def test_spatial_knn_correctness(self):
point_rdd = PointRDD(self.sc, input_location, offset, splitter, True,
StorageLevel.MEMORY_ONLY, "epsg:4326",
"epsg:3005")
result_no_index = KNNQuery.SpatialKnnQuery(point_rdd, query_point,
top_k, False)
point_rdd.buildIndex(IndexType.RTREE, False)
result_with_index = KNNQuery.SpatialKnnQuery(point_rdd, query_point,
top_k, True)
sorted_result_no_index = sorted(
result_no_index,
key=lambda geo_data: distance_sorting_functions(
geo_data, query_point))
sorted_result_with_index = sorted(
result_with_index,
key=lambda geo_data: distance_sorting_functions(
geo_data, query_point))
difference = 0
for x in range(top_k):
difference += sorted_result_no_index[x].geom.distance(
sorted_result_with_index[x].geom)
assert difference == 0
def test_spatial_knn_query_using_index(self):
point_rdd = PointRDD(self.sc, input_location, offset, splitter, False)
point_rdd.buildIndex(IndexType.RTREE, False)
for i in range(self.loop_times):
result = KNNQuery.SpatialKnnQuery(point_rdd, self.query_point,
self.top_k, False)
assert result.__len__() > -1
assert result[0].getUserData() is not None
def test_build_index_without_set_grid(self):
spatial_rdd = PointRDD(sparkContext=self.sc,
InputLocation=input_location,
Offset=offset,
splitter=splitter,
carryInputData=True,
partitions=num_partitions,
newLevel=StorageLevel.MEMORY_ONLY)
spatial_rdd.buildIndex(IndexType.RTREE, False)
def test_knn_query_with_index(self):
object_rdd = PointRDD(sparkContext=self.sc,
InputLocation=point_rdd_input_location,
Offset=point_rdd_offset,
splitter=point_rdd_splitter,
carryInputData=False)
object_rdd.buildIndex(point_rdd_index_type, False)
for i in range(each_query_loop_times):
result = KNNQuery.SpatialKnnQuery(object_rdd, knn_query_point,
1000, True)
def test_spatial_knn_query_using_index(self):
point_rdd = PointRDD(self.sc, input_location, offset, splitter, True,
StorageLevel.MEMORY_ONLY, "epsg:4326",
"epsg:3005")
point_rdd.buildIndex(IndexType.RTREE, False)
for i in range(loop_times):
result = KNNQuery.SpatialKnnQuery(point_rdd, query_point, top_k,
False)
assert result.__len__() > 0
assert result[0].getUserData() is not None
def test_range_query_using_index(self):
object_rdd = PointRDD(sparkContext=self.sc,
InputLocation=point_rdd_input_location,
Offset=point_rdd_offset,
splitter=point_rdd_splitter,
carryInputData=False)
object_rdd.buildIndex(point_rdd_index_type, False)
for i in range(each_query_loop_times):
result_size = RangeQuery.SpatialRangeQuery(object_rdd,
range_query_window,
False, True).count
def test_empty_constructor(self):
spatial_rdd = PointRDD(sparkContext=self.sc,
InputLocation=input_location,
Offset=offset,
splitter=splitter,
carryInputData=True,
partitions=num_partitions,
newLevel=StorageLevel.MEMORY_ONLY)
spatial_rdd.buildIndex(IndexType.RTREE, False)
spatial_rdd_copy = PointRDD()
spatial_rdd_copy.rawJvmSpatialRDD = spatial_rdd.rawJvmSpatialRDD
spatial_rdd_copy.analyze()
def test_spatial_range_query_using_index(self):
spatial_rdd = PointRDD(self.sc, input_location, offset, splitter,
False)
spatial_rdd.buildIndex(IndexType.RTREE, False)
for i in range(self.loop_times):
result_size = RangeQuery.\
SpatialRangeQuery(spatial_rdd, self.query_envelope, False, False)\
.count()
assert result_size == 2830
assert RangeQuery.SpatialRangeQuery(
spatial_rdd, self.query_envelope, False, False).take(10)[1].\
getUserData() is not None
def test_crs_tranformed_spatial_range_query_using_index(self):
object_rdd = PointRDD(sparkContext=self.sc,
InputLocation=point_rdd_input_location,
Offset=point_rdd_offset,
splitter=point_rdd_splitter,
carryInputData=False,
newLevel=StorageLevel.DISK_ONLY,
sourceEpsgCRSCode="epsg:4326",
targetEpsgCode="epsg:3005")
object_rdd.buildIndex(point_rdd_index_type, False)
for i in range(each_query_loop_times):
result_size = RangeQuery.SpatialRangeQuery(object_rdd,
range_query_window,
False, True).count
def test_spatial_range_query_using_index(self):
spatial_rdd = PointRDD(self.sc, input_location, offset, splitter, True,
StorageLevel.MEMORY_ONLY, "epsg:4326",
"epsg:3005")
spatial_rdd.buildIndex(IndexType.RTREE, False)
for i in range(loop_times):
result_size = RangeQuery.SpatialRangeQuery(spatial_rdd,
query_envelope, False,
False).count()
assert result_size == 3127
assert RangeQuery.SpatialRangeQuery(
spatial_rdd, query_envelope, False,
False).take(10)[1].getUserData() is not None
def test_distance_join_query_using_index(self):
object_rdd = PointRDD(sparkContext=self.sc,
InputLocation=point_rdd_input_location,
Offset=point_rdd_offset,
splitter=point_rdd_splitter,
carryInputData=False)
query_window_rdd = CircleRDD(object_rdd, 0.1)
object_rdd.analyze()
object_rdd.spatialPartitioning(GridType.QUADTREE)
query_window_rdd.spatialPartitioning(object_rdd.getPartitioner())
object_rdd.buildIndex(IndexType.RTREE, True)
for i in range(each_query_loop_times):
result_size = JoinQuery.DistanceJoinQuery(object_rdd,
query_window_rdd, True,
True).count
def test_spatial_join_using_index(self):
query_window = 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.spatialPartitioning(object_rdd.getPartitioner())
object_rdd.buildIndex(point_rdd_index_type, True)
for i in range(each_query_loop_times):
result_size = JoinQuery.SpatialJoinQuery(object_rdd, query_window,
True, False).count()
def test_spatial_join_query_with_polygon_rdd_using_index(self):
query_rdd = PolygonRDD(self.sc, input_location_query_polygon, splitter,
True, num_partitions, StorageLevel.MEMORY_ONLY,
"epsg:4326", "epsg:3005")
spatial_rdd = PointRDD(self.sc, input_location, offset, splitter, True,
num_partitions, StorageLevel.MEMORY_ONLY,
"epsg:4326", "epsg:3005")
spatial_rdd.spatialPartitioning(grid_type)
spatial_rdd.buildIndex(IndexType.RTREE, True)
query_rdd.spatialPartitioning(spatial_rdd.grids)
result = JoinQuery.SpatialJoinQuery(spatial_rdd, query_rdd, False,
True).collect()
assert result[1][0].getUserData() is not None
for data in result:
if data[1].__len__() != 0:
for right_data in data[1]:
assert right_data.getUserData() is not None
