Merge pull request #423 from constantinpape/release-prep

Bump to 0.7.5

test/util/test_modelzoo.py

@@ -26,6 +26,7 @@ def __call__(self, labels):
         return labels
 
 
+@unittest.skip("BioImage.IO libraries are broken.")
 class TestModelzoo(unittest.TestCase):
     checkpoint_folder = "./checkpoints"
     log_folder = "./logs"

torch_em/__version__.py

@@ -1 +1 @@
-__version__ = "0.7.4"
+__version__ = "0.7.5"

torch_em/transform/raw.py

@@ -160,18 +160,16 @@ def __call__(self, img):
 
 
 class GaussianBlur():
+    """Blur the image with a randomly drawn sigma / bandwidth value.
     """
-    Blur the image.
-    """
-    def __init__(self, kernel_size=(2, 12), sigma=(0, 2.5)):
-        self.kernel_size = kernel_size
+    def __init__(self, sigma=(0, 3.0)):
         self.sigma = sigma
 
     def __call__(self, img):
-        # sample kernel_size and make sure it is odd
-        kernel_size = 2 * (np.random.randint(self.kernel_size[0], self.kernel_size[1]) // 2) + 1
-        # switch boundaries to make sure 0 is excluded from sampling
+        # Sample the sigma value. Note that we switch the bounds to ensure zero is excluded from sampling.
         sigma = np.random.uniform(self.sigma[1], self.sigma[0])
+        # Determine the kernel size based on the sigma value.
+        kernel_size = int(2 * np.ceil(3 * sigma) + 1)
         if isinstance(img, np.ndarray):
             img = torch.from_numpy(img)
         out = transforms.GaussianBlur(kernel_size, sigma=sigma)(img)

torch_em/util/modelzoo.py

@@ -630,7 +630,9 @@ def _load_model(model_spec, device):
     model = PytorchModelAdapter.get_network(weight_spec)
     weight_file = weight_spec.source.path
     if not os.path.exists(weight_file):
-        weight_file = os.path.join(model_spec.root, weight_file)
+        root_folder = f"{model_spec.root.filename}.unzip"
+        assert os.path.exists(root_folder), root_folder
+        weight_file = os.path.join(root_folder, weight_file)
     assert os.path.exists(weight_file), weight_file
     state = torch.load(weight_file, map_location=device)
     model.load_state_dict(state)