[Ready for Review] Pandas 2.0 compatibility (#569)

fixes #512

Authors:
  - Ajay Thorve (https://github.com/AjayThorve)

Approvers:
  - GALI PREM SAGAR (https://github.com/galipremsagar)
  - Jake Awe (https://github.com/AyodeAwe)
  - Allan (https://github.com/exactlyallan)

URL: #569

conda/environments/all_cuda-118_arch-x86_64.yaml

@@ -24,7 +24,7 @@ dependencies:
 - jupyter_sphinx
 - libwebp
 - nbsphinx
-- nodejs>=14
+- nodejs>=18
 - notebook>=0.5.0
 - numba>=0.57
 - numpy>=1.21

conda/environments/all_cuda-120_arch-x86_64.yaml

@@ -23,7 +23,7 @@ dependencies:
 - jupyter_sphinx
 - libwebp
 - nbsphinx
-- nodejs>=14
+- nodejs>=18
 - notebook>=0.5.0
 - numba>=0.57
 - numpy>=1.21

dependencies.yaml

@@ -173,7 +173,7 @@ dependencies:
       - output_types: conda
         packages:
           - cupy>=12.0.0
-          - nodejs>=14
+          - nodejs>=18
           - libwebp
       - output_types: [requirements, pyproject]
         packages:

python/cuxfilter/assets/geojson_mapper.py

@@ -1,6 +1,7 @@
 from urllib.request import urlopen
 import geopandas as gpd
 import pandas as pd
+from io import StringIO
 
 
 def geo_json_mapper(
@@ -16,7 +17,7 @@ def geo_json_mapper(
 
     temp_gpd_df = gpd.read_file(data).to_crs(epsg=projection)
 
-    df = pd.read_json(temp_gpd_df.to_json())
+    df = pd.read_json(StringIO(temp_gpd_df.to_json()))
 
     x_range = (
         temp_gpd_df.geometry.bounds.minx.min(),

python/cuxfilter/charts/bokeh/plots/bar.py

@@ -111,7 +111,7 @@ def calculate_source(self, data=None):
         Output:
             cudf.DataFrame
         """
-        data = data or self.source
+        data = self.source if data is None else data
         return calc_groupby(self, data)
 
     def reload_chart(self, data):

python/cuxfilter/charts/bokeh/plots/histogram.py

@@ -112,7 +112,7 @@ def calculate_source(self, data=None):
         Output:
             cudf.DataFrame
         """
-        data = data or self.source
+        data = self.source if data is None else data
         return calc_value_counts(
             data[self.x],
             self.stride,

python/cuxfilter/charts/core/aggregate/core_aggregate.py

@@ -187,12 +187,12 @@ def reset_callback(resetting):
     def get_box_select_callback(self, dashboard_cls):
         def cb(bounds, x_selection, y_selection):
             self.box_selected_range, self.selected_indices = None, None
-            if type(x_selection) == tuple:
+            if isinstance(x_selection, tuple):
                 self.box_selected_range = {
                     self.x + "_min": x_selection[0],
                     self.x + "_max": x_selection[1],
                 }
-            elif type(x_selection) == list:
+            elif isinstance(x_selection, list):
                 self.selected_indices = (
                     dashboard_cls._cuxfilter_df.data[self.x]
                     .isin(x_selection)

python/cuxfilter/charts/core/non_aggregate/utils.py

@@ -10,4 +10,4 @@ def point_in_polygon(df, x, y, polygons):
     polygons = cuspatial.GeoSeries(
         gpd.GeoSeries(Polygon(polygons)), index=["selection"]
     )
-    return cuspatial.point_in_polygon(points, polygons)
+    return cuspatial.point_in_polygon(points, polygons).selection

python/cuxfilter/charts/deckgl/bindings/panel_deck.py

@@ -45,7 +45,7 @@ class PanelDeck(param.Parameterized):
 
     @property
     def valid_indices(self):
-        return self.indices.intersection(self.data.index)
+        return list(self.indices.intersection(self.data.index))
 
     def get_tooltip_html(self):
         """
@@ -110,7 +110,8 @@ def click_event(self):
             temp_colors = self.colors.copy()
             if len(self.indices) > 0:
                 temp_colors.loc[
-                    set(self.data.index) - self.indices, self.colors.columns
+                    list(set(self.data.index) - self.indices),
+                    self.colors.columns,
                 ] = self.default_color
             self.data[self.colors.columns] = temp_colors
         self._update_layer_data(self.data)

python/cuxfilter/charts/panel_widgets/plots.py

@@ -1,3 +1,4 @@
+import datetime
 from ..core import BaseWidget
 from ..core.aggregate import BaseNumberChart
 from ..constants import (
@@ -160,7 +161,10 @@ def compute_query_dict(self, query_str_dict, query_local_variables_dict):
             reference to dashboard.__cls__.query_dict
         """
         if self.chart.value != (self.chart.start, self.chart.end):
-            min_temp, max_temp = self.chart.value
+            min_temp, max_temp = (
+                datetime.datetime.fromordinal(x.toordinal())
+                for x in self.chart.value
+            )
             query = f"@{self.x}_min<={self.x}<=@{self.x}_max"
             query_str_dict[self.name] = query
             query_local_variables_dict[self.x + "_min"] = min_temp
@@ -328,9 +332,7 @@ def initiate_chart(self, dashboard_cls):
         self.min_value, self.max_value = get_min_max(
             dashboard_cls._cuxfilter_df.data, self.x
         )
-        self.source = dashboard_cls._cuxfilter_df.data[self.x].reset_index(
-            drop=True
-        )
+        self.source = dashboard_cls._cuxfilter_df.data[self.x]
         self.calc_list_of_values(dashboard_cls._cuxfilter_df.data)
         self.generate_widget()
         self.add_events(dashboard_cls)
@@ -396,17 +398,20 @@ def compute_query_dict(self, query_str_dict, query_local_variables_dict):
         if len(self.chart.value) == 0:
             query_str_dict.pop(self.name, None)
         else:
-            df_module = (
-                cudf if isinstance(self.source, cudf.Series) else dask_cudf
-            )
 
-            if self.source.dtype == "object":
-                query_str_dict[self.name] = df_module.DataFrame(
-                    self.source.str.contains("|".join(self.chart.value))
+            def filter_source(s: cudf.Series, v: list):
+                if s.dtype == "object":
+                    return s.str.contains("|".join(v))
+                else:
+                    return s.isin(v)
+
+            if isinstance(self.source, dask_cudf.Series):
+                query_str_dict[self.name] = self.source.map_partitions(
+                    filter_source, self.chart.value
                 )
-            else:
-                query_str_dict[self.name] = df_module.DataFrame(
-                    self.source.isin(self.chart.value)
+            elif isinstance(self.source, cudf.Series):
+                query_str_dict[self.name] = filter_source(
+                    self.source, self.chart.value
                 )
 
     def apply_theme(self, theme):

python/cuxfilter/dashboard.py

@@ -119,7 +119,7 @@ def queried_indices(self):
         """
         Read-only propery queried_indices returns a merged index
         of all queried index columns present in self._query_str_dict
-        as a cudf.Series.
+        as a `cudf.Series` or `dask_cudf.Series`.
 
         Returns None if no index columns are present.
 
@@ -132,13 +132,20 @@ def queried_indices(self):
             else dask_cudf
         )
         selected_indices = {
-            key: value
+            key: value.reset_index(drop=True)
             for (key, value) in self._query_str_dict.items()
-            if type(value) in [cudf.DataFrame, dask_cudf.DataFrame]
+            if type(value)
+            in [
+                cudf.DataFrame,
+                dask_cudf.DataFrame,
+                cudf.Series,
+                dask_cudf.Series,
+            ]
         }
         if len(selected_indices) > 0:
             result = (
                 df_module.concat(list(selected_indices.values()), axis=1)
+                .set_index(self._cuxfilter_df.data.index)
                 .fillna(False)
                 .all(axis=1)
             )

python/cuxfilter/tests/charts/deckgl/test_deckgl.py

@@ -81,6 +81,7 @@ def test_init(self, cux_df):
         )
 
         assert choropleth3d_chart.chart.indices == set()
+        assert choropleth3d_chart.chart.valid_indices == list()
 
         assert choropleth3d_chart.chart.multi_select is False