Make polygon fixture data valid

python/cuspatial/cuspatial/testing/helpers.py

@@ -0,0 +1,15 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+from itertools import chain
+
+from shapely import get_coordinates
+
+
+def geometry_to_coords(geom, geom_types):
+    points_list = geom[geom.apply(lambda x: isinstance(x, geom_types))]
+    # flatten multigeometries, then geometries, then coordinates
+    points = list(chain(points_list.apply(get_coordinates)))
+    coords_list = list(chain(*points))
+    xy = list(chain(*coords_list))
+    x = xy[::2]
+    y = xy[1::2]
+    return xy, x, y

python/cuspatial/cuspatial/tests/conftest.py

@@ -19,48 +19,46 @@
 @pytest.fixture
 def gs():
     g0 = Point(-1, 0)
-    g1 = MultiPoint(((1, 2), (3, 4)))
-    g2 = MultiPoint(((5, 6), (7, 8)))
-    g3 = Point(9, 10)
+    g1 = Point(9, 10)
+    g2 = MultiPoint(((1, 2), (3, 4)))
+    g3 = MultiPoint(((5, 6), (7, 8)))
     g4 = LineString(((11, 12), (13, 14)))
     g5 = MultiLineString((((15, 16), (17, 18)), ((19, 20), (21, 22))))
     g6 = MultiLineString((((23, 24), (25, 26)), ((27, 28), (29, 30))))
     g7 = LineString(((31, 32), (33, 34)))
     g8 = Polygon(
-        ((35, 36), (37, 38), (39, 40), (41, 42)),
+        ((35, 36), (38, 36), (41, 39), (41, 42)),
     )
-    # TODO: g9, g10, g11 are invalid
-    # https://github.com/libgeos/geos/issues/1177
     g9 = MultiPolygon(
         [
             (
-                ((43, 44), (45, 46), (47, 48)),
-                [((49, 50), (51, 52), (53, 54))],
+                ((43, 44), (48, 44), (47, 48)),
+                [((45, 45), (46, 46), (47, 45))],
             ),
             (
-                ((55, 56), (57, 58), (59, 60)),
-                [((61, 62), (63, 64), (65, 66))],
+                ((55, 56), (60, 56), (59, 60)),
+                [((57, 57), (58, 58), (59, 57))],
             ),
         ]
     )
     g10 = MultiPolygon(
         [
             (
-                ((67, 68), (69, 70), (71, 72)),
-                [((73, 74), (75, 76), (77, 78))],
+                ((67, 68), (72, 68), (71, 72)),
+                [((69, 69), (70, 70), (71, 69))],
             ),
             (
-                ((79, 80), (81, 82), (83, 84)),
+                ((79, 80), (90, 82), (83, 90)),
                 [
-                    ((85, 86), (87, 88), (89, 90)),
-                    ((91, 92), (93, 94), (95, 96)),
+                    ((80, 81), (82, 84), (84, 82)),
+                    ((85, 85), (88, 82), (86, 82)),
                 ],
             ),
         ]
     )
     g11 = Polygon(
         ((97, 98), (99, 101), (102, 103), (101, 108)),
-        [((106, 107), (108, 109), (110, 111), (113, 108))],
+        [((99, 102), (100, 103), (101, 103), (100, 102))],
     )
     gs = gpd.GeoSeries([g0, g1, g2, g3, g4, g5, g6, g7, g8, g9, g10, g11])
     return gs
@@ -72,8 +70,10 @@ def gpdf(gs):
     random_col = int_col
     np.random.shuffle(random_col)
     str_col = [str(x) for x in int_col]
-    key_col = np.repeat(np.arange(4), len(int_col) // 4)
+    key_col = np.repeat(np.arange(4), (len(int_col) // 4) + 1)
+    key_col = key_col[: len(int_col)]
     np.random.shuffle(key_col)
+
     result = gpd.GeoDataFrame(
         {
             "geometry": gs,
@@ -87,65 +87,6 @@ def gpdf(gs):
     return result
 
 
-@pytest.fixture
-def polys():
-    return np.array(
-        (
-            (35, 36),
-            (37, 38),
-            (39, 40),
-            (41, 42),
-            (35, 36),
-            (43, 44),
-            (45, 46),
-            (47, 48),
-            (43, 44),
-            (49, 50),
-            (51, 52),
-            (53, 54),
-            (49, 50),
-            (55, 56),
-            (57, 58),
-            (59, 60),
-            (55, 56),
-            (61, 62),
-            (63, 64),
-            (65, 66),
-            (61, 62),
-            (67, 68),
-            (69, 70),
-            (71, 72),
-            (67, 68),
-            (73, 74),
-            (75, 76),
-            (77, 78),
-            (73, 74),
-            (79, 80),
-            (81, 82),
-            (83, 84),
-            (79, 80),
-            (85, 86),
-            (87, 88),
-            (89, 90),
-            (85, 86),
-            (91, 92),
-            (93, 94),
-            (95, 96),
-            (91, 92),
-            (97, 98),
-            (99, 101),
-            (102, 103),
-            (101, 108),
-            (97, 98),
-            (106, 107),
-            (108, 109),
-            (110, 111),
-            (113, 108),
-            (106, 107),
-        )
-    )
-
-
 @pytest.fixture
 def gs_sorted(gs):
     result = pd.concat(

