Question

我有两个文件，一个带点和它的id。另一个有多边形及其id和更多列。我试图获得与每个点相关联的多边形id。我有近170万个点和50000个多边形。多边形检查中的点在IPython笔记本中永远运行，有没有办法在这里优化我的代码，或者可以使用多处理模块并进行并行处理以加快多边形检查中的点？

首先阅读多边形文件。

import shapefile
import shapely
polygons_sf = shapefile.Reader("ploygonshapefile.shp")
polygon_shapes = polygons_sf.shapes()
polygon_points = [q.points for q in polygon_shapes]
from shapely.geometry import Polygon
polygons = [Polygon(q) for q in polygon_points]
polygons_sf_records = polygons_sf.records() #To get the polygon id

然后为多边形

创建rtree索引

from rtree import index
idx = index.Index() #creating an rtree
count = -1
for a,b in zip(polygon_shapes,polygons_sf_records):
    count +=1
    idx.insert(count,a.bbox,obj=b[1])

然后加载点文件。

points_sf = shapefile.Reader("pointshapefile.shp")
point_shapes = points_sf.iterShapes()
#Creating points as list to iterate
point_coords= [q.points[0] for q in point_shapes]
from shapely.geometry import Point
points = [Point(q.points[0]) for q in point_shapes]
points_sf_records = points_sf.records() #To get the point id

然后尝试找出每个点的多边形id。

matches = []
for i in range(len(points)): #Iterate through each point
    temp= None
    #Iterate only through the bounding boxes which contain the point
    for j in idx.intersection(point_coords[i]):
    #Verify that point is within the polygon itself not just the bounding box
        if points[i].within(polygons[j]):
            temp=j
            break
            matches.append({'Point_id' : points_sf_recods[i][0],'Polygon_id' : polygons_sf_records[j][1]}) #Either the first match found, or None for no matches

有什么方法可以加速这个python代码（点多边形）？

0 个答案: