def testOneHotRound(self):
self.assertTrue(
np.allclose(strict_onehot_round(np.array([0.1, 0.5, 0.3])),
np.array([0, 1, 0])))
# One item should be set to one at random.
self.assertEqual(
np.count_nonzero(
np.isclose(
randomized_onehot_round(np.array([0.0, 0.0, 0.0])),
np.array([1, 1, 1]),
)),
1,
)
def untransform_observation_features(
self, observation_features: List[ObservationFeatures]
) -> List[ObservationFeatures]:
for obsf in observation_features:
for p_name in self.encoder.keys():
x = np.array([
obsf.parameters.pop(p)
for p in self.encoded_parameters[p_name]
])
if self.rounding == "strict":
x = strict_onehot_round(x)
else:
x = randomized_onehot_round(x)
val = self.encoder[p_name].inverse_transform(
encoded_labels=x[None, :])[0]
if isinstance(val, np.bool_):
val = bool(val) # Numpy bools don't serialize
obsf.parameters[p_name] = val
return observation_features