def test_bdf_stim_channel():
"""Test BDF stim channel."""
# test if last channel is detected as STIM by default
raw_py = _test_raw_reader(read_raw_edf, input_fname=bdf_path,
stim_channel='auto')
assert channel_type(raw_py.info, raw_py.info["nchan"] - 1) == 'stim'
# test BDF file with wrong scaling info in header - this should be ignored
# for BDF stim channels
events = [[242, 0, 4],
[310, 0, 2],
[952, 0, 1],
[1606, 0, 1],
[2249, 0, 1],
[2900, 0, 1],
[3537, 0, 1],
[4162, 0, 1],
[4790, 0, 1]]
with pytest.deprecated_call(match='stim_channel'):
raw = read_raw_edf(bdf_stim_channel_path, preload=True)
bdf_events = find_events(raw)
assert_array_equal(events, bdf_events)
raw = read_raw_edf(bdf_stim_channel_path, preload=False,
stim_channel='auto')
bdf_events = find_events(raw)
assert_array_equal(events, bdf_events)
def test_edf_data():
"""Test edf files"""
_test_raw_reader(read_raw_edf, input_fname=edf_path, stim_channel=None)
raw_py = read_raw_edf(edf_path, preload=True)
# Test saving and loading when annotations were parsed.
tempdir = _TempDir()
raw_file = op.join(tempdir, 'test-raw.fif')
raw_py.save(raw_file, overwrite=True, buffer_size_sec=1)
Raw(raw_file, preload=True)
edf_events = find_events(raw_py, output='step', shortest_event=0,
stim_channel='STI 014')
# onset, duration, id
events = [[0.1344, 0.2560, 2],
[0.3904, 1.0000, 2],
[2.0000, 0.0000, 3],
[2.5000, 2.5000, 2]]
events = np.array(events)
events[:, :2] *= 512 # convert time to samples
events = np.array(events, dtype=int)
events[:, 1] -= 1
events[events[:, 1] <= 0, 1] = 1
events[:, 1] += events[:, 0]
onsets = events[:, [0, 2]]
offsets = events[:, [1, 2]]
events = np.zeros((2 * events.shape[0], 3), dtype=int)
events[0::2, [0, 2]] = onsets
events[1::2, [0, 1]] = offsets
assert_array_equal(edf_events, events)
def test_edf_data():
"""Test edf files."""
raw = _test_raw_reader(read_raw_edf, input_fname=edf_path,
stim_channel=None, exclude=['Ergo-Left', 'H10'])
raw_py = read_raw_edf(edf_path, preload=True)
assert_equal(len(raw.ch_names) + 2, len(raw_py.ch_names))
# Test saving and loading when annotations were parsed.
edf_events = find_events(raw_py, output='step', shortest_event=0,
stim_channel='STI 014')
# onset, duration, id
events = [[0.1344, 0.2560, 2],
[0.3904, 1.0000, 2],
[2.0000, 0.0000, 3],
[2.5000, 2.5000, 2]]
events = np.array(events)
events[:, :2] *= 512 # convert time to samples
events = np.array(events, dtype=int)
events[:, 1] -= 1
events[events[:, 1] <= 0, 1] = 1
events[:, 1] += events[:, 0]
onsets = events[:, [0, 2]]
offsets = events[:, [1, 2]]
events = np.zeros((2 * events.shape[0], 3), dtype=int)
events[0::2, [0, 2]] = onsets
events[1::2, [0, 1]] = offsets
assert_array_equal(edf_events, events)
def test_edf_annotations():
"""Test if events are detected correctly in a typical MNE workflow."""
# test an actual file
raw = read_raw_edf(edf_path, preload=True)
edf_events = find_events(raw, output='step', shortest_event=0,
stim_channel='STI 014')
# onset, duration, id
events = [[0.1344, 0.2560, 2],
[0.3904, 1.0000, 2],
[2.0000, 0.0000, 3],
[2.5000, 2.5000, 2]]
events = np.array(events)
events[:, :2] *= 512 # convert time to samples
events = np.array(events, dtype=int)
events[:, 1] -= 1
events[events[:, 1] <= 0, 1] = 1
events[:, 1] += events[:, 0]
onsets = events[:, [0, 2]]
offsets = events[:, [1, 2]]
events = np.zeros((2 * events.shape[0], 3), dtype=int)
events[0::2, [0, 2]] = onsets
events[1::2, [0, 1]] = offsets
assert_array_equal(edf_events, events)
def test_stim_channel():
"""Test reading raw edf files with stim channel."""
raw_py = read_raw_edf(edf_path,
misc=range(-4, 0),
stim_channel=139,
preload=True)
picks = pick_types(raw_py.info,
meg=False,
eeg=True,
exclude=['EDF Annotations'])
data_py, _ = raw_py[picks]
print(raw_py) # to test repr
print(raw_py.info) # to test Info repr
# this .mat was generated using the EEG Lab Biosemi Reader
raw_eeglab = io.loadmat(edf_eeglab_path)
raw_eeglab = raw_eeglab['data'] * 1e-6 # data are stored in microvolts
data_eeglab = raw_eeglab[picks]
assert_array_almost_equal(data_py, data_eeglab, 10)
events = find_edf_events(raw_py)
assert_true(len(events) - 1 == len(find_events(raw_py))) # start not found
# Test uneven sampling
raw_py = read_raw_edf(edf_uneven_path, stim_channel=None)
data_py, _ = raw_py[0]
# this .mat was generated using the EEG Lab Biosemi Reader
raw_eeglab = io.loadmat(edf_uneven_eeglab_path)
raw_eeglab = raw_eeglab['data']
data_eeglab = raw_eeglab[0]
# match upsampling
upsample = len(data_eeglab) / len(raw_py)
data_py = np.repeat(data_py, repeats=upsample)
assert_array_equal(data_py, data_eeglab)
assert_raises(RuntimeError,
read_raw_edf,
edf_path,
preload=False,
stim_channel=-1)
with warnings.catch_warnings(record=True) as w:
warnings.simplefilter('always')
raw = read_raw_edf(edf_stim_resamp_path, verbose=True, stim_channel=-1)
assert_equal(len(w), 1)
assert_true('Events may jitter' in str(w[0].message))
with warnings.catch_warnings(record=True) as w:
warnings.simplefilter('always')
raw[:]
assert_equal(len(w), 0)
events = raw_py.find_edf_events()
assert_true(len(events) == 0)
def test_edf_data():
"""Test edf files."""
raw = _test_raw_reader(read_raw_edf,
input_fname=edf_path,
stim_channel=None,
exclude=['Ergo-Left', 'H10'],
verbose='error')
raw_py = read_raw_edf(edf_path, stim_channel='auto', preload=True)
assert_equal(len(raw.ch_names) + 2, len(raw_py.ch_names))
# Test saving and loading when annotations were parsed.
edf_events = find_events(raw_py,
output='step',
shortest_event=0,
stim_channel='STI 014')
# onset, duration, id
events = [[0.1344, 0.2560, 2], [0.3904, 1.0000, 2], [2.0000, 0.0000, 3],
[2.5000, 2.5000, 2]]
events = np.array(events)
events[:, :2] *= 512 # convert time to samples
events = np.array(events, dtype=int)
events[:, 1] -= 1
events[events[:, 1] <= 0, 1] = 1
events[:, 1] += events[:, 0]
onsets = events[:, [0, 2]]
offsets = events[:, [1, 2]]
events = np.zeros((2 * events.shape[0], 3), dtype=int)
events[0::2, [0, 2]] = onsets
events[1::2, [0, 1]] = offsets
assert_array_equal(edf_events, events)
# Test with number of records not in header (-1).
tempdir = _TempDir()
broken_fname = op.join(tempdir, 'broken.edf')
with open(edf_path, 'rb') as fid_in:
fid_in.seek(0, 2)
n_bytes = fid_in.tell()
fid_in.seek(0, 0)
rbytes = fid_in.read(int(n_bytes * 0.4))
with open(broken_fname, 'wb') as fid_out:
fid_out.write(rbytes[:236])
fid_out.write(bytes('-1 '.encode()))
fid_out.write(rbytes[244:])
with pytest.warns(RuntimeWarning,
match='records .* not match the file size'):
raw = read_raw_edf(broken_fname, preload=True, stim_channel='auto')
read_raw_edf(broken_fname,
exclude=raw.ch_names[:132],
preload=True,
stim_channel='auto')
def create_mdm(raw, event_id):
tmin, tmax = -1., 4.
events = find_events(raw, shortest_event=0, stim_channel='STI 014')
epochs = Epochs(raw, events, event_id, tmin, tmax, proj=True,
baseline=None, preload=True, verbose=False)
labels = epochs.events[:, -1]
epochs_data_train = epochs.get_data()[:, :-1]
cov_data_train = Covariances().transform(epochs_data_train)
mdm = MDM(metric=dict(mean='riemann', distance='riemann'))
mdm.fit(cov_data_train, labels)
return mdm
def test_bdf_stim_channel():
"""Test BDF stim channel."""
# test if last channel is detected as STIM by default
raw_py = _test_raw_reader(read_raw_edf,
input_fname=bdf_path,
stim_channel='auto')
assert channel_type(raw_py.info, raw_py.info["nchan"] - 1) == 'stim'
# test BDF file with wrong scaling info in header - this should be ignored
# for BDF stim channels
events = [[242, 0, 4], [310, 0, 2], [952, 0, 1], [1606, 0, 1],
[2249, 0, 1], [2900, 0, 1], [3537, 0, 1], [4162, 0, 1],
[4790, 0, 1]]
raw = read_raw_edf(bdf_stim_channel_path, preload=True)
bdf_events = find_events(raw)
assert_array_equal(events, bdf_events)
raw = read_raw_edf(bdf_stim_channel_path,
preload=False,
stim_channel='auto')
bdf_events = find_events(raw)
assert_array_equal(events, bdf_events)
def test_stim_channel():
"""Test reading raw edf files with stim channel."""
raw_py = read_raw_edf(edf_path, misc=range(-4, 0), stim_channel=139,
preload=True)
picks = pick_types(raw_py.info, meg=False, eeg=True,
exclude=['EDF Annotations'])
data_py, _ = raw_py[picks]
print(raw_py) # to test repr
print(raw_py.info) # to test Info repr
# this .mat was generated using the EEG Lab Biosemi Reader
raw_eeglab = io.loadmat(edf_eeglab_path)
raw_eeglab = raw_eeglab['data'] * 1e-6 # data are stored in microvolts
data_eeglab = raw_eeglab[picks]
assert_array_almost_equal(data_py, data_eeglab, 10)
events = find_edf_events(raw_py)
assert_true(len(events) - 1 == len(find_events(raw_py))) # start not found
# Test uneven sampling
raw_py = read_raw_edf(edf_uneven_path, stim_channel=None)
data_py, _ = raw_py[0]
# this .mat was generated using the EEG Lab Biosemi Reader
raw_eeglab = io.loadmat(edf_uneven_eeglab_path)
raw_eeglab = raw_eeglab['data']
data_eeglab = raw_eeglab[0]
# match upsampling
upsample = len(data_eeglab) / len(raw_py)
data_py = np.repeat(data_py, repeats=upsample)
assert_array_equal(data_py, data_eeglab)
assert_raises(RuntimeError, read_raw_edf, edf_path, preload=False,
stim_channel=-1)
with warnings.catch_warnings(record=True) as w:
warnings.simplefilter('always')
raw = read_raw_edf(edf_stim_resamp_path, verbose=True, stim_channel=-1)
assert_equal(len(w), 2)
assert_true(any('Events may jitter' in str(ww.message) for ww in w))
assert_true(any('truncated' in str(ww.message) for ww in w))
with warnings.catch_warnings(record=True) as w:
warnings.simplefilter('always')
raw[:]
assert_equal(len(w), 0)
events = raw_py.find_edf_events()
assert_true(len(events) == 0)
def test_edf_data():
"""Test edf files."""
