def mkdict(row):
"""Turn a geocoder response row into a well-formed dict"""
d = dict(zip(geocoder_header, row))
if len(row) > 3:
try:
try:
d['lat'], d['lon'] = d['latlon'].split(',')
except ValueError as e:
d['lat'], d['lon'] = '',''
d['tract_geoid'] = str(Tract( int(d['state_fips']), int(d['county_fips']), int(d['tract_fips']) ))
try:
del d['latlon']
except Exception as e:
pass
except Exception as e:
# These appear to be errors associated with quote characters in the addresses, like
# 366426709,"8430 I AVENUE""", HESPERIA, CA," 92345""",No_Match. There aren't many of
# them, but they are a problem
print("ERROR for ", row, e, file=sys.stderr)
d['input_address'] = ''
d['match'] = 'Parse Error'
return d
def parse_tiger_tract(self, v):
from geoid.census import Tract
if isinstance(v, (float, int)):
v = '{:0>11}'.format(int(v))
return Tract.parse(v)
def extract_geoid(self, v, row):
from geoid.census import Place, County, State, Cosub, Tract, Zcta
from geoid.civick import GVid
from ambry.valuetype import FailedValue, GeoAcsVT
CA_STATE = 6
if row.geotype == 'PL':
r = Place(CA_STATE, int(row.geotypevalue)).convert(GVid)
elif row.geotype == 'CO':
gt = row.geotypevalue
assert int(gt[0:2]) == CA_STATE
r = County(CA_STATE, int(gt[2:])).convert(GVid)
elif row.geotype == 'CA' or row.geotype == 'ST':
r = State(CA_STATE).convert(GVid)
elif row.geotype == 'CD':
r = Cosub.parse(row.geotypevalue).convert(GVid)
elif row.geotype == 'CT':
try:
r = Tract.parse(row.geotypevalue).convert(GVid)
except ValueError:
r = Tract.parse('06'+row.geotypevalue).convert(GVid)
elif row.geotype == 'ZC':
r = Zcta.parse(row.geotypevalue).convert(GVid)
elif row.geotype == 'NA': # Sub-state region, not a census area
r = None
elif row.geotype == 'RE': # Sub-state region, not a census area
r = None
elif row.geotype == 'R4': # Sub-state region, not a census area
r = None
elif row.geotype == 'MS': # Probably an MSA or similar
r = None
else:
self.error("Unknown geotype {} in row {}".format(row.geotype, row))
r = None
if r is None:
return None
return GeoAcsVT(r)
def extract_geoid(self, v, row):
from geoid.census import Place, County, State, Cosub, Tract, Zcta
from geoid.civick import GVid
import ambry.valuetype
CA_STATE = 6
if row.geotype == 'PL':
r = Place(CA_STATE, int(row.geotypevalue)).convert(GVid)
elif row.geotype == 'CO':
gt = row.geotypevalue
assert int(gt[0:2]) == CA_STATE
r = County(CA_STATE, int(gt[2:])).convert(GVid)
elif row.geotype == 'CA' or row.geotype == 'ST':
r = State(CA_STATE).convert(GVid)
elif row.geotype == 'CD':
r = Cosub.parse(row.geotypevalue).convert(GVid)
elif row.geotype == 'CT':
try:
r = Tract.parse(row.geotypevalue).convert(GVid)
except ValueError:
r = Tract.parse('06' + row.geotypevalue).convert(GVid)
elif row.geotype == 'ZC':
r = Zcta.parse(row.geotypevalue).convert(GVid)
elif row.geotype == 'NA': # Sub-state region, not a census area
r = None
elif row.geotype == 'RE': # Sub-state region, not a census area
r = None
elif row.geotype == 'R4': # Sub-state region, not a census area
r = None
elif row.geotype == 'MS': # Probably an MSA or similar
r = None
else:
self.error("Unknown geotype {} in row {}".format(row.geotype, row))
r = None
if r is None:
return ambry.valuetype.FailedValue(None)
return ambry.valuetype.Geoid(r)
from geoid.census import Tract
for row_n, was_geocoded, row in chunked_geocode(doc):
if not row.get('tract_geoid'):
row['tract_geoid'] = zip_to_tract(fac_zip[int(row['unique_id'])], int(row['unique_id']))
row['side_of_street'] = None
row['tiger_id'] = None
if row['tract_geoid']:
if len(row['tract_geoid']) != 18:
# It's probably still a Census Tract, so convert it to an Acs tract.
row['tract_geoid'] = str(CensusTract.parse(row['tract_geoid'].zfill(11)).convert(AcsTract))
assert(len(row['tract_geoid'])) == 18, row['tract_geoid']
t = AcsTract.parse(row['tract_geoid'])
#print(str(t), file=sys.stderr)
row['state_fips'] = t.state
row['county_fips'] = t.county
row['tract_fips'] = t.tract
if row.get('state_fips'):
row['state_fips'] = str(row['state_fips']).zfill(2)