将2D空间中的点分组为python中的预定义笛卡尔网格单元

Question

I have huge no. of points (about million) in 2D plane. I want to group them into cells defined by a square grid of fixed length (each cell side is 3). I don't need no. of points in a given cell rather I need the information about the cell in which given point is. What would be the best and fast way to do this? Is there a python package which can be used?

Answer 1

You can use np.searchsorted to do this. For example:

import numpy as np

xgrid = np.arange(0,33,3)
ygrid = np.arange(0,33,3)

mypoints = np.array([[4,8],[13,19],[21,1]])

icells = np.searchsorted(xgrid,mypoints[:,0])-1
jcells = np.searchsorted(ygrid,mypoints[:,1])-1

print icells,jcells
# [1 4 6] [2 6 0]  

print xgrid[icells],ygrid[jcells]
# [ 3 12 18] [ 6 18  0]

For each point, you find the i and j index for the x and y grid, which you can then use to find "information" about the cell which that given point is in.