feat: Implement multi-source weather data collection with caching, rate limiting, and disk persistence.

config/config.yaml

@@ -1,76 +1,97 @@
-# Weather API Configuration
 weather:
   timeout: 30
-
-# Target Cities
 cities:
-  - id: "london"
+- id: london
-    city: "London"
+  city: London
-    country: "UK"
+  country: UK
-    latitude: 51.5074
+  latitude: 51.5074
-    longitude: -0.1278
+  longitude: -0.1278
-  - id: "paris"
+- id: paris
-    city: "Paris"
+  city: Paris
-    country: "France"
+  country: France
-    latitude: 48.8566
+  latitude: 48.8566
-    longitude: 2.3522
+  longitude: 2.3522
-  - id: "ankara"
+- id: ankara
-    city: "Ankara"
+  city: Ankara
-    country: "Turkey"
+  country: Turkey
-    latitude: 39.9334
+  latitude: 39.9334
-    longitude: 32.8597
+  longitude: 32.8597
-  - id: "new_york"
+- id: new_york
-    city: "New York"
+  city: New York
-    country: "USA"
+  country: USA
-    latitude: 40.7128
+  latitude: 40.7128
-    longitude: -74.0060
+  longitude: -74.006
-  - id: "chicago"
+- id: chicago
-    city: "Chicago"
+  city: Chicago
-    country: "USA"
+  country: USA
-    latitude: 41.8781
+  latitude: 41.8781
-    longitude: -87.6298
+  longitude: -87.6298
-  - id: "lucknow"
+- id: lucknow
-    city: "Lucknow"
+  city: Lucknow
-    country: "India"
+  country: India
-    latitude: 26.7606
+  latitude: 26.7606
-    longitude: 80.8893
+  longitude: 80.8893
-  - id: "sao paulo"
+- id: sao paulo
-    city: "São Paulo"
+  city: São Paulo
-    country: "Brazil"
+  country: Brazil
-    latitude: -23.4356
+  latitude: -23.4356
-    longitude: -46.4731
+  longitude: -46.4731
-  - id: "munich"
+- id: munich
-    city: "Munich"
+  city: Munich
-    country: "Germany"
+  country: Germany
-    latitude: 48.3538
+  latitude: 48.3538
-    longitude: 11.7861
+  longitude: 11.7861
-  - id: "hong_kong"
+- id: hong_kong
-    city: "Hong Kong"
+  city: Hong Kong
-    country: "China"
+  country: China
-    latitude: 22.3080
+  latitude: 22.308
-    longitude: 113.9185
+  longitude: 113.9185
-  - id: "shanghai"
+- id: shanghai
-    city: "Shanghai"
+  city: Shanghai
-    country: "China"
+  country: China
-    latitude: 31.1434
+  latitude: 31.1434
-    longitude: 121.8052
+  longitude: 121.8052
-  - id: "singapore"
+- id: singapore
-    city: "Singapore"
+  city: Singapore
-    country: "Singapore"
+  country: Singapore
-    latitude: 1.3644
+  latitude: 1.3644
-    longitude: 103.9915
+  longitude: 103.9915
-  - id: "tokyo"
+- id: tokyo
-    city: "Tokyo"
+  city: Tokyo
-    country: "Japan"
+  country: Japan
-    latitude: 35.5523
+  latitude: 35.5523
-    longitude: 139.7798
+  longitude: 139.7798
-  - id: "tel_aviv"
+- id: tel_aviv
-    city: "Tel Aviv"
+  city: Tel Aviv
-    country: "Israel"
+  country: Israel
-    latitude: 32.0114
+  latitude: 32.0114
-    longitude: 34.8867
+  longitude: 34.8867
-# Logging
+- id: chengdu
+  city: Chengdu
+  country: China
+  latitude: 30.5785
+  longitude: 103.9471
+- id: chongqing
+  city: Chongqing
+  country: China
+  latitude: 29.7196
+  longitude: 106.6416
+- id: shenzhen
+  city: Shenzhen
+  country: China
+  latitude: 22.6393
+  longitude: 113.8107
+- id: beijing
+  city: Beijing
+  country: China
+  latitude: 40.0801
+  longitude: 116.5846
+- id: wuhan
+  city: Wuhan
+  country: China
+  latitude: 30.7838
+  longitude: 114.2081
 logging:
-  level: "INFO"
+  level: INFO
-  rotation: "10 MB"
+  rotation: 10 MB
-  retention: "10 days"
+  retention: 10 days

src/analysis/deb_algorithm.py

@@ -2,7 +2,7 @@ import os
 import json
 from datetime import datetime, timedelta
 import requests
-from src.analysis.settlement_rounding import wu_round
+from src.analysis.settlement_rounding import wu_round, apply_city_settlement, is_exact_settlement_city
 
