def markov_gauss():
cov_params = rng.randn(num_data + 1, D_in, 2 * D_in) / 2.0 # (N+1)xDx2D
Xcov = cov_params @ np.transpose(cov_params, (0, 2, 1)) # (N+1)xDxD
Xcross = cov_params[:-1] @ np.transpose(cov_params[1:], (0, 2, 1)) # NxDxD
Xcross = np.concatenate((Xcross, np.zeros((1, D_in, D_in))),
0) # (N+1)xDxD
Xcov = np.stack([Xcov, Xcross]) # 2x(N+1)xDxD
return MarkovGaussian(Xmu_markov, ctt(Xcov))
import gpflow
from gpflow import inducing_variables, kernels
from gpflow import mean_functions as mf
from gpflow.config import default_float
from gpflow.expectations import expectation, quadrature_expectation
from gpflow.probability_distributions import DiagonalGaussian, Gaussian, MarkovGaussian
rng = np.random.RandomState(1)
RTOL = 1e-6
num_data = 5
num_ind = 4
D_in = 2
D_out = 2
Xmu = ctt(rng.randn(num_data, D_in))
Xmu_markov = ctt(rng.randn(num_data + 1, D_in)) # (N+1)xD
Xcov = rng.randn(num_data, D_in, D_in)
Xcov = ctt(Xcov @ np.transpose(Xcov, (0, 2, 1)))
Z = rng.randn(num_ind, D_in)
def markov_gauss():
cov_params = rng.randn(num_data + 1, D_in, 2 * D_in) / 2.0 # (N+1)xDx2D
Xcov = cov_params @ np.transpose(cov_params, (0, 2, 1)) # (N+1)xDxD
Xcross = cov_params[:-1] @ np.transpose(cov_params[1:], (0, 2, 1)) # NxDxD
Xcross = np.concatenate((Xcross, np.zeros((1, D_in, D_in))),
0) # (N+1)xDxD
Xcov = np.stack([Xcov, Xcross]) # 2x(N+1)xDxD
return MarkovGaussian(Xmu_markov, ctt(Xcov))