def test_slice_browser_display(_slice_browser):
"""Test that the slice browser GUI displays properly."""
# test no seghead, fsaverage doesn't have seghead
with pytest.warns(RuntimeWarning, match='`seghead` not found'):
with catch_logging() as log:
_slice_browser(subject='fsaverage', subjects_dir=subjects_dir,
verbose=True)
log = log.getvalue()
assert 'using marching cubes' in log
# test functions
with pytest.warns(RuntimeWarning, match='`pial` surface not found'):
gui = _slice_browser(subject=subject, subjects_dir=subjects_dir)
# test RAS
gui._RAS_textbox.setPlainText('10 10 10\n')
assert_allclose(gui._ras, [10, 10, 10])
# test vox
gui._VOX_textbox.setPlainText('150, 150, 150\n')
assert_allclose(gui._ras, [23, 22, 23])
# test click
with use_log_level('debug'):
_fake_click(gui._figs[2], gui._figs[2].axes[0],
[137, 140], xform='data', kind='release')
assert_allclose(gui._ras, [10, 12, 23])
def test_manifest_check_download(tmpdir, n_have, monkeypatch):
"""Test our manifest downloader."""
monkeypatch.setattr(datasets.utils, '_fetch_file', _fake_zip_fetch)
destination = op.join(str(tmpdir), 'empty')
manifest_path = op.join(str(tmpdir), 'manifest.txt')
with open(manifest_path, 'w') as fid:
for fname in _zip_fnames:
fid.write('%s\n' % fname)
assert n_have in range(len(_zip_fnames) + 1)
assert not op.isdir(destination)
if n_have > 0:
os.makedirs(op.join(destination, 'foo'))
assert op.isdir(op.join(destination, 'foo'))
for fname in _zip_fnames:
assert not op.isfile(op.join(destination, fname))
for fname in _zip_fnames[:n_have]:
with open(op.join(destination, fname), 'w'):
pass
with catch_logging() as log:
with use_log_level(True):
url = hash_ = '' # we mock the _fetch_file so these are not used
_manifest_check_download(manifest_path, destination, url, hash_)
log = log.getvalue()
n_missing = 3 - n_have
assert ('%d file%s missing from' % (n_missing, _pl(n_missing))) in log
for want in ('Extracting missing', 'Successfully '):
if n_missing > 0:
assert want in log
else:
assert want not in log
assert op.isdir(destination)
for fname in _zip_fnames:
assert op.isfile(op.join(destination, fname))
def test_progressbar_parallel_more(capsys):
"""Test ProgressBar with parallel computing, advanced version."""
assert capsys.readouterr().out == ''
# This must be "1" because "capsys" won't get stdout properly otherwise
parallel, p_fun, _ = parallel_func(_identity_block_wide,
n_jobs=1,
verbose=False)
arr = np.arange(10)
with use_log_level(True):
with ProgressBar(len(arr) * 2) as pb:
out = parallel(
p_fun(x, pb.subset(pb_idx))
for pb_idx, x in array_split_idx(arr, 2, n_per_split=2))
idxs = np.concatenate([o[1] for o in out])
assert_array_equal(idxs, np.arange(len(arr) * 2))
out = np.concatenate([o[0] for o in out])
assert op.isfile(pb._mmap_fname)
sum_ = np.memmap(pb._mmap_fname,
dtype='bool',
mode='r',
shape=len(arr) * 2).sum()
assert sum_ == len(arr) * 2
assert not op.isfile(pb._mmap_fname), '__exit__ not called?'
cap = capsys.readouterr()
out = cap.err
assert '100%' in out
def test_manifest_check_download(tmpdir, n_have, monkeypatch):
"""Test our manifest downloader."""
monkeypatch.setattr(datasets.utils, '_fetch_file', _fake_zip_fetch)
destination = op.join(str(tmpdir), 'empty')
manifest_path = op.join(str(tmpdir), 'manifest.txt')
with open(manifest_path, 'w') as fid:
for fname in _zip_fnames:
fid.write('%s\n' % fname)
assert n_have in range(len(_zip_fnames) + 1)
assert not op.isdir(destination)
if n_have > 0:
os.makedirs(op.join(destination, 'foo'))
assert op.isdir(op.join(destination, 'foo'))
for fname in _zip_fnames:
assert not op.isfile(op.join(destination, fname))
for fname in _zip_fnames[:n_have]:
with open(op.join(destination, fname), 'w'):
pass
with catch_logging() as log:
with use_log_level(True):
url = hash_ = '' # we mock the _fetch_file so these are not used
_manifest_check_download(manifest_path, destination, url, hash_)
log = log.getvalue()
n_missing = 3 - n_have
assert ('%d file%s missing from' % (n_missing, _pl(n_missing))) in log
for want in ('Extracting missing', 'Successfully '):
if n_missing > 0:
assert want in log
else:
assert want not in log
assert op.isdir(destination)
for fname in _zip_fnames:
assert op.isfile(op.join(destination, fname))
def test_progressbar():
"""Test progressbar class."""
a = np.arange(10)
pbar = ProgressBar(a)
assert a is pbar.iterable
assert pbar.max_value == 10
pbar = ProgressBar(10)
assert pbar.max_value == 10
assert pbar.iterable is None
# Make sure that non-iterable input raises an error
def iter_func(a):
for ii in a:
pass
pytest.raises(Exception, iter_func, ProgressBar(20))
# Make sure different progress bars can be used
with catch_logging() as log, modified_env(MNE_TQDM='tqdm'), \
use_log_level('debug'), ProgressBar(np.arange(3)) as pbar:
for p in pbar:
pass
log = log.getvalue()
assert 'Using ProgressBar with tqdm\n' in log
with modified_env(MNE_TQDM='broken'), pytest.raises(ValueError):
ProgressBar(np.arange(3))
with modified_env(MNE_TQDM='tqdm.broken'), pytest.raises(AttributeError):
ProgressBar(np.arange(3))
def test_progressbar_parallel_basic(capsys):
"""Test ProgressBar with parallel computing, basic version."""
assert capsys.readouterr().out == ''
parallel, p_fun, _ = parallel_func(_identity, total=10, n_jobs=1,
verbose=True)
with use_log_level(True):
out = parallel(p_fun(x) for x in range(10))
assert out == list(range(10))
cap = capsys.readouterr()
out = cap.err
assert '100%' in out
def test_frame_info(capsys, monkeypatch):
"""Test _frame_info."""
stack = _frame_info(100)
assert 2 < len(stack) < 100
this, pytest_line = stack[:2]
assert re.match('^test_logging:[1-9][0-9]$', this) is not None, this
assert 'pytest' in pytest_line
capsys.readouterr()
with use_log_level('debug', add_frames=4):
_fun()
out, _ = capsys.readouterr()
out = out.replace('\n', ' ')
assert re.match(
'.*pytest'
'.*test_logging:[2-9][0-9] '
'.*test_logging:[1-9][0-9] :.*Test', out) is not None, this
monkeypatch.setattr('inspect.currentframe', lambda: None)
assert _frame_info(1) == ['unknown']
def test_generalization_across_time():
"""Test time generalization decoding
"""
from sklearn.svm import SVC
from sklearn.base import is_classifier
# KernelRidge is used for testing 1) regression analyses 2) n-dimensional
# predictions.
from sklearn.kernel_ridge import KernelRidge
from sklearn.preprocessing import LabelEncoder
from sklearn.metrics import roc_auc_score, mean_squared_error
epochs = make_epochs()
y_4classes = np.hstack((epochs.events[:7, 2], epochs.events[7:, 2] + 1))
if check_version('sklearn', '0.18'):
from sklearn.model_selection import (KFold, StratifiedKFold,
ShuffleSplit, LeaveOneLabelOut)
cv_shuffle = ShuffleSplit()
cv = LeaveOneLabelOut()
# XXX we cannot pass any other parameters than X and y to cv.split
# so we have to build it before hand
cv_lolo = [(train, test) for train, test in cv.split(
X=y_4classes, y=y_4classes, labels=y_4classes)]
# With sklearn >= 0.17, `clf` can be identified as a regressor, and
# the scoring metrics can therefore be automatically assigned.
scorer_regress = None
else:
from sklearn.cross_validation import (KFold, StratifiedKFold,
ShuffleSplit, LeaveOneLabelOut)
cv_shuffle = ShuffleSplit(len(epochs))
cv_lolo = LeaveOneLabelOut(y_4classes)
# With sklearn < 0.17, `clf` cannot be identified as a regressor, and
# therefore the scoring metrics cannot be automatically assigned.
scorer_regress = mean_squared_error
# Test default running
gat = GeneralizationAcrossTime(picks='foo')
assert_equal("<GAT | no fit, no prediction, no score>", "%s" % gat)
assert_raises(ValueError, gat.fit, epochs)
with warnings.catch_warnings(record=True):
# check classic fit + check manual picks
gat.picks = [0]
gat.fit(epochs)
# check optional y as array
gat.picks = None
gat.fit(epochs, y=epochs.events[:, 2])
# check optional y as list
gat.fit(epochs, y=epochs.events[:, 2].tolist())
assert_equal(len(gat.picks_), len(gat.ch_names), 1)
assert_equal("<GAT | fitted, start : -0.200 (s), stop : 0.499 (s), no "
"prediction, no score>", '%s' % gat)
assert_equal(gat.ch_names, epochs.ch_names)
# test different predict function:
gat = GeneralizationAcrossTime(predict_method='decision_function')
gat.fit(epochs)
# With classifier, the default cv is StratifiedKFold
assert_true(gat.cv_.__class__ == StratifiedKFold)
gat.predict(epochs)
assert_array_equal(np.shape(gat.y_pred_), (15, 15, 14, 1))
gat.predict_method = 'predict_proba'
gat.predict(epochs)
assert_array_equal(np.shape(gat.y_pred_), (15, 15, 14, 2))
gat.predict_method = 'foo'
assert_raises(NotImplementedError, gat.predict, epochs)
gat.predict_method = 'predict'
gat.predict(epochs)
assert_array_equal(np.shape(gat.y_pred_), (15, 15, 14, 1))
assert_equal("<GAT | fitted, start : -0.200 (s), stop : 0.499 (s), "
"predicted 14 epochs, no score>",
"%s" % gat)
gat.score(epochs)
assert_true(gat.scorer_.__name__ == 'accuracy_score')
# check clf / predict_method combinations for which the scoring metrics
# cannot be inferred.
gat.scorer = None
gat.predict_method = 'decision_function'
assert_raises(ValueError, gat.score, epochs)
# Check specifying y manually
gat.predict_method = 'predict'
gat.score(epochs, y=epochs.events[:, 2])
gat.score(epochs, y=epochs.events[:, 2].tolist())
assert_equal("<GAT | fitted, start : -0.200 (s), stop : 0.499 (s), "
"predicted 14 epochs,\n scored "
"(accuracy_score)>", "%s" % gat)
with warnings.catch_warnings(record=True):
gat.fit(epochs, y=epochs.events[:, 2])
old_mode = gat.predict_mode
gat.predict_mode = 'super-foo-mode'
assert_raises(ValueError, gat.predict, epochs)
gat.predict_mode = old_mode
gat.score(epochs, y=epochs.events[:, 2])
assert_true("accuracy_score" in '%s' % gat.scorer_)
epochs2 = epochs.copy()
# check _DecodingTime class
assert_equal("<DecodingTime | start: -0.200 (s), stop: 0.499 (s), step: "
"0.050 (s), length: 0.050 (s), n_time_windows: 15>",
"%s" % gat.train_times_)
assert_equal("<DecodingTime | start: -0.200 (s), stop: 0.499 (s), step: "
"0.050 (s), length: 0.050 (s), n_time_windows: 15 x 15>",
"%s" % gat.test_times_)
# the y-check
gat.predict_mode = 'mean-prediction'
epochs2.events[:, 2] += 10
gat_ = copy.deepcopy(gat)
with use_log_level('error'):
assert_raises(ValueError, gat_.score, epochs2)
gat.predict_mode = 'cross-validation'
# Test basics
# --- number of trials
assert_true(gat.y_train_.shape[0] ==
gat.y_true_.shape[0] ==
len(gat.y_pred_[0][0]) == 14)
# --- number of folds
assert_true(np.shape(gat.estimators_)[1] == gat.cv)
# --- length training size
assert_true(len(gat.train_times_['slices']) == 15 ==
np.shape(gat.estimators_)[0])
# --- length testing sizes
assert_true(len(gat.test_times_['slices']) == 15 ==
np.shape(gat.scores_)[0])
assert_true(len(gat.test_times_['slices'][0]) == 15 ==
np.shape(gat.scores_)[1])
# Test score_mode
gat.score_mode = 'foo'
assert_raises(ValueError, gat.score, epochs)
gat.score_mode = 'fold-wise'
scores = gat.score(epochs)
assert_array_equal(np.shape(scores), [15, 15, 5])
gat.score_mode = 'mean-sample-wise'
scores = gat.score(epochs)
assert_array_equal(np.shape(scores), [15, 15])
gat.score_mode = 'mean-fold-wise'
scores = gat.score(epochs)
assert_array_equal(np.shape(scores), [15, 15])
gat.predict_mode = 'mean-prediction'
with warnings.catch_warnings(record=True) as w:
gat.score(epochs)
assert_true(any("score_mode changed from " in str(ww.message)
for ww in w))
# Test longer time window
gat = GeneralizationAcrossTime(train_times={'length': .100})
with warnings.catch_warnings(record=True):
gat2 = gat.fit(epochs)
assert_true(gat is gat2) # return self
assert_true(hasattr(gat2, 'cv_'))
assert_true(gat2.cv_ != gat.cv)
with warnings.catch_warnings(record=True): # not vectorizing
scores = gat.score(epochs)
assert_true(isinstance(scores, np.ndarray)) # type check
assert_equal(len(scores[0]), len(scores)) # shape check
assert_equal(len(gat.test_times_['slices'][0][0]), 2)
# Decim training steps
gat = GeneralizationAcrossTime(train_times={'step': .100})
with warnings.catch_warnings(record=True):
gat.fit(epochs)
gat.score(epochs)
assert_true(len(gat.scores_) == len(gat.estimators_) == 8) # training time
assert_equal(len(gat.scores_[0]), 15) # testing time
# Test start stop training & test cv without n_fold params
y_4classes = np.hstack((epochs.events[:7, 2], epochs.events[7:, 2] + 1))
train_times = dict(start=0.090, stop=0.250)
gat = GeneralizationAcrossTime(cv=cv_lolo, train_times=train_times)
# predict without fit
assert_raises(RuntimeError, gat.predict, epochs)
with warnings.catch_warnings(record=True):
gat.fit(epochs, y=y_4classes)
gat.score(epochs)
assert_equal(len(gat.scores_), 4)
assert_equal(gat.train_times_['times'][0], epochs.times[6])
assert_equal(gat.train_times_['times'][-1], epochs.times[9])
# Test score without passing epochs & Test diagonal decoding
gat = GeneralizationAcrossTime(test_times='diagonal')
with warnings.catch_warnings(record=True): # not vectorizing
gat.fit(epochs)
assert_raises(RuntimeError, gat.score)
with warnings.catch_warnings(record=True): # not vectorizing
gat.predict(epochs)
scores = gat.score()
assert_true(scores is gat.scores_)
assert_equal(np.shape(gat.scores_), (15, 1))
assert_array_equal([tim for ttime in gat.test_times_['times']
for tim in ttime], gat.train_times_['times'])
# Test generalization across conditions
gat = GeneralizationAcrossTime(predict_mode='mean-prediction', cv=2)
with warnings.catch_warnings(record=True):
gat.fit(epochs[0:6])
with warnings.catch_warnings(record=True):
# There are some empty test folds because of n_trials
gat.predict(epochs[7:])
gat.score(epochs[7:])
# Test training time parameters
gat_ = copy.deepcopy(gat)
# --- start stop outside time range
gat_.train_times = dict(start=-999.)
