ArcGIS Queries

Intro to ArcGIS

TODO

Official Documentation (Python API)

https://esri.github.io/arcgis-python-api/apidoc/html/index.html

Helper Functions

These helper functions are available at the gis_utils.py file.

Getting features from the ArcGIS server

from arcgis.gis import *
from arcgis.features import FeatureLayer
from arcgis import geometry
from pprint import pprint  # for printing JSON nicely

gis = GIS()
URL = "https://services.arcgis.com/sFnw0xNflSi8J0uh/ArcGIS/rest/services/Supermarkets_GroceryStores/FeatureServer/0/"
QUERY = { 'where': '1=1', 'out_sr': '4326' }


def get_features_from_feature_server(url, query):
  """
  Given a url to a City of Boston Feature Server, return a list
  of Features (for example, parking lots that are not full)

  :param url: url for Feature Server
  :param query: query to select features (example: "Spaces > 0")
  :return: list of all features returned from the query
  """
  features = []
  f = FeatureLayer(url = url)
  feature_set = f.query(**query)
  for feature in feature_set:
      features.append(feature.as_dict)
  return features


grocery_stores = get_features_from_feature_server(URL, QUERY)
pprint(grocery_stores[0])

Geocoding an address

Translating and address such as 220 Clarendon Street to a long, lat JSON object {'x': -71.07545238040102, 'y': 42.3509583718135}

from arcgis.gis import *
from arcgis.geocoding import geocode

gis = GIS()


def geocode_address(m_address):
    """
    :param m_address: address of interest in street form
    :return: address in coordinate (X and Y) form
    """
    m_address = m_address + ", City: Boston, State: MA"
    m_location = geocode(address=m_address)[0]
    return m_location['location']

Calculating the distance between two features

from arcgis.gis import *
from arcgis import geometry

gis = GIS()
FEATURE1 = grocery_stores[0]  # from helper function above
FEATURE2 = grocery_stores[1]


def get_geodesic_distance(feature1, feature2):
  geometry1 = feature1['geometry']  # if feature1 is an address, use just feature1 (not `['geometry']`)
  geometry2 = feature2['geometry']
  spation_ref = {"wkid" : 4326}
  return geometry.distance(spation_ref, 
                           geometry1, 
                           geometry2,
                           distance_unit='', 
                           geodesic=True)['distance']


distance = get_geodesic_distance(FEATURE1, FEATURE2)
print(distance)  # in meters

Putting it all together: Use cases

Calculating the distance from an address to a feature

MILE = 1600  # meters
DAY = datetime.datetime.today().strftime('%A')
food_trucks_url = 'https://services.arcgis.com/sFnw0xNflSi8J0uh/arcgis/rest/services/food_trucks_schedule/FeatureServer/0'
address = geocode_address('120 Boylston St')

# This will return a list of all available food trucks in Boston
trucks = get_feature_location(food_trucks_url, QUERY)

# Filter features by attributes (in this case, Day = Today and Time = Lunch)
trucks_today = []
for t in trucks:
  if(t['attributes']['Day'] == DAY and t['attributes']['Time'] == 'Lunch'):
    trucks_today.append(t)

# Print only features that are within a mile
for t in trucks_today:
  distance = get_geodesic_distance(address, t)
  if distance <= MILE:
    print(t['attributes']['Truck'] + ' is located at ' + \
          t['attributes']['Loc'] + ' between ' + \
          t['attributes']['Start_time'] + ' and ' + \
          t['attributes']['End_time'] + '\n')