 # Cross-platform file locking
 import sys
@@ -442,8 +442,8 @@ def get_deb_accuracy(city_name):
             continue
 
         total += 1
-        deb_wu = wu_round(deb_pred)
+        deb_wu = apply_city_settlement(city_name, deb_pred)
-        actual_wu = wu_round(actual)
+        actual_wu = apply_city_settlement(city_name, actual)
         if deb_wu == actual_wu:
             hits += 1
         errors.append(abs(deb_pred - actual))
@@ -507,13 +507,13 @@ def get_mu_accuracy(city_name):
 
         total += 1
         mu_errors.append(abs(mu_val - actual))
-        if wu_round(mu_val) == wu_round(actual):
+        if apply_city_settlement(city_name, mu_val) == apply_city_settlement(city_name, actual):
             mu_hits += 1
 
         # Brier Score from probability snapshot
         prob_snap = record.get("prob_snapshot", [])
         if prob_snap:
-            actual_wu = wu_round(actual)
+            actual_wu = apply_city_settlement(city_name, actual)
             bs = 0.0
             for entry in prob_snap:
                 predicted_p = entry.get("p", 0)

src/analysis/market_alert_engine.py

@@ -9,7 +9,7 @@ import re
 from datetime import datetime, timezone
 from typing import Any, Dict, List, Optional, Tuple
 
-from src.analysis.settlement_rounding import wu_round
+from src.analysis.settlement_rounding import wu_round, apply_city_settlement, is_exact_settlement_city
 
 
 def _sf(v: Any) -> Optional[float]:
@@ -716,7 +716,7 @@ def _extract_market_snapshot(city_weather: Dict[str, Any]) -> Dict[str, Any]:
             market_url = f"https://polymarket.com/market/{slug}"
 
     anchor_today_high_c, anchor_model = _extract_multi_model_anchor_high_c(city_weather)
-    anchor_settlement = wu_round(anchor_today_high_c)
+    anchor_settlement = apply_city_settlement(city, anchor_today_high_c)
     forecast_bucket = _pick_bucket_for_forecast(
         rows=all_bucket_rows,
         forecast_settlement=anchor_settlement,

src/analysis/settlement_rounding.py

@@ -18,3 +18,24 @@ def wu_round(value: Optional[Number]) -> Optional[int]:
         return int(math.floor(x + 0.5))
     return int(math.ceil(x - 0.5))
 
+
+def is_exact_settlement_city(city: str) -> bool:
+    """是否为不四舍五入的精确结算城市"""
+    if not city:
+        return False
+    c = str(city).lower().strip()
+    return c in ["hong kong", "hk", "taipei", "tpe", "臺北", "台北", "香港"]
+
+
+def apply_city_settlement(city: str, value: Optional[Number]) -> Optional[int]:
+    """
+    根据城市返回最终的结算值：
+    - 香港/台北: 向下取整 (e.g. 28.9 -> 28)
+    - 其他: WU 规则四舍五入
+    """
+    if value is None:
+        return None
+    if is_exact_settlement_city(city):
+        return int(math.floor(float(value)))
+    return wu_round(value)
+

src/analysis/trend_engine.py

@@ -15,7 +15,7 @@ from src.analysis.deb_algorithm import (
     update_daily_record,
     _is_excluded_model_name,
 )
-from src.analysis.settlement_rounding import wu_round
+from src.analysis.settlement_rounding import wu_round, apply_city_settlement, is_exact_settlement_city
 from src.data_collection.city_registry import CITY_REGISTRY
 from src.data_collection.city_risk_profiles import get_city_risk_profile
 
@@ -429,7 +429,7 @@ def analyze_weather_trend(
     forecast_miss_deg = 0.0
 
     if is_dead_market:
-        settled_wu = wu_round(max_so_far) if max_so_far is not None else 0
+        settled_wu = apply_city_settlement(city_name, max_so_far) if max_so_far is not None else 0
         dead_msg = (
             f"🎲 <b>结算预测</b>：已锁定 {settled_wu}{temp_symbol} "
             f"({settlement_source_label} 死盘确认)"
@@ -481,7 +481,7 @@ def analyze_weather_trend(
 
         # Probability Engine
         probs_result = calculate_prob_distribution(
-            mu, sigma, max_so_far, temp_symbol
+            mu, sigma, max_so_far, temp_symbol, city_name
         )
         mu = probs_result.get("mu", mu)
         probabilities = probs_result.get("probabilities", [])
@@ -514,7 +514,7 @@ def analyze_weather_trend(
 
