def geoclient_intersection(streetNumber, streetName, boroughName):
''' retrieve intersection street1 and street2 with main street'''
g = Geoclient('799db7eb', '02b0bed977c344cb27b77e549eb69ed8')
response_dict = {}
dataGeo = g.address(streetNumber, streetName, boroughName)
sideOfStreet = dataGeo['firstStreetNameNormalized']
fromStreet = dataGeo['highCrossStreetName1']
toStreet = dataGeo['lowCrossStreetName1']
borough = dataGeo['firstBoroughName']
response_dict.update({'sideOfStreet': sideOfStreet})
response_dict.update({'fromStreet': fromStreet})
response_dict.update({'toStreet': toStreet})
response_dict.update({'borough': borough})
return response_dict
def geoclient_intersection(streetNumber, streetName, boroughName):
''' retrieve intersection street1 and street2 with main street'''
g = Geoclient('799db7eb', '02b0bed977c344cb27b77e549eb69ed8')
response_dict = {}
dataGeo = g.address(streetNumber, streetName, boroughName)
sideOfStreet = dataGeo['firstStreetNameNormalized']
fromStreet = dataGeo['highCrossStreetName1']
toStreet = dataGeo['lowCrossStreetName1']
borough = dataGeo['firstBoroughName']
response_dict.update({'sideOfStreet': sideOfStreet})
response_dict.update({'fromStreet': fromStreet})
response_dict.update({'toStreet': toStreet})
response_dict.update({'borough': borough})
return response_dict
count+=1
if count == 1 or count == 2:
continue
### For testing code purposes to make sure we can get data w/o pulling it all at once ###
#if count == 5:
# break;
# Get the address row and split the number from the street for API purposes
address = row[7]
address = address.split(' ', 1)
if len(address) == 1:
address = ['', address[0]]
# Call the API and store in jsonData
jsonData = g.address(address[0], address[1], row[5])
# Write to the TSV storing file the original data, only with address column replaced with separated address,
# plus the longitude, latitude
try:
lat = jsonData['latitude']
lon = jsonData['longitude']
except:
lat, lon = '', ''
print "error in %s on line %d" % (file, count)
print row
print jsonData
csvout.writerow(row[0:6] + [s, address[0], address[1]] + row[8:11] + [lat, lon] + row[12:])
jsonout.write(json.dumps(jsonData) + '\n')
my_app_ID = id_key[1][0]
my_app_key = id_key[1][1]
g = Geoclient(my_app_ID, my_app_key)
with open('../out/nyc_sales_clean.csv', 'rb') as f:
reader = csv.reader(f)
nyc_sales = map(tuple, reader)
nrow_nyc = len(nyc_sales)
with open("../out/coords_nyc_api.csv", "wb") as csv_file:
writer = csv.writer(csv_file, delimiter=',')
writer.writerow([
'id_sale', 'lat', 'long', 'returned_street_name', 'returned_zip_code'
])
for i in range(1, len(nyc_sales)):
print(str(i) + ' / ' + str(nrow_nyc))
# address(houseNumber, street, borough)
addr = g.address(nyc_sales[i][24], nyc_sales[i][23], nyc_sales[i][25])
out = [
nyc_sales[i][22],
addr.get('latitudeInternalLabel'),
addr.get('longitudeInternalLabel'),
addr.get('streetName1In'),
addr.get('zipCode')
]
print(out)
print('\n\n')
writer.writerow(out)
class GeoHelper:
def __init__(self, app_id, key):
# empty for now
self._g = Geoclient(app_id, key)
def _find_id_column(self, col_list):
ids_by_pref = [
'BIN', 'BUILDINGIDENTIFICATIONNUMBER', 'GEOCODEDBIN', 'BBL',
'GEOCODEDBBL', 'ADDRESS'
]
col_map = dict(zip(map(str.upper, col_list), col_list))
return next(
(col_map[i] for i in ids_by_pref if i in map(str.upper, col_list)))
def _inferredGeocoder(self, input_str):
'''
Attempts to infer the format of the input provided for geocoding, either BIN, BBL, or Street Address (very rudimentary). Will return the complete geoclient object.
