Assign Linestring to Polygon based on Max length - pandas

Question

I have two geopandas dataframes one is Linestring and other is Polygon. I need to assign the Linestring to Polygon based on Max length. The plot of them looks below.The two polygons are separated by edge color which is Balck.
I am using the following code to assign Linestring to Polygon
well_segments = gpd.overlay(Polygons,Linestring, how='intersection')
well_segments['segment_length'] = well_segments.length
well_segments["geometry"] = well_segments.geometry.to_wkt()
well_segments_df = spark.createDataFrame(well_segments)
windowSpec = Window.partitionBy("api12").orderBy(col("segment_length").desc())
well_segments_valid_df = well_segments_df.select("API", "ID", f.row_number().over(windowSpec).alias("rn"), "segment_length", "geometry").filter(f.col("rn") == 1)
Is there any most efiicient way of doing it in Geopandas or Pandas

Answer 1

you have not provided any sample data. So have used some polygons from natural earth dataset and generated 5 lines which will be of different lengths in each to these polygons
the actual solution is:
use sjoin() instead of overlay()
filter down line with greatest length for each polygon (index_right)
longest = (
gpd.sjoin(linestrings, polygons, predicate="intersects")
.assign(len=lambda d: d["geometry"].length)
.sort_values(["index_right", "len"])
.groupby("index_right")["geometry"]
.last()
)
### full working code ###
import geopandas as gpd
from shapely.geometry import LineString
import numpy as np
import folium
import warnings
r = np.random.RandomState(22)
polygons = (
gpd.read_file(gpd.datasets.get_path("naturalearth_lowres"))
.loc[lambda d: d["geometry"].type.eq("Polygon") & d["continent"].eq("Africa")]
.sample(5, random_state=r)
.loc[:, ["geometry"]]
)
lss = (
polygons.exterior.apply(
lambda g: np.array(g.coords)[r.choice(len(g.coords), [5, 2])]
)
.explode()
.apply(LineString)
)
linestrings = gpd.GeoDataFrame(geometry=lss, crs=polygons.crs).reset_index(drop=True)
# find the longest line in each polygon
with warnings.catch_warnings():
warnings.simplefilter("ignore")
longest = (
gpd.sjoin(linestrings, polygons, predicate="intersects")
.assign(len=lambda d: d["geometry"].length)
.sort_values(["index_right", "len"])
.groupby("index_right")["geometry"]
.last()
)
longest = gpd.GeoSeries(longest, crs=polygons.crs)
# visualise it...
m = polygons.explore(height=300, width=600, color="cyan", name="polys")
m = linestrings.explore(m=m, name="all lines", color="blue", style_kwds={"weight":.8})
m = longest.explore(m=m, name="longest", color="red")
folium.LayerControl().add_to(m)
m