如何在特定纬度范围内的地理框架中选择线链

Question

我正在使用以下数据集进行研究工作。

https://xfer.services.ncdot.gov/gisdot/DistDOTData/NCRoutes_SHP. zip

我正在尝试从上面数据集提取衬里。线条必须在纬度-80.86和-80.83之间。

首先，我使用地理器加载了数据集。

import geopandas as gpd
graph = gpd.read_file("NCRoutes.shp")
graph = graph.to_crs(4236)
print(len(graph['geometry'])) # number of rows in geodataframe
graph.head()

输出 -

373370
Division    MaintCntyC  RouteID RouteClass  RouteNumbe  RouteQuali  RouteInven  RouteName   BeginMP EndMP   RouteMaint  Shape_Leng  geometry
0   13  011 10000026011 1   26  0   0   I-26    0.000063    29.164804   System  56252.469013    MULTILINESTRING Z ((-82.53991 35.79167 0.00000...
1   14  045 10000026045 1   26  0   0   I-26    0.000000    17.459971   System  89044.485587    MULTILINESTRING Z ((-82.53591 35.44044 664.953...
2   13  057 10000026057 1   26  0   0   I-26    0.000000    12.606171   System  66456.002756    MULTILINESTRING Z ((-82.56037 35.95480 1145.01...
3   14  075 10000026075 1   26  0   0   I-26    0.000000    13.121029   System  69207.803704    LINESTRING Z (-82.34883 35.25453 565.22225, -8...
4   7   001 10000040001 1   40  0   0   I-40    0.000000    16.013000   System  84492.598738    LINESTRING Z (-79.53887 36.06290 179.46660, -7...

如上所述，Graph geodataframe具有373370行。 几何列的每一行中都有一个多线或统一性。因此，我编写了下面的代码以提取所需的线条，并将其保留在GeodataFrame 区域中。

from IPython.display import clear_output
area = gpd.GeoDataFrame(columns=['geometry'], geometry='geometry') # geodataframe for saving roads of interest
itr = 0 # to keep track of which row I'm going through
for String in graph['geometry']: # iterating through each row in geometry column of geodataframe
    print("iterating through",itr,"th row in geodataframe")
    itr += 1
    if String.geom_type=='MultiLineString': # if the current row contains a multilinestring
        for ls in String: # iterating through linestrings in the multilinestring
            for i in range(len(ls.xy[0])): # iterating through points in the linestring
                #print("longitude:",ls.xy[0][i],"latitude:",ls.xy[1][i])
                if(-80.86<=ls.xy[0][i]<=-80.83):
                    temp = gpd.GeoDataFrame({'geometry':[ls]},geometry='geometry',crs=4236) #creating a temporary geodataframe for storing the road of interest
                    area = area.append(temp)
                    break # no need to keep looking if linestring is already added to area geodataframe
    else: # if the current row contains a linestring
        for i in range(len(String.xy[0])):
            if(-80.86<=String.xy[0][i]<=-80.83):
                temp = gpd.GeoDataFrame({'geometry':[String]},geometry='geometry',crs=4236)
                area = area.append(temp)
                break
    clear_output(wait=True)

但是，我的笔记本电脑中经过150行的迭代大约需要12分钟。数据框架中有373370行。因此，通过整个Graph GeodataFrame在我的计算机中迭代大约需要500个小时。有更快的方法吗？

Answer 1

您要使用cx，坐标索引器。

subset = graph.cx[:, -80.86:-80.83]

请参阅文档中的更多信息 htttps：//geopandas.org/geopandas.org/en/stable/docs /user_guide/indexing.html

关于多线链vs linestring，您可以在索引之前使用爆炸>爆炸以确保使用各个零件，而不是整个几何形状。

subset = graph.explode().cx[:, -80.86:-80.83]