def SpatialJoinQueryFlat(cls, spatialRDD: SpatialRDD, queryRDD: SpatialRDD,
useIndex: bool,
considerBoundaryIntersection: bool) -> RDD:
"""
Function takes SpatialRDD and other SpatialRDD and based on two parameters
- useIndex
- considerBoundaryIntersection
creates RDD with result of Spatial Join operation. It Returns RDD[GeoData, GeoData]
:param spatialRDD: SpatialRDD
:param queryRDD: SpatialRDD
:param useIndex: bool
:param considerBoundaryIntersection: bool
:return: RDD
>> spatial_join_result = JoinQuery.SpatialJoinQueryFlat(
>> spatialRDD, queryRDD, useIndex, considerBoundaryIntersection
>> )
>> spatial_join_result.collect()
[[GeoData(Polygon, ), GeoData()], [GeoData(), GeoData()], [GeoData(), GeoData()]]
"""
pair_rdd = JoinQueryRaw.SpatialJoinQueryFlat(
spatialRDD, queryRDD, useIndex, considerBoundaryIntersection)
return pair_rdd.to_rdd()
def test_spatial_join_query_flat_to_df(self):
poi_point_rdd = WktReader.readToGeometryRDD(self.sc, bank_csv_path, 1,
False, False)
areas_polygon_rdd = WktReader.readToGeometryRDD(
self.sc, areas_csv_path, 1, False, False)
poi_point_rdd.analyze()
areas_polygon_rdd.analyze()
poi_point_rdd.spatialPartitioning(GridType.QUADTREE)
areas_polygon_rdd.spatialPartitioning(poi_point_rdd.getPartitioner())
jvm_sedona_rdd = JoinQueryRaw.SpatialJoinQueryFlat(
poi_point_rdd, areas_polygon_rdd, False, True)
pois_within_areas_with_default_column_names = Adapter.toDf(
jvm_sedona_rdd, self.spark)
assert pois_within_areas_with_default_column_names.count() == 5
pois_within_areas_with_passed_column_names = Adapter.toDf(
jvm_sedona_rdd, ["area_id", "area_name"], ["poi_id", "poi_name"],
self.spark)
assert pois_within_areas_with_passed_column_names.count() == 5
assert pois_within_areas_with_passed_column_names.columns == [
"leftgeometry", "area_id", "area_name", "rightgeometry", "poi_id",
"poi_name"
]
assert pois_within_areas_with_default_column_names.schema == StructType(
[
StructField("leftgeometry", GeometryType()),
StructField("rightgeometry", GeometryType()),
])
left_geometries_raw = pois_within_areas_with_default_column_names. \
selectExpr("ST_AsText(leftgeometry)"). \
collect()
left_geometries = self.__row_to_list(left_geometries_raw)
right_geometries_raw = pois_within_areas_with_default_column_names. \
selectExpr("ST_AsText(rightgeometry)"). \
collect()
right_geometries = self.__row_to_list(right_geometries_raw)
# Ignore the ordering of these
assert set(geom[0] for geom in left_geometries) == set([
'POLYGON ((0 4, -3 3, -8 6, -6 8, -2 9, 0 4))',
'POLYGON ((10 3, 10 6, 14 6, 14 3, 10 3))',
'POLYGON ((2 2, 2 4, 3 5, 7 5, 9 3, 8 1, 4 1, 2 2))',
'POLYGON ((-1 -1, -1 -3, -2 -5, -6 -8, -5 -2, -3 -2, -1 -1))',
'POLYGON ((-1 -1, -1 -3, -2 -5, -6 -8, -5 -2, -3 -2, -1 -1))'
])
assert set(geom[0] for geom in right_geometries) == set([
'POINT (-3 5)', 'POINT (11 5)', 'POINT (4 3)', 'POINT (-1 -1)',
'POINT (-4 -5)'
])
def test_spatial_join_to_spatial_rdd(self):
poi_point_rdd = WktReader.readToGeometryRDD(self.sc, bank_csv_path, 1, False, False)
areas_polygon_rdd = WktReader.readToGeometryRDD(self.sc, areas_csv_path, 1, False, False)
poi_point_rdd.analyze()
areas_polygon_rdd.analyze()
poi_point_rdd.spatialPartitioning(GridType.QUADTREE)
areas_polygon_rdd.spatialPartitioning(poi_point_rdd.getPartitioner())
jvm_sedona_rdd = JoinQueryRaw.spatialJoin(poi_point_rdd, areas_polygon_rdd, JoinParams(considerBoundaryIntersection=True))
sedona_rdd = jvm_sedona_rdd.to_rdd().collect()
assert sedona_rdd.__len__() == 5
def test_spatial_join_query_flat_to_df(self):
poi_point_rdd = WktReader.readToGeometryRDD(self.sc, bank_csv_path, 1, False, False)
