Add code style via ruff

.codespellignore

@@ -1,4 +1,6 @@
+assertIn
 mapp
 padd
 struc
+TE
 warmup

.github/workflows/code-style.yml

@@ -3,9 +3,9 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.event.number }}-${{ github.event.ref }}
   cancel-in-progress: true
 on:
-#  pull_request:
-#  push:
-#    branches: [dev]
+  pull_request:
+  push:
+    branches: [dev]
   workflow_dispatch:
 
 jobs:
@@ -32,9 +32,9 @@ jobs:
         with:
           check_filenames: true
           check_hidden: true
-          skip: './.git,./build,./.mypy_cache,./.pytest_cache'
+          skip: './build,./doc/images,./Tutorial,./.git,./.mypy_cache,./.pytest_cache'
           ignore_words_file: ./.codespellignore
-      - name: Run pydocstyle
-        run: pydocstyle .
-      - name: Run bibclean
-        run: bibclean-check doc/references.bib
+#      - name: Run pydocstyle
+#        run: pydocstyle .
+#      - name: Run bibclean
+#        run: bibclean-check doc/references.bib

CerebNet/apply_warp.py

@@ -1,4 +1,7 @@
 import argparse
+from os.path import join
+
+import nibabel as nib
 
 # Copyright 2022 Image Analysis Lab, German Center for Neurodegenerative Diseases (DZNE), Bonn
 #
@@ -13,12 +16,9 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-
 # IMPORTS
 import numpy as np
-import nibabel as nib
 
-from os.path import join
 from CerebNet.datasets import utils
 
 

CerebNet/config/__init__.py

@@ -15,6 +15,7 @@
 # IMPORTS
 from CerebNet.config.cerebnet import get_cfg_cerebnet
 from CerebNet.config.dataset import get_cfg_dataset
+
 __all__ = [
     "cerebnet",
     "dataset",

CerebNet/config/cerebnet.py

@@ -116,7 +116,7 @@
 # Data Augmentation options
 # ---------------------------------------------------------------------------- #
 
-# Augmentation for traning
+# Augmentation for training
 _C.AUGMENTATION = CN()
 
 # list of augmentations to use for training

CerebNet/data_loader/augmentation.py

@@ -16,21 +16,19 @@
 # IMPORTS
 import numbers
 import random
-from typing import Optional
 
 import numpy as np
 import torch
 from numpy import random as npr
-from scipy.ndimage import gaussian_filter, affine_transform
+from scipy.ndimage import affine_transform, gaussian_filter
 from scipy.stats import median_abs_deviation
 from torchvision import transforms
 
 from CerebNet.data_loader.data_utils import FLIPPED_LABELS
 
 
-
 # Transformations for training
-class ToTensor(object):
+class ToTensor:
     """
     Convert ndarrays in sample to Tensors.
     """
@@ -66,7 +64,7 @@ def _apply_img(self, img):
         return super()._apply_img(img.transpose((2, 0, 1)))
 
 
-class RandomAffine(object):
+class RandomAffine:
     """
     Apply a random affine transformation to
     images, label and weight
@@ -99,15 +97,15 @@ def _get_random_affine(self):
             degrees = (-self.degree, self.degree)
         else:
             assert (
-                isinstance(self.degree, (tuple, list)) and len(self.degree) == 2
+                isinstance(self.degree, tuple | list) and len(self.degree) == 2
             ), "degrees should be a list or tuple and it must be of length 2."
         if isinstance(self.translate, numbers.Number):
             if not (0.0 <= self.translate <= 1.0):
                 raise ValueError("translation values should be between 0 and 1")
             translate = (self.translate, self.translate)
         else:
             assert (
-                isinstance(self.translate, (tuple, list)) and len(self.translate) == 2
+                isinstance(self.translate, tuple | list) and len(self.translate) == 2
             ), "translate should be a list or tuple and it must be of length 2."
             for t in self.translate:
                 if not (0.0 <= t <= 1.0):
@@ -159,7 +157,7 @@ def __call__(self, sample):
         return sample
 
 
-class RandomFlip(object):
+class RandomFlip:
     """
     Random horizontal flipping.
     """
@@ -196,7 +194,7 @@ class RandomBiasField:
     def __init__(
         self,
         cfg,
-        seed: Optional[int] = None,
+        seed: int | None = None,
     ):
         """
         Initialize the RandomBiasField object with configuration and optional seed.
@@ -287,7 +285,7 @@ def __call__(self, sample):
         return sample
 
 
-class RandomLabelsToImage(object):
+class RandomLabelsToImage:
     """
     Generate image from segmentation
     using the dataset intensity priors.

