def test_laplace(mu_b: Tuple[float, float], hybridize: bool) -> None:
'''
Test to check that maximizing the likelihood recovers the parameters
'''
# test instance
mu, b = mu_b
# generate samples
mus = mx.nd.zeros((NUM_SAMPLES, )) + mu
bs = mx.nd.zeros((NUM_SAMPLES, )) + b
laplace_distr = Laplace(mu=mus, b=bs)
samples = laplace_distr.sample()
init_biases = [
mu - START_TOL_MULTIPLE * TOL * mu,
inv_softplus(b + START_TOL_MULTIPLE * TOL * b),
]
mu_hat, b_hat = maximum_likelihood_estimate_sgd(LaplaceOutput(),
samples,
hybridize=hybridize,
init_biases=init_biases)
assert (np.abs(mu_hat - mu) <
TOL * mu), f"mu did not match: mu = {mu}, mu_hat = {mu_hat}"
assert (np.abs(b_hat - b) <
TOL * b), f"b did not match: b = {b}, b_hat = {b_hat}"
def test_deterministic_l1(mu: float, hybridize: bool) -> None:
'''
Test to check that maximizing the likelihood recovers the parameters.
This tests uses the Laplace distribution with fixed variance and sample mean.
This essentially reduces to determistic L1.
'''
# generate samples
mu = mu
mus = mx.nd.zeros(NUM_SAMPLES) + mu
class LaplaceFixedVarianceOutput(LaplaceOutput):
@classmethod
def domain_map(cls, F, mu, b):
b = 0.1 * F.ones_like(b)
return mu.squeeze(axis=-1), b.squeeze(axis=-1)
deterministic_distr = Laplace(mu=mus, b=0.1 * mx.nd.ones_like(mus))
samples = deterministic_distr.sample()
mu_hat, _ = maximum_likelihood_estimate_sgd(
LaplaceFixedVarianceOutput(),
samples,
init_biases=[3 * mu, 0.1],
learning_rate=PositiveFloat(1e-3),
hybridize=hybridize,
)
assert (np.abs(mu_hat - mu) <
TOL * mu), f"mu did not match: mu = {mu}, mu_hat = {mu_hat}"