with use_log_level('error'):
assert_raises(ValueError, gat_.fit, epochs)
gat_.train_times = dict(start=999.)
assert_raises(ValueError, gat_.fit, epochs)
# --- impossible slices
gat_.train_times = dict(step=.000001)
assert_raises(ValueError, gat_.fit, epochs)
gat_.train_times = dict(length=.000001)
assert_raises(ValueError, gat_.fit, epochs)
gat_.train_times = dict(length=999.)
assert_raises(ValueError, gat_.fit, epochs)
# Test testing time parameters
# --- outside time range
gat.test_times = dict(start=-999.)
with warnings.catch_warnings(record=True): # no epochs in fold
assert_raises(ValueError, gat.predict, epochs)
gat.test_times = dict(start=999.)
with warnings.catch_warnings(record=True): # no test epochs
assert_raises(ValueError, gat.predict, epochs)
# --- impossible slices
gat.test_times = dict(step=.000001)
with warnings.catch_warnings(record=True): # no test epochs
assert_raises(ValueError, gat.predict, epochs)
gat_ = copy.deepcopy(gat)
gat_.train_times_['length'] = .000001
gat_.test_times = dict(length=.000001)
with warnings.catch_warnings(record=True): # no test epochs
assert_raises(ValueError, gat_.predict, epochs)
# --- test time region of interest
gat.test_times = dict(step=.150)
with warnings.catch_warnings(record=True): # not vectorizing
gat.predict(epochs)
assert_array_equal(np.shape(gat.y_pred_), (15, 5, 14, 1))
# --- silly value
gat.test_times = 'foo'
with warnings.catch_warnings(record=True): # no test epochs
assert_raises(ValueError, gat.predict, epochs)
assert_raises(RuntimeError, gat.score)
# --- unmatched length between training and testing time
gat.test_times = dict(length=.150)
assert_raises(ValueError, gat.predict, epochs)
# --- irregular length training and testing times
# 2 estimators, the first one is trained on two successive time samples
# whereas the second one is trained on a single time sample.
train_times = dict(slices=[[0, 1], [1]])
# The first estimator is tested once, the second estimator is tested on
# two successive time samples.
test_times = dict(slices=[[[0, 1]], [[0], [1]]])
gat = GeneralizationAcrossTime(train_times=train_times,
test_times=test_times)
gat.fit(epochs)
with warnings.catch_warnings(record=True): # not vectorizing
gat.score(epochs)
assert_array_equal(np.shape(gat.y_pred_[0]), [1, len(epochs), 1])
assert_array_equal(np.shape(gat.y_pred_[1]), [2, len(epochs), 1])
# check cannot Automatically infer testing times for adhoc training times
gat.test_times = None
assert_raises(ValueError, gat.predict, epochs)
svc = SVC(C=1, kernel='linear', probability=True)
gat = GeneralizationAcrossTime(clf=svc, predict_mode='mean-prediction')
with warnings.catch_warnings(record=True):
gat.fit(epochs)
# sklearn needs it: c.f.
# https://github.com/scikit-learn/scikit-learn/issues/2723
# and http://bit.ly/1u7t8UT
with use_log_level('error'):
assert_raises(ValueError, gat.score, epochs2)
gat.score(epochs)
assert_true(0.0 <= np.min(scores) <= 1.0)
assert_true(0.0 <= np.max(scores) <= 1.0)
# Test that gets error if train on one dataset, test on another, and don't
# specify appropriate cv:
gat = GeneralizationAcrossTime(cv=cv_shuffle)
gat.fit(epochs)
with warnings.catch_warnings(record=True):
gat.fit(epochs)
gat.predict(epochs)
assert_raises(ValueError, gat.predict, epochs[:10])
# Make CV with some empty train and test folds:
# --- empty test fold(s) should warn when gat.predict()
gat._cv_splits[0] = [gat._cv_splits[0][0], np.empty(0)]
with warnings.catch_warnings(record=True) as w:
gat.predict(epochs)
assert_true(len(w) > 0)
assert_true(any('do not have any test epochs' in str(ww.message)
for ww in w))
# --- empty train fold(s) should raise when gat.fit()
gat = GeneralizationAcrossTime(cv=[([0], [1]), ([], [0])])
assert_raises(ValueError, gat.fit, epochs[:2])
# Check that still works with classifier that output y_pred with
# shape = (n_trials, 1) instead of (n_trials,)
if check_version('sklearn', '0.17'): # no is_regressor before v0.17
gat = GeneralizationAcrossTime(clf=KernelRidge(), cv=2)
epochs.crop(None, epochs.times[2])
gat.fit(epochs)
# With regression the default cv is KFold and not StratifiedKFold
assert_true(gat.cv_.__class__ == KFold)
gat.score(epochs)
# with regression the default scoring metrics is mean squared error
assert_true(gat.scorer_.__name__ == 'mean_squared_error')
# Test combinations of complex scenarios
# 2 or more distinct classes
n_classes = [2, 4] # 4 tested
# nicely ordered labels or not
le = LabelEncoder()
y = le.fit_transform(epochs.events[:, 2])
y[len(y) // 2:] += 2
ys = (y, y + 1000)
# Univariate and multivariate prediction
svc = SVC(C=1, kernel='linear', probability=True)
reg = KernelRidge()
def scorer_proba(y_true, y_pred):
return roc_auc_score(y_true, y_pred[:, 0])
# We re testing 3 scenario: default, classifier + predict_proba, regressor
scorers = [None, scorer_proba, scorer_regress]
predict_methods = [None, 'predict_proba', None]
clfs = [svc, svc, reg]
# Test all combinations
for clf, predict_method, scorer in zip(clfs, predict_methods, scorers):
for y in ys:
for n_class in n_classes:
for predict_mode in ['cross-validation', 'mean-prediction']:
# Cannot use AUC for n_class > 2
if (predict_method == 'predict_proba' and n_class != 2):
continue
y_ = y % n_class
with warnings.catch_warnings(record=True):
gat = GeneralizationAcrossTime(
cv=2, clf=clf, scorer=scorer,
predict_mode=predict_mode)
gat.fit(epochs, y=y_)
gat.score(epochs, y=y_)
# Check that scorer is correctly defined manually and
# automatically.
scorer_name = gat.scorer_.__name__
if scorer is None:
if is_classifier(clf):
assert_equal(scorer_name, 'accuracy_score')
else:
assert_equal(scorer_name, 'mean_squared_error')
else:
assert_equal(scorer_name, scorer.__name__)
def test_brain_time_viewer(renderer_interactive, pixel_ratio, brain_gc):
"""Test time viewer primitives."""
