tommy/libpostal
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[addresses] Roman numerals can be returned by Floor.random, relaxing the Zipfian distribution on floors so we get higher floors

Browse Source
This commit is contained in:
Al
2016-06-28 19:47:23 -04:00
parent 0b87a5387d
commit 6fc18b9adb
2 changed files with 10 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
scripts/geodata/addresses/floors.py
View File

@@ -6,6 +6,7 @@ from geodata.addresses.config import address_config
from geodata.addresses.numbering import NumberedComponent, sample_alphabet, latin_alphabet from geodata.addresses.numbering import NumberedComponent, sample_alphabet, latin_alphabet
from geodata.encoding import safe_decode from geodata.encoding import safe_decode
from geodata.math.sampling import weighted_choice, zipfian_distribution, cdf from geodata.math.sampling import weighted_choice, zipfian_distribution, cdf
from geodata.numbers.spellout import numeric_expressions
class Floor(NumberedComponent): class Floor(NumberedComponent):
@@ -15,7 +16,7 @@ class Floor(NumberedComponent):
max_floors = 10 max_floors = 10
max_basements = 2 max_basements = 2
numbered_floors = range(max_floors + 1) + range(-1, -max_basements - 1, -1) numbered_floors = range(max_floors + 1) + range(-1, -max_basements - 1, -1)
floor_probs = zipfian_distribution(len(numbered_floors), 2.0) floor_probs = zipfian_distribution(len(numbered_floors), 0.75)
floor_probs_cdf = cdf(floor_probs) floor_probs_cdf = cdf(floor_probs)
# For use with letters e.g. A0 is probably not as common # For use with letters e.g. A0 is probably not as common
@@ -53,6 +54,12 @@ class Floor(NumberedComponent):
if num_type == cls.NUMERIC: if num_type == cls.NUMERIC:
return safe_decode(number) return safe_decode(number)
elif num_type == cls.ROMAN_NUMERAL:
roman_numeral = numeric_expressions.roman_numeral(number)
if roman_numeral is not None:
return roman_numeral
else:
return safe_decode(number)
else: else:
alphabet = address_config.get_property('alphabet', language, country=country, default=latin_alphabet) alphabet = address_config.get_property('alphabet', language, country=country, default=latin_alphabet)
letter = sample_alphabet(alphabet) letter = sample_alphabet(alphabet)

3
scripts/geodata/addresses/numbering.py
View File

@@ -117,6 +117,7 @@ class NumberedComponent(object):
ALPHA = 'alpha' ALPHA = 'alpha'
ALPHA_PLUS_NUMERIC = 'alpha_plus_numeric' ALPHA_PLUS_NUMERIC = 'alpha_plus_numeric'
NUMERIC_PLUS_ALPHA = 'numeric_plus_alpha' NUMERIC_PLUS_ALPHA = 'numeric_plus_alpha'
ROMAN_NUMERAL = 'roman_numeral'
@classmethod @classmethod
def choose_alphanumeric_type(cls, key, language, country=None): def choose_alphanumeric_type(cls, key, language, country=None):
@@ -127,7 +128,7 @@ class NumberedComponent(object):
values = [] values = []
probs = [] probs = []
for num_type in (cls.NUMERIC, cls.ALPHA, cls.ALPHA_PLUS_NUMERIC, cls.NUMERIC_PLUS_ALPHA): for num_type in (cls.NUMERIC, cls.ALPHA, cls.ALPHA_PLUS_NUMERIC, cls.NUMERIC_PLUS_ALPHA, cls.ROMAN_NUMERAL):
key = '{}_probability'.format(num_type) key = '{}_probability'.format(num_type)
prob = alphanumeric_props.get(key) prob = alphanumeric_props.get(key)
if prob is not None: if prob is not None: