bug categorical annotatoins points (#304)

* bug categorical annotatoins points

* added test for continuous column

src/napari_spatialdata/_widgets.py

@@ -198,14 +198,35 @@ def _(self, vec: pd.Series, **kwargs: Any) -> dict[str, Any]:
         if isinstance(layer, Labels):
             element_indices = element_indices[element_indices != 0]
         # When merging if the row is not present in the other table it will be nan so we can give it a default color
-        if (vec_color_name := vec.name + "_color") not in self.model.adata.uns:
-            colorer = AnnData(shape=(len(vec), 0), obs=pd.DataFrame(index=vec.index, data={"vec": vec}))
-            _set_colors_for_categorical_obs(colorer, "vec", palette="tab20")
-            colors = colorer.uns["vec_colors"]
-            color_dict = dict(zip(vec.cat.categories, colors))
-            color_dict.update({np.nan: "#808080ff"})
+        vec_color_name = vec.name + "_color"
+        if self._attr != "columns_df":
+            if vec_color_name not in self.model.adata.uns:
+                colorer = AnnData(shape=(len(vec), 0), obs=pd.DataFrame(index=vec.index, data={"vec": vec}))
+                _set_colors_for_categorical_obs(colorer, "vec", palette="tab20")
+                colors = colorer.uns["vec_colors"]
+                color_dict = dict(zip(vec.cat.categories, colors))
+                color_dict.update({np.nan: "#808080ff"})
+            else:
+                color_dict = self.model.adata.uns[vec_color_name]
         else:
-            color_dict = self.model.adata.uns[vec_color_name]
+            df = layer.metadata["_columns_df"]
+            if vec_color_name not in df.columns:
+                colorer = AnnData(shape=(len(vec), 0), obs=pd.DataFrame(index=vec.index, data={"vec": vec}))
+                _set_colors_for_categorical_obs(colorer, "vec", palette="tab20")
+                colors = colorer.uns["vec_colors"]
+                color_dict = dict(zip(vec.cat.categories, colors))
+                color_dict.update({np.nan: "#808080ff"})
+                color_column = vec.apply(lambda x: color_dict[x])
+                df[vec_color_name] = color_column
+            else:
+                unique_colors = df[[vec.name, vec_color_name]].drop_duplicates()
+                unique_colors.set_index(vec.name, inplace=True)
+                if not unique_colors.index.is_unique:
+                    raise ValueError(
+                        f"The {vec_color_name} column must have unique values for the each {vec.name} level. Found:\n"
+                        f"{unique_colors}"
+                    )
+                color_dict = unique_colors.to_dict()["genes_color"]
 
         if self.model.instance_key is not None and self.model.instance_key == vec.index.name:
             merge_df = pd.merge(

tests/test_view.py

@@ -4,6 +4,7 @@
 from napari.utils.events import EventedList
 from napari_spatialdata._sdata_widgets import SdataWidget
 from napari_spatialdata._view import QtAdataViewWidget
+from napari_spatialdata.utils._test_utils import click_list_widget_item, get_center_pos_listitem
 from spatialdata.datasets import blobs
 
 
@@ -33,3 +34,32 @@ def test_channel_slider_images(qtbot, make_napari_viewer: any, widget: Any, n_ch
     qtbot.wait(50)  # wait for a short time to simulate user interaction
 
     viewer.close()
+
+
+@pytest.mark.parametrize("widget", [QtAdataViewWidget])
+@pytest.mark.parametrize("column", ["genes", "instance_id"])
+def test_plot_dataframe_annotation_on_points(qtbot, make_napari_viewer: any, widget: Any, column: str):
+    sdata_blobs = blobs()
+    viewer = make_napari_viewer()
+    sdata_widget = SdataWidget(viewer, EventedList([sdata_blobs]))
+
+    viewer.window.add_dock_widget(sdata_widget, name="SpatialData")
+
+    # init the adata view widget
+    widget = widget(viewer)
+
+    sdata_widget.viewer_model.add_sdata_points(sdata_blobs, "blobs_points", "global", False)
+
+    # plot dataframe annotations on the points
+    center_pos = get_center_pos_listitem(widget.dataframe_columns_widget, "instance_id")
+    # TODO: the double click doesn't trigger the signal, so below we are calling _onAction directly (looking at a
+    #  screenshot of the qtbot, the interface shows to be correctly clicked)
+    click_list_widget_item(
+        qtbot,
+        widget=widget.dataframe_columns_widget,
+        position=center_pos,
+        wait_signal="currentItemChanged",
+        click="double",
+    )
+    widget.dataframe_columns_widget._onAction([column])
+    viewer.close()