if renderer_interactive._get_3d_backend() != 'pyvista':
pytest.skip('TimeViewer tests only supported on PyVista')
with pytest.raises(ValueError, match="between 0 and 1"):
_create_testing_brain(hemi='lh', show_traces=-1.0)
with pytest.raises(ValueError, match="got unknown keys"):
_create_testing_brain(hemi='lh',
surf='white',
src='volume',
volume_options={'foo': 'bar'})
brain = _create_testing_brain(
hemi='both',
show_traces=False,
brain_kwargs=dict(silhouette=dict(decimate=0.95)))
# test sub routines when show_traces=False
brain._on_pick(None, None)
brain._configure_vertex_time_course()
brain._configure_label_time_course()
brain.setup_time_viewer() # for coverage
brain.callbacks["time"](value=0)
assert "renderer" not in brain.callbacks
brain.callbacks["orientation"](value='lat', update_widget=True)
brain.callbacks["orientation"](value='medial', update_widget=True)
brain.callbacks["time"](
value=0.0,
time_as_index=False,
)
# Need to process events for old Qt
brain.callbacks["smoothing"](value=1)
_assert_brain_range(brain, [0.1, 0.3])
from mne.utils import use_log_level
print('\nCallback fmin\n')
with use_log_level('debug'):
brain.callbacks["fmin"](value=12.0)
assert brain._data["fmin"] == 12.0
brain.callbacks["fmax"](value=4.0)
_assert_brain_range(brain, [4.0, 4.0])
brain.callbacks["fmid"](value=6.0)
_assert_brain_range(brain, [4.0, 6.0])
brain.callbacks["fmid"](value=4.0)
brain.callbacks["fplus"]()
brain.callbacks["fminus"]()
brain.callbacks["fmin"](value=12.0)
brain.callbacks["fmid"](value=4.0)
_assert_brain_range(brain, [4.0, 12.0])
brain._shift_time(op=lambda x, y: x + y)
brain._shift_time(op=lambda x, y: x - y)
brain._rotate_azimuth(15)
brain._rotate_elevation(15)
brain.toggle_interface()
brain.toggle_interface(value=False)
brain.callbacks["playback_speed"](value=0.1)
brain.toggle_playback()
brain.toggle_playback(value=False)
brain.apply_auto_scaling()
brain.restore_user_scaling()
brain.reset()
plt.close('all')
brain.help()
assert len(plt.get_fignums()) == 1
plt.close('all')
assert len(plt.get_fignums()) == 0
# screenshot
# Need to turn the interface back on otherwise the window is too wide
# (it keeps the window size and expands the 3D area when the interface
# is toggled off)
brain.toggle_interface(value=True)
brain.show_view(view=dict(azimuth=180., elevation=90.))
img = brain.screenshot(mode='rgb')
want_shape = np.array([300 * pixel_ratio, 300 * pixel_ratio, 3])
assert_allclose(img.shape, want_shape)
brain.close()
def test_generalization_across_time():
"""Test time generalization decoding
"""
from sklearn.svm import SVC
from sklearn.base import is_classifier
# KernelRidge is used for testing 1) regression analyses 2) n-dimensional
# predictions.
from sklearn.kernel_ridge import KernelRidge
from sklearn.preprocessing import LabelEncoder
from sklearn.metrics import roc_auc_score, mean_squared_error
epochs = make_epochs()
y_4classes = np.hstack((epochs.events[:7, 2], epochs.events[7:, 2] + 1))
if check_version('sklearn', '0.18'):
from sklearn.model_selection import (KFold, StratifiedKFold,
ShuffleSplit, LeaveOneLabelOut)
cv_shuffle = ShuffleSplit()
cv = LeaveOneLabelOut()
# XXX we cannot pass any other parameters than X and y to cv.split
# so we have to build it before hand
cv_lolo = [(train, test) for train, test in cv.split(
X=y_4classes, y=y_4classes, labels=y_4classes)]
# With sklearn >= 0.17, `clf` can be identified as a regressor, and
# the scoring metrics can therefore be automatically assigned.
scorer_regress = None
else:
from sklearn.cross_validation import (KFold, StratifiedKFold,
ShuffleSplit, LeaveOneLabelOut)
cv_shuffle = ShuffleSplit(len(epochs))
cv_lolo = LeaveOneLabelOut(y_4classes)
# With sklearn < 0.17, `clf` cannot be identified as a regressor, and
# therefore the scoring metrics cannot be automatically assigned.
scorer_regress = mean_squared_error
# Test default running
gat = GeneralizationAcrossTime(picks='foo')
assert_equal("<GAT | no fit, no prediction, no score>", "%s" % gat)
assert_raises(ValueError, gat.fit, epochs)
with warnings.catch_warnings(record=True):
# check classic fit + check manual picks
gat.picks = [0]
gat.fit(epochs)
# check optional y as array
gat.picks = None
gat.fit(epochs, y=epochs.events[:, 2])
# check optional y as list
gat.fit(epochs, y=epochs.events[:, 2].tolist())
assert_equal(len(gat.picks_), len(gat.ch_names), 1)
assert_equal(
"<GAT | fitted, start : -0.200 (s), stop : 0.499 (s), no "
"prediction, no score>", '%s' % gat)
assert_equal(gat.ch_names, epochs.ch_names)
# test different predict function:
gat = GeneralizationAcrossTime(predict_method='decision_function')
gat.fit(epochs)
# With classifier, the default cv is StratifiedKFold
assert_true(gat.cv_.__class__ == StratifiedKFold)
gat.predict(epochs)
assert_array_equal(np.shape(gat.y_pred_), (15, 15, 14, 1))
gat.predict_method = 'predict_proba'
gat.predict(epochs)
assert_array_equal(np.shape(gat.y_pred_), (15, 15, 14, 2))
gat.predict_method = 'foo'
assert_raises(NotImplementedError, gat.predict, epochs)
gat.predict_method = 'predict'
gat.predict(epochs)
assert_array_equal(np.shape(gat.y_pred_), (15, 15, 14, 1))
assert_equal(
"<GAT | fitted, start : -0.200 (s), stop : 0.499 (s), "
"predicted 14 epochs, no score>", "%s" % gat)
gat.score(epochs)
assert_true(gat.scorer_.__name__ == 'accuracy_score')
