11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278 | class ResidentLinkedDistributor(BaseDistributor):
"""
Generic distributor that links 'visitors' to 'venues' based on the residents
already present in those venues. Matches residents to visitor units
(individuals or households) geographically.
This replicates the JUNE 1 care home linking behavior while remaining generic.
"""
def __init__(self, config_file: str = None, config_dict: Dict = None):
super().__init__(config_file, config_dict)
# Configuration
self.target_venue_type = self.config.get("target_venue_type", "care_home")
self.resident_subset = self.config.get("resident_subset", "resident")
self.activity_map_key = self.config.get("activity_map_key", "leisure")
self.link_level = self.config.get(
"link_level", "household"
) # 'person' or 'household'
self.multiplier = self.config.get("multiplier", 1) # Visitors per resident
# Component managers
self._allocation = VenueAllocation(self)
self.visitor_filters = self.config.get("visitor_eligibility", {}).get(
"global_filters", []
)
self._pre_processed_filters = self._pre_process_filters(self.visitor_filters)
logger.info(
f"Initialized ResidentLinkedDistributor for '{self.target_venue_type}'"
)
logger.info(f" Link level: {self.link_level}, Multiplier: {self.multiplier}")
def allocate(self, world):
"""
Main allocation logic for large-scale populations.
"""
start_time = time.time()
logger.info("=" * 60)
logger.info(f"Starting ResidentLinkedDistributor: {self.target_venue_type}")
logger.info("=" * 60)
venues = world.venues_by_type(self.target_venue_type)
if not venues:
return {"total_links": 0}
# 1. Prepare vectorized population data
logger.info(f" Preparing population data...")
self._prepare_vectorized_data(world)
# 2. Setup filtering manager
logger.info(f" Setting up filtering manager...")
# 3. Batch processing by geography level
geo_level = self.config.get("geography_level", self.batch_geo_level)
geo_units = list(world.geography.get_units_by_level(geo_level).values())
logger.info(
f" Building '{self.target_venue_type}' links for {len(geo_units)} {geo_level}s"
)
# Calculate total residents for summary stats
total_residents = 0
for venue in venues:
res_subset = venue.subsets.get(self.resident_subset)
if res_subset:
total_residents += len(res_subset.members)
logger.info(
f" Found {len(venues)} target venues with {total_residents:,} total residents"
)
total_links = 0
total_venues_processed = 0
unit_count = len(geo_units)
report_interval = max(1, unit_count // 10)
for i, geo_unit in enumerate(geo_units):
people_in_unit = list(geo_unit.get_people())
if not people_in_unit:
continue
# B. Filter by eligibility
eligible_people = self._allocation.apply_global_filters(people_in_unit)
if not eligible_people:
continue
# C. Group by household units if configured
visitor_units = self._group_visitors(eligible_people)
if not visitor_units:
continue
# C. Get venues in this geo unit
unit_venues = [
v for v in venues if self._venue_matches_geo_unit(v, geo_unit)
]
if not unit_venues:
continue
links_created = self._allocate_batch(visitor_units, unit_venues)
total_links += links_created
total_venues_processed += len(unit_venues)
# Progress reporting
if (i + 1) % report_interval == 0 or (i + 1) == unit_count:
percent = ((i + 1) / unit_count) * 100
logger.info(
f" Progress: {i+1:,}/{unit_count:,} {geo_level}s processed ({percent:.1f}%)"
)
elapsed = time.time() - start_time
avg_links = total_links / total_residents if total_residents > 0 else 0
logger.info(
f"Built {total_links:,} total links (avg {avg_links:.1f} per resident) in {elapsed:.2f}s"
)
return {"total_links": total_links}
def _prepare_vectorized_data(self, world):
"""Build population arrays needed for vectorized filtering/grouping."""
if self.population_arrays:
return
needed_attrs = ["age", "sex", "residence.id"]
# Add attributes from filters
for f in self._pre_processed_filters:
if f.get("attribute"):
needed_attrs.append(f["attribute"])
needed_attrs = list(set(needed_attrs))
self._build_population_arrays(world.people, needed_attrs)
def _allocate_batch(self, visitor_units: List[List], venues: List) -> int:
"""
Fast pairing logic for a batch of venues and visitors.
"""
link_data = [] # List of (visitor_unit, venue)
# 1. Flatten residents from all venues in batch
# We just need to know how many visits are available across all venues
venue_capacities = []
for venue in venues:
res_subset = venue.subsets.get(self.resident_subset)
if not res_subset:
continue
num_residents = len(res_subset.members)
if num_residents == 0:
continue
venue_capacities.append((venue, num_residents))
logger.debug(
f" Batch matching: {len(visitor_units)} visitor units, {len(venue_capacities)} eligible venues"
)
if not venue_capacities:
return 0
# 2. Match with visitor units using multiplier
multiplier = self.multiplier
v_idx = 0
logger.debug(f" Multiplier: {multiplier}")
for venue, num_residents in venue_capacities:
# For each resident, we link 'multiplier' units
for _ in range(num_residents):
num_to_link = int(multiplier)
if multiplier % 1 > 0 and random.random() < (multiplier % 1):
num_to_link += 1
for _ in range(num_to_link):
if v_idx >= len(visitor_units):
break
link_data.append((visitor_units[v_idx], venue))
v_idx += 1
if v_idx >= len(visitor_units):
break
if v_idx >= len(visitor_units):
break
# 3. Bulk apply links
if link_data:
self._bulk_link_units_to_venues(link_data)
return len(link_data)
def _bulk_link_units_to_venues(self, link_data: List):
"""
Apply links in bulk to subsets and activity maps.
"""
from may.population import Subset
venue_subset_to_people = {} # (venue, subset_key) -> [list of people]
subset_key = self.config.get("subset_key", "visitor")
activity_key = self.activity_map_key
venue_type = self.target_venue_type
for unit, venue in link_data:
key = (venue, subset_key)
if key not in venue_subset_to_people:
venue_subset_to_people[key] = []
venue_subset_to_people[key].extend(unit)
# Apply in bulk
for (venue, subset_key), people in venue_subset_to_people.items():
# Get or create subset
if subset_key not in venue.subsets:
subset = Subset(
venue=venue, subset_index=len(venue.subsets), subset_name=subset_key
)
venue.subsets[subset_key] = subset
else:
subset = venue.subsets[subset_key]
# Bulk add members
subset.members.update(people)
# Bulk update people
for p in people:
if activity_key not in p.activity_map:
p.activity_map[activity_key] = {}
if venue_type not in p.activity_map[activity_key]:
p.activity_map[activity_key][venue_type] = []
# IMPORTANT: Use a list of unique subsets per (activity, venue_type)
if subset not in p.activity_map[activity_key][venue_type]:
p.activity_map[activity_key][venue_type].append(subset)
if activity_key not in p.activities:
p.add_activity(activity_key)
def _venue_matches_geo_unit(self, venue, geo_unit) -> bool:
"""Efficiently check if venue belongs to geo_unit."""
v_geo = venue.geographical_unit
while v_geo:
if v_geo.id == geo_unit.id:
return True
if v_geo.level == geo_unit.level: # Reached same level, no match
return False
v_geo = v_geo.parent
return False
def _group_visitors(self, people: List) -> List[List]:
"""Group people into visitor units."""
if self.link_level != "household":
return [[p] for p in people]
# Read household IDs from the prepared population arrays, then group
# the original Person objects by those IDs.
p_indices = [self.person_id_to_index[p.id] for p in people]
hh_ids = self.population_arrays["residence.id"][p_indices]
# The dictionary groups the original Person objects for this batch.
hh_to_members = {}
for i, person in enumerate(people):
hh_id = hh_ids[i]
if hh_id == -1:
continue
if hh_id not in hh_to_members:
hh_to_members[hh_id] = []
hh_to_members[hh_id].append(person)
return list(hh_to_members.values())
|