Efficient way to calculate geographic density in P

问题:

I have a large list of longitude and latidue data corresponding to fast food places in the U.S. For each fast food place, I want to know how many other fast food places are within 5 miles. I could calculate this in Pandas using Geopy like so (each row in the DataFrame is a different fast food place):

import pandas as pd
import geopy.distance

df = pd.DataFrame({'Fast Food Place':[1,2,3], 'Lat':[33,34,35], 'Lon':[42,43,44]})

for index1, row1 in df.iterrows():
    num_fastfood = 0

    for index2, row2 in df.iterrows():
        # calculate distance in miles between longitude and latitude
        dist = geopy.distance.VincentyDistance(row1[['Lat','Lon']],
                                               row2[['Lat','Lon']]).miles

        # if fast food is within 5 miles, increment num_fastfood
        if dist < 5: # if less than five miles
            num_fastfood = num_fastfood + 1

    df.loc[index1, 'num_fastfood_5miles'] = num_fastfood - 1 # (subtract 1 to exclude self)

But this is extremely slow on very large data sets (i.e. 50,000 rows). I considered using a KDTree for the search, but curious if other people have a much quicker method?

回答1:

Implementation with scipy.spatial.cKDTree:

from scipy.spatial import cKDTree

def find_neighbours_within_radius(xy, radius):
    tree = cKDTree(xy)
    within_radius = tree.query_ball_tree(tree, r=radius)
    return within_radius

def flatten_nested_list(nested_list):
    return [item for sublist in nested_list for item in sublist]

def total_neighbours_within_radius(xy, radius):
    neighbours = find_neighbours_within_radius(xy, radius)
    return len(flatten_nested_list(neighbours))