CerebNet/data_loader/data_utils.py

@@ -402,7 +402,7 @@ def uncrop_volume(vol, uncrop_shape, roi):
 
 
 def get_binary_map(lbl_map, class_names):
-    bin_map = np.logical_or.reduce(list(map(lambda l: lbl_map == l, class_names)))
+    bin_map = np.logical_or.reduce(list(map(lambda lb: lbl_map == lb, class_names)))
     return bin_map
 
 

CerebNet/data_loader/dataset.py

@@ -13,31 +13,30 @@
 # limitations under the License.
 
 # IMPORTS
-from typing import Tuple, Literal, TypeVar, Dict
 from numbers import Number
+from typing import Literal, TypeVar
 
+import h5py
 import nibabel as nib
-import torch
 import numpy as np
+import torch
 from numpy import typing as npt
-import h5py
 from torch.utils.data.dataset import Dataset
 from torchvision.transforms import Compose
 
-from FastSurferCNN.utils import logging, Plane
+from CerebNet.data_loader import data_utils as utils
+from CerebNet.data_loader.augmentation import ToTensor
+from CerebNet.datasets.load_data import SubjectLoader
+from CerebNet.datasets.utils import bounding_volume_offset, crop_transform
 from FastSurferCNN.data_loader.data_utils import (
     get_thick_slices,
     transform_axial,
     transform_sagittal,
 )
-
-from CerebNet.data_loader import data_utils as utils
-from CerebNet.data_loader.augmentation import ToTensor
-from CerebNet.datasets.load_data import SubjectLoader
-from CerebNet.datasets.utils import crop_transform, bounding_volume_offset
+from FastSurferCNN.utils import Plane, logging
 
 ROIKeys = Literal["source_shape", "offsets", "target_shape"]
-LocalizerROI = Dict[ROIKeys, Tuple[int, ...]]
+LocalizerROI = dict[ROIKeys, tuple[int, ...]]
 
 NT = TypeVar("NT", bound=Number)
 
@@ -118,11 +117,7 @@ def __init__(self, dataset_path, cfg, transforms, load_aux_data):
         del self.dataset["subject"]
 
         logger.info(
-            "Successfully loaded {} slices in {} plane from {}".format(
-                self.count,
-                cfg.DATA.PLANE,
-                dataset_path,
-            )
+            f"Successfully loaded {self.count} slices in {cfg.DATA.PLANE} plane from {dataset_path}"
         )
 
         logger.info(
@@ -242,7 +237,7 @@ def __init__(
         self,
         img_org: nib.analyze.SpatialImage,
         brain_seg: nib.analyze.SpatialImage,
-        patch_size: Tuple[int, ...],
+        patch_size: tuple[int, ...],
         slice_thickness: int,
         primary_slice: str,
     ):
@@ -298,10 +293,10 @@ def __init__(
             "coronal": img,
             "sagittal": transform_sagittal(img),
         }
-        for plane, data in data.items():
+        for plane, data_i in data.items():
             # data is transformed to 'plane'-direction in axis 2
             thick_slices = get_thick_slices(
-                data, self.slice_thickness
+                data_i, self.slice_thickness
             )  # [H, W, n_slices, C]
             # it seems x and y are flipped with respect to expectations here
             self.images_per_plane[plane] = np.transpose(
@@ -315,7 +310,7 @@ def locate_mask_bbox(self, mask: npt.NDArray[bool]):
             bbox of min0, min1, ..., max0, max1, ...
         """
         # filter disconnected components
-        from skimage.measure import regionprops, label
+        from skimage.measure import label, regionprops
 
         label_image = label(mask, connectivity=3)
         regions = regionprops(label_image)
@@ -341,7 +336,7 @@ def plane(self) -> Plane:
         """Returns the active plane"""
         return self._plane
 
-    def __getitem__(self, index: int) -> Tuple[Plane, np.ndarray]:
+    def __getitem__(self, index: int) -> tuple[Plane, np.ndarray]:
         """Get the plane and data belonging to indices given."""
 
         if not (0 <= index < self.images_per_plane[self.plane].shape[0]):

CerebNet/data_loader/loader.py

@@ -13,13 +13,12 @@
 # limitations under the License.
 
