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

[osm/parser] Making a much lower probability of generating sub-building components for named venues (usually on the ground floor, etc.)

Browse Source
This commit is contained in:
Al
2016-07-31 20:40:44 -04:00
parent b727078be5
commit afbb79b81d
2 changed files with 4 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
resources/parser/data_sets/osm.yaml
View File

@@ -24,6 +24,7 @@ venues:
address_probability: 0.4 address_probability: 0.4
abbreviate_probability: 0.3 abbreviate_probability: 0.3
separate_probability: 0.0 separate_probability: 0.0
sub_building_probability: 0.05
countries: countries:
jp: jp:

5
scripts/geodata/osm/formatter.py
View File

@@ -689,7 +689,7 @@ class OSMAddressFormatter(object):
building_tags = self.normalize_address_components(building_tags) building_tags = self.normalize_address_components(building_tags)
for k, v in six.iteritems(building_tags): for k, v in six.iteritems(building_tags):
if k not in revised_tags and k in (AddressFormatter.HOUSE_NUMBER, AddressFormatter.ROAD): if k not in revised_tags and k in (AddressFormatter.HOUSE_NUMBER, AddressFormatter.ROAD, AddressFormatter.POSTCODE):
revised_tags[k] = v revised_tags[k] = v
elif k == AddressFormatter.HOUSE: elif k == AddressFormatter.HOUSE:
building_venue_names.append(v) building_venue_names.append(v)
@@ -698,7 +698,8 @@ class OSMAddressFormatter(object):
if subdivision_components: if subdivision_components:
zone = self.zone(subdivision_components) zone = self.zone(subdivision_components)
add_sub_building_components = AddressFormatter.HOUSE_NUMBER in revised_tags venue_sub_building_prob = float(nested_get(self.config, ('venues', 'sub_building_probability'), default=0.0))
add_sub_building_components = AddressFormatter.HOUSE_NUMBER in revised_tags and (AddressFormatter.HOUSE not in revised_tags or random.random() < venue_sub_building_prob)
address_components, country, language = self.components.expanded(revised_tags, latitude, longitude, language=namespaced_language, address_components, country, language = self.components.expanded(revised_tags, latitude, longitude, language=namespaced_language,
num_floors=num_floors, num_basements=num_basements, num_floors=num_floors, num_basements=num_basements,