# check clf / predict_method combinations for which the scoring metrics
# cannot be inferred.
gat.scorer = None
gat.predict_method = 'decision_function'
assert_raises(ValueError, gat.score, epochs)
# Check specifying y manually
gat.predict_method = 'predict'
gat.score(epochs, y=epochs.events[:, 2])
gat.score(epochs, y=epochs.events[:, 2].tolist())
assert_equal(
"<GAT | fitted, start : -0.200 (s), stop : 0.499 (s), "
"predicted 14 epochs,\n scored "
"(accuracy_score)>", "%s" % gat)
with warnings.catch_warnings(record=True):
gat.fit(epochs, y=epochs.events[:, 2])
old_mode = gat.predict_mode
gat.predict_mode = 'super-foo-mode'
assert_raises(ValueError, gat.predict, epochs)
gat.predict_mode = old_mode
gat.score(epochs, y=epochs.events[:, 2])
assert_true("accuracy_score" in '%s' % gat.scorer_)
epochs2 = epochs.copy()
# check _DecodingTime class
assert_equal(
"<DecodingTime | start: -0.200 (s), stop: 0.499 (s), step: "
"0.050 (s), length: 0.050 (s), n_time_windows: 15>",
"%s" % gat.train_times_)
assert_equal(
"<DecodingTime | start: -0.200 (s), stop: 0.499 (s), step: "
"0.050 (s), length: 0.050 (s), n_time_windows: 15 x 15>",
"%s" % gat.test_times_)
# the y-check
gat.predict_mode = 'mean-prediction'
epochs2.events[:, 2] += 10
gat_ = copy.deepcopy(gat)
with use_log_level('error'):
assert_raises(ValueError, gat_.score, epochs2)
gat.predict_mode = 'cross-validation'
# Test basics
# --- number of trials
assert_true(gat.y_train_.shape[0] == gat.y_true_.shape[0] == len(
gat.y_pred_[0][0]) == 14)
# --- number of folds
assert_true(np.shape(gat.estimators_)[1] == gat.cv)
# --- length training size
assert_true(
len(gat.train_times_['slices']) == 15 == np.shape(gat.estimators_)[0])
# --- length testing sizes
assert_true(
len(gat.test_times_['slices']) == 15 == np.shape(gat.scores_)[0])
assert_true(
len(gat.test_times_['slices'][0]) == 15 == np.shape(gat.scores_)[1])
# Test score_mode
gat.score_mode = 'foo'
assert_raises(ValueError, gat.score, epochs)
gat.score_mode = 'fold-wise'
scores = gat.score(epochs)
assert_array_equal(np.shape(scores), [15, 15, 5])
gat.score_mode = 'mean-sample-wise'
scores = gat.score(epochs)
assert_array_equal(np.shape(scores), [15, 15])
gat.score_mode = 'mean-fold-wise'
scores = gat.score(epochs)
assert_array_equal(np.shape(scores), [15, 15])
gat.predict_mode = 'mean-prediction'
with warnings.catch_warnings(record=True) as w:
gat.score(epochs)
assert_true(
any("score_mode changed from " in str(ww.message) for ww in w))
# Test longer time window
gat = GeneralizationAcrossTime(train_times={'length': .100})
with warnings.catch_warnings(record=True):
gat2 = gat.fit(epochs)
assert_true(gat is gat2) # return self
assert_true(hasattr(gat2, 'cv_'))
assert_true(gat2.cv_ != gat.cv)
with warnings.catch_warnings(record=True): # not vectorizing
scores = gat.score(epochs)
assert_true(isinstance(scores, np.ndarray)) # type check
assert_equal(len(scores[0]), len(scores)) # shape check
assert_equal(len(gat.test_times_['slices'][0][0]), 2)
# Decim training steps
gat = GeneralizationAcrossTime(train_times={'step': .100})
with warnings.catch_warnings(record=True):
gat.fit(epochs)
gat.score(epochs)
assert_true(len(gat.scores_) == len(gat.estimators_) == 8) # training time
assert_equal(len(gat.scores_[0]), 15) # testing time
# Test start stop training & test cv without n_fold params
y_4classes = np.hstack((epochs.events[:7, 2], epochs.events[7:, 2] + 1))
train_times = dict(start=0.090, stop=0.250)
gat = GeneralizationAcrossTime(cv=cv_lolo, train_times=train_times)
# predict without fit
assert_raises(RuntimeError, gat.predict, epochs)
with warnings.catch_warnings(record=True):
gat.fit(epochs, y=y_4classes)
gat.score(epochs)
assert_equal(len(gat.scores_), 4)
assert_equal(gat.train_times_['times'][0], epochs.times[6])
assert_equal(gat.train_times_['times'][-1], epochs.times[9])
# Test score without passing epochs & Test diagonal decoding
gat = GeneralizationAcrossTime(test_times='diagonal')
with warnings.catch_warnings(record=True): # not vectorizing
gat.fit(epochs)
assert_raises(RuntimeError, gat.score)
with warnings.catch_warnings(record=True): # not vectorizing
gat.predict(epochs)
scores = gat.score()
assert_true(scores is gat.scores_)
assert_equal(np.shape(gat.scores_), (15, 1))
assert_array_equal(
[tim for ttime in gat.test_times_['times'] for tim in ttime],
gat.train_times_['times'])
# Test generalization across conditions
gat = GeneralizationAcrossTime(predict_mode='mean-prediction', cv=2)
with warnings.catch_warnings(record=True):
gat.fit(epochs[0:6])
with warnings.catch_warnings(record=True):
# There are some empty test folds because of n_trials
gat.predict(epochs[7:])
gat.score(epochs[7:])
# Test training time parameters
gat_ = copy.deepcopy(gat)
# --- start stop outside time range
gat_.train_times = dict(start=-999.)
with use_log_level('error'):
assert_raises(ValueError, gat_.fit, epochs)
gat_.train_times = dict(start=999.)
assert_raises(ValueError, gat_.fit, epochs)
# --- impossible slices
gat_.train_times = dict(step=.000001)
assert_raises(ValueError, gat_.fit, epochs)
gat_.train_times = dict(length=.000001)
assert_raises(ValueError, gat_.fit, epochs)
gat_.train_times = dict(length=999.)
assert_raises(ValueError, gat_.fit, epochs)
# Test testing time parameters
# --- outside time range
gat.test_times = dict(start=-999.)
with warnings.catch_warnings(record=True): # no epochs in fold
assert_raises(ValueError, gat.predict, epochs)
gat.test_times = dict(start=999.)
