def _get_model_and_data(
self,
iteration_fidelity,
data_fidelity,
batch_shape,
m,
lin_truncated,
outcome_transform=None,
input_transform=None,
**tkwargs,
):
n_fidelity = (iteration_fidelity is not None) + (data_fidelity
is not None)
train_X, train_Y = _get_random_data_with_fidelity(
batch_shape=batch_shape, m=m, n_fidelity=n_fidelity, **tkwargs)
train_Yvar = torch.full_like(train_Y, 0.01)
model_kwargs = {
"train_X": train_X,
"train_Y": train_Y,
"train_Yvar": train_Yvar,
"iteration_fidelity": iteration_fidelity,
"data_fidelity": data_fidelity,
"linear_truncated": lin_truncated,
}
if outcome_transform is not None:
model_kwargs["outcome_transform"] = outcome_transform
if input_transform is not None:
model_kwargs["input_transform"] = input_transform
model = FixedNoiseMultiFidelityGP(**model_kwargs)
return model, model_kwargs
def test_init_error(self):
train_X = torch.rand(2, 2, device=self.device)
train_Y = torch.rand(2, 1)
train_Yvar = torch.full_like(train_Y, 0.01)
for lin_truncated in (True, False):
with self.assertRaises(UnsupportedError):
FixedNoiseMultiFidelityGP(
train_X, train_Y, train_Yvar, linear_truncated=lin_truncated
)