Color circles and polygons with annotations (#73)

* added support for circles

* added affine transformations

* minor fix

* cleanup

* added color circles by obs

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* added polygons

* pre-commit fix

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* cleanup

* cleanup

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

examples/spatialdata_visium.py

@@ -0,0 +1,12 @@
+# In this script, we are reading the Visium dataset into a SpatialData object and visualising
+# it with the Interactive class from napari_spatialdata.
+
+# The dataset can be downloaded from https://spatialdata.scverse.org/en/latest/tutorials/notebooks/datasets/README.html
+
+from napari_spatialdata import Interactive
+from spatialdata import SpatialData
+
+if __name__ == "__main__":
+    sdata = SpatialData.read("../data/visium/data.zarr")  # Change this path!
+    i = Interactive(sdata)
+    i.run()

src/napari_spatialdata/_interactive.py

@@ -97,13 +97,15 @@ def _add_circles(self, key: str) -> None:
                     self._sdata.table.obs[self._sdata.table.uns["spatialdata_attrs"]["region_key"]] == key
                 ],
                 "shapes_key": self._sdata.table.uns["spatialdata_attrs"]["region_key"],
+                "shapes_type": "circles",
             },
         )
 
     def _add_polygons(self, key: str) -> None:
         polygons = []
         df = self._sdata.shapes[key]
         affine = _get_transform(self._sdata.shapes[key], self.coordinate_system_widget._system)
+
         # when mulitpolygons are present, we select the largest ones
         if "MultiPolygon" in np.unique(df.geometry.type):
             logger.info("Multipolygons are present in the data. Only the largest polygon per cell is retained.")
@@ -133,6 +135,7 @@ def _add_polygons(self, key: str) -> None:
                     self._sdata.table.obs[self._sdata.table.uns["spatialdata_attrs"]["region_key"]] == key
                 ],
                 "shapes_key": self._sdata.table.uns["spatialdata_attrs"]["region_key"],
+                "shapes_type": "polygons",
             },
         )
 
@@ -185,6 +188,7 @@ def _add_points(self, key: str) -> None:
             logger.info("Subsampling points because the number of points exceeds the currently supported 100 000.")
             gen = np.random.default_rng()
             subsample = gen.choice(len(points), size=100000, replace=False)
+
         self._viewer.add_points(
             points[["y", "x"]].values[subsample],
             name=key,

src/napari_spatialdata/_view.py

@@ -244,14 +244,24 @@ def _select_layer(self) -> None:
         # if layer is not None and "adata" in layer.metadata:
         self.model.adata = layer.metadata["adata"]
 
-        self.model.coordinates = np.insert(self.model.adata.obsm[Key.obsm.spatial][:, ::-1][:, :2], 0, values=0, axis=1)
+        if self.model.adata.shape == (0, 0):
+            return
+
+        if "spatial" in self.model.adata.obsm:
+            self.model.coordinates = np.insert(
+                self.model.adata.obsm[Key.obsm.spatial][:, ::-1][:, :2], 0, values=0, axis=1
+            )
+
         if "points" in layer.metadata:
+            # TODO: Check if this can be removed
             self.model.points_coordinates = layer.metadata["points"].X
             self.model.points_var = layer.metadata["points"].obs["gene"]
             self.model.point_diameter = np.array([0.0] + [layer.metadata["point_diameter"]] * 2) * self.model.scale
+
         self.model.spot_diameter = np.array([0.0, 10.0, 10.0])
         self.model.labels_key = layer.metadata["labels_key"] if isinstance(layer, Labels) else None
         self.model.system_name = self.model.layer.name
+
         if "colormap" in layer.metadata:
             self.model.cmap = layer.metadata["colormap"]
         if hasattr(

src/napari_spatialdata/_widgets.py

@@ -9,7 +9,7 @@
 import numpy as np
 import pandas as pd
 from loguru import logger
-from napari.layers import Labels, Layer, Points
+from napari.layers import Labels, Layer, Points, Shapes
 from napari.viewer import Viewer
 from qtpy import QtCore, QtWidgets
 from qtpy.QtCore import Qt, Signal
@@ -108,8 +108,8 @@ def __init__(self, viewer: Viewer | None, model: ImageModel, attr: str, **kwargs
         self._model = model
 
         self._attr = attr
-        self._getter = getattr(self.model, f"get_{attr}")
 
+        self._getter = getattr(self.model, f"get_{attr}")
         self.layerChanged.connect(self._onChange)
         self._onChange()
 
@@ -125,6 +125,7 @@ def _onAction(self, items: Iterable[str]) -> None:
             except Exception as e:  # noqa: BLE001
                 logger.error(e)
                 continue
+
             if vec.ndim == 2:
                 self.viewer.add_points(
                     vec,
@@ -136,7 +137,8 @@ def _onAction(self, items: Iterable[str]) -> None:
                 )
             else:
                 properties = self._get_points_properties(vec, key=item, layer=self.model.layer)
-                if isinstance(self.model.layer, Points):
+
+                if isinstance(self.model.layer, (Points, Shapes)):
                     self.model.layer.name = (
                         "" if self.model.system_name is None else self.model.system_name + ":"
                     ) + item
@@ -201,11 +203,26 @@ def _(self, vec: NDArrayA, **kwargs: Any) -> dict[str, Any]:
         norm_vec = _min_max_norm(vec)
         color_vec = cmap(norm_vec)
         if layer is not None and isinstance(layer, Labels):
+            cmap = plt.get_cmap(self.model.cmap)
+            norm_vec = _min_max_norm(vec)
+            color_vec = cmap(norm_vec)
+
             return {
                 "color": dict(zip(self.model.adata.obs[self.model.labels_key].values, color_vec)),
                 "properties": {"value": vec},
                 "text": None,
             }
+
+        if layer is not None and isinstance(layer, Shapes):
+            cmap = plt.get_cmap(self.model.cmap)
+            norm_vec = _min_max_norm(vec)
+            color_vec = cmap(norm_vec)
+
+            return {
+                "text": None,
+                "face_color": color_vec,
+            }
+
         return {
             "text": None,
             "face_color": color_vec,
@@ -217,8 +234,12 @@ def _(self, vec: pd.Series, key: str, layer: Layer) -> dict[str, Any]:
         face_color = _get_categorical(
             self.model.adata, key=key, palette=self.model.palette, colordict=colortypes, vec=vec
         )
+
         if layer is not None and isinstance(layer, Labels):
-            return {"color": dict(zip(self.model.adata.obs[self.model.labels_key].values, face_color))}
+            return {"color": dict(zip(self.model.adata.obs[self.model.labels_key].values, face_color)), "text": None}
+
+        if layer is not None and isinstance(layer, Shapes):
+            return {"face_color": face_color, "metadata": None, "text": None}
 
         cluster_labels = _position_cluster_labels(self.model.coordinates, vec)
         return {