raw = _test_raw_reader(read_raw_edf, input_fname=edf_path,
stim_channel=None, exclude=['Ergo-Left', 'H10'],
verbose='error')
raw_py = read_raw_edf(edf_path, stim_channel='auto', preload=True)
assert_equal(len(raw.ch_names) + 2, len(raw_py.ch_names))
# Test saving and loading when annotations were parsed.
edf_events = find_events(raw_py, output='step', shortest_event=0,
stim_channel='STI 014')
# onset, duration, id
events = [[0.1344, 0.2560, 2],
[0.3904, 1.0000, 2],
[2.0000, 0.0000, 3],
[2.5000, 2.5000, 2]]
events = np.array(events)
events[:, :2] *= 512 # convert time to samples
events = np.array(events, dtype=int)
events[:, 1] -= 1
events[events[:, 1] <= 0, 1] = 1
events[:, 1] += events[:, 0]
onsets = events[:, [0, 2]]
offsets = events[:, [1, 2]]
events = np.zeros((2 * events.shape[0], 3), dtype=int)
events[0::2, [0, 2]] = onsets
events[1::2, [0, 1]] = offsets
assert_array_equal(edf_events, events)
# Test with number of records not in header (-1).
tempdir = _TempDir()
broken_fname = op.join(tempdir, 'broken.edf')
with open(edf_path, 'rb') as fid_in:
fid_in.seek(0, 2)
n_bytes = fid_in.tell()
fid_in.seek(0, 0)
rbytes = fid_in.read(int(n_bytes * 0.4))
with open(broken_fname, 'wb') as fid_out:
fid_out.write(rbytes[:236])
fid_out.write(b'-1 ')
fid_out.write(rbytes[244:])
with pytest.warns(RuntimeWarning,
match='records .* not match the file size'):
raw = read_raw_edf(broken_fname, preload=True, stim_channel='auto')
read_raw_edf(broken_fname, exclude=raw.ch_names[:132], preload=True,
stim_channel='auto')
def test_stim_channel():
"""Test reading raw edf files with stim channel."""
raw_py = read_raw_edf(edf_path, misc=range(-4, 0), stim_channel=139,
preload=True)
picks = pick_types(raw_py.info, meg=False, eeg=True,
exclude=['EDF Annotations'])
data_py, _ = raw_py[picks]
print(raw_py) # to test repr
print(raw_py.info) # to test Info repr
# this .mat was generated using the EEG Lab Biosemi Reader
raw_eeglab = io.loadmat(edf_eeglab_path)
raw_eeglab = raw_eeglab['data'] * 1e-6 # data are stored in microvolts
data_eeglab = raw_eeglab[picks]
assert_array_almost_equal(data_py, data_eeglab, 10)
events = find_edf_events(raw_py)
assert (len(events) - 1 == len(find_events(raw_py))) # start not found
# Test uneven sampling
raw_py = read_raw_edf(edf_uneven_path, stim_channel=None)
data_py, _ = raw_py[0]
# this .mat was generated using the EEG Lab Biosemi Reader
raw_eeglab = io.loadmat(edf_uneven_eeglab_path)
raw_eeglab = raw_eeglab['data']
data_eeglab = raw_eeglab[0]
# match upsampling
upsample = len(data_eeglab) / len(raw_py)
data_py = np.repeat(data_py, repeats=upsample)
assert_array_equal(data_py, data_eeglab)
pytest.raises(RuntimeError, read_raw_edf, edf_path, preload=False,
stim_channel=-1)
with pytest.warns(RuntimeWarning,
match='Interpolating stim .* Events may jitter'):
raw = read_raw_edf(edf_stim_resamp_path, verbose=True, stim_channel=-1)
with pytest.warns(None) as w:
raw[:]
assert len(w) == 0
events = raw_py.find_edf_events()
assert (len(events) == 0)
def test_find_events_and_events_from_annot_are_the_same():
"""Test that old behaviour and new produce the same events."""
EXPECTED_EVENTS = [[68, 0, 2], [199, 0, 2], [1024, 0, 3], [1280, 0, 2]]
raw = read_raw_edf(edf_path, preload=True)
raw_shell = _get_empty_raw_with_valid_annot(edf_path)
assert raw_shell.info['meas_date'] == raw.info['meas_date']
assert raw_shell.info['sfreq'] == raw.info['sfreq']
assert raw_shell.first_samp == raw.first_samp
events_from_find_events = find_events(raw)
assert_array_equal(events_from_find_events, EXPECTED_EVENTS)
annot = read_annotations_edf(edf_path)
raw_shell.set_annotations(annot)
event_id = _get_edf_default_event_id(annot.description)
event_id.pop('start')
events_from_EFA, _ = events_from_annotations(raw_shell,
event_id=event_id,
use_rounding=False)
assert_array_equal(events_from_EFA, events_from_find_events)
def test_find_events_and_events_from_annot_are_the_same():
"""Test that old behaviour and new produce the same events."""
