Intersection of two GeoPandas Dataframes

Question

I have two GeoPandas dataframes titre_polyG and df which I have intersected as follows:

titre_polyG is as follows:

0      MULTIPOLYGON (((-7.64026 33.59134, -7.63996 33...
1      MULTIPOLYGON (((-7.63987 33.58992, -7.63988 33...
2      MULTIPOLYGON (((-7.63643 33.59023, -7.63670 33...
3      MULTIPOLYGON (((-7.63868 33.59040, -7.63874 33...
4      MULTIPOLYGON (((-7.63892 33.59003, -7.63903 33...
                             ...                        
133    MULTIPOLYGON (((-7.63924 33.59013, -7.63933 33...
134    MULTIPOLYGON (((-7.63958 33.59210, -7.63939 33...
135    MULTIPOLYGON (((-7.63958 33.59210, -7.63939 33...
136    MULTIPOLYGON (((-7.63958 33.59210, -7.63939 33...
137    MULTIPOLYGON (((-7.63937 33.59163, -7.63937 33...
Name: geometry, Length: 138, dtype: geometry

df is as follows:

      apt_psqm           geometry
0   15010.000000    POLYGON ((-7.61096 33.55602, -7.61095 33.55602...
1   12675.000000    POLYGON ((-7.59432 33.53328, -7.59432 33.53328...
2   13810.000000    POLYGON ((-7.65112 33.52504, -7.65112 33.52504...
3   16070.000000    POLYGON ((-7.64377 33.54902, -7.64377 33.54902...
4   13930.000000    POLYGON ((-7.64605 33.52386, -7.64605 33.52387...
... ... ...
85  10732.000000    POLYGON ((-7.52905 33.61630, -7.52905 33.61630...
86  9900.000000 POLYGON ((-7.49753 33.59832, -7.49753 33.59832...
87  9170.000000 POLYGON ((-7.51582 33.59657, -7.51582 33.59657...
88  8509.166667 MULTIPOLYGON (((-7.55564 33.57855, -7.55563 33...
89  9495.280899 POLYGON ((-7.57822 33.56580, -7.57821 33.56580...

The code related to the intersection is below:

# iterate over each row in df and identify intersecting polygons

for index, row in df.iterrows():   
    match_poly = df[df.geometry.intersects(titre_polyG['geometry'])]

match_poly shows the rows of df where there is an intersection with titre_polyG.

My aim is to find the row numbers in titre_polyG where that intersection happens. titre_polyG has 138 rows and if intersection happens for instance at row 1 and 60 then I need to find a way to display those numbers.

Any idea?

Answer 1

Unless you need the intersection geometries you dont need to intersect. You can use spatial join instead:

import geopandas as gpd
import numpy as np

titre = gpd.read_file("/home/bera/Desktop/GIStest/titre.shp")
titre["titrerow"] = np.arange(0, titre.shape[0]).astype(int) #Create a row number, to later groupby
df = gpd.read_file("/home/bera/Desktop/GIStest/df.shp")
df["row"] = np.arange(0, df.shape[0]).astype(int).astype(str) #Create a row number column

join = gpd.sjoin(left_df=titre[["titrerow","geometry"]], right_df=df[["row","geometry"]], how="left", predicate="intersects")
#print(join[["titrerow","row"]])
#    titrerow row
#         0   1
#         0   2
#         0   0
#         1   4
#         2   4
#So titrerow 0 is intersecting 1, 2 and 0 ...

#Groupby titrerow and concatenate the df rows
intersecting = join.groupby("titrerow")["row"].apply(lambda x: ','.join(x)) 
#0    1,2,0
#1        4
#2        4

#Join the result to input
titre = titre.merge(intersecting, left_on="titrerow", right_index=True)

Answer 2

@BERA Thank you. Just like you said, by using spatial join we can retrieve all attributes about the intersecting polygons which is what I was looking for at the first place.

Here is the new code changed accordingly:

join = gpd.sjoin(left_df=df, right_df=titre_polyG,  how="left", predicate="intersects")
    
#Groupby by the required column

intersecting=join.groupby('quartier').agg({'Montant_x':'mean','Titre':'-'.join, 'Adresse':'-'.join}).reset_index()