class Geocoder():
"""
A class that provides functionality for getting street stretches, etc.
Exposes the following functions:
normalize_street_name
Returns a street in the format the LION and Geosupport use.
get_street_code
Given a street name, return the street code.
street_code_exists
Return whether the given street code exists in LION's streets
autocomplete
Given a text search, return possible street name matches.
address
Find the street segment and side of street of an address.
"""
def __init__(self, lion_version, crs='epsg:2263',
lion_loader=load_lion_gdf, force_rebuild=False,
network_type='cscl',
include_spatial_index=False, spatial_index_crs=None, **kwargs):
self.crs = crs
self.lion_version = lion_version.lower()
self.lion_loader = lion_loader
self.force_rebuild = force_rebuild
self.cache_path = MICHI_HOME / self.lion_version / 'geocoder'
if not self.cache_path.exists():
self.cache_path.mkdir(parents=True)
# ensure CRS parameters are parsable before instantiation
pyproj.crs.CRS(crs)
if spatial_index_crs:
pyproj.crs.CRS(spatial_index_crs)
# Instantiate Geosupport
try:
if geosupport_version is None:
geosupport_version = self.lion_version
self.geosupport = Geosupport(geosupport_version=version)
except:
# Use default geosupport
self.geosupport = Geosupport()
warn(
"Using default Geosupport. "
"May not match LION %s" % self.lion_version
)
(
self.lion_df, self.nodes_df,
self.lion_segments, self.streets,
self.street_code_map
) = self._load_lion_data()
self.street_names = self._build_search_data()
self.network_type = network_type
if network_type == 'cscl':
self.segment_column = 'physical_id'
self.segment_dict = 'cscl_segments'
elif network_type == 'lion':
self.segment_column = 'segment_id'
self.segment_dict = 'lion_segments'
# A regex to parse geometry strings, to be used by `self.parse_geometry`
self.geometry_regex = re.compile(r'([a-z_]*):*(\d+)([A-Z]{0,1})')
(
self.cscl_segments_df, self.cscl_segments,
self.node_network, self.segment_network,
self.nodes_df, self.nodes
) = self._build_networks(self.network_type)
self.segments = getattr(self, self.segment_dict)
# SearchMixin
# TODO Don't skip this, testing!
#super().__init__(clear_cache=clear_cache, **kwargs)
# get cache key to uniquely recognize spatial indexes
#cache_key = get_cache_key(('_load_lion_data',) + (self,) + ({},))
# load spatial indexes
self.include_spatial_index = include_spatial_index
self.spatial_index_crs = spatial_index_crs
if self.include_spatial_index:
self.lion_index = _get_spatial_index(
self.lion_df, self.crs, self.spatial_index_crs, lion_version,
'segment_index', 'segment_id'
)
self.node_index = _get_spatial_index(
self.nodes_df, self.crs, self.spatial_index_crs, lion_version,
'node_index', 'node'
)
@method_file_cache('lion.pkl')
def _load_lion_data(self):
print("Downloading and Building Geocoder Data...")
#lion_df = _load_lion_df(
# self.lion_version, self.crs
#)
lion_df = self.lion_loader(self.lion_version)
'''return (
lion_df,# nodes_df,
None, None, None, None
#segments, streets,
#street_code_map
)'''
lion_df = _clean_lion_df(lion_df, self.crs)
# Create nodes_df
nodes_df = _get_nodes_df(lion_df)
# Get dead ends
dead_end_nodes = _get_dead_end_nodes(nodes_df)
nodes_df['dead_end'] = nodes_df['node'].isin(dead_end_nodes)
# Convert to GeoDataFrame
nodes_df = gp.GeoDataFrame(
nodes_df, geometry='geometry', crs={'init': self.crs}
)
# Cache stuff for quick access
streets = _get_streets_dict(lion_df)
segments = _get_segments_dict(lion_df)
# Add dead end nodes to `streets`
dead_end_street_codes = self._get_dead_end_street_codes()
streets['dead_end'] = {'nodes': dead_end_nodes, 'df': None}
# Create a mapping of street codes to internal street codes
street_code_map = dict(lion_df[[
'original_street_code', 'street_code'
]].drop_duplicates().values)
# Add dead ends to street_code_map
street_code_map.update(dict([
(i, 'dead_end') for i in dead_end_street_codes
]))
# Add mapping from new values to themselves
for key, value in list(street_code_map.items()):
if value not in street_code_map:
street_code_map[value] = value
return (
lion_df, nodes_df,
segments, streets,
street_code_map
)
@method_file_cache('search.pkl')
def _build_search_data(self):
print("Building Search Data...")
