def test_hk_shape(pts, dims):
n_bins = 10
x = get_input(pts, dims)
hk = HeatKernel(sigma=1, n_bins=n_bins)
num_dimensions = len(np.unique(dims))
x_t = hk.fit(x).transform(x)
assert x_t.shape == (x.shape[0], num_dimensions, n_bins, n_bins)
def test_hk_big_sigma(pts, dims):
""" We expect that with a huge sigma, the diagrams are so diluted that
they are almost 0. Effectively, verifies that the smoothing is applied."""
n_bins = 10
x = get_input(pts, dims)
hk = HeatKernel(sigma=100*np.max(np.abs(x)), n_bins=n_bins)
x_t = hk.fit(x).transform(x)
assert np.all(np.abs(x_t) <= 1e-4)
def test_hk_positive(pts, dims):
""" We expect the points above the PD-diagonal to be non-negative,
(up to a numerical error)"""
n_bins = 10
hk = HeatKernel(sigma=1, n_bins=n_bins)
x = get_input(pts, dims)
x_t = hk.fit(x).transform(x)
assert np.all((np.tril(x_t[:, :, ::-1, :]) + 1e-13) >= 0.)
def test_hk_with_diag_points(pts):
"""Add points on the diagonal, and verify that we have the same results
(on the same fitted values)."""
n_bins = 10
hk = HeatKernel(sigma=1, n_bins=n_bins)
X = get_input(pts, np.zeros((pts.shape[0], pts.shape[1], 1)))
diag_points = np.array([[[2, 2, 0], [3, 3, 0], [7, 7, 0]]])
X_with_diag_points = np.concatenate([X, diag_points], axis=1)
hk = hk.fit(X_with_diag_points)
X_t, X_with_diag_points_t = [hk.transform(X_)
for X_ in [X, X_with_diag_points]]
assert_almost_equal(X_with_diag_points_t, X_t, decimal=13)
def test_all_pts_the_same():
X = np.zeros((1, 4, 3))
hk = HeatKernel(sigma=1)
with pytest.raises(IndexError):
_ = hk.fit(X).transform(X)