with warnings.catch_warnings(record=True): # no test epochs
assert_raises(ValueError, gat.predict, epochs)
# --- impossible slices
gat.test_times = dict(step=.000001)
with warnings.catch_warnings(record=True): # no test epochs
assert_raises(ValueError, gat.predict, epochs)
gat_ = copy.deepcopy(gat)
gat_.train_times_['length'] = .000001
gat_.test_times = dict(length=.000001)
with warnings.catch_warnings(record=True): # no test epochs
assert_raises(ValueError, gat_.predict, epochs)
# --- test time region of interest
gat.test_times = dict(step=.150)
with warnings.catch_warnings(record=True): # not vectorizing
gat.predict(epochs)
assert_array_equal(np.shape(gat.y_pred_), (15, 5, 14, 1))
# --- silly value
gat.test_times = 'foo'
with warnings.catch_warnings(record=True): # no test epochs
assert_raises(ValueError, gat.predict, epochs)
assert_raises(RuntimeError, gat.score)
# --- unmatched length between training and testing time
gat.test_times = dict(length=.150)
assert_raises(ValueError, gat.predict, epochs)
# --- irregular length training and testing times
# 2 estimators, the first one is trained on two successive time samples
# whereas the second one is trained on a single time sample.
train_times = dict(slices=[[0, 1], [1]])
# The first estimator is tested once, the second estimator is tested on
# two successive time samples.
test_times = dict(slices=[[[0, 1]], [[0], [1]]])
gat = GeneralizationAcrossTime(train_times=train_times,
test_times=test_times)
gat.fit(epochs)
with warnings.catch_warnings(record=True): # not vectorizing
gat.score(epochs)
assert_array_equal(np.shape(gat.y_pred_[0]), [1, len(epochs), 1])
assert_array_equal(np.shape(gat.y_pred_[1]), [2, len(epochs), 1])
# check cannot Automatically infer testing times for adhoc training times
gat.test_times = None
assert_raises(ValueError, gat.predict, epochs)
svc = SVC(C=1, kernel='linear', probability=True)
gat = GeneralizationAcrossTime(clf=svc, predict_mode='mean-prediction')
with warnings.catch_warnings(record=True):
gat.fit(epochs)
# sklearn needs it: c.f.
# https://github.com/scikit-learn/scikit-learn/issues/2723
# and http://bit.ly/1u7t8UT
with use_log_level('error'):
assert_raises(ValueError, gat.score, epochs2)
gat.score(epochs)
assert_true(0.0 <= np.min(scores) <= 1.0)
assert_true(0.0 <= np.max(scores) <= 1.0)
# Test that gets error if train on one dataset, test on another, and don't
# specify appropriate cv:
gat = GeneralizationAcrossTime(cv=cv_shuffle)
gat.fit(epochs)
with warnings.catch_warnings(record=True):
gat.fit(epochs)
gat.predict(epochs)
assert_raises(ValueError, gat.predict, epochs[:10])
# Make CV with some empty train and test folds:
# --- empty test fold(s) should warn when gat.predict()
gat._cv_splits[0] = [gat._cv_splits[0][0], np.empty(0)]
with warnings.catch_warnings(record=True) as w:
gat.predict(epochs)
assert_true(len(w) > 0)
assert_true(
any('do not have any test epochs' in str(ww.message) for ww in w))
# --- empty train fold(s) should raise when gat.fit()
gat = GeneralizationAcrossTime(cv=[([0], [1]), ([], [0])])
assert_raises(ValueError, gat.fit, epochs[:2])
# Check that still works with classifier that output y_pred with
# shape = (n_trials, 1) instead of (n_trials,)
if check_version('sklearn', '0.17'): # no is_regressor before v0.17
gat = GeneralizationAcrossTime(clf=KernelRidge(), cv=2)
epochs.crop(None, epochs.times[2])
gat.fit(epochs)
# With regression the default cv is KFold and not StratifiedKFold
assert_true(gat.cv_.__class__ == KFold)
gat.score(epochs)
# with regression the default scoring metrics is mean squared error
assert_true(gat.scorer_.__name__ == 'mean_squared_error')
# Test combinations of complex scenarios
# 2 or more distinct classes
n_classes = [2, 4] # 4 tested
# nicely ordered labels or not
le = LabelEncoder()
y = le.fit_transform(epochs.events[:, 2])
y[len(y) // 2:] += 2
ys = (y, y + 1000)
# Univariate and multivariate prediction
svc = SVC(C=1, kernel='linear', probability=True)
reg = KernelRidge()
def scorer_proba(y_true, y_pred):
return roc_auc_score(y_true, y_pred[:, 0])
# We re testing 3 scenario: default, classifier + predict_proba, regressor
scorers = [None, scorer_proba, scorer_regress]
predict_methods = [None, 'predict_proba', None]
clfs = [svc, svc, reg]
# Test all combinations
for clf, predict_method, scorer in zip(clfs, predict_methods, scorers):
for y in ys:
for n_class in n_classes:
for predict_mode in ['cross-validation', 'mean-prediction']:
# Cannot use AUC for n_class > 2
if (predict_method == 'predict_proba' and n_class != 2):
continue
y_ = y % n_class
with warnings.catch_warnings(record=True):
gat = GeneralizationAcrossTime(
cv=2,
clf=clf,
scorer=scorer,
predict_mode=predict_mode)
gat.fit(epochs, y=y_)
gat.score(epochs, y=y_)
# Check that scorer is correctly defined manually and
# automatically.
scorer_name = gat.scorer_.__name__
if scorer is None:
if is_classifier(clf):
assert_equal(scorer_name, 'accuracy_score')
else:
assert_equal(scorer_name, 'mean_squared_error')
else:
assert_equal(scorer_name, scorer.__name__)
def test_logging_options(tmpdir):
"""Test logging (to file)."""
with use_log_level(None): # just ensure it's set back
with pytest.raises(ValueError, match="Invalid value for the 'verbose"):
set_log_level('foo')
tempdir = str(tmpdir)
test_name = op.join(tempdir, 'test.log')
with open(fname_log, 'r') as old_log_file:
# [:-1] used to strip an extra "No baseline correction applied"
old_lines = clean_lines(old_log_file.readlines())
old_lines.pop(-1)
with open(fname_log_2, 'r') as old_log_file_2:
old_lines_2 = clean_lines(old_log_file_2.readlines())
old_lines_2.pop(14)
old_lines_2.pop(-1)
if op.isfile(test_name):
os.remove(test_name)
# test it one way (printing default off)
set_log_file(test_name)
set_log_level('WARNING')
# should NOT print
evoked = read_evokeds(fname_evoked, condition=1)
with open(test_name) as fid:
assert (fid.readlines() == [])
# should NOT print
evoked = read_evokeds(fname_evoked, condition=1, verbose=False)
with open(test_name) as fid:
assert (fid.readlines() == [])
# should NOT print
evoked = read_evokeds(fname_evoked, condition=1, verbose='WARNING')
with open(test_name) as fid:
assert (fid.readlines() == [])
# SHOULD print
evoked = read_evokeds(fname_evoked, condition=1, verbose=True)
with open(test_name, 'r') as new_log_file:
new_lines = clean_lines(new_log_file.readlines())
assert new_lines == old_lines
set_log_file(None) # Need to do this to close the old file
os.remove(test_name)
# now go the other way (printing default on)
set_log_file(test_name)
set_log_level('INFO')
# should NOT print
evoked = read_evokeds(fname_evoked, condition=1, verbose='WARNING')
with open(test_name) as fid:
assert (fid.readlines() == [])
# should NOT print
evoked = read_evokeds(fname_evoked, condition=1, verbose=False)
with open(test_name) as fid:
assert (fid.readlines() == [])
# SHOULD print
evoked = read_evokeds(fname_evoked, condition=1)
with open(test_name, 'r') as new_log_file:
new_lines = clean_lines(new_log_file.readlines())
assert new_lines == old_lines
# check to make sure appending works (and as default, raises a warning)
set_log_file(test_name, overwrite=False)
with pytest.warns(RuntimeWarning, match='appended to the file'):
set_log_file(test_name)
evoked = read_evokeds(fname_evoked, condition=1)
with open(test_name, 'r') as new_log_file:
new_lines = clean_lines(new_log_file.readlines())
assert new_lines == old_lines_2
# make sure overwriting works
set_log_file(test_name, overwrite=True)
# this line needs to be called to actually do some logging
evoked = read_evokeds(fname_evoked, condition=1)
del evoked
with open(test_name, 'r') as new_log_file:
new_lines = clean_lines(new_log_file.readlines())
assert new_lines == old_lines
def test_ieeg_elec_locate_gui_display(_locate_ieeg, _fake_CT_coords):
"""Test that the intracranial location GUI displays properly."""
