s)" % (key, metrics[key][0])
def print_summary(profile_file):
stats = pstats.Stats(profile_file)
stats.strip_dirs().sort_stats('cumulative').print_stats(
r'^spike_train_metrics.py:\d+\([^_<].*(?<!compute)\)')
def profile_metrics(trains, to_profile):
for key in to_profile:
