auto updates

monai/bundle/config_item.py

@@ -64,9 +64,7 @@ def _find_module_names(self) -> List[str]:
         Find all the modules start with MOD_START and don't contain any of `excludes`.
 
         """
-        return [
-            m for m in sys.modules.keys() if m.startswith(self.MOD_START) and all(s not in m for s in self.excludes)
-        ]
+        return [m for m in sys.modules if m.startswith(self.MOD_START) and all(s not in m for s in self.excludes)]
 
     def _find_classes_or_functions(self, modnames: Union[Sequence[str], str]) -> Dict[str, List]:
         """

monai/bundle/scripts.py

@@ -384,7 +384,7 @@ def get_all_bundles_list(
 
     bundles_info = _get_all_bundles_info(repo=repo, tag=tag, auth_token=auth_token)
     bundles_list = []
-    for bundle_name in bundles_info.keys():
+    for bundle_name in bundles_info:
         latest_version = sorted(bundles_info[bundle_name].keys())[-1]
         bundles_list.append((bundle_name, latest_version))
 

monai/losses/multi_scale.py

@@ -60,7 +60,7 @@ def __init__(
             kernel: gaussian or cauchy.
         """
         super().__init__(reduction=LossReduction(reduction).value)
-        if kernel not in kernel_fn_dict.keys():
+        if kernel not in kernel_fn_dict:
             raise ValueError(f"got unsupported kernel type: {kernel}", "only support gaussian and cauchy")
         self.kernel_fn = kernel_fn_dict[kernel]
         self.loss = loss

monai/networks/nets/efficientnet.py

@@ -529,7 +529,7 @@ def __init__(
         ]
 
         # check if model_name is valid model
-        if model_name not in efficientnet_params.keys():
+        if model_name not in efficientnet_params:
             model_name_string = ", ".join(efficientnet_params.keys())
             raise ValueError(f"invalid model_name {model_name} found, must be one of {model_name_string} ")
 
@@ -586,7 +586,7 @@ def __init__(
         ]
 
         # check if model_name is valid model
-        if model_name not in efficientnet_params.keys():
+        if model_name not in efficientnet_params:
             model_name_string = ", ".join(efficientnet_params.keys())
             raise ValueError(f"invalid model_name {model_name} found, must be one of {model_name_string} ")
 
@@ -720,7 +720,7 @@ def get_efficientnet_image_size(model_name: str) -> int:
 
     """
     # check if model_name is valid model
-    if model_name not in efficientnet_params.keys():
+    if model_name not in efficientnet_params:
         model_name_string = ", ".join(efficientnet_params.keys())
         raise ValueError(f"invalid model_name {model_name} found, must be one of {model_name_string} ")
 

monai/transforms/compose.py

@@ -158,7 +158,7 @@ def flatten(self):
         """
         new_transforms = []
         for t in self.transforms:
-            if type(t) is Compose:
+            if type(t) is Compose:  # nopep8
                 new_transforms += t.flatten().transforms
             else:
                 new_transforms.append(t)

monai/transforms/signal/array.py

@@ -140,7 +140,7 @@ def __call__(self, signal: NdarrayOrTensor) -> NdarrayOrTensor:
         self.randomize(None)
         self.magnitude = self.R.uniform(low=self.boundaries[0], high=self.boundaries[1])
 
-        length = signal.shape[len(signal.shape) - 1]
+        length = signal.shape[-1]
         mask = torch.zeros(round(self.magnitude * length))
         trange = torch.arange(length)
         loc = trange[torch.randint(0, trange.size(0), (1,))]
@@ -265,7 +265,7 @@ def __call__(self, signal: NdarrayOrTensor) -> NdarrayOrTensor:
         self.fracs = self.R.uniform(low=self.fraction[0], high=self.fraction[1])
         self.freqs = self.R.uniform(low=self.frequencies[0], high=self.frequencies[1])
 
-        length = signal.shape[len(signal.shape) - 1]
+        length = signal.shape[-1]
 
         time_partial = np.arange(0, round(self.fracs * length), 1)
         data = convert_to_tensor(self.freqs * time_partial)
@@ -347,7 +347,7 @@ def __call__(self, signal: NdarrayOrTensor) -> NdarrayOrTensor:
         self.fracs = self.R.uniform(low=self.fraction[0], high=self.fraction[1])
         self.freqs = self.R.uniform(low=self.frequencies[0], high=self.frequencies[1])
 
-        length = signal.shape[len(signal.shape) - 1]
+        length = signal.shape[-1]
 
         time_partial = np.arange(0, round(self.fracs * length), 1)
         squaredpulse_partial = self.magnitude * squarepulse(self.freqs * time_partial)

monai/transforms/utils.py

@@ -1746,7 +1746,7 @@ def paste_slices(tup):
     given a tuple (pos,w,max_w), return a tuple of slices
     """
     pos, w, max_w = tup
-    max_w = max_w.shape[len(max_w.shape) - 1]
+    max_w = max_w.shape[-1]
     orig_min = max(pos, 0)
     orig_max = min(pos + w, max_w)
     block_min = -min(pos, 0)

tests/test_flatten_sub_keysd.py

@@ -23,7 +23,6 @@
 D1 = {"a": A, "b": B}
 D2 = {"a": A, "b": B, "c": C}
 
-
 TEST_CASE_0 = [{"keys": "pred"}, {"image": I, "pred": D1}, {"a": A, "b": B, "image": I}]
 TEST_CASE_1 = [{"keys": "pred"}, {"image": I, "pred": D2}, {"a": A, "b": B, "c": C, "image": I}]
 TEST_CASE_2 = [{"keys": "pred", "sub_keys": ["a", "b"]}, {"image": I, "pred": D1}, {"a": A, "b": B, "image": I}]

tests/test_global_mutual_information_loss.py

@@ -94,7 +94,7 @@ def transformation(translate_params=(0.0, 0.0, 0.0), rotate_params=(0.0, 0.0, 0.
         a1 = transformation()
         a1 = torch.tensor(a1).unsqueeze(0).unsqueeze(0).to(device)
 
-        for mode in transform_params_dict.keys():
+        for mode in transform_params_dict:
             transform_params_list = transform_params_dict[mode]
             expected_value_list = EXPECTED_VALUE[mode]
             for transform_params, expected_value in zip(transform_params_list, expected_value_list):