# Create dataframe with street names for autocomplete
street_names = self.lion_df[
self.lion_df['feature_type'].isin(['0', '6', '8', 'A', 'W'])
].groupby([
'street', 'street_code', 'borough_code'
]).count()['segment_id'].reset_index()
# Add options for a dead end for each borough to the dataframe.
dead_end_street_codes = self._get_dead_end_street_codes()
street_names = street_names.append([{
'street': 'DEAD END',
'street_code': 'dead_end',
'borough_code': street_code[0],
'segment_id': 100
} for street_code in dead_end_street_codes])
# A function to get the geometry of each "street" used for the
# centroid distance functionality of autocomplete.
def get_geometry(street_code):
if street_code == 'dead_end':
return self.nodes_df[
self.nodes_df['node'].isin(self.streets['dead_end']['nodes'])
]['geometry'].unary_union
return self.streets[street_code]['df']['geometry'].unary_union
street_names['geometry'] = street_names['street_code'].apply(get_geometry)
street_names = gp.GeoDataFrame(street_names.sort_values(
'segment_id', ascending=False
).reset_index(drop=True))
return street_names
def _get_street_code(self, borough, street):
return self.geosupport.get_street_code({
'street_name': street, 'borough_code': borough
})['B10SC - First Borough and Street Code']
def _get_dead_end_street_codes(self):
return [self._get_street_code(b, 'DEAD END') for b in BOROUGHS]
def _get_intersection(self, street_code_1, street_code_2):
street_code_1 = self.normalize_street_code(street_code_1)
street_code_2 = self.normalize_street_code(street_code_2)
return self.streets[street_code_1]['nodes'].intersection(
self.streets[street_code_2]['nodes']
)
def get_street_code(self, borough, street):
"""
Given a borough and street name, return the street code.
Parameters
----------
street : str
borough : str
The borough code, abbreviation or name.
Returns
-------
str
The street code
Raises
------
NotAStreetError
If the "street" is a valid identifier in Geosupport, but isn't a
drivable street in LION.
StreetNameNotFoundError
If the street name isn't recognized raise this error. The error
has an attribute `options` with up to 10 alternate street names
that are similar to the given one.
"""
try:
street_code = self.normalize_street_code(
self._get_street_code(borough, street)
)
if self.street_code_exists(street_code):
return street_code
else:
raise NotAStreetError(street, street_code)
except GeosupportError as error:
# If geosupport raised an error, include it's suggested street
# names in the raised StreetNameNotFoundError.
options = []
for s in error.result['List of Street Names']:
street_code = self.normalize_street_code(
self._get_street_code(borough, s)
)
# Only return valid streets
if self.street_code_exists(street_code):
# Append the name and the code.
options.append((s, street_code))
raise StreetNameNotFoundError(street, options)
def normalize_street_code(self, street_code):
if street_code in self.street_code_map:
return self.street_code_map[street_code]
street_code = str(street_code)[:6].zfill(6)
return self.street_code_map.get(street_code, street_code)
def street_code_exists(self, street_code):
return self.normalize_street_code(street_code) in self.streets
def normalize_street_name(self, street):
"""
Return the street name normalized into the format used by LION and
Geosupport.
Parameters
----------
street : str
The raw street name.
Returns
-------
str
"""
return re.sub(
r'\s+', ' ',
self.geosupport.normalize_street_name(
street=street
)['First Street Name Normalized']
)
def autocomplete(self, text, borough=None, cross_street_code=None,
return_top_n=10, centroid=None):
"""
Given a string and optional filter parameters, return the most likely
streets.
Parameters
----------
text : str
The street name or partial street name.
borough : str, optional
Optionally constrain the search to a single borough which can be
provided as a borough code, abbreviation or full name.
cross_street_code : str, optional
Only return streets that intersect with the given street.
return_top_n : int, optional
The number of results to return. (Default 10)
centroid : shapely.geometry.Point, optional
A point in the same crs as geocoder.crs. If given, sort the results
by distance from the centroid.
"""
text = text.strip().upper()
text_normalized = self.normalize_street_name(text)
df = self.street_names.copy()
if borough:
df = df[df['borough_code'] == str(BOROUGH_CODES[str(borough)])]
if cross_street_code:
cross_street_code = self.normalize_street_code(cross_street_code)
# Get all the nodes that are on the given street code.
nodes = self.nodes_df[self.nodes_df['node'].isin(
self.streets[cross_street_code]['nodes']
)]
# Get a list of street codes that intersect with those nodes.
street_codes = nodes['street_code'].unique().tolist()
# If any of the nodes are a dead end, add that street.
if nodes['dead_end'].any():
street_codes.append('dead_end')
boroughs = nodes['borough_code'].unique()
df = df[
df['street_code'].isin(street_codes) &
df['borough_code'].isin(boroughs) # Handles the extra dead ends
]
query = (
# Starts with the text
df['street'].str.startswith(text) |
df['street'].str.startswith(text_normalized) |
# Contains the full word
df['street'].str.contains(r'\b%s\b' % text) |
df['street'].str.contains(r'\b%s\b' % text_normalized)
)
# If the string is 5 or more characters, search anywhere in the string
if len(text) >= 5:
query = (
query | df['street'].str.contains(text) |
df['street'].str.contains(text_normalized)
)
df = df[query].copy()
if centroid:
# If a cross street is given, find the distance to the place
# where the two streets intersect, and not just to anywhere along
# the street.
if cross_street_code:
func = lambda s: centroid.distance(
self.nodes[
self._get_intersection(s, cross_street_code).pop()
]['geometry']
)
df['distance'] = df['street_code'].apply(func)
else:
df['distance'] = df['geometry'].distance(centroid)
# By default, the options are sorted by "segment_id", which is the
# number of segments with that street name in the city.
# i.e, show common streets like "Broadway" before "Broad Street"
# When sorting by distance, we want to take into account both
# the distance and how common the street is.
# The log of the count will map the count into a much smaller but
# still increasing number. "+ e - 1" ensures that the devisor starts
# at 1 for streets that only have 1 segment.
# After division, if two streets are the same distance from the
# centroid, the one with a higher count will have a lower
# "distance." But the count won't overpower the actual distance.
# After, the options are sorted by distance in ascending order.
df['distance'] = df['distance'] / np.log(df['segment_id'] + np.e - 1)
df = df.sort_values('distance')
df = df.head(return_top_n)
# Convert the resulting dataframe into a list of dictionaries
results = []
for i,row in df.iterrows():
results.append({
'street': row['street'],
'street_code': row['street_code'],
'borough_code': row['borough_code'],
'node': list(self._get_intersection(
cross_street_code, row['street_code']
)) if cross_street_code else None
})
return results
def address(self, house_number, street, borough, drivable=True):
"""
Given an address as house number, street and borough, return the
segment id, physical id, blockface id and side of street of that
address.
Parameters
----------
house_number : str or int
The house number, including hyphens for Queens addresses.
street : str
borough : str
The borough code, abbreviation or name.
drivable : bool, optional
Whether to only return drivable segments. (Default True)
Returns
-------
dict
dict of segment id, physical id, blockface id and side of street
"""
street_code = self._get_street_code(borough, street)
street_normalized = self.normalize_street_name(street)
# Create a list of possible geographic identifiers for this address.
# Sometimes the physical location of a building is not reflected
# in its address, so we'll use GeoSupport to identify other options.
# The first one is simply the house number and street code.
lgis = [(house_number, street_code)]
# Then pass the address to GeoSupport and iterate through its
# list of geographic identifiers and add all of them.
for lgi in self.geosupport.address(
borough=borough, street=str(house_number) + ' ' + street_normalized
)['LIST OF GEOGRAPHIC IDENTIFIERS']:
lgis.append((
lgi['High House Number'],
lgi['Borough Code'] + lgi['5-Digit Street Code']
))
# Eliminate generic segments and, if drivable is True, only include
# drivable segments.
df = self.lion_df[
(
self.lion_df['traffic_direction'].isin(['W', 'A', 'T'])
if drivable else True
) & (~self.lion_df['segment_type'].isin(['G']))
]
# A function to determine if an address is on the given side of the
# street.
def same_parity(a, b):
return (a % 2) == (b % 2)
for house, street_code in lgis:
# First, normalize the house number.
if type(house) == str:
if '-' in house:
if street_code[0] == '4':
# If the street is in Queens and the number contains
# a hyphen, then convert it into the format tha LION
# uses by multiplying the first part by 1000 and adding
# the second.
a,b = house.split('-')
house = 1000*int(a) + int(b)
else:
# If not Queens, treat it as a range and use the first.
house = int(house.split('-')[0])
else:
house = int(house)
# Get a dataframe of the segments that could match the given
# street code and house number.
segments = df[
(df['original_street_code'] == street_code[:6]) &
(
(
(df['from_left'] <= house) & (df['to_left'] >= house) &
same_parity(df['from_left'], house)
) | (
(df['from_right'] <= house) & (df['to_right'] >= house) &
same_parity(df['from_right'], house)
)
)
]
# If there are matches, check whether the address matches the
# left or right side.
for i,row in segments.iterrows():
if (row['from_left'] != 0) and (
(row['from_left'] % 2) == (house % 2)
):
return {
'segment_id': row['segment_id'],
'physical_id': row['physical_id'],
'blockface_id': row['left_blockface_id'],
'side': 'L'
}
if (row['from_right'] != 0) and (
(row['from_right'] % 2) == (house % 2)
):
return {
'segment_id': row['segment_id'],
'physical_id': row['physical_id'],
'blockface_id': row['right_blockface_id'],
'side': 'R'
}
def _get_terminators(self, nodes):
"""
Terminator segments are segments which exist at the point when a
multi-roadbed street becomes a single roadbed.
Geocoder uses these to prevent u-turns from one roadbed onto another
at these nodes where the terminator segments meet.
This function returns a dictionary which has a key for each
physical_id that is a terminator where the value is a set of all
terminator physical_ids which connect to that segment.
Parameters
----------
nodes : dict
The nodes dictionary from `_get_nodes_dict`
Returns
-------
dict of physical_id -> set of physical_ids
"""
terminators = {}
for node in nodes:
# Get all the physical_ids that are terminator segments that
# connect to the given nodes.
pids = [
self.lion_segments[sid]['physical_id']
for sid in nodes[node]['segments']
if self.lion_segments[sid]['segment_type'] == 'T'
]
# Create a set of the physical_ids connected to this node and then
# update that group with any other existing groups that overlap.
group = set(pids)
for p in pids:
if p in terminators:
group.update(terminators[p])
# TODO: I think this is supposed to be
# `for p in group:`
# instead of in pids.
for p in pids: # Add the group to the dictionary of terminators.
terminators[p] = group
return terminators
@method_file_cache('network.pkl')
def _build_networks(self, network_type):
print("Building Routing Networks...")
# Create a basic network for both LION and CSCL
# This is needed to create the physical_id geometry, at least.
lion_network = build_monodirectional_network(self.lion_df, 'segment_id')
cscl_network = build_monodirectional_network(self.lion_df, 'physical_id')
drop_internal_nodes(cscl_network)
# Update nodes with whether it's an internal node or not.
cscl_nodes = [n.split(':')[1] for n in cscl_network.nodes if 'node' in n]
self.nodes_df['internal'] = ~self.nodes_df['node'].isin(cscl_nodes)
nodes = _get_nodes_dict(self.nodes_df)
def merge_geometry(group):
"""
Merge the segments into a physical ID to get attributes that depend
on the order of the segments.
Parameters
----------
group : pandas.DataFrame
A subset of lion_df with all the segments of a single physical_id
Returns
-------
shapely.LineString
"""
# Get all the segments and nodes that are part of the given physical_id
nodes = set(
['segment_id:%s' % i for i in group['segment_id']] +
['node:%s' % i for i in group['node_from']] +
['node:%s' % i for i in group['node_to']]
)
# Get a subgraph of lion_network with all the nodes and segments
# and all the edges between them.
subnetwork = lion_network.subgraph(nodes)
# Next, order all of the nodes in subnetwork so that the geometry
# will be in the right order to be merged.
try:
# Most physical_ids can be sorted via topological_sort
stretch = nx.topological_sort(subnetwork)
segments = [s.split(':')[1] for s in stretch if 'segment_id' in s]
except:
# But a few physical ids have cicrles/cycles, so in that case,
# get an order by trying all combinations of start/end point
# and use the first one that contains all the nodes.
for a, b in itertools.combinations(nodes, 2):
try:
stretch = nx.shortest_path(subnetwork, a, b)
if len(stretch) == len(nodes):
break
except:
pass
segments = [s.split(':')[1] for s in stretch if 'segment_id' in s]
# Create a LineString from the individual coordinates in the
# ordered list of segments.
coords = [
c for s in segments
for c in self.lion_segments[s]['geometry'].coords
]
coords = drop_consecutive_duplicates(coords)
return LineString(coords)
geometry = self.lion_df.groupby('physical_id').apply(
merge_geometry
).rename('geometry')
terminators = self._get_terminators(nodes)
terminators = self.lion_df.groupby('physical_id').apply(
lambda g: terminators.get(g.iloc[0]['physical_id'])
).rename('terminator_group')
# Create the DataFrame and dict for physical_ids
cscl_segments_df = _get_cscl_segments_df(self.lion_df)
cscl_segments_df = cscl_segments_df.join(
geometry, how='left'
).join(terminators, how='left')
cscl_segments = cscl_segments_df.to_dict('index')
if self.network_type == 'cscl':
segments = cscl_segments
network = cscl_network
else:
segments = self.lion_segments
network = lion_network
# Create a directional network of the given type
# This network still has nodes
node_network = build_directional_network(network, segments)
# Create a network where segments connect directly to segments.
segment_network = build_segment_network(
node_network, default_cost_function(
segments, nodes, turn_cost=100000, intersection_cost=0
)
)
return (
cscl_segments_df, cscl_segments, node_network, segment_network,
self.nodes_df, nodes
)
def get_segment(self, segment):
"""
Given a segment_id string, return a dictionary from `self.segments`.
Parameters
----------
segment : str
A segment_id in a format accepted by `parse_geometry`.
Returns
-------
dict
"""
type_, id_, side = self.geometry_regex.match(segment).groups()
return self.segments.get(id_, None)
def normalize_segment_id(self, id_):
if self.segment_column == 'segment_id':
return id_.zfill(7)
else:
return str(int(id_))
def parse_geometry(self, geometry):
"""
Parse a "geometry" string and return a standardize geometry string,
the type, id, and side of street.
A geometry string is the type of geometry, followed by the id and
optionally a side of street.
For example:
node:0055555M
segment_id:0005555L
physical_id:555R
Parameters
----------
geometry : str
Returns
-------
geometry : str
The geometry in a standardized format
type : str
node, segment_id or physical_id
id : str
The geometry ID
side_of_street : str
A letter for the side of street. One of: '', 'R', 'L', 'E', 'B'
"""
try:
type_, id_, letter = self.geometry_regex.match(geometry).groups()
if not type_:
if (id_.zfill(7) + letter) in self.nodes:
type_ = 'node'
elif self.normalize_segment_id(id_) in self.segments:
type_ = self.segment_column
assert type_ in ['node', self.segment_column]
if type_ == 'node':
id_ = id_.zfill(7) + letter
letter = ''
assert id_ in self.nodes
else:
assert letter in ['', 'R', 'L', 'E', 'B']
id_ = self.normalize_segment_id(id_)
assert id_ in self.segments
return '%s:%s%s' % (type_, id_, letter), type_, id_, letter
except:
raise ValueError("Unrecognized geometry: %s" % geometry)
def get_street_stretch_by_geometry(self, geometry_1, geometry_2,
on_street_code=None):
"""
Given two endpoint geometries, return a shortest path street stretch.
Geometries can either be nodes or segments or a combination of the two.
For nodes, if an optional on_street_code is provided, only start and
end on streets on the given street code.
This function works by adding temporary nodes to the Geocoder's
segment_network called START and END. START connects to
geometry_1 and END connects to geometry_2.
Then find a shortest path from START to END and return it as a
StreetStretch.
This function will always return a result if there is a possible path
from geometry_1 to geometry_2 even if it is not actually a "stretch"
(not all along one on street). You can use the StreetStretch object's
attributes to determine if the stretch is valid for your use case.
Parameters
----------
geometry_1, geometry_2 : str
Start and endpoints for the stretch
on_street_code : str, optional
An on street code to start and end on.
Returns
-------
StreetStretch
"""
geometry_1, type_1, id_1, side_1 = self.parse_geometry(geometry_1)
geometry_2, type_2, id_2, side_2 = self.parse_geometry(geometry_2)
# Add start node (START -> geometry_1)
if type_1 == 'node':
# Since we will be routing on the segment network, for nodes connect
# START to all segments that the node connects to.
for segment in self.node_network[geometry_1]:
# If on_street_code is provided, only connect to segments
# on the given street.
if (
(on_street_code is None) or
(on_street_code in self.get_segment(segment)['street_code'])
):
self.segment_network.add_edge('START', segment, weight=1)
elif type_1 == self.segment_column:
# For segments, connect the segment itself.
# If side isn't given, allow both sides.
if not side_1:
for side in ['L', 'R']:
self.segment_network.add_edge('START', geometry_1 + side, weight=1)
else:
self.segment_network.add_edge('START', geometry_1, weight=1)
# Add End Node in the same fashion, but connecting geometry_2 -> END
if type_2 == 'node':
for segment in self.node_network.predecessors(geometry_2):
if (
(on_street_code is None) or
(on_street_code in self.get_segment(segment)['street_code'])
):
self.segment_network.add_edge(segment, 'END', weight=1)
elif type_2 == self.segment_column:
if not side_2:
for side in ['L', 'R']:
self.segment_network.add_edge(geometry_2 + side, 'END', weight=1)
else:
self.segment_network.add_edge(geometry_2, 'END', weight=1)
try:
path = nx.bidirectional_dijkstra(
self.segment_network, 'START', 'END', weight='weight'
)[1]
return StreetStretch(self, path[1:-1])
finally:
# Even if there's an error, make sure to remove START and END from
# the network.
for n in ['START', 'END']:
self.segment_network.remove_node(n)
def get_street_stretch_by_code(self, on_street_code, from_street_code,
to_street_code):
"""
Given an on street code, from street code and to street code,
return a list of possible stretches.
There can be more than one because sometime streets intersect multiple
times.
Parameters
----------
on_street_code, from_street_code, to_street_code : str
The street codes of the on, from and to streets
Returns
-------
list of StreetStretch
A list of stretches for all combinations of from and to
intersections sorted from shortest to longest.
"""
on_street_code = self.normalize_street_code(on_street_code)
from_street_code = self.normalize_street_code(from_street_code)
to_street_code = self.normalize_street_code(to_street_code)
nodes_from = self._get_intersection(on_street_code, from_street_code)
nodes_to = self._get_intersection(on_street_code, to_street_code)
stretches = []
for node_from in nodes_from:
for node_to in nodes_to:
if node_from != node_to:
try:
stretches.append(self.get_street_stretch_by_geometry(
'node:' + node_from, 'node:' + node_to,
on_street_code=on_street_code
))
except:
pass
return sorted(stretches, key=len)
def __str__(self):
return "Geocoder (lion_version=%s, crs=%s, index_crs=%s)" % (
self.lion_version, self.crs, self.spatial_index_crs
)
def __init__(self, lion_version, crs='epsg:2263',
lion_loader=load_lion_gdf, force_rebuild=False,
network_type='cscl',
include_spatial_index=False, spatial_index_crs=None, **kwargs):
self.crs = crs
self.lion_version = lion_version.lower()
self.lion_loader = lion_loader
self.force_rebuild = force_rebuild
self.cache_path = MICHI_HOME / self.lion_version / 'geocoder'
if not self.cache_path.exists():
self.cache_path.mkdir(parents=True)
# ensure CRS parameters are parsable before instantiation
pyproj.crs.CRS(crs)
if spatial_index_crs:
pyproj.crs.CRS(spatial_index_crs)
# Instantiate Geosupport
try:
if geosupport_version is None:
geosupport_version = self.lion_version
self.geosupport = Geosupport(geosupport_version=version)
except:
# Use default geosupport
self.geosupport = Geosupport()
warn(
"Using default Geosupport. "
"May not match LION %s" % self.lion_version
)
(
self.lion_df, self.nodes_df,
self.lion_segments, self.streets,
self.street_code_map
) = self._load_lion_data()
self.street_names = self._build_search_data()
self.network_type = network_type
if network_type == 'cscl':
self.segment_column = 'physical_id'
self.segment_dict = 'cscl_segments'
elif network_type == 'lion':
self.segment_column = 'segment_id'
self.segment_dict = 'lion_segments'
# A regex to parse geometry strings, to be used by `self.parse_geometry`
self.geometry_regex = re.compile(r'([a-z_]*):*(\d+)([A-Z]{0,1})')
(
self.cscl_segments_df, self.cscl_segments,
self.node_network, self.segment_network,
self.nodes_df, self.nodes
) = self._build_networks(self.network_type)
self.segments = getattr(self, self.segment_dict)
# SearchMixin
# TODO Don't skip this, testing!
#super().__init__(clear_cache=clear_cache, **kwargs)
# get cache key to uniquely recognize spatial indexes
#cache_key = get_cache_key(('_load_lion_data',) + (self,) + ({},))
# load spatial indexes
self.include_spatial_index = include_spatial_index
self.spatial_index_crs = spatial_index_crs
if self.include_spatial_index:
self.lion_index = _get_spatial_index(
self.lion_df, self.crs, self.spatial_index_crs, lion_version,
'segment_index', 'segment_id'
)
self.node_index = _get_spatial_index(
self.nodes_df, self.crs, self.spatial_index_crs, lion_version,
'node_index', 'node'
)
from multiprocessing import Pool, cpu_count
from sqlalchemy import create_engine
from geosupport import Geosupport, GeosupportError
from pathlib import Path
import pandas as pd
import usaddress
import json
import re
import os
g = Geosupport()
def geocode(inputs):
bin = inputs.get('bin', '')
try:
geo = g['BN'](bin=bin, mode='tpad')
#geo = g['BN'](bin=bin, mode_switch='X')
#geo = g.BN(mode='extended', bin=bin)
#print(geo)
except GeosupportError as e:
geo = e.result
geo = geo_parser(geo)
geo.update(inputs)
return geo
def geo_parser(geo):
#million_bins = ['1000000', '2000000', '3000000', '4000000', '5000000']
# -*- coding: utf-8 -*- """ Created on Sun Dec 8 20:46:17 2019 @author: ASchwenker """ from geosupport import Geosupport g = Geosupport() g.help()
def setUpClass(cls):
g = Geosupport()
cls.suggest = GeosupportSuggest(g)
def setUpClass(cls):
g = Geosupport()
cls.suggest = GeosupportSuggest(g)
cls.app = api.app.test_client()
def setUpClass(cls):
cls.geosupport = Geosupport()
