def test_computes_diag_eval(self):
a = torch.tensor([4, 2, 8], dtype=torch.float).view(3, 1)
variances = torch.randn(3)
kernel = WhiteNoiseKernel(variances=variances)
kernel.eval()
actual = torch.diag(variances)
res = kernel(a).evaluate()
self.assertLess(torch.norm(res - actual), 1e-5)
def test_computes_diag_eval_batch(self):
a = torch.Tensor([[4, 2, 8], [4, 2, 8]]).view(2, 3, 1)
variances = torch.randn(2, 3, 1)
kernel = WhiteNoiseKernel(variances=variances)
kernel.eval()
actual = torch.cat((torch.diag(variances[0].squeeze(-1)).unsqueeze(0),
torch.diag(variances[1].squeeze(-1)).unsqueeze(0)))
res = kernel(a).evaluate()
self.assertLess(torch.norm(res - actual), 1e-5)
def test_computes_zero_eval(self):
a = torch.tensor([4, 2, 8], dtype=torch.float).view(3, 1)
b = torch.tensor([3, 7], dtype=torch.float).view(2, 1)
variances = torch.randn(3)
kernel = WhiteNoiseKernel(variances=variances)
kernel.eval()
actual_one = torch.zeros(3, 2)
actual_two = torch.zeros(2, 3)
res_one = kernel(a, b).evaluate()
res_two = kernel(b, a).evaluate()
self.assertLess(torch.norm(res_one - actual_one), 1e-5)
self.assertLess(torch.norm(res_two - actual_two), 1e-5)
def test_computes_diag_train(self):
a = torch.Tensor([4, 2, 8]).view(3, 1)
variances = torch.randn(3)
kernel = WhiteNoiseKernel(variances=variances)
actual = torch.diag(variances)
res = kernel(a).evaluate()
self.assertLess(torch.norm(res - actual), 1e-5)
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.grid_covar_module = GridInterpolationKernel(self.base_covar_module, grid_size=50, num_dims=1)
self.noise_covar_module = WhiteNoiseKernel(variances=torch.ones(100) * 0.001)
self.covar_module = self.grid_covar_module + self.noise_covar_module
def __init__(self, train_inputs, train_targets, likelihood):
super(ExactGPModel, self).__init__(train_inputs, train_targets,
likelihood)
self.mean_module = ConstantMean(constant_bounds=(-1, 1))
self.rbf_covar_module = RBFKernel(log_lengthscale_bounds=(-3, 3))
self.noise_covar_module = WhiteNoiseKernel(variances=torch.ones(11) *
0.001)
self.covar_module = self.rbf_covar_module + self.noise_covar_module
def __init__(self, train_inputs, train_targets, likelihood):
super(ExactGPModel, self).__init__(train_inputs, train_targets, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1, 1))
self.rbf_covar_module = RBFKernel(
log_lengthscale_prior=SmoothedBoxPrior(exp(-3), exp(3), sigma=0.1, log_transform=True)
)
self.noise_covar_module = WhiteNoiseKernel(variances=torch.ones(11) * 0.001)
self.covar_module = ScaleKernel(self.rbf_covar_module + self.noise_covar_module)
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(constant_bounds=[-1e-5, 1e-5])
self.base_covar_module = RBFKernel(log_lengthscale_bounds=(-5, 6))
self.grid_covar_module = GridInterpolationKernel(
self.base_covar_module, grid_size=50, grid_bounds=[(0, 1)])
self.noise_covar_module = WhiteNoiseKernel(variances=torch.ones(100) *
0.001)
self.covar_module = self.grid_covar_module + self.noise_covar_module
def __init__(self,
train_x,
train_y,
likelihood,
var=None,
latent=None,
kernel_params=None,
latent_params=None):
super(ExactGPModel, self).__init__(train_x, train_y, likelihood)
if latent_params is None:
latent_params = {'input_dim': train_x.size(-1)}
self._set_latent_function(latent, latent_params)
self.mean_module = ZeroMean()
ard_num_dims = self.latent_func.embed_dim if self.latent_func.embed_dim is not None else train_x.size(
-1)
kernel = kernel_params['type'] if kernel_params is not None else 'rbf'
if kernel is None or kernel == 'rbf':
self.kernel_covar_module = ScaleKernel(
RBFKernel(ard_num_dims=ard_num_dims))
elif kernel == 'matern':
self.kernel_covar_module = ScaleKernel(
MaternKernel(nu=1.5, ard_num_dims=ard_num_dims))
# without scale kernel: very poor performance
# matern 0.5, 1.5 and 2.5 all have similar performance
elif kernel == 'spectral_mixture':
self.kernel_covar_module = SpectralMixtureKernel(
num_mixtures=kernel_params['n_mixtures'],
ard_num_dims=train_x.size(-1))
self.kernel_covar_module.initialize_from_data(train_x, train_y)
else:
raise NotImplementedError
# set covariance module
if var is not None:
self.noise_covar_module = WhiteNoiseKernel(var)
self.covar_module = self.kernel_covar_module + self.noise_covar_module
else:
self.covar_module = self.kernel_covar_module
