This is an automated email from the ASF dual-hosted git repository.
jiayu pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/sedona.git
The following commit(s) were added to refs/heads/master by this push:
new 30cc3b14ea [GH-2020] Geopandas.GeoSeries: Implement .length and
refactor process_geometry_column() (#2021)
30cc3b14ea is described below
commit 30cc3b14ea494c2717d502e83ea26841da73aa1c
Author: Peter Nguyen <[email protected]>
AuthorDate: Tue Jul 1 14:52:41 2025 -0700
[GH-2020] Geopandas.GeoSeries: Implement .length and refactor
process_geometry_column() (#2021)
* Refactor process_geometry_column to create a more flexible
query_geometry_column()
* Implement length()
---
python/sedona/geopandas/geoseries.py | 129 +++++++++++++++------
python/tests/geopandas/test_geoseries.py | 19 ++-
.../tests/geopandas/test_match_geopandas_series.py | 6 +-
3 files changed, 116 insertions(+), 38 deletions(-)
diff --git a/python/sedona/geopandas/geoseries.py
b/python/sedona/geopandas/geoseries.py
index 066cacb75a..4dc612d171 100644
--- a/python/sedona/geopandas/geoseries.py
+++ b/python/sedona/geopandas/geoseries.py
@@ -340,6 +340,7 @@ class GeoSeries(GeoFrame, pspd.Series):
) -> "GeoSeries":
"""
Helper method to process a single geometry column with a specified
operation.
+ This method wraps the _query_geometry_column method for simpler
convenient use.
Parameters
----------
@@ -360,45 +361,66 @@ class GeoSeries(GeoFrame, pspd.Series):
# Find the first column with BinaryType or GeometryType
first_col = self.get_first_geometry_column() # TODO: fixme
- if first_col:
- data_type = self._internal.spark_frame.schema[first_col].dataType
+ # Handle both positional and keyword arguments
+ all_args = list(args)
+ for k, v in kwargs.items():
+ all_args.append(v)
- # Handle both positional and keyword arguments
- all_args = list(args)
- for k, v in kwargs.items():
- all_args.append(v)
+ # Join all arguments as comma-separated values
+ params = ""
+ if all_args:
+ params_list = [
+ str(arg) if isinstance(arg, (int, float)) else repr(arg)
+ for arg in all_args
+ ]
+ params = f", {', '.join(params_list)}"
- # Join all arguments as comma-separated values
- params = ""
- if all_args:
- params_list = [
- str(arg) if isinstance(arg, (int, float)) else repr(arg)
- for arg in all_args
- ]
- params = f", {', '.join(params_list)}"
+ sql_expr = f"{operation}(`{first_col}`{params})"
- rename = first_col if not rename else rename
+ return self._query_geometry_column(sql_expr, first_col, rename)
- if isinstance(data_type, BinaryType):
- sql_expr = (
- f"{operation}(ST_GeomFromWKB(`{first_col}`){params}) as
`{rename}`"
- )
- else:
- sql_expr = f"{operation}(`{first_col}`{params}) as `{rename}`"
-
- sdf = self._internal.spark_frame.selectExpr(sql_expr)
- internal = InternalFrame(
- spark_frame=sdf,
- index_spark_columns=None,
- column_labels=[self._column_label],
- data_spark_columns=[scol_for(sdf, rename)],
- data_fields=[self._internal.data_fields[0]],
- column_label_names=self._internal.column_label_names,
- )
- return
_to_geo_series(first_series(PandasOnSparkDataFrame(internal)))
- else:
+ def _query_geometry_column(
+ self, query: str, col: Union[str, None], rename: str
+ ) -> "GeoSeries":
+ """
+ Helper method to query a single geometry column with a specified
operation.
+
+ Parameters
+ ----------
+ query : str
+ The query to apply to the geometry column.
+ col : str
+ The name of the column to query.
+ rename : str
+ The name of the resulting column.
+
+ Returns
+ -------
+ GeoSeries
+ A GeoSeries with the operation applied to the geometry column.
+ """
+ if not col:
raise ValueError("No valid geometry column found.")
+ data_type = self._internal.spark_frame.schema[col].dataType
+
+ if isinstance(data_type, BinaryType):
+ # the backticks here are important so we don't match strings that
happen to be the same as the column name
+ query = query.replace(f"`{col}`", f"ST_GeomFromWKB(`{col}`)")
+
+ sql_expr = f"{query} as `{rename}`"
+
+ sdf = self._internal.spark_frame.selectExpr(sql_expr)
+ internal = InternalFrame(
+ spark_frame=sdf,
+ index_spark_columns=None,
+ column_labels=[self._column_label],
+ data_spark_columns=[scol_for(sdf, rename)],
+ data_fields=[self._internal.data_fields[0]],
+ column_label_names=self._internal.column_label_names,
+ )
+ return _to_geo_series(first_series(PandasOnSparkDataFrame(internal)))
+
@property
def dtypes(self) -> Union[gpd.GeoSeries, pd.Series, Dtype]:
# Implementation of the abstract method
@@ -511,9 +533,44 @@ class GeoSeries(GeoFrame, pspd.Series):
raise NotImplementedError("This method is not implemented yet.")
