File size: 3,103 Bytes
ae2aa11 |
1 2 3 4 5 6 7 8 9 10 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 |
import numpy as np
import pandas as pd
DAMAGE_STATES = ['none', 'minor', 'severe', 'collapse']
REPORT_SOURCES = {
'automated_sensor': {'reliability': 0.85, 'weight': 0.25},
'phone_call': {'reliability': 0.60, 'weight': 0.50},
'social_media': {'reliability': 0.40, 'weight': 0.20},
'inspector': {'reliability': 0.95, 'weight': 0.05}
}
def generate_noisy_report(true_state, reliability):
if np.random.random() < reliability:
return true_state
other_states = [s for s in DAMAGE_STATES if s != true_state]
true_idx = DAMAGE_STATES.index(true_state)
weights = []
for state in other_states:
state_idx = DAMAGE_STATES.index(state)
distance = abs(state_idx - true_idx)
weight = 1.0 / (distance + 1)
weights.append(weight)
weights = np.array(weights) / sum(weights)
return np.random.choice(other_states, p=weights)
def generate_building_reports(building, lambda_rates, max_time_hours=3, seed=None):
if seed is not None:
np.random.seed(seed)
true_damage = building['true_damage']
lambda_rate = lambda_rates[true_damage]
reports = []
current_time = 0
max_time_minutes = max_time_hours * 60
while current_time < max_time_minutes:
lambda_per_minute = lambda_rate / 60.0
time_to_next = np.random.exponential(1.0 / lambda_per_minute)
current_time += time_to_next
if current_time >= max_time_minutes:
break
source_types = list(REPORT_SOURCES.keys())
source_weights = [REPORT_SOURCES[s]['weight'] for s in source_types]
source = np.random.choice(source_types, p=source_weights)
reliability = REPORT_SOURCES[source]['reliability']
reported_state = generate_noisy_report(true_damage, reliability)
reports.append({
'time_minutes': round(current_time, 2),
'source': source,
'reported_state': reported_state,
'building_id': building['building_id'],
'building_type': building['building_type'],
'true_damage': building['true_damage']
})
return reports
def generate_all_reports(buildings_df, lambda_rates=None, max_time_hours=3, seed=42):
if lambda_rates is None:
lambda_rates = {
'collapse': 8.0,
'severe': 3.0,
'minor': 0.8,
'none': 0.2
}
np.random.seed(seed)
all_reports = []
for idx, building in buildings_df.iterrows():
building_seed = seed + idx if seed is not None else None
reports = generate_building_reports(
building.to_dict(),
lambda_rates,
max_time_hours,
building_seed
)
all_reports.extend(reports)
reports_df = pd.DataFrame(all_reports)
if len(reports_df) > 0:
reports_df = reports_df.sort_values('time_minutes').reset_index(drop=True)
return reports_df
def get_report_reliability(source):
return REPORT_SOURCES[source]['reliability'] |