def test_st_exterior_ring(self):
polygon_df = create_simple_polygons_df(self.spark, 5)
additional_wkt = "POLYGON((0 0, 1 1, 1 2, 1 1, 0 0))"
additional_wkt_df = self.spark.createDataFrame(
[[wkt.loads(additional_wkt)]], self.geo_schema)
polygons_df = polygon_df.union(additional_wkt_df)
other_geometry_df = create_sample_lines_df(self.spark, 5).union(
create_sample_points_df(self.spark, 5))
linestring_df = polygons_df.selectExpr(
"ST_ExteriorRing(geom) as geom").filter("geom IS NOT NULL")
empty_df = other_geometry_df.selectExpr(
"ST_ExteriorRing(geom) as geom").filter("geom IS NOT NULL")
linestring_wkt = [
wkt_row[0] for wkt_row in linestring_df.selectExpr(
"ST_AsText(geom)").collect()
]
assert (linestring_wkt == [
"LINESTRING (0 0, 0 1, 1 1, 1 0, 0 0)",
"LINESTRING (0 0, 1 1, 1 2, 1 1, 0 0)"
])
assert (not empty_df.count())
def test_st_start_point(self):
point_df = create_sample_points_df(self.spark, 5)
polygon_df = create_sample_polygons_df(self.spark, 5)
linestring_df = create_sample_lines_df(self.spark, 5)
expected_points = [
"POINT (-112.506968 45.98186)", "POINT (-112.519856 45.983586)",
"POINT (-112.504872 45.919281)", "POINT (-112.574945 45.987772)",
"POINT (-112.520691 42.912313)"
]
points = point_df.selectExpr("ST_StartPoint(geom) as geom").filter(
"geom IS NOT NULL")
polygons = polygon_df.selectExpr("ST_StartPoint(geom) as geom").filter(
"geom IS NOT NULL")
linestrings = linestring_df.selectExpr(
"ST_StartPoint(geom) as geom").filter("geom IS NOT NULL")
assert ([line[0] for line in linestrings.collect()
] == [wkt.loads(el) for el in expected_points])
assert (not points.count())
assert (not polygons.count())
def test_st_end_point(self):
linestring_dataframe = create_sample_lines_df(self.spark, 5)
other_geometry_dataframe = create_sample_points_df(self.spark, 5). \
union(create_sample_points_df(self.spark, 5))
point_data_frame = linestring_dataframe.selectExpr("ST_EndPoint(geom) as geom"). \
filter("geom IS NOT NULL")
expected_ending_points = [
"POINT (-112.504872 45.98186)", "POINT (-112.506968 45.983586)",
"POINT (-112.41643 45.919281)", "POINT (-112.519856 45.987772)",
"POINT (-112.442664 42.912313)"
]
empty_dataframe = other_geometry_dataframe.selectExpr("ST_EndPoint(geom) as geom"). \
filter("geom IS NOT NULL")
assert ([
wkt_row[0] for wkt_row in point_data_frame.selectExpr(
"ST_AsText(geom)").collect()
] == expected_ending_points)
assert (empty_dataframe.count() == 0)
def test_st_interior_ring_n(self):
polygon_df = self.__wkt_list_to_data_frame(
["POLYGON((0 0, 0 5, 5 5, 5 0, 0 0), (1 1, 2 1, 2 2, 1 2, 1 1), (1 3, 2 3, 2 4, 1 4, 1 3), (3 3, 4 3, 4 4, 3 4, 3 3))"]
)
other_geometry = create_sample_points_df(self.spark, 5).union(create_sample_lines_df(self.spark, 5))
wholes = [polygon_df.selectExpr(f"ST_InteriorRingN(geom, {i}) as geom").
selectExpr("ST_AsText(geom)").collect()[0][0]
for i in range(3)]
empty_df = other_geometry.selectExpr("ST_InteriorRingN(geom, 1) as geom").filter("geom IS NOT NULL")
assert(not empty_df.count())
assert(wholes == ["LINESTRING (1 1, 2 1, 2 2, 1 2, 1 1)",
"LINESTRING (1 3, 2 3, 2 4, 1 4, 1 3)",
"LINESTRING (3 3, 4 3, 4 4, 3 4, 3 3)"])
def test_st_y(self):
point_df = create_sample_points_df(self.spark, 5)
polygon_df = create_sample_polygons_df(self.spark, 5)
linestring_df = create_sample_lines_df(self.spark, 5)
points = point_df \
.selectExpr("ST_Y(geom)").collect()
polygons = polygon_df.selectExpr("ST_Y(geom) as y").filter("y IS NOT NULL")
linestrings = linestring_df.selectExpr("ST_Y(geom) as y").filter("y IS NOT NULL")
assert([point[0] for point in points] == [42.28787, 32.324142, 32.324142, 5.3324324, -88.331492])
assert(not linestrings.count())
assert(not polygons.count())
def test_st_x(self):
point_df = create_sample_points_df(self.spark, 5)
polygon_df = create_sample_polygons_df(self.spark, 5)
linestring_df = create_sample_lines_df(self.spark, 5)
points = point_df \
.selectExpr("ST_X(geom)").collect()
polygons = polygon_df.selectExpr("ST_X(geom) as x").filter("x IS NOT NULL")
linestrings = linestring_df.selectExpr("ST_X(geom) as x").filter("x IS NOT NULL")
assert([point[0] for point in points] == [-71.064544, -88.331492, 88.331492, 1.0453, 32.324142])
assert(not linestrings.count())
assert(not polygons.count())
def test_st_dump_points(self):
expected_points = [
"POINT (-112.506968 45.98186)", "POINT (-112.506968 45.983586)",
"POINT (-112.504872 45.983586)", "POINT (-112.504872 45.98186)",
"POINT (-71.064544 42.28787)", "POINT (0 0)", "POINT (0 1)",
"POINT (1 1)", "POINT (1 0)", "POINT (0 0)"
]
geometry_df = create_sample_lines_df(self.spark, 1) \
.union(create_sample_points_df(self.spark, 1)) \
.union(create_simple_polygons_df(self.spark, 1))
dumped_points = geometry_df.selectExpr("ST_DumpPoints(geom) as geom") \
.select(explode(col("geom")).alias("geom"))
assert (dumped_points.count() == 10)
collected_points = [
geom_row[0] for geom_row in dumped_points.selectExpr(
"ST_AsText(geom)").collect()
]
assert (collected_points == expected_points)