🤖 AI Suggested Fixes (#1046)

This PR updates documentation and improves code clarity across the project. It revises Python version references to reflect supported/testing versions, fixes typos and formatting issues, and enhances several docstrings. Minor code quality improvements include clearer error messages, corrected variable names, and removal of redundant statements. The NuClick post‑processing logic is updated for compatibility with newer scikit‑image behavior. No functional or API-breaking changes are introduced.

Author: shaneahmed

CONTRIBUTING.rst

@@ -98,7 +98,7 @@ Before you submit a pull request, check that it meets these guidelines:
 
 1. The pull request should include tests.
 2. If the pull request adds functionality, the docs should be updated. Put your new functionality into a function with a docstring, and add the feature to the pull request description.
-3. The pull request should work for Python 3.11, 3.12, and 3.13 and for PyPy. Check https://github.com/TissueImageAnalytics/tiatoolbox/actions/workflows/python-package.yml and make sure that the tests pass for all supported Python versions.
+3. The pull request should work for all Python versions supported by the project (including CPython and PyPy). Check https://github.com/TissueImageAnalytics/tiatoolbox/actions/workflows/python-package.yml and make sure that the tests pass for all supported Python versions.
 
 Tips
 ----

docs/installation.rst

@@ -59,7 +59,7 @@ MacPorts
 Installing Stable Release
 =========================
 
-Please note that TIAToolbox is tested for Python versions 3.11, 3.12, and 3.13.
+Please note that TIAToolbox is currently tested with Python versions 3.11, 3.12, and 3.13. For the full range of supported Python versions, please refer to the package metadata (for example, the ``python_requires`` setting in ``setup.py`` or ``pyproject.toml``)
 
 Recommended
 -----------

docs/pretrained.rst

@@ -173,8 +173,8 @@ The input output configuration is as follows:
                 {'units': 'mpp', 'resolution': 0.25},
                 {'units': 'mpp', 'resolution': 0.25}
             ],
-            margin=128
-            tile_shape=[1024, 1024]
+            margin=128,
+            tile_shape=[1024, 1024],
             patch_input_shape=(256, 256),
             patch_output_shape=(164, 164),
             stride_shape=(164, 164),
@@ -256,8 +256,8 @@ The input output configuration is as follows:
                 {'units': 'mpp', 'resolution': 0.25},
                 {'units': 'mpp', 'resolution': 0.25}
             ],
-            margin=128
-            tile_shape=[1024, 1024]
+            margin=128,
+            tile_shape=[1024, 1024],
             patch_input_shape=(270, 270),
             patch_output_shape=(80, 80),
             stride_shape=(80, 80),
@@ -289,7 +289,7 @@ The input output configuration is as follows:
             output_resolutions=[
                 {'units': 'mpp', 'resolution': 0.25}
             ],
-            tile_shape=[2048, 2048]
+            tile_shape=[2048, 2048],
             patch_input_shape=(252, 252),
             patch_output_shape=(252, 252),
             stride_shape=(150, 150),
@@ -323,8 +323,8 @@ The input output configuration is as follows:
                 {'units': 'mpp', 'resolution': 0.25},
                 {'units': 'mpp', 'resolution': 0.25}
             ],
-            margin=128
-            tile_shape=[1024, 1024]
+            margin=128,
+            tile_shape=[1024, 1024],
             patch_input_shape=(270, 270),
             patch_output_shape=(80, 80),
             stride_shape=(80, 80),

tiatoolbox/__init__.py

@@ -19,7 +19,7 @@
 __version__ = "2.0.1"
 
 # This will set the tiatoolbox external data
-# default to be the user home folder, should work on both Window and Unix/Linux
+# default to be the user home folder, should work on both Windows and Unix/Linux
 # C:\Users\USER\.tiatoolbox
 # /home/USER/.tiatoolbox
 
@@ -113,7 +113,10 @@ def read_registry_files(path_to_registry: str | Path) -> dict:
 def _lazy_import(name: str, module_location: Path) -> ModuleType:
     spec = importlib.util.spec_from_file_location(name, module_location)
     if spec is None or spec.loader is None:
-        raise ModuleNotFoundError(name=name, path=str(module_location))
+        msg = f"Cannot find module {name!r} at path {module_location}"
+        raise ModuleNotFoundError(
+            msg,
+        )
     loader = importlib.util.LazyLoader(spec.loader)
     spec.loader = loader
     module = importlib.util.module_from_spec(spec)

tiatoolbox/annotation/storage.py

@@ -230,7 +230,7 @@ def to_wkb(self: Annotation) -> bytes:
         return copy.copy(self.wkb)
 
     def to_wkt(self: Annotation) -> str:
-        """Returns the geometry as Well-Know Text (WKT).
+        """Returns the geometry as Well-Known Text (WKT).
 
         Returns:
             Annotation:
@@ -1877,7 +1877,7 @@ def _load_cases(
     @classmethod
     def from_geojson(
         cls: type[AnnotationStore],
-        fp: IO | str,
+        fp: IO | str | Path,
         scale_factor: tuple[float, float] = (1, 1),
         origin: tuple[float, float] = (0, 0),
         transform: Callable[[Annotation], Annotation] | None = None,
@@ -1897,7 +1897,7 @@ def from_geojson(
                 A function to apply to each annotation after loading. Should take an
                 annotation as input and return an annotation. Defaults to None.
                 Intended to facilitate modifying the way annotations are loaded to
-                accomodate the specifics of different annotation formats.
+                accommodate the specifics of different annotation formats.
 
         Returns:
             AnnotationStore:
@@ -1939,7 +1939,7 @@ def transform(annotation: Annotation) -> Annotation:
 
     def add_from_geojson(
         self: AnnotationStore,
-        fp: IO | str,
+        fp: IO | str | Path,
         scale_factor: tuple[float, float] = (1, 1),
         origin: tuple[float, float] = (0, 0),
         transform: Callable[[Annotation], Annotation] = lambda x: x,

tiatoolbox/models/architecture/nuclick.py

@@ -108,7 +108,7 @@ def get_block(
         out_channels: int,
         kernel_size: int | tuple[int, int],
         strides: IntPair,
-        dilation_rate: int or IntPair,
+        dilation_rate: int | IntPair,
         activation: str,
         *,
         do_batchnorm: bool,
@@ -306,7 +306,7 @@ def __init__(
             num_output_channels,
             kernel_size=kernel_size,
             strides=strides,
-            activation="None",
+            activation=None,
             use_bias=use_bias,
             dilation_rate=dilation_rate,
             do_batchnorm=True,
@@ -316,7 +316,7 @@ def __init__(
             num_output_channels,
             kernel_size=kernel_size,
             strides=strides,
-            activation="None",
+            activation=None,
             use_bias=use_bias,
             dilation_rate=dilation_rate,
             do_batchnorm=True,
@@ -599,27 +599,39 @@ def postproc(
         """Post-processing.
 
         Args:
-            preds (ndarray): list of prediction output of each patch and
+            preds (ndarray):
+                List of prediction output of each patch and
                 assumed to be in the order of (no.patch, h, w) (match with the output
                 of `infer_batch`).
-            thresh (float): Threshold value. If a pixel has a predicted value larger
+            thresh (float):
+                Threshold value. If a pixel has a predicted value larger
                 than the threshold, it will be classified as nuclei.
-            min_size (int): The smallest allowable object size.
-            min_hole_size (int):  The maximum area, in pixels, of a contiguous hole
+            min_size (int):
+                The smallest allowable object size (objects with
+                fewer pixels than this will be removed).
+            min_hole_size (int):
+                The maximum area, in pixels, of a contiguous hole
                 that will be filled.
-            do_reconstruction (bool): Whether to perform a morphological reconstruction
+            do_reconstruction (bool):
+                Whether to perform a morphological reconstruction
                 of an image.
-            nuc_points (ndarray): In the order of (no.patch, h, w).
+            nuc_points (ndarray):
+                In the order of (no.patch, h, w).
                 In each patch, The pixel that has been 'clicked' is set to 1 and the
                 rest pixels are set to 0.
 
         Returns:
-            masks (ndarray): pixel-wise nuclei instance segmentation
-                prediction, shape:(no.patch, h, w).
+            masks (ndarray):
+                Pixel-wise nuclei instance segmentation prediction,
+                shape:(no.patch, h, w).
 