@property
- def length(self):
- # Implementation of the abstract method
- raise NotImplementedError("This method is not implemented yet.")
+ def length(self) -> pspd.Series:
+ """
+ Returns a Series containing the length of each geometry in the
GeoSeries.
+
+ In the case of a (Multi)Polygon it measures the length of its exterior
(i.e. perimeter).
+
+ For a GeometryCollection it measures sums the values for each of the
individual geometries.
+
+ Returns
+ -------
+ Series
+ A Series containing the length of each geometry.
+
+ Examples
+ --------
+ >>> from shapely.geometry import Polygon
+ >>> import geopandas as gpd
+ >>> from sedona.geopandas import GeoSeries
+
+ >>> gs = GeoSeries([Point(0, 0), LineString([(0, 0), (1, 1)]),
Polygon([(0, 0), (1, 0), (1, 1)]), GeometryCollection([Point(0, 0),
LineString([(0, 0), (1, 1)]), Polygon([(0, 0), (1, 0), (1, 1)])])])
+ >>> gs.length
+ 0 0.000000
+ 1 1.414214
+ 2 3.414214
+ 3 4.828427
+ dtype: float64
+ """
+ col = self.get_first_geometry_column()
+ select = f"""
+ CASE
+ WHEN GeometryType(`{col}`) IN ('LINESTRING',
'MULTILINESTRING') THEN ST_Length(`{col}`)
+ WHEN GeometryType(`{col}`) IN ('POLYGON', 'MULTIPOLYGON') THEN
ST_Perimeter(`{col}`)
+ WHEN GeometryType(`{col}`) IN ('POINT', 'MULTIPOINT') THEN 0.0
+ WHEN GeometryType(`{col}`) IN ('GEOMETRYCOLLECTION') THEN
ST_Length(`{col}`) + ST_Perimeter(`{col}`)
+ END"""
+ return self._query_geometry_column(
+ select, col, rename="length"
+ ).to_spark_pandas()
@property
def is_valid(self):
diff --git a/python/tests/geopandas/test_geoseries.py
b/python/tests/geopandas/test_geoseries.py
index 3b8020c513..0c13e92673 100644
--- a/python/tests/geopandas/test_geoseries.py
+++ b/python/tests/geopandas/test_geoseries.py
@@ -19,6 +19,7 @@ import pytest
import pandas as pd
import geopandas as gpd
import sedona.geopandas as sgpd
+from sedona.geopandas import GeoSeries
from tests.test_base import TestBase
from shapely import wkt
from shapely.geometry import Point, LineString, Polygon, GeometryCollection
@@ -154,7 +155,23 @@ class TestGeoSeries(TestBase):
pass
def test_length(self):
- pass
+ geoseries = GeoSeries(
+ [
+ Point(0, 0),
+ LineString([(0, 0), (1, 1)]),
+ Polygon([(0, 0), (1, 0), (1, 1)]),
+ GeometryCollection(
+ [
+ Point(0, 0),
+ LineString([(0, 0), (1, 1)]),
+ Polygon([(0, 0), (1, 0), (1, 1)]),
+ ]
+ ),
+ ]
+ )
+ result = geoseries.length.to_pandas()
+ expected = pd.Series([0.000000, 1.414214, 3.414214, 4.828427])
+ assert_series_equal(result, expected)
def test_is_valid(self):
pass
diff --git a/python/tests/geopandas/test_match_geopandas_series.py
b/python/tests/geopandas/test_match_geopandas_series.py
index 1f6943624a..fee2fa565d 100644
--- a/python/tests/geopandas/test_match_geopandas_series.py
+++ b/python/tests/geopandas/test_match_geopandas_series.py
@@ -291,7 +291,11 @@ class TestMatchGeopandasSeries(TestBase):
pass
def test_length(self):
- pass
+ for _, geom in self.geoms:
+ sgpd_result = GeoSeries(geom).length
+ assert isinstance(sgpd_result, ps.Series)
+ gpd_result = gpd.GeoSeries(geom).length
+ self.check_pd_series_equal(sgpd_result, gpd_result)
def test_is_valid(self):
pass