areas_polygon_rdd = WktReader.readToGeometryRDD(self.sc, areas_csv_path, 1, False, False)
poi_point_rdd.analyze()
areas_polygon_rdd.analyze()
poi_point_rdd.spatialPartitioning(GridType.QUADTREE)
areas_polygon_rdd.spatialPartitioning(poi_point_rdd.getPartitioner())
jvm_sedona_rdd = JoinQueryRaw.SpatialJoinQueryFlat(
poi_point_rdd, areas_polygon_rdd, False, True)
assert jvm_sedona_rdd.to_rdd().collect().__len__() == 5
def test_distance_join_query_flat_to_df(self):
poi_point_rdd = WktReader.readToGeometryRDD(self.sc, bank_csv_path, 1, False, False)
circle_rdd = CircleRDD(poi_point_rdd, 2.0)
circle_rdd.analyze()
poi_point_rdd.analyze()
poi_point_rdd.spatialPartitioning(GridType.QUADTREE)
circle_rdd.spatialPartitioning(poi_point_rdd.getPartitioner())
jvm_sedona_rdd = JoinQueryRaw.DistanceJoinQueryFlat(poi_point_rdd, circle_rdd, False, True)
assert jvm_sedona_rdd.to_rdd().collect().__len__() == 10
def spatialJoin(cls, queryWindowRDD: SpatialRDD, objectRDD: SpatialRDD,
joinParams: JoinParams) -> RDD:
"""
:param queryWindowRDD: SpatialRDD
:param objectRDD: SpatialRDD
:param joinParams: JoinParams
:return:
"""
pair_rdd = JoinQueryRaw.spatialJoin(queryWindowRDD, objectRDD,
joinParams)
return pair_rdd.to_rdd()
def DistanceJoinQuery(cls, spatialRDD: SpatialRDD, queryRDD: SpatialRDD,
useIndex: bool,
considerBoundaryIntersection: bool) -> RDD:
"""
:param spatialRDD: SpatialRDD
:param queryRDD: SpatialRDD
:param useIndex: bool
:param considerBoundaryIntersection: bool
:return:
"""
pair_rdd = JoinQueryRaw.DistanceJoinQuery(
spatialRDD, queryRDD, useIndex, considerBoundaryIntersection)
return pair_rdd.to_rdd()
def DistanceJoinQueryFlat(cls, spatialRDD: SpatialRDD,
queryRDD: SpatialRDD, useIndex: bool,
considerBoundaryIntersection: bool) -> RDD:
"""
:param spatialRDD: SpatialRDD
:param queryRDD: SpatialRDD
:param useIndex: bool
:param considerBoundaryIntersection: bool
>> spatial_rdd =
>> query_rdd =
>> spatial_join_result = JoinQuery.DistanceJoinQueryFlat(spatial_rdd, query_rdd, True, True)
>> spatial_join_result.collect()
[GeoData(), GeoData()]
:return:
"""
pair_rdd = JoinQueryRaw.DistanceJoinQueryFlat(
spatialRDD, queryRDD, useIndex, considerBoundaryIntersection)
return pair_rdd.to_rdd()
def test_distance_join_query_flat_to_df(self):
poi_point_rdd = WktReader.readToGeometryRDD(self.sc, bank_csv_path, 1,
False, False)
circle_rdd = CircleRDD(poi_point_rdd, 2.0)
circle_rdd.analyze()
poi_point_rdd.analyze()
poi_point_rdd.spatialPartitioning(GridType.QUADTREE)
circle_rdd.spatialPartitioning(poi_point_rdd.getPartitioner())
jvm_sedona_rdd = JoinQueryRaw.DistanceJoinQueryFlat(
poi_point_rdd, circle_rdd, False, True)
df_sedona_rdd = Adapter.toDf(jvm_sedona_rdd,
["poi_from_id", "poi_from_name"],
["poi_to_id", "poi_to_name"], self.spark)
assert df_sedona_rdd.count() == 10
assert df_sedona_rdd.columns == [
"leftgeometry", "poi_from_id", "poi_from_name", "rightgeometry",
"poi_to_id", "poi_to_name"
]
def test_spatial_join_to_df(self):
poi_point_rdd = WktReader.readToGeometryRDD(self.sc, bank_csv_path, 1,
False, False)
areas_polygon_rdd = WktReader.readToGeometryRDD(
self.sc, areas_csv_path, 1, False, False)
poi_point_rdd.analyze()
areas_polygon_rdd.analyze()
poi_point_rdd.spatialPartitioning(GridType.QUADTREE)
areas_polygon_rdd.spatialPartitioning(poi_point_rdd.getPartitioner())
jvm_sedona_rdd = JoinQueryRaw.spatialJoin(poi_point_rdd,
areas_polygon_rdd,
JoinParams())
sedona_df = Adapter.toDf(jvm_sedona_rdd, ["area_id", "area_name"],
["poi_id", "poi_name"], self.spark)
assert sedona_df.count() == 5
assert sedona_df.columns == [
"leftgeometry", "area_id", "area_name", "rightgeometry", "poi_id",
"poi_name"
]