         """
         masks = preds > thresh
-        masks = remove_small_objects(masks, max_size=min_size - 1)
+        # For scikit-image >=0.26.0, `remove_small_objects` uses `max_size` to
+        # specify the largest object size to be removed. To keep objects with
+        # at least `min_size` pixels, we remove objects with size <= min_size - 1.
+        max_size = max(min_size - 1, 0)
+        masks = remove_small_objects(masks, max_size=max_size)
         masks = remove_small_holes(masks, area_threshold=min_hole_size)
         if do_reconstruction:
             for i in range(len(masks)):

tiatoolbox/models/engine/engine_abc.py

@@ -536,7 +536,6 @@ def infer_patches(
 
         if self.return_labels:
             keys.append("labels")
-            labels = []
 
         if return_coordinates:
             keys.append("coordinates")
@@ -1208,7 +1207,6 @@ def _validate_input_numbers(
             raise ValueError(
                 msg,
             )
-        return
 
     def _update_run_params(
         self: EngineABC,

tiatoolbox/models/engine/multi_task_segmentor.py

@@ -210,7 +210,7 @@ class MultiTaskSegmentorRunParams(SemanticSegmentorRunParams, total=False):
             Shape of output patches (height, width).
         return_labels (bool):
             Whether to return labels with predictions.
-        return_predictions (tuple(bool, ...):
+        return_predictions (tuple[bool, ...]):
             Whether to return array predictions for individual tasks.
         return_probabilities (bool):
             Whether to return per-class probabilities.
@@ -533,7 +533,7 @@ def infer_wsi(
                         Shape of output patches (height, width).
                     return_labels (bool):
                         Whether to return labels with predictions.
-                    return_predictions (tuple(bool, ...):
+                    return_predictions (tuple[bool, ...]):
                         Whether to return array predictions for individual tasks.
                     return_probabilities (bool):
                         Whether to return per-class probabilities.
@@ -783,7 +783,7 @@ def post_process_patches(  # skipcq: PYL-R0201
                         Shape of output patches (height, width).
                     return_labels (bool):
                         Whether to return labels with predictions.
-                    return_predictions (tuple(bool, ...):
+                    return_predictions (tuple[bool, ...]):
                         Whether to return array predictions for individual tasks.
                     return_probabilities (bool):
                         Whether to return per-class probabilities.
@@ -893,7 +893,7 @@ def post_process_wsi(  # skipcq: PYL-R0201
                         Shape of output patches (height, width).
                     return_labels (bool):
                         Whether to return labels with predictions.
-                    return_predictions (tuple(bool, ...):
+                    return_predictions (tuple[bool, ...]):
                         Whether to return array predictions for individual tasks.
                     return_probabilities (bool):
                         Whether to return per-class probabilities.

tiatoolbox/utils/misc.py

@@ -234,7 +234,7 @@ def load_stain_matrix(stain_matrix_input: np.ndarray | PathLike) -> np.ndarray:
 
     """
     if isinstance(stain_matrix_input, (str, Path)):
-        _, __, suffixes = split_path_name_ext(stain_matrix_input)
+        _dir, _file, suffixes = split_path_name_ext(stain_matrix_input)
         if suffixes[-1] not in [".csv", ".npy"]:
             msg = (
                 "If supplying a path to a stain matrix, use either a npy or a csv file"
@@ -914,12 +914,12 @@ def get_bounding_box(img: np.ndarray) -> np.ndarray:
     rows = np.any(img, axis=1)
     cols = np.any(img, axis=0)
     r_min, r_max = np.where(rows)[0][[0, -1]]
-    c_min, cmax = np.where(cols)[0][[0, -1]]
+    c_min, c_max = np.where(cols)[0][[0, -1]]
     # due to python indexing, need to add 1 to max
     # else accessing will be 1px in the box, not out
     r_max += 1
-    cmax += 1
-    return np.array([c_min, r_min, cmax, r_max])
+    c_max += 1
+    return np.array([c_min, r_min, c_max, r_max])
 
 
 def string_to_tuple(in_str: str) -> tuple[str, ...]:

tiatoolbox/visualization/tileserver.py

@@ -470,7 +470,6 @@ def update_renderer(self: TileServer, prop: str) -> str:
         if val in ["None", "null"]:
             val = None
         self.renderers[session_id].__setattr__(prop, val)
-        self.renderers[session_id].__setattr__(prop, val)
         if prop == "blur_radius":
             self.overlaps[session_id] = int(1.5 * val)
             self.get_ann_layer(session_id).overlap = self.overlaps[session_id]
@@ -633,8 +632,10 @@ def unpack_qupath(ann: Annotation) -> Annotation:
                 return ann
 
             sq = SQLiteStore.from_geojson(overlay_path, transform=unpack_qupath)
-        elif overlay_path.suffix == ".dat":
+
+        if overlay_path.suffix == ".dat":
             sq = store_from_dat(overlay_path)
+
         if overlay_path.suffix == ".db":
             sq = SQLiteStore(overlay_path, auto_commit=False)
         else: