Engineering proximity spatial features

In the previous section, we covered summary spatial features that are derived from preexisting spatial relationships. In this section, we’ll be covering proximity spatial features that are derived based on the proximity, or distance, between two or more observations. To calculate these features, let’s import the NYC Airbnb dataset that we’ve worked with previously. To do that, you’ll run the following code cell:

1. Import the data:

   # Reading in the data

   listings = pd.read_csv(data_path + r'NY Airbnb June 2020\listings.csv.gz', compression='gzip', low_memory=False)

   # Converting it to a GeoPandas DataFrame

   listings_gpdf = gpd.GeoDataFrame(

       listings,

       geometry=gpd.points_from_xy(listings['longitude'],

                           &...