raw = mne.io.read_raw_fif(raw_path, preload=True)
raw.pick_types(eeg=True)
ch_dict = {
'EEG 001': 'LAMY 1',
'EEG 002': 'LAMY 2',
'EEG 003': 'LSTN 1',
'EEG 004': 'LSTN 2'
}
raw.pick_channels(list(ch_dict.keys()))
raw.rename_channels(ch_dict)
raw.set_eeg_reference('average')
raw.set_channel_types({name: 'seeg' for name in raw.ch_names})
raw.set_montage(None)
aligned_ct, coords = _fake_CT_coords
trans = mne.read_trans(fname_trans)
# test no seghead, fsaverage doesn't have seghead
with pytest.warns(RuntimeWarning, match='`seghead` not found'):
with catch_logging() as log:
_locate_ieeg(raw.info,
trans,
aligned_ct,
subject='fsaverage',
subjects_dir=subjects_dir,
verbose=True)
log = log.getvalue()
assert 'Using marching cubes' in log
# test functions
with pytest.warns(RuntimeWarning, match='`pial` surface not found'):
gui = _locate_ieeg(raw.info,
trans,
aligned_ct,
subject=subject,
subjects_dir=subjects_dir,
verbose=True)
with pytest.raises(ValueError, match='read-only'):
gui._ras[:] = coords[0] # start in the right position
gui._set_ras(coords[0])
gui._mark_ch()
assert not gui._lines and not gui._lines_2D # no lines for one contact
for ci, coord in enumerate(coords[1:], 1):
coord_vox = apply_trans(gui._ras_vox_t, coord)
with use_log_level('debug'):
_fake_click(gui._figs[2],
gui._figs[2].axes[0],
coord_vox[:-1],
xform='data',
kind='release')
assert_allclose(coord[:2],
gui._ras[:2],
atol=0.1,
err_msg=f'coords[{ci}][:2]')
assert_allclose(coord[2],
gui._ras[2],
atol=2,
err_msg=f'coords[{ci}][2]')
gui._mark_ch()
# ensure a 3D line was made for each group
assert len(gui._lines) == 2
# test snap to center
gui._ch_index = 0
gui._set_ras(coords[0]) # move to first position
gui._mark_ch()
assert_allclose(coords[0], gui._chs['LAMY 1'], atol=0.2)
gui._snap_button.click()
assert gui._snap_button.text() == 'Off'
# now make sure no snap happens
gui._ch_index = 0
gui._set_ras(coords[1] + 1)
gui._mark_ch()
assert_allclose(coords[1] + 1, gui._chs['LAMY 1'], atol=0.01)
# check that it turns back on
gui._snap_button.click()
assert gui._snap_button.text() == 'On'
# test remove
gui._ch_index = 1
gui._update_ch_selection()
gui._remove_ch()
assert np.isnan(gui._chs['LAMY 2']).all()
# check that raw object saved
assert not np.isnan(raw.info['chs'][0]['loc'][:3]).any() # LAMY 1
assert np.isnan(raw.info['chs'][1]['loc'][:3]).all() # LAMY 2 (removed)
# move sliders
gui._alpha_slider.setValue(75)
assert gui._ch_alpha == 0.75
gui._radius_slider.setValue(5)
assert gui._radius == 5
ct_sum_before = np.nansum(gui._images['ct'][0].get_array().data)
gui._ct_min_slider.setValue(500)
assert np.nansum(gui._images['ct'][0].get_array().data) < ct_sum_before
# test buttons
gui._toggle_show_brain()
assert 'mri' in gui._images
assert 'local_max' not in gui._images
gui._toggle_show_max()
assert 'local_max' in gui._images
assert 'mip' not in gui._images
gui._toggle_show_mip()
assert 'mip' in gui._images
assert 'mip_chs' in gui._images
assert len(gui._lines_2D) == 1 # LAMY only has one contact
def save_epochs(p, subjects, in_names, in_numbers, analyses, out_names,
out_numbers, must_match, decim, run_indices):
"""Generate epochs from raw data based on events.
Can only complete after preprocessing is complete.
Parameters
----------
p : instance of Parameters
Analysis parameters.
subjects : list of str
Subject names to analyze (e.g., ['Eric_SoP_001', ...]).
in_names : list of str
Names of input events.
in_numbers : list of list of int
Event numbers (in scored event files) associated with each name.
analyses : list of str
Lists of analyses of interest.
out_names : list of list of str
Event types to make out of old ones.
out_numbers : list of list of int
Event numbers to convert to (e.g., [[1, 1, 2, 3, 3], ...] would create
three event types, where the first two and last two event types from
the original list get collapsed over).
must_match : list of int
Indices from the original in_names that must match in event counts
before collapsing. Should eventually be expanded to allow for
ratio-based collapsing.
decim : int | list of int
Amount to decimate.
run_indices : array-like | None
Run indices to include.
"""
in_names = np.asanyarray(in_names)
old_dict = dict()
for n, e in zip(in_names, in_numbers):
old_dict[n] = e
# let's do some sanity checks
if len(in_names) != len(in_numbers):
raise RuntimeError('in_names (%d) must have same length as '
'in_numbers (%d)'
% (len(in_names), len(in_numbers)))
if np.any(np.array(in_numbers) <= 0):
raise ValueError('in_numbers must all be > 0')
if len(out_names) != len(out_numbers):
raise RuntimeError('out_names must have same length as out_numbers')
for name, num in zip(out_names, out_numbers):
num = np.array(num)
if len(name) != len(np.unique(num[num > 0])):
raise RuntimeError('each entry in out_names must have length '
'equal to the number of unique elements in the '
'corresponding entry in out_numbers:\n%s\n%s'
% (name, np.unique(num[num > 0])))
if len(num) != len(in_names):
raise RuntimeError('each entry in out_numbers must have the same '
'length as in_names')
if (np.array(num) == 0).any():
raise ValueError('no element of out_numbers can be zero')
ch_namess = list()
drop_logs = list()
sfreqs = set()
for si, subj in enumerate(subjects):
if p.disp_files:
print(' Loading raw files for subject %s.' % subj)
epochs_dir = op.join(p.work_dir, subj, p.epochs_dir)
if not op.isdir(epochs_dir):
os.mkdir(epochs_dir)
evoked_dir = op.join(p.work_dir, subj, p.inverse_dir)
if not op.isdir(evoked_dir):
os.mkdir(evoked_dir)
# read in raw files
raw_names = get_raw_fnames(p, subj, 'pca', False, False,
run_indices[si])
raw, ratios = _concat_resamp_raws(p, subj, raw_names)
assert ratios.shape == (len(raw_names),)
# optionally calculate autoreject thresholds
this_decim = _handle_decim(decim[si], raw.info['sfreq'])
new_sfreq = raw.info['sfreq'] / this_decim
if p.disp_files:
print(' Epoching data (decim=%s -> sfreq=%0.1f Hz).'
