def dynamic_rtree_int(self, query_rdd, num_partitions, use_legacy_apis,
grid_type, index_type, expected_count):
spatial_rdd = self.create_point_rdd(input_location, splitter,
num_partitions)
self.partition_rdds(query_rdd, spatial_rdd, grid_type, use_legacy_apis)
join_params = JoinParams(True, index_type, JoinBuildSide.LEFT)
results = JoinQuery.spatialJoin(query_rdd, spatial_rdd,
join_params).collect()
self.sanity_check_flat_join_results(results)
assert expected_count == results.__len__()
def test_dynamic_index_int(self, num_partitions, use_legacy_apis, grid_type, index_type, intersects):
query_rdd = self.create_polygon_rdd(query_polygon_set, splitter, num_partitions)
spatial_rdd = self.create_polygon_rdd(input_location, splitter, num_partitions)
self.partition_rdds(query_rdd, spatial_rdd, grid_type, use_legacy_apis)
join_params = JoinParams(intersects, index_type, JoinBuildSide.LEFT)
result = JoinQuery.spatialJoin(query_rdd, spatial_rdd, join_params).collect()
self.sanity_check_flat_join_results(result)
expected_count = self.get_expected_with_original_duplicates_count(intersects) \
if self.expect_to_preserve_original_duplicates(grid_type) else self.get_expected_count(intersects)
assert expected_count == result.__len__()
def test_spatial_join_query_and_build_index_on_polygons_on_the_fly(self):
query_window_rdd = PolygonRDD(self.sc, polygon_rdd_input_location,
polygon_rdd_start_offset,
polygon_rdd_end_offset,
polygon_rdd_splitter, True)
object_rdd = PointRDD(sparkContext=self.sc,
InputLocation=point_rdd_input_location,
Offset=point_rdd_offset,
splitter=point_rdd_splitter,
carryInputData=False)
object_rdd.analyze()
object_rdd.spatialPartitioning(join_query_partitionin_type)
query_window_rdd.spatialPartitioning(object_rdd.getPartitioner())
for i in range(each_query_loop_times):
join_params = JoinParams(False, polygon_rdd_index_type,
JoinBuildSide.LEFT)
resultSize = JoinQuery.spatialJoin(query_window_rdd, object_rdd,
join_params).count()