 # IMPORTS
-from torchvision import transforms
 from torch.utils.data import DataLoader
-
-from FastSurferCNN.utils import logging
+from torchvision import transforms
 
 from CerebNet.data_loader import dataset as dset
 from CerebNet.data_loader.augmentation import ToTensor, get_transform
+from FastSurferCNN.utils import logging
 
 logger = logging.get_logger(__name__)
 

CerebNet/datasets/generate_hdf5.py

@@ -14,10 +14,10 @@
 
 
 # IMPORTS
+import time
 import warnings
 from collections import defaultdict
-from os.path import join, isfile
-import time
+from os.path import isfile, join
 
 import h5py
 import numpy as np
@@ -65,7 +65,7 @@ def _read_warp_dict(self):
         """
         subj2warps = defaultdict(list)
         all_imgs = []
-        with open(join(self.cfg.REG_DATA_DIR, self.cfg.REG_DATA_CSV), "r") as f:
+        with open(join(self.cfg.REG_DATA_DIR, self.cfg.REG_DATA_CSV)) as f:
             for line in f.readlines():
                 line = line.strip()
                 ids = line.split(",")
@@ -84,7 +84,7 @@ def _read_warp_dict(self):
                         all_imgs.append(img_path)
                         subj2warps[i].append((img_path, lbl_path))
                     else:
-                        warnings.warn(f"Warp field at {img_path} not found.")
+                        warnings.warn(f"Warp field at {img_path} not found.", stacklevel=2)
         return subj2warps
 
     def create_hdf5_dataset(
@@ -123,9 +123,7 @@ def create_hdf5_dataset(
             # try:
             start = time.time()
             print(
-                "Volume Nr: {}/{} Processing MRI Data from {}".format(
-                    idx + 1, len(subjects_list), current_subject
-                )
+                f"Volume Nr: {idx + 1}/{len(subjects_list)} Processing MRI Data from {current_subject}"
             )
 
             in_data = self.subj_loader.load_subject(
@@ -145,9 +143,7 @@ def create_hdf5_dataset(
             end = time.time() - start
             print("Number of Cerebellum classes", len(np.unique(in_data["label"])))
             print(
-                "Volume: {} Finished Data Reading and Appending in {:.3f} seconds.".format(
-                    idx + 1, end
-                )
+                f"Volume: {idx + 1} Finished Data Reading and Appending in {end:.3f} seconds."
             )
 
             # except Exception as e:
@@ -156,4 +152,4 @@ def create_hdf5_dataset(
 
         self._save_hdf5_file(datasets, dataset_name)
         end_d = time.time() - start_d
-        print("Successfully written {} in {:.3f} seconds.".format(dataset_name, end_d))
+        print(f"Successfully written {dataset_name} in {end_d:.3f} seconds.")

CerebNet/datasets/load_data.py

@@ -14,8 +14,8 @@
 
 
 # IMPORTS
-from os.path import join, isfile
 from functools import partial
+from os.path import isfile, join
 
 import numpy as np
 
@@ -91,9 +91,9 @@ def _load_volumes(self, subject_path, store_talairach=False):
 
         img_meta_data = {}
         orig, _ = utils.load_reorient_rescale_image(orig_path)
-        print("Orig image {}".format(orig_path))
+        print(f"Orig image {orig_path}")
 
-        print("Loading from {}".format(subseg_path))
+        print(f"Loading from {subseg_path}")
         subseg_file = utils.load_reorient(subseg_path)
         cereb_subseg = np.asarray(subseg_file.get_fdata(), dtype=np.int16)
         img_meta_data["affine"] = subseg_file.affine