'''
input_str = str(input_str).replace('.0', '')
if input_str.isdigit():
if len(input_str) == 10: # BBL
#print '10 digit number: attempting BBL'
out = self._g.bbl(int(input_str[0]),int(input_str[1:6].lstrip("0")) , \
int(float(input_str[6:].lstrip("0"))))
elif len(input_str) == 7: # BIN
#print '7 digit number: attempting BIN'
out = self._g.bin(input_str)
else:
#print 'Unrecognized number of digits, no ID possible'
out = None
else: # try to split addresses
#print "Attempting to split text into house number / street name / borough."
split = str.split(input_str)
house_num = split[0]
street_name = " ".join(split[1:-1])
boro_name = split[-1]
try:
out = self._g.address(house_num, street_name, boro_name)
except:
print 'Format not recognized'
out = None
return out
def _checkGeoclientValidity(self, geoclient_output):
if 'returnCode1a' in geoclient_output:
if str(geoclient_output['returnCode1a'][0]) == '0':
return True
elif geoclient_output['message']:
return 'Error Code: ' + str(geoclient_output['message'])
else:
return 'error returned with no message'
def _addressGeocoder(df):
'''
private function to make a generic call to NYC geoclientBatch.
'''
try:
x = self._g.address(df[house_num], df[street], df[boro])
BBL = x['bbl']
BIN = x['buildingIdentificationNumber']
except:
e = _sys.exc_info()[0]
BBL = ("Error: %s" % e)
BIN = BBL
return BBL, BIN
def get_BINandBBL(self, df, identifier_col=None):
'''
Uses DOITT's GeoClient (the web API to DCP's GeoSupport)
via the python wrapper https://github.com/talos/nyc-geoclient
to geocode a dataframe df with columns number, street, and boro.
Returns the dataframe df with two additional columns: geocodedBBL and geocodedBIN
'''
if identifier_col:
print 'using provided ID column: ' + identifier_col
else:
identifier_col = self._find_id_column(df.columns)
print 'found ID column: ' + identifier_col
def wrapper_func(x):
out = self._inferredGeocoder(x[identifier_col])
log = self._checkGeoclientValidity(out)
if log is True:
return out['bbl'], out['buildingIdentificationNumber']
else:
return log, log
df[['geocodedBBL',
'geocodedBIN']] = df.apply(lambda x: wrapper_func(x),
axis=1).apply(_pd.Series)
return df
def GetLatLong(self, df, identifier_col=None):
if identifier_col:
print 'using provided ID column: ' + identifier_col
else:
identifier_col = self._find_id_column(df.columns)
print 'found ID column: ' + identifier_col
df['Latitude'] = _np.nan
df['Longitude'] = _np.nan
def wrapper_func(x):
out = self._inferredGeocoder(x[identifier_col])
log = self._checkGeoclientValidity(out)
if log is True:
return out['latitudeInternalLabel'], out[
'longitudeInternalLabel']
else:
return log, log
df[['Latitude', 'Longitude']] = df.apply(lambda x: wrapper_func(x),
axis=1).apply(_pd.Series)
return df
reader = csv.reader(i)
print "opening file: %s" % outfile
with open(outfile, 'w') as o:
writer = csv.writer(o, lineterminator='\n')
all = []
row = next(reader, None)
row.append('bbl')
all.append(row)
try:
for row in reader:
street_num = row[0]
street_name = row[1]
#print street_num, street_name
row_gc = g.address(street_num, street_name, 'Manhattan')
row_bbl = row_gc.get('bbl', '0')
print "The BBL for %s %s is %s" % (street_num, street_name,
row_bbl)
row.append(row_bbl)
all.append(row)
writer.writerows(all)
except csv.Error as e:
sys.exit('file %s, line %d: %s' % (infile, reader.line_num, e))
