Handle empty annotations (#198)

dlup/data/dataset.py

@@ -19,7 +19,7 @@
 from numpy.typing import NDArray
 
 from dlup import BoundaryMode, SlideImage
-from dlup.annotations import WsiAnnotations
+from dlup.annotations import Point, Polygon, WsiAnnotations
 from dlup.background import is_foreground
 from dlup.experimental_backends import ImageBackend  # type: ignore
 from dlup.tiling import Grid, GridOrder, TilingMode
@@ -28,6 +28,8 @@
 
 MaskTypes = Union["SlideImage", npt.NDArray[np.int_], "WsiAnnotations"]
 
+_AnnotationsTypes = Point | Polygon
+
 T_co = TypeVar("T_co", covariant=True)
 T = TypeVar("T")
 _BaseAnnotationTypes = Union[SlideImage, WsiAnnotations]
@@ -42,7 +44,7 @@ class TileSample(TypedDict):
     path: pathlib.Path
     region_index: int
     labels: dict[str, Any] | None
-    annotations: Any | None
+    annotations: Optional[Iterable[_AnnotationsTypes]]
 
 
 PointType = tuple[float, float]

dlup/data/transforms.py

@@ -120,6 +120,15 @@ class ConvertAnnotationsToMask:
 
     def __init__(self, *, roi_name: str | None, index_map: dict[str, int], default_value: int = 0):
         """
+        Converts annotations given my `dlup.annotations.Polygon` or `dlup.annotations.Point` to a mask and a dictionary
+        of points. The mask is initialized with `default_value`, (i.e., background). The values in the mask are
+        subsequently determined by `index_map`, where each value is written to the mask according to this map, in the
+        order of the elements in the annotations. This means that if you have overlapping polygons, the last polygon
+        will overwrite the previous one. The sorting can be handled in the `dlup.annotations.WsiAnnotation` class.
+
+        In case there are no annotations present (i.e. the "annotations" key is None) a `ValueError` is
+        raised.
+
         Parameters
         ----------
         roi_name : str, optional
@@ -134,10 +143,30 @@ def __init__(self, *, roi_name: str | None, index_map: dict[str, int], default_v
         self._default_value = default_value
 
     def __call__(self, sample: TileSample) -> TileSampleWithAnnotationData:
-        if not sample["annotations"]:
-            raise ValueError("No annotations found to convert to mask.")
+        """
+        Convert the annotations to a mask.
+
+        Parameters
+        ----------
+        sample : TileSample
+            The input sample.
+
+        Raises
+        ------
+        ValueError
+            If no annotations are found.
+
+        Returns
+        -------
+        TileSampleWithAnnotationData
+            The input sample with the annotation data added.
+
+        """
 
         _annotations = sample["annotations"]
+        if _annotations is None:
+            raise ValueError("No annotations found to convert to mask.")
+
         points, boxes, mask, roi = convert_annotations(
             _annotations,
             sample["image"].size[::-1],