EXPECTED_EVENTS = [[68, 0, 2],
[199, 0, 2],
[1024, 0, 3],
[1280, 0, 2]]
raw = read_raw_edf(edf_path, preload=True, stim_channel='auto')
raw_shell = _get_empty_raw_with_valid_annot(edf_path)
assert raw_shell.info['meas_date'] == raw.info['meas_date']
assert raw_shell.info['sfreq'] == raw.info['sfreq']
assert raw_shell.first_samp == raw.first_samp
events_from_find_events = find_events(raw)
assert_array_equal(events_from_find_events, EXPECTED_EVENTS)
annot = read_annotations(edf_path)
raw_shell.set_annotations(annot)
event_id = _get_edf_default_event_id(annot.description)
event_id.pop('start')
events_from_EFA, _ = events_from_annotations(raw_shell, event_id=event_id,
use_rounding=False)
assert_array_equal(events_from_EFA, events_from_find_events)
def get_score(subject=7, runs=[6, 10, 14], event_id=dict(hands=2, feet=3)):
tmin, tmax = -1., 4.
raw = get_raw(subject, runs)
events = find_events(raw, shortest_event=0, stim_channel='STI 014')
epochs = Epochs(raw, events, event_id, tmin, tmax, proj=True,
baseline=None, preload=True, verbose=False)
labels = epochs.events[:, -1]
# cv = KFold(len(labels), 10, shuffle=True, random_state=42)
epochs_data_train = epochs.get_data()[:, :-1]
cov_data_train = Covariances().transform(epochs_data_train)
###############################################################################
# Classification with Minimum distance to mean
mdm = MDM(metric=dict(mean='riemann', distance='riemann'))
pl = Pipeline([("mdm", mdm)])
params = {"mdm__metric": [dict(mean='riemann', distance='riemann')]}
clf = GridSearchCV(pl, params, n_jobs=-1, cv=5, return_train_score=True)
clf.fit(cov_data_train, labels)
df = pd.DataFrame(clf.cv_results_)
return df
# cue onset.
tmin, tmax = -2., 6.
event_id = dict(hands=2, feet=3)
subject = 1
runs = [6, 10, 14] # motor imagery: hands vs feet
raw_files = [
read_raw_edf(f, preload=True, verbose=False)
for f in eegbci.load_data(subject, runs)
]
raw = concatenate_raws(raw_files)
# strip channel names
raw.info['ch_names'] = [chn.strip('.') for chn in raw.info['ch_names']]
events = find_events(raw, shortest_event=0, stim_channel='STI 014')
picks = pick_types(raw.info,
meg=False,
eeg=True,
stim=False,
eog=False,
exclude='bads')
raw.filter(7., 35., method='iir', picks=picks)
epochs = Epochs(raw,
events,
event_id,
tmin,
tmax,
proj=True,
def get_score(subject=7, runs=[6, 10, 14], event_id=dict(hands=2, feet=3)):
if subject in EXCLUDE_SUBJECTS:
return
tmin, tmax = -1., 4.
weights = np.arange(0.1, 1.0, 0.1)
for weight in weights:
first_sub = 2 if subject == 1 else 1
raw = get_raw(subject, runs)
scores = []
for i in range(first_sub, TRANS_SUBJECT_COUNT):
print(i)
if i == subject or (i in EXCLUDE_SUBJECTS):
continue
raw.append(get_raw(i, runs))
events = find_events(raw, shortest_event=0, stim_channel='STI 014')
epochs = Epochs(raw, events, event_id, tmin, tmax, proj=True,
baseline=None, preload=True, verbose=False)
labels = epochs.events[:, -1]
epochs_data_train = 1e6*epochs.get_data()[:, :-1]
cov_data_train = Covariances().transform(epochs_data_train)
target_sample_weight_base = np.ones(EPOCH_COUNT)*weight
others_sample_weight_base = np.ones(
len(epochs)-EPOCH_COUNT)*(1.-weight)
sample_weight = np.hstack(
(target_sample_weight_base, others_sample_weight_base))
others_size = others_sample_weight_base.size
others_index = np.arange(others_size)
mdm = MDM(metric=dict(mean='riemann', distance='riemann'))
cv = KFold(n_splits=5, shuffle=True, random_state=42)
train_scores = []
test_scores = []
dumy_array = np.ones(EPOCH_COUNT)
for train_index, test_index in cv.split(dumy_array):
train_index = np.hstack(
(others_index, train_index+others_size))
x = cov_data_train[train_index]
y = labels[train_index]
mdm.fit(x, y, sample_weight=sample_weight[train_index])
score = (mdm.predict(x) == y).sum()/len(train_index)
train_scores.append(score)
test_index = test_index + others_size
y = mdm.predict(cov_data_train[test_index])
score = (y == labels[test_index]).sum()/len(test_index)
test_scores.append(score)
train_score = np.mean(train_scores)
test_score = np.mean(test_scores)
scores.append([subject, i, train_score, test_score])
df = pd.DataFrame(
scores, columns=["subject", "transfer_count", "train_score", "test_score"])
df.to_excel("data/riemann/gradually/test_subject_%d_weight_%e.xlsx" %
(subject, weight), index=False)
def get_score(subject=7):
###############################################################################
# Set parameters and read data
# avoid classification of evoked responses by using epochs that start 1s after
# cue onset.
tmin, tmax = 1., 2.
event_id = dict(hands=2, feet=3)
runs = [6, 10, 14] # motor imagery: hands vs feet
raw_files = [
read_raw_edf(f, preload=True) for f in eegbci.load_data(subject, runs)
]
raw = concatenate_raws(raw_files)
picks = pick_types(raw.info,
meg=False,
eeg=True,
stim=False,
eog=False,
exclude='bads')
# subsample elecs
picks = picks[::2]
# Apply band-pass filter
raw.filter(7., 35., method='iir', picks=picks)
events = find_events(raw, shortest_event=0, stim_channel='STI 014')
# Read epochs (train will be done only between 1 and 2s)
# Testing will be done with a running classifier
epochs = Epochs(raw,
events,
event_id,
tmin,
tmax,
proj=True,
picks=picks,
baseline=None,
preload=True,
verbose=False)
labels = epochs.events[:, -1] - 2
# cross validation
cv = KFold(len(labels), 10, shuffle=True, random_state=42)
# get epochs
epochs_data_train = 1e6 * epochs.get_data()
# compute covariance matrices
cov_data_train = Covariances().transform(epochs_data_train)
###############################################################################
# Classification with Minimum distance to mean
mdm = MDM(metric=dict(mean='riemann', distance='riemann'))
# Use scikit-learn Pipeline with cross_val_score function
mdm.fit(cov_data_train, labels)
print(123)
###############################################################################
# Classification with Tangent Space Logistic Regression
clf = TSclassifier()
# Use scikit-learn Pipeline with cross_val_score function
scores = cross_val_score(clf, cov_data_train, labels, cv=cv, n_jobs=1)
# Printing the results
class_balance = np.mean(labels == labels[0])
class_balance = max(class_balance, 1. - class_balance)
ts_score = np.mean(scores)
print("Tangent space Classification accuracy: %f / Chance level: %f" %
(ts_score, class_balance))
###############################################################################
return [subject, mdm_score, ts_score]
def edf_loader(self, single_file):
raw = read_raw_edf(single_file, preload=True)
# print "raw: "
# print raw
# print "number of channels: %d, type: %s, raw.ch_names: " % (len(raw.ch_names), type(raw.ch_names))
# print raw.ch_names
start, stop = raw.time_as_index([0, 150])
ch_start = 2
ch_end = 16
data, times = raw[ch_start:ch_end, start:stop]
# print data.shape
# print type(data)
# print raw.ch_names[2:16]
# print "TTTTTTTTTTTTTTTTTTTTTTT"
# print type(times)
# print np.shape(times)
# print times
# print "TTTTTTTTTTTTTTTTTTTTTTT"
cha_names = []
for n in raw.ch_names:
str_n = str(n)
# str_n = str(unicodedata.normalize('NFKD', n).encode('ascii', 'ignore'))
cha_names.append(str_n)
# print type(raw.ch_names[0])
# print "cha_names: "
# print cha_names
data_dict = {i[0]: i[1:] for i in zip(cha_names[2:16], data)}
# print "data: "
# print data
# print "data_dict['AF3']: "
# print len(data_dict['AF3'][0])
d_dict = {}
for k, v in data_dict.iteritems():
d_dict[k] = v[0]
# print "d_dict['AF3']:"
# print len(d_dict['AF3'])
raw.info['ch_names'] = [chn.strip('.') for chn in raw.info['ch_names']]
raw.filter(2., 30., method='iir')
events = find_events(raw, shortest_event=0, stim_channel='STI 014')
print "events: "
print events
picks = pick_types(raw.info,
meg=False,
eeg=True,
stim=False,
exclude='bads')
"""
epochs = Epochs(raw, events, self.event_id, self.tmin, self.tmax, proj=True, picks=picks, baseline=None, preload=True, add_eeg_ref=False)
epoch_train = epochs.crop(tmin=1., tmax=2., copy=True)
labels = epochs.event[:, -1] - 2
"""
# raw.info['ch_names'] = [chn.strip('.') for chn in raw.info['ch_names']]
# epochs = Epochs()
# print "picks: "
# print picks
# print raw[picks]
# print type(raw)
# print raw.info
# print d_dict
# print d_dict.keys()
return d_dict
def get_score(subject=7, runs=[6, 10, 14], event_id=dict(hands=2, feet=3)):
tmin, tmax = -1., 4.
# learn all suject exclude target subject. #############################
first_sub = 2 if subject == 1 else 1
raw = get_raw(first_sub, runs)
for i in range(first_sub + 1, 3):
if i != subject and not (i in [88, 89, 92, 100]):
# print(i)
raw.append(get_raw(i, runs))
raw.append(get_raw(subject, runs))
events = find_events(raw, shortest_event=0, stim_channel='STI 014')
epochs = Epochs(raw,
events,
event_id,
tmin,
tmax,
proj=True,
baseline=None,
preload=True,
verbose=False)
labels = epochs.events[:, -1]
epochs_data_train = 1e6 * epochs.get_data()[:, :-1]
cov_data_train = Covariances().transform(epochs_data_train)
weights = np.arange(0.1, 1.0, 0.1)
scores = []
for weight in weights:
mdm = MDM(metric=dict(mean='riemann', distance='riemann'))
others_sample_weight_base = np.ones(len(epochs) -
EPOCH_COUNT) * (1. - weight)
target_sample_weight_base = np.ones(EPOCH_COUNT) * weight
sample_weight = np.hstack(
(others_sample_weight_base, target_sample_weight_base))
others_size = others_sample_weight_base.size
others_index = np.arange(others_size)
cv = KFold(n_splits=5, shuffle=True, random_state=42)
train_scores = []
test_scores = []
dumy_array = np.ones(EPOCH_COUNT)
for train_index, test_index in cv.split(dumy_array):
train_index = np.hstack((others_index, train_index + others_size))
x = cov_data_train[train_index]
y = labels[train_index]
mdm.fit(x, y, sample_weight=sample_weight[train_index])
score = (mdm.predict(x) == y).sum() / len(train_index)
train_scores.append(score)
test_index = test_index + others_size
y = mdm.predict(cov_data_train[test_index])
score = (y == labels[test_index]).sum() / len(test_index)
test_scores.append(score)
train_score = np.mean(train_scores)
test_score = np.mean(test_scores)
scores.append([subject, weight, train_score, test_score])
# print("train:%s test:%s" % (train_score, test_score))
return scores
for s,subject in enumerate(subjects):
print('Processing Subject %s' %(subject))
raw_files = [read_raw_edf(f, preload=True,verbose=False) for f in eegbci.load_data(subject, runs)]
raw = concatenate_raws(raw_files)
picks = pick_types(raw.info, meg=False, eeg=True, stim=False, eog=False,
exclude='bads')
# subsample elecs
picks = picks[::2]
# Apply band-pass filter
raw.filter(7., 35., method='iir',picks=picks)
events = find_events(raw, shortest_event=0, stim_channel='STI 014',verbose=False)
# Read epochs (train will be done only between 1 and 2s)
# Testing will be done with a running classifier
epochs = Epochs(raw, events, event_id, tmin, tmax, proj=True, picks=picks,
baseline=None, preload=True, add_eeg_ref=False,verbose = False)
labels = epochs.events[:, -1] - 2
X = epochs.get_data()
for i,c in enumerate(classifiers):
r = cross_val_score(classifiers[c],X,labels,cv=10,n_jobs=-1)
results[s,i] = np.mean(r)
def get_event_data(self,
event_id,
tmin,
tmax,
picks=None,
stim_channel=None,
min_duration=0):
"""Simulate the data for a particular event-id.
