This example uses the Natural Earth Dataset introduced in Chapter 4, Data Types, Storage, and Conversion, under the Reading and writing vector data with GeoPandas section. We'll use latitude-longitude coordinates to create a spatial filter in the form of a bounding box. This box selects only the data inside of this box. This is a way to work with a subset of our data. We'll use OGR's SpatialFilterRec method, which takes four values—minx, miny, maxx and maxy, to create a bounding box. Our (random) example is to select the cities in our bounding box (which shows the state of Texas, as well as parts of Oklahoma and Mexico). To filter our results even further, we only want the cities in the US. This means we have to filter our search results with an extra if/else statement in our for loop.

The website www.mapsofworld.com gives the following four values for our example code: -102 (minx), 26 (miny), -94 (maxx), and 36 (maxy) for the state of Texas. Here is the script:

In: # import the modules
    from osgeo import ogr
    import os
    # reference the shapefile and specify driver type
    shapefile =                                              
    r"C:datagdalNE10m_cultural
e_10m_populated_places.shp"
    driver = ogr.GetDriverByName("ESRI Shapefile")
    # open the data source with driver, zero means open in read-only 
    mode
    dataSource = driver.Open(shapefile, 0)
    # use the GetLayer() function for referencing the layer that holds 
    the data
    layer = dataSource.GetLayer()
    # pass in the coordinates for the data frame to the                     
    SetSpatialFilterRect() function. This filter creates a rectangular     
    extent and selects the features
      inside the extent
      layer.SetSpatialFilterRect(-102, 26, -94, 36)
      for feature in layer:
      # select only the cities inside of the USA
      # we can do this through a SQL query:
      # we skip the cities that are not in the USA,
      # and print the names of the cities that are
          if feature.GetField("ADM0NAME") != "United States of                              
      America":
              continue
          else:
              print(feature.GetField("NAME"))

Out:    Ardmore
        McAlester
        Bryan
        San Marcos
        Longview
        …