def test_multivariate_normal() -> None:
num_samples = 2000
dim = 2
mu = np.arange(0, dim) / float(dim)
L_diag = np.ones((dim,))
L_low = 0.1 * np.ones((dim, dim)) * np.tri(dim, k=-1)
L = np.diag(L_diag) + L_low
Sigma = L.dot(L.transpose())
distr = MultivariateNormal(loc=torch.Tensor(mu), scale_tril=torch.Tensor(L))
samples = distr.sample((num_samples,))
mu_hat, L_hat = maximum_likelihood_estimate_sgd(
MultivariateNormalOutput(dim=dim),
samples,
init_biases=None, # todo we would need to rework biases a bit to use it in the multivariate case
learning_rate=0.01,
num_epochs=10,
)
distr = MultivariateNormal(loc=torch.tensor(mu_hat), scale_tril=torch.tensor(L_hat))
Sigma_hat = distr.covariance_matrix.numpy()
assert np.allclose(
mu_hat, mu, atol=0.1, rtol=0.1
), f"mu did not match: mu = {mu}, mu_hat = {mu_hat}"
assert np.allclose(
Sigma_hat, Sigma, atol=0.1, rtol=0.1
), f"Sigma did not match: sigma = {Sigma}, sigma_hat = {Sigma_hat}"
),
(
LowRankMultivariateNormalOutput(dim=target_dim, rank=2),
10,
estimator,
True,
),
(
LowRankMultivariateNormalOutput(dim=target_dim, rank=2),
10,
estimator,
False,
),
(None, 10, estimator, True),
(
MultivariateNormalOutput(dim=target_dim),
10,
estimator,
True,
),
(
MultivariateNormalOutput(dim=target_dim),
10,
estimator,
False,
),
],
)
def test_deepvar(
distr_output, num_batches_per_epoch, Estimator, use_marginal_transformation,
):
# @pytest.mark.timeout(10)
@pytest.mark.parametrize(
"distr_output, num_batches_per_epoch, Estimator, use_marginal_transformation",
[
(NormalOutput(dim=target_dim), 10, estimator, True,),
(NormalOutput(dim=target_dim), 10, estimator, False,),
(LowRankMultivariateNormalOutput(dim=target_dim, rank=2), 10, estimator, True,),
(
LowRankMultivariateNormalOutput(dim=target_dim, rank=2),
10,
estimator,
False,
),
(None, 10, estimator, True),
(MultivariateNormalOutput(dim=target_dim), 10, estimator, True,),
(MultivariateNormalOutput(dim=target_dim), 10, estimator, False,),
],
)
def test_deepvar(
distr_output, num_batches_per_epoch, Estimator, use_marginal_transformation,
):
estimator = Estimator(
input_size=47,
num_cells=20,
num_layers=1,
dropout_rate=0.0,
pick_incomplete=True,
target_dim=target_dim,
prediction_length=metadata.prediction_length,