def test_ERPcovariances():
"""Test fit ERPCovariances"""
x = np.random.randn(10, 3, 100)
labels = np.array([0, 1]).repeat(5)
cov = ERPCovariances()
cov.fit_transform(x, labels)
cov = ERPCovariances(classes=[0])
cov.fit_transform(x, labels)
# assert raise svd
assert_raises(TypeError, ERPCovariances, svd='42')
cov = ERPCovariances(svd=1)
assert_equal(cov.get_params(), dict(classes=None, estimator='scm', svd=1))
def test_ERPcovariances():
"""Test fit ERPCovariances"""
x = np.random.randn(10, 3, 100)
labels = np.array([0, 1]).repeat(5)
cov = ERPCovariances()
cov.fit_transform(x, labels)
cov = ERPCovariances(classes=[0])
cov.fit_transform(x, labels)
# assert raise svd
assert_raises(TypeError, ERPCovariances, svd='42')
cov = ERPCovariances(svd=1)
assert_equal(cov.get_params(), dict(classes=None, estimator='scm',
svd=1))
def test_erp_covariances(estimator, svd, rndstate, get_labels):
"""Test fit ERPCovariances"""
n_classes, n_matrices, n_channels, n_times = 2, 4, 3, 100
x = rndstate.randn(n_matrices, n_channels, n_times)
labels = get_labels(n_matrices, n_classes)
cov = ERPCovariances(estimator=estimator, svd=svd)
covmats = cov.fit_transform(x, labels)
if svd is None:
covsize = (n_classes + 1) * n_channels
else:
covsize = n_classes * svd + n_channels
assert cov.get_params() == dict(classes=None, estimator=estimator, svd=svd)
assert covmats.shape == (n_matrices, covsize, covsize)
assert is_spsd(covmats)