import tensorflow as tf
import gpflow.kernels as kernels
rng = np.random.RandomState(0)
class Datum:
num_data = 100
D = 100
X = rng.rand(num_data, D) * 100
kernel_list = [
kernels.Matern12(),
kernels.Matern32(),
kernels.Matern52(),
kernels.Exponential(),
kernels.Cosine()
]
@pytest.mark.parametrize('kernel', kernel_list)
def test_kernel_euclidean_distance(kernel):
'''
Tests output & gradients of kernels that are a function of the (scaled) euclidean distance
of the points. We test on a high dimensional space, which can generate very small distances
causing the scaled_square_dist to generate some negative values.
'''
K = kernel(Datum.X)
assert not np.isnan(
def matern_kern():
return kernels.Matern32(Data.D_in, variance=rng.rand())
"dirac_diag":
DiagonalGaussian(Xmu, np.zeros((num_data, D_in))),
"dirac_markov_gauss":
MarkovGaussian(Xmu_markov, np.zeros((2, num_data + 1, D_in, D_in))),
"markov_gauss":
markov_gauss(),
}
_kerns = {
"rbf":
kernels.SquaredExponential(variance=rng.rand(),
lengthscales=rng.rand() + 1.0),
"lin":
kernels.Linear(variance=rng.rand()),
"matern":
kernels.Matern32(variance=rng.rand()),
"rbf_act_dim_0":
kernels.SquaredExponential(variance=rng.rand(),
lengthscales=rng.rand() + 1.0,
active_dims=[0]),
"rbf_act_dim_1":
kernels.SquaredExponential(variance=rng.rand(),
lengthscales=rng.rand() + 1.0,
active_dims=[1]),
"lin_act_dim_0":
kernels.Linear(variance=rng.rand(), active_dims=[0]),
"lin_act_dim_1":
kernels.Linear(variance=rng.rand(), active_dims=[1]),
"rbf_lin_sum":
kernels.Sum([
kernels.SquaredExponential(variance=rng.rand(),