Mongo-Loves-Data part 2 Plotting a region on a map with individual markers

map-region.py

from pymongo import MongoClient
import pandas as pd
   
def get_data( neighborhood ):
   data = {}
   client = MongoClient('mongodb+srv://restaurantAppUser:restaurantAppPwd@mflix.beal2.mongodb.net/test?authSource=admin&replicaSet=atlas-a7tqy4-shard-0&readPreference=primary&appname=MongoDB%20Compass&ssl=true')
   restaurants = client['sample_restaurants']['restaurants']
#  Sometimes all the work is in getting your data into the format your program requires
#  In this example there are two queries: a) get the region coordinates for the neighborhood selected, and b) use that set of coordinates to 
#  find the restaurants within that region.

# Within Mongo Aggregation the command for the first is a trivial match by name, followed by some projection to get the coordinates
# into an immediate sub-child of the returned record.  One less step for the python program code that follow
   records  = client['sample_restaurants']['neighborhoods'].aggregate([
    {
        '$match': {
            'name': neighborhood
        }
    }, {
        '$project': {
            'name': 1, 
           # the coords are wrapped in an array so unwrap it
            'coord': {
                '$arrayElemAt': [
                    '$geometry.coordinates', 0
                ]
            }, 
            '_id': 0
        }
    }, {
        '$addFields': {
           # want the longitude and latitude as one array each.  Use the $map
            'long': {
                '$map': {
                    'input': '$coord', 
                    'in': {
                        '$arrayElemAt': [
                            '$$this', 0
                        ]
                    }
                }
            }, 
            'lat': {
                '$map': {
                    'input': '$coord', 
                    'in': {
                        '$arrayElemAt': [
                            '$$this', 1
                        ]
                    }
                }
            }
        }
    }]).next()
   
   
   coord  = records['coord']
   name = records['name']

# To find the restaurants within the neighborhood is a $geoWithin query using the Polygon cordinates from the neighborhood query
# The query returns a list of restaurant collection documents

   resultList = restaurants.find( {"address.coord" : {'$geoWithin':
       { '$geometry': {'type': 'Polygon', 'coordinates': [coord]  }}}})

# Use pandas DataFrame to uppack the results of geoWithin query into a DataFrame
   results = pd.DataFrame(resultList )
   data['name'] = name
   rests = pd.DataFrame(results)
   data['restaurants'] = rests

   data['region_lng'] = records['long']
   data['region_lat'] = records['lat']

# extract just the address section from the restaurants object
   address  = pd.DataFrame( rests['address'] )

# Initialize the long and lat for the individial restaurant markers
   long = []
   lat = []

# We want the address column from the addresses
# This could also be done as above with a $map operator
   for i in  address['address']:
     long.append( i['coord'][0]  )
     lat.append( i['coord'][1]  )
    
    
    #add the long, lat to the result    
   data['long'] = long
   data['lat'] = lat

    #data is an object that contains objects:
      # restaurants - the restaurant collection items contained in the region
      # name - the name of the region which was passed into the function
      # long and lat: an array for the restaurant marker data
      # region_long, region_lat:  two arrays for the longitude and lattitude data for the region coordinates "geometry.coordinates"
   return data

# Test the code when run from python at the command line
print ( get_data( 'Middle Village' ) )