def test_with_zero_gamma(self):
transform = RandomGamma(log_gamma=0)
transformed = transform(self.sample_subject)
self.assertTensorAlmostEqual(self.sample_subject.t1.data,
transformed.t1.data)
def test_wrong_gamma_type(self):
with self.assertRaises(ValueError):
RandomGamma(log_gamma='wrong')
def test_negative_values(self):
with self.assertWarns(UserWarning):
RandomGamma()(torch.rand(1, 3, 3, 3) - 1)
def test_with_low_gamma(self):
transform = RandomGamma(log_gamma=(-100, -100))
transformed = transform(self.sample_subject)
self.assertTensorAlmostEqual(self.sample_subject.t1.data > 0,
transformed.t1.data)
def test_with_non_zero_gamma(self):
transform = RandomGamma(log_gamma=(0.1, 0.3))
transformed = transform(self.sample_subject)
self.assertTensorNotEqual(self.sample_subject.t1.data,
transformed.t1.data)
def test_with_low_gamma(self):
transform = RandomGamma(log_gamma=(-100, -100))
tensor = self.get_random_tensor_zero_one()
transformed = transform(tensor)
self.assertTensorAlmostEqual(tensor > 0, transformed)
def test_with_non_zero_gamma(self):
transform = RandomGamma(log_gamma=(0.1, 0.3))
tensor = self.get_random_tensor_zero_one()
transformed = transform(tensor)
self.assertTensorNotEqual(tensor, transformed)
def test_with_high_gamma(self):
transform = RandomGamma(log_gamma=(100, 100))
transformed = transform(self.sample)
self.assertTensorAlmostEqual(self.sample.t1.data == 1,
transformed.t1.data)