% (this_decim, new_sfreq))
if new_sfreq not in sfreqs:
if len(sfreqs) > 0:
warnings.warn('resulting new sampling frequency %s not equal '
'to previous values %s' % (new_sfreq, sfreqs))
sfreqs.add(new_sfreq)
epochs_fnames, evoked_fnames = get_epochs_evokeds_fnames(p, subj,
analyses)
mat_file, fif_file = epochs_fnames
# handle regex-based reject_epochs_by_annot defs
reject_epochs_by_annot = _check_reject_annot_regex(p, raw)
if p.autoreject_thresholds:
assert len(p.autoreject_types) > 0
assert all(a in ('mag', 'grad', 'eeg', 'ecg', 'eog')
for a in p.autoreject_types)
from autoreject import get_rejection_threshold
picker = p.pick_events_autoreject
if type(picker) is str:
assert picker == 'restrict', \
'Only "restrict" is valid str for p.pick_events_autoreject'
events = _read_events(
p, subj, run_indices[si], raw, ratios, picker=picker)
print(' Computing autoreject thresholds', end='')
rtmin = p.reject_tmin if p.reject_tmin is not None else p.tmin
rtmax = p.reject_tmax if p.reject_tmax is not None else p.tmax
temp_epochs = Epochs(
raw, events, event_id=None, tmin=rtmin, tmax=rtmax,
baseline=_get_baseline(p), proj=True, reject=None,
flat=None, preload=True, decim=this_decim,
reject_by_annotation=reject_epochs_by_annot)
kwargs = dict()
if 'verbose' in get_args(get_rejection_threshold):
kwargs['verbose'] = False
new_dict = get_rejection_threshold(temp_epochs, **kwargs)
use_reject = dict()
msgs = list()
for k in p.autoreject_types:
msgs.append('%s=%d %s'
% (k, DEFAULTS['scalings'][k] * new_dict[k],
DEFAULTS['units'][k]))
use_reject[k] = new_dict[k]
print(': ' + ', '.join(msgs))
hdf5_file = fif_file.replace('-epo.fif', '-reject.h5')
assert hdf5_file.endswith('.h5')
write_hdf5(hdf5_file, use_reject, overwrite=True)
else:
use_reject = _handle_dict(p.reject, subj)
# read in events and create epochs
events = _read_events(p, subj, run_indices[si], raw, ratios,
picker='restrict')
if len(events) == 0:
raise ValueError('No valid events found')
flat = _handle_dict(p.flat, subj)
use_reject, use_flat = _restrict_reject_flat(use_reject, flat, raw)
epochs = Epochs(raw, events, event_id=old_dict, tmin=p.tmin,
tmax=p.tmax, baseline=_get_baseline(p),
reject=use_reject, flat=use_flat, proj=p.epochs_proj,
preload=True, decim=this_decim,
on_missing=p.on_missing,
reject_tmin=p.reject_tmin, reject_tmax=p.reject_tmax,
reject_by_annotation=reject_epochs_by_annot)
if epochs.events.shape[0] < 1:
_raise_bad_epochs(raw, epochs, events, plot=p.plot_drop_logs)
del raw
drop_logs.append(epochs.drop_log)
ch_namess.append(epochs.ch_names)
# only kept trials that were not dropped
sfreq = epochs.info['sfreq']
# now deal with conditions to save evoked
if p.disp_files:
print(' Matching trial counts and saving data to disk.')
for var, name in ((out_names, 'out_names'),
(out_numbers, 'out_numbers'),
(must_match, 'must_match'),
(evoked_fnames, 'evoked_fnames')):
if len(var) != len(analyses):
raise ValueError('len(%s) (%s) != len(analyses) (%s)'
% (name, len(var), len(analyses)))
for analysis, names, numbers, match, fn in zip(analyses, out_names,
out_numbers, must_match,
evoked_fnames):
# do matching
numbers = np.asanyarray(numbers)
nn = numbers[numbers >= 0]
new_numbers = []
for num in numbers:
if num > 0 and num not in new_numbers:
# Eventually we could relax this requirement, but not
# having it in place is likely to cause people pain...
if any(num < n for n in new_numbers):
raise RuntimeError('each list of new_numbers must be '
' monotonically increasing')
new_numbers.append(num)
new_numbers = np.array(new_numbers)
in_names_match = list(in_names[match])
# use some variables to allow safe name re-use
offset = max(epochs.events[:, 2].max(), new_numbers.max()) + 1
safety_str = '__mnefun_copy__'
assert len(new_numbers) == len(names) # checked above
if p.match_fun is None:
e = None
else: # use custom matching
args = [epochs.copy(), analysis, nn, in_names_match, names]
if len(get_args(p.match_fun)) > 5:
args = args + [numbers]
e = p.match_fun(*args)
if e is None:
# first, equalize trial counts (this will make a copy)
e = epochs[list(in_names[numbers > 0])]
# some could be missing
in_names_match = [
name for name in in_names_match if name in e.event_id]
if len(in_names_match) > 1:
print(f' Equalizing: {in_names_match}')
e.equalize_event_counts(in_names_match)
# second, collapse relevant types
for num, name in zip(new_numbers, names):
collapse = [x for x in in_names[num == numbers]
if x in e.event_id]
combine_event_ids(e, collapse,
{name + safety_str: num + offset},
copy=False)
for num, name in zip(new_numbers, names):
e.events[e.events[:, 2] == num + offset, 2] -= offset
e.event_id[name] = num
del e.event_id[name + safety_str]
# now make evoked for each out type
evokeds = list()
n_standard = 0
kinds = ['standard']
if p.every_other:
kinds += ['even', 'odd']
for kind in kinds:
for name in names:
this_e = e[name]
if kind == 'even':
this_e = this_e[::2]
elif kind == 'odd':
this_e = this_e[1::2]
else:
assert kind == 'standard'
with use_log_level('error'):
with warnings.catch_warnings(record=True):
warnings.simplefilter('ignore')
ave = this_e.average(picks='all')
ave.comment = name
stde = this_e.standard_error(picks='all')
stde.comment = name
if kind != 'standard':
ave.comment += ' %s' % (kind,)
stde.comment += ' %s' % (kind,)
evokeds.append(ave)
evokeds.append(stde)
if kind == 'standard':
n_standard += 2
write_evokeds(fn, evokeds)
naves = [str(n) for n in sorted(set([
evoked.nave for evoked in evokeds[:n_standard]]))]
bad = [evoked.comment for evoked in evokeds[:n_standard:2]
if evoked.nave == 0]
if bad:
print(f' Got 0 epochs for: {bad}')
naves = ', '.join(naves)
if p.disp_files:
print(' Analysis "%s": %s epochs / condition'
% (analysis, naves))
if p.disp_files:
print(' Saving epochs to disk.')
if 'mat' in p.epochs_type:
spio.savemat(mat_file, dict(epochs=epochs.get_data(),
events=epochs.events, sfreq=sfreq,
drop_log=epochs.drop_log),
do_compression=True, oned_as='column')
if 'fif' in p.epochs_type:
epochs.save(fif_file, **_get_epo_kwargs())
if p.plot_drop_logs:
for subj, drop_log in zip(subjects, drop_logs):
plot_drop_log(drop_log, threshold=p.drop_thresh, subject=subj)
