-
Notifications
You must be signed in to change notification settings - Fork 0
/
GMGeocoder.py
154 lines (130 loc) · 6.46 KB
/
GMGeocoder.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
import pandas as pd
import requests
import logging
import time
logger = logging.getLogger("root")
logger.setLevel(logging.DEBUG)
# create console handler
ch = logging.StreamHandler()
ch.setLevel(logging.DEBUG)
logger.addHandler(ch)
#------------------ CONFIGURATION -------------------------------
# Set your Google API key here.
# Even if using the free 2500 queries a day, its worth getting an API key since the rate limit is 50 / second.
# With API_KEY = None, you will run into a 2 second delay every 10 requests or so.
# With a "Google Maps Geocoding API" key from https://console.developers.google.com/apis/,
# the daily limit will be 2500, but at a much faster rate.
# Example: API_KEY = 'AIzaSyC9azed9tLdjpZNjg2_kVePWvMIBq154eA'
API_KEY = 'AIzaSyBpYTT6Y9ZMP8yORjF4YmSwnPCAcIE9jiI'
# Backoff time sets how many minutes to wait between google pings when your API limit is hit
BACKOFF_TIME = 30
# Set your output file name here.
output_filename = 'output.csv'
# Set your input file here
input_filename = "input.csv"
# Specify the column name in your input data that contains addresses here
address_column_name = "Address"
# Return Full Google Results? If True, full JSON results from Google are included in output
RETURN_FULL_RESULTS = True
#------------------ DATA LOADING --------------------------------
# Read the data to a Pandas Dataframe
data = pd.read_csv(input_filename, encoding='utf8')
if address_column_name not in data.columns:
raise ValueError("Missing Address column in input data")
# Form a list of addresses for geocoding:
# Make a big list of all of the addresses to be processed.
addresses = data[address_column_name].tolist()
#------------------ FUNCTION DEFINITIONS ------------------------
def get_google_results(address, api_key=None, return_full_response=False):
"""
Get geocode results from Google Maps Geocoding API.
Note, that in the case of multiple google geocode reuslts, this function returns details of the FIRST result.
@param address: String address as accurate as possible. For Example "18 Grafton Street, Dublin, Ireland"
@param api_key: String API key if present from google.
If supplied, requests will use your allowance from the Google API. If not, you
will be limited to the free usage of 2500 requests per day.
@param return_full_response: Boolean to indicate if yod like to return the full response from google. This
is useful if yod like additional location details for storage or parsing later.
"""
# Set up your Geocoding url
geocode_url = "https://maps.googleapis.com/maps/api/geocode/json?address={}".format(address)
if api_key is not None:
geocode_url = geocode_url + "&key={}".format(api_key)
# Ping google for the reuslts:
results = requests.get(geocode_url)
# Results will be in JSON format - convert to dict using requests functionality
results = results.json()
# if there's no results or an error, return empty results.
if len(results['results']) == 0:
output = {
"formatted_address" : None,
"latitude": None,
"longitude": None,
"accuracy": None,
"google_place_id": None,
"type": None,
"postcode": None
}
else:
answer = results['results'][0]
output = {
"formatted_address" : answer.get('formatted_address'),
"latitude": answer.get('geometry').get('location').get('lat'),
"longitude": answer.get('geometry').get('location').get('lng'),
"accuracy": answer.get('geometry').get('location_type'),
"google_place_id": answer.get("place_id"),
"type": ",".join(answer.get('types')),
"postcode": ",".join([x['long_name'] for x in answer.get('address_components')
if 'postal_code' in x.get('types')])
}
# Append some other details:
output['input_string'] = address
output['number_of_results'] = len(results['results'])
output['status'] = results.get('status')
if return_full_response is True:
output['response'] = results
return output
#------------------ PROCESSING LOOP -----------------------------
# Ensure, before we start, that the API key is ok/valid, and internet access is ok
test_result = get_google_results("London, England", API_KEY, RETURN_FULL_RESULTS)
if (test_result['status'] != 'OK') or (test_result['formatted_address'] != 'London, UK'):
logger.warning("There was an error when testing the Google Geocoder.")
raise ConnectionError('Problem with test results from Google Geocode - check your API key and internet connection.')
# Create a list to hold results
results = []
# Go through each address in turn
for address in addresses:
# While the address geocoding is not finished:
geocoded = False
while geocoded is not True:
# Geocode the address with google
try:
geocode_result = get_google_results(address, API_KEY, return_full_response=RETURN_FULL_RESULTS)
except Exception as e:
logger.exception(e)
logger.error("Major error with {}".format(address))
logger.error("Skipping!")
geocoded = True
# If we're over the API limit, backoff for a while and try again later.
if geocode_result['status'] == 'OVER_QUERY_LIMIT':
logger.info("Hit Query Limit! Backing off for a bit.")
time.sleep(BACKOFF_TIME * 60) # sleep for 30 minutes
geocoded = False
else:
# If we're ok with API use, save the results
# Note that the results might be empty / non-ok - log this
if geocode_result['status'] != 'OK':
logger.warning("Error geocoding {}: {}".format(address, geocode_result['status']))
logger.debug("Geocoded: {}: {}".format(address, geocode_result['status']))
results.append(geocode_result)
geocoded = True
# Print status every 100 addresses
if len(results) % 100 == 0:
logger.info("Completed {} of {} address".format(len(results), len(addresses)))
# Every 500 addresses, save progress to file
if len(results) % 500 == 0:
pd.DataFrame(results).to_csv("{}_bak".format(output_filename))
# All done
logger.info("Finished geocoding all addresses")
# Write the full results to csv using the pandas library.
pd.DataFrame(results).to_csv(output_filename, encoding='utf8')