The epochs corresponding to a particular event-id are returned. The
method remembers the epoch that was returned in the previous call and
returns the next epoch in sequence. Once all epochs corresponding to
an event-id have been exhausted, the method returns None.
Parameters
----------
event_id : int
The id of the event to consider.
tmin : float
Start time before event.
tmax : float
End time after event.
%(picks_all)s
stim_channel : None | string | list of string
Name of the stim channel or all the stim channels
affected by the trigger. If None, the config variables
'MNE_STIM_CHANNEL', 'MNE_STIM_CHANNEL_1', 'MNE_STIM_CHANNEL_2',
etc. are read. If these are not found, it will default to
'STI 014'.
min_duration : float
The minimum duration of a change in the events channel required
to consider it as an event (in seconds).
Returns
-------
data : 2D array with shape [n_channels, n_times]
The epochs that are being simulated
"""
# Get the list of all events
picks = _picks_to_idx(self.info, picks, 'all', exclude=())
events = find_events(self.raw,
stim_channel=stim_channel,
verbose=False,
output='onset',
consecutive='increasing',
min_duration=min_duration)
# Get the list of only the specified event
idx = np.where(events[:, -1] == event_id)[0]
event_samp = events[idx, 0]
# Only do this the first time for each event type
if event_id not in self._current:
# Initialize pointer for the event to 0
self._current[event_id] = 0
self._last[event_id] = len(event_samp)
# relative start and stop positions in samples
tmin_samp = int(round(self.info['sfreq'] * tmin))
tmax_samp = int(round(self.info['sfreq'] * tmax)) + 1
if self._current[event_id] < self._last[event_id]:
# Select the current event from the events list
ev_samp = event_samp[self._current[event_id]]
# absolute start and stop positions in samples
start = ev_samp + tmin_samp - self.raw.first_samp
stop = ev_samp + tmax_samp - self.raw.first_samp
self._current[event_id] += 1 # increment pointer
data, _ = self.raw[picks, start:stop]
return data
else:
return None
# cue onset.
tmin, tmax = 1., 3.
event_id = dict(hands=2, feet=3)
subject = 1
runs = [6, 10, 14] # motor imagery: hands vs feet
raw_files = [read_raw_edf(f, preload=True,verbose=False) for f in eegbci.load_data(subject, runs) ]
raw = concatenate_raws(raw_files)
# strip channel names
raw.info['ch_names'] = [chn.strip('.') for chn in raw.info['ch_names']]
# Apply band-pass filter
raw.filter(7., 35., method='iir')
events = find_events(raw, shortest_event=0, stim_channel='STI 014')
picks = pick_types(raw.info, meg=False, eeg=True, stim=False, eog=False,
exclude='bads')
# Read epochs (train will be done only between 1 and 2s)
# Testing will be done with a running classifier
epochs = Epochs(raw, events, event_id, tmin, tmax, proj=True, picks=picks,
baseline=None, preload=True, add_eeg_ref=False,verbose=False)
labels = epochs.events[:, -1] - 2
# get epochs
epochs_data = epochs.get_data()
# compute covariance matrices
covmats = Covariances().fit_transform(epochs_data)
raw = concatenate_raws(raw_files)
picks = pick_types(raw.info,
meg=False,
eeg=True,
stim=False,
eog=False,
exclude='bads')
# subsample elecs
picks = picks[::2]
# Apply band-pass filter
raw.filter(7., 35., method='iir', picks=picks)
events = find_events(raw,
shortest_event=0,
stim_channel='STI 014',
verbose=False)
# Read epochs (train will be done only between 1 and 2s)
# Testing will be done with a running classifier
epochs = Epochs(raw,
events,
event_id,
tmin,
tmax,
proj=True,
picks=picks,
baseline=None,
preload=True,
add_eeg_ref=False,
verbose=False)