geopandas looping with lambda function

Question

I have a geojson document 'streets' with a column "geometry" filled with data of type shapely Point. I have another geojson document 'quartiers' with a column 'geometry" with data of type shapely Polygon and in same document I have another column called "l_qu". What I am trying to achieve is looping through all streets['geometry] with a lambda in quartiers['geometry'] and assign the corresponding 'l_qu' match value in a new column 'matchquartier' in 'streets'. There are 150K lines in streets (the adresses) and 80 lines in quartiers (the neighboorhoods) The lambda for I built is :

def checkquartier(point):
    return quartiers[quartiers["geometry"].apply(point.within)]["l_qu"]

When running some test fo the lambda, say: checkquartier(streets['geometry'][19]) , no issue, it returns:

19 La Chapelle Name: l_qu, dtype: object

which is the expected result. The problem is when I try to do it for my whole doc with

streets["matchquartier"]=streets['geometry'].apply(lambda x: checkquartier(x))

It returns the error: ValueError: Wrong number of items passed 80, placement implies 1

I know this is a broader pandas issue but the other fixes/answers I found on SO do not work in this case so wondering if there is something from the GeoPandas side that I am doing wrong ? Thanks


import pandas as pd
import geopandas
import os
import shapely
from shapely.geometry import Polygon, LineString, Point,box
import matplotlib.pyplot as plt
from shapely import wkt

quartiers = geopandas.read_file("https://parisdata.opendatasoft.com/explore/dataset/quartier_paris/download/?format=geojson&timezone=Europe/Berlin&lang=fr")
streets=geopandas.read_file("https://opendata.paris.fr/explore/dataset/adresse_paris/download/?format=geojson&timezone=Europe/Berlin&lang=fr")

def checkquartier(point):
    return quartiers[quartiers["geometry"].apply(point.within)]["l_qu"]   

streets["matchquartier"]=streets['geometry'].apply(lambda x: checkquartier(x))

user_7327483 · Answer

IIUC, you can just use a spatial join. This is working for me.

import os
import pandas as pd
import geopandas as gpd
import shapely
from shapely import wkt
from shapely.geometry import Polygon, LineString, Point, box, shape
import json

quartiers = gpd.read_file("quartier_paris.csv")
streets=gpd.read_file("adresse_paris.csv")

quartiers["geometry"] = quartiers["geom"].apply(lambda x: shape(json.loads(x)))

streets["geometry"] = streets["geom"].apply(lambda x: shape(json.loads(x)))

sj = gpd.sjoin(streets, quartiers)
streets.join(sj["l_qu"])

geopandas looping with lambda function

1 Answers