def test_vectorizer():
"""Test Vectorizer."""
data = np.random.rand(150, 18, 6)
vect = Vectorizer()
result = vect.fit_transform(data)
assert_equal(result.ndim, 2)
# check inverse_trasnform
orig_data = vect.inverse_transform(result)
assert_equal(orig_data.ndim, 3)
assert_array_equal(orig_data, data)
assert_array_equal(vect.inverse_transform(result[1:]), data[1:])
# check with different shape
assert_equal(vect.fit_transform(np.random.rand(150, 18, 6, 3)).shape,
(150, 324))
assert_equal(vect.fit_transform(data[1:]).shape, (149, 108))
# check if raised errors are working correctly
vect.fit(np.random.rand(105, 12, 3))
assert_raises(ValueError, vect.transform, np.random.rand(105, 12, 3, 1))
assert_raises(ValueError, vect.inverse_transform,
np.random.rand(102, 12, 12))
def test_vectorizer():
"""Test Vectorizer."""
data = np.random.rand(150, 18, 6)
vect = Vectorizer()
result = vect.fit_transform(data)
assert_equal(result.ndim, 2)
# check inverse_trasnform
orig_data = vect.inverse_transform(result)
assert_equal(orig_data.ndim, 3)
assert_array_equal(orig_data, data)
assert_array_equal(vect.inverse_transform(result[1:]), data[1:])
# check with different shape
assert_equal(vect.fit_transform(np.random.rand(150, 18, 6, 3)).shape,
(150, 324))
assert_equal(vect.fit_transform(data[1:]).shape, (149, 108))
# check if raised errors are working correctly
vect.fit(np.random.rand(105, 12, 3))
assert_raises(ValueError, vect.transform, np.random.rand(105, 12, 3, 1))
assert_raises(ValueError, vect.inverse_transform,
np.random.rand(102, 12, 12))
labels = np.concatenate(labels)
# get the data
# sacale the data to (0,1)
data = []
for tf in glob(os.path.join(working_dir, '*-tfr.h5')):
tfcs = mne.time_frequency.read_tfrs(tf)[0]
data_ = tfcs.data
# define a (0,1) scaler
scaler = MinMaxScaler(feature_range=(0, 1))
# define a vectorizer so we can transform the data from 3D to 2D
vectorizer = Vectorizer()
data_vec = vectorizer.fit_transform(data_)
data_scaled = scaler.fit_transform(data_vec)
# after we scale the data to (0,1), we transform the data from 2D back to 3D
data_scaled = vectorizer.inverse_transform(data_scaled)
del tfcs
del data_, data_vec
data.append(data_scaled)
del data_scaled
data = np.concatenate(data, axis=0)
# shuffle the order of the feature matrix and the labels
for _ in range(10):
data, labels = utils.shuffle(data, labels)
# customized the temporal decoding process
# define 10-fold cross validation
cv = StratifiedShuffleSplit(n_splits=10, random_state=12345)
coefs = []