python/cuspatial/cuspatial/tests/test_cudf_integration.py

@@ -2,17 +2,10 @@
 import geopandas as gpd
 import numpy as np
 import pandas as pd
-import pytest
 
 import cuspatial
 
-reason = (
-    "gs fixture contains invalid Polygons/MultiPolygons: "
-    "https://github.com/libgeos/geos/issues/1177"
-)
 
-
-@pytest.mark.xfail(reason=reason)
 def test_sort_index_series(gs):
     gs.index = np.random.permutation(len(gs))
     cugs = cuspatial.from_geopandas(gs)
@@ -21,7 +14,6 @@ def test_sort_index_series(gs):
     gpd.testing.assert_geoseries_equal(got, expected)
 
 
-@pytest.mark.xfail(reason=reason)
 def test_sort_index_dataframe(gpdf):
     gpdf.index = np.random.permutation(len(gpdf))
     cugpdf = cuspatial.from_geopandas(gpdf)
@@ -30,7 +22,6 @@ def test_sort_index_dataframe(gpdf):
     gpd.testing.assert_geodataframe_equal(got, expected)
 
 
-@pytest.mark.xfail(reason=reason)
 def test_sort_values(gpdf):
     cugpdf = cuspatial.from_geopandas(gpdf)
     expected = gpdf.sort_values("random")

python/cuspatial/cuspatial/tests/test_from_geopandas.py

@@ -36,15 +36,8 @@ def test_dataframe_column_access(gs):
 
 def test_from_geoseries_complex(gs):
     cugs = cuspatial.from_geopandas(gs)
-    assert cugs.points.xy.sum() == 18
-    assert cugs.lines.xy.sum() == 540
-    assert cugs.multipoints.xy.sum() == 36
-    assert cugs.polygons.xy.sum() == 7436
-    assert cugs._column.polygons._column.base_children[0].sum() == 15
-    assert (
-        cugs._column.polygons._column.base_children[1].base_children[0].sum()
-        == 38
-    )
+    gs_roundtrip = cugs.to_geopandas()
+    gpd.testing.assert_geoseries_equal(gs_roundtrip, gs)
 
 
 def test_from_geopandas_point():

python/cuspatial/cuspatial/tests/test_geodataframe.py

@@ -19,6 +19,7 @@
 import cudf
 
 import cuspatial
+from cuspatial.testing.helpers import geometry_to_coords
 
 np.random.seed(0)
 
@@ -114,7 +115,7 @@ def test_type_persistence(gpdf):
     assert type(cugpdf["geometry"]) is cuspatial.GeoSeries
 
 
-def test_interleaved_point(gpdf, polys):
+def test_interleaved_point(gpdf):
     cugpdf = cuspatial.from_geopandas(gpdf)
     cugs = cugpdf["geometry"]
     gs = gpdf["geometry"]
@@ -128,7 +129,7 @@ def test_interleaved_point(gpdf, polys):
     )
 
 
-def test_interleaved_multipoint(gpdf, polys):
+def test_interleaved_multipoint(gpdf):
     cugpdf = cuspatial.from_geopandas(gpdf)
     cugs = cugpdf["geometry"]
     gs = gpdf["geometry"]
@@ -156,7 +157,7 @@ def test_interleaved_multipoint(gpdf, polys):
     )
 
 
-def test_interleaved_lines(gpdf, polys):
+def test_interleaved_lines(gpdf):
     cugpdf = cuspatial.from_geopandas(gpdf)
     cugs = cugpdf["geometry"]
     cudf.testing.assert_series_equal(
@@ -175,16 +176,19 @@ def test_interleaved_lines(gpdf, polys):
     )
 
 
-def test_interleaved_polygons(gpdf, polys):
+def test_interleaved_polygons(gpdf):
     cugpdf = cuspatial.from_geopandas(gpdf)
     cugs = cugpdf["geometry"]
+    gs = gpdf["geometry"]
+    xy, x, y = geometry_to_coords(gs, (Polygon, MultiPolygon))
+
     cudf.testing.assert_series_equal(
         cudf.Series.from_arrow(cugs.polygons.x.to_arrow()),
-        cudf.Series(polys[:, 0], dtype="float64"),
+        cudf.Series(x, dtype="float64"),
     )
     cudf.testing.assert_series_equal(
         cudf.Series.from_arrow(cugs.polygons.y.to_arrow()),
-        cudf.Series(polys[:, 1], dtype="float64"),
+        cudf.Series(y, dtype="float64"),
     )