Eliminate DataFrame copies and double data loading in datashader path

Two additional performance fixes on top of the datashader speedups:

1. Replace .assign() + .rename() with direct column assignment when
   attaching the color column to the transformed element. Avoids two
   full DataFrame copies (~320MB saved for 10M points).

2. Add preloaded_color_data parameter to _set_color_source_vec so
   _render_points can pass already-loaded color data from get_values()
   instead of triggering a redundant second load from the table.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: timtreis

src/spatialdata_plot/pl/render.py

@@ -768,6 +768,10 @@ def _render_points(
     # from the registered points (see above) avoids duplicate-origin ambiguities.
     color_table_name = table_name
 
+    # When color was already loaded from a table (line 690), pass it directly
+    # to avoid a redundant get_values() call inside _set_color_source_vec.
+    _preloaded = points_pd_with_color[col_for_color] if added_color_from_table and col_for_color else None
+
     color_source_vector, color_vector, _ = _set_color_source_vec(
         sdata=sdata_filt,
         element=color_element,
@@ -781,6 +785,7 @@ def _render_points(
         table_name=color_table_name,
         render_type="points",
         coordinate_system=coordinate_system,
+        preloaded_color_data=_preloaded,
     )
 
     if added_color_from_table and col_for_color is not None:
@@ -856,7 +861,7 @@ def _render_points(
                     if isinstance(color_source_vector, pd.Series)
                     else pd.Series(color_source_vector, index=series_index)
                 )
-                transformed_element = transformed_element.assign(col_for_color=source_series)
+                transformed_element[col_for_color] = source_series
             else:
                 if isinstance(color_vector, dd.Series):
                     color_vector = color_vector.compute()
@@ -865,8 +870,7 @@ def _render_points(
                     if isinstance(color_vector, pd.Series)
                     else pd.Series(color_vector, index=series_index)
                 )
-                transformed_element = transformed_element.assign(col_for_color=color_series)
-            transformed_element = transformed_element.rename(columns={"col_for_color": col_for_color})
+                transformed_element[col_for_color] = color_series
 
         color_dtype = transformed_element[col_for_color].dtype if col_for_color is not None else None
         color_by_categorical = col_for_color is not None and (

src/spatialdata_plot/pl/utils.py

@@ -1019,6 +1019,7 @@ def _set_color_source_vec(
     table_layer: str | None = None,
     render_type: Literal["points", "labels"] | None = None,
     coordinate_system: str | None = None,
+    preloaded_color_data: pd.Series | None = None,
 ) -> tuple[ArrayLike | pd.Series | None, ArrayLike, bool]:
     if value_to_plot is None and element is not None:
         color = np.full(len(element), na_color.get_hex_with_alpha())
@@ -1046,13 +1047,16 @@ def _set_color_source_vec(
                 element_name=element_name,
                 table_name=table_name,
             )
-        color_source_vector = get_values(
-            value_key=value_to_plot,
-            sdata=sdata,
-            element_name=element_name,
-            table_name=table_name,
-            table_layer=table_layer,
-        )[value_to_plot]
+        if preloaded_color_data is not None:
+            color_source_vector = preloaded_color_data
+        else:
+            color_source_vector = get_values(
+                value_key=value_to_plot,
+                sdata=sdata,
+                element_name=element_name,
+                table_name=table_name,
+                table_layer=table_layer,
+            )[value_to_plot]
 
         color_series = (
             color_source_vector if isinstance(color_source_vector, pd.Series) else pd.Series(color_source_vector)