     # === Settlement boundary ===
     if max_so_far is not None:
-        settled = wu_round(max_so_far)
+        settled = apply_city_settlement(city_name, max_so_far)
         fractional = max_so_far - int(max_so_far)
         dist_to_boundary = abs(fractional - 0.5)
         if dist_to_boundary <= 0.3:
@@ -585,7 +585,7 @@ def analyze_weather_trend(
     if max_so_far is not None:
         ai_features.append(
             f"🏔️ 今日实测最高温: {max_so_far}{temp_symbol} "
-            f"({settlement_source_label}结算={wu_round(max_so_far)}{temp_symbol})。"
+            f"({settlement_source_label}结算={apply_city_settlement(city_name, max_so_far)}{temp_symbol})。"
         )
     if city_name:
         _profile = get_city_risk_profile(city_name)
@@ -634,7 +634,7 @@ def analyze_weather_trend(
                 for t, p in sorted_probs[:4]
             ]
         elif is_dead_market and max_so_far is not None:
-            _prob_list = [{"value": wu_round(max_so_far), "probability": 1.0}]
+            _prob_list = [{"value": apply_city_settlement(city_name, max_so_far), "probability": 1.0}]
 
         update_daily_record(
             city_name,
@@ -672,14 +672,14 @@ def analyze_weather_trend(
         "forecast_miss_deg": forecast_miss_deg,
         "max_so_far": max_so_far,
         "cur_temp": cur_temp,
-        "wu_settle": wu_round(max_so_far) if max_so_far is not None else None,
+        "wu_settle": apply_city_settlement(city_name, max_so_far) if max_so_far is not None else None,
     }
     display_str = "\n".join(insights) if insights else ""
     return display_str, "\n".join(ai_features), structured
 
 
 def calculate_prob_distribution(
-    mu: float, sigma: float, max_so_far: Optional[float], temp_symbol: str
+    mu: float, sigma: float, max_so_far: Optional[float], temp_symbol: str, city_name: str = ""
 ) -> Dict[str, Any]:
     """
     Generalized Gaussian probability distribution calculation.
@@ -691,17 +691,26 @@ def calculate_prob_distribution(
         # 0.5 * (1 + erf( (x-m)/(s*sqrt(2)) ))
         return 0.5 * (1 + math.erf((x - m) / (sigma * math.sqrt(2))))
 
-    min_possible_wu = wu_round(max_so_far) if max_so_far is not None else -999
+    min_possible_wu = apply_city_settlement(city_name, max_so_far) if max_so_far is not None else -999
     probs = {}
     
     # Range: mu +/- 3 sigma or at least +/- 2 degrees
     search_range = max(2, int(sigma * 2.5))
-    target_mu = wu_round(mu)
+    is_exact = is_exact_settlement_city(city_name)
+    target_mu = apply_city_settlement(city_name, mu)
+    if is_exact:
+        target_mu = int(math.floor(mu))
     
     for n in range(target_mu - search_range, target_mu + search_range + 1):
         if n < min_possible_wu:
             continue
-        p = _norm_cdf(n + 0.5, mu, sigma) - _norm_cdf(n - 0.5, mu, sigma)
+        if is_exact:
+            # 向下取整的概率区间为 [n, n + 1)
+            p = _norm_cdf(n + 1.0, mu, sigma) - _norm_cdf(n, mu, sigma)
+        else:
+            # 常规四舍五入的概率区间为 [n - 0.5, n + 0.5)
+            p = _norm_cdf(n + 0.5, mu, sigma) - _norm_cdf(n - 0.5, mu, sigma)
+            
         if p > 0.01:
             probs[n] = p
 
@@ -713,9 +722,10 @@ def calculate_prob_distribution(
         norm_probs = {k: v / total_p for k, v in probs.items()}
         sorted_probs = sorted(norm_probs.items(), key=lambda x: x[1], reverse=True)
         for t, p in sorted_probs[:4]:
+            rng_str = f"[{t}.0~{t+1}.0)" if is_exact else f"[{t-0.5}~{t+0.5})"
             probabilities.append({
                 "value": int(t),
-                "range": f"[{t-0.5}~{t+0.5})",
+                "range": rng_str,
                 "probability": round(p, 3)
             })
 

src/data_collection/city_registry.py

@@ -357,6 +357,77 @@ CITY_REGISTRY = {
         "distance_km": 13.0,
         "warning": "机场位于东北侧开阔区，午后混合层增强时与核心城区体感温度可能出现偏差。",
     },
+
+    "chengdu": {
+        "name": "Chengdu",
+        "lat": 30.5785,
+        "lon": 103.9471,
+        "icao": "ZUUU",
+        "tz_offset": 28800,
+        "use_fahrenheit": False,
+        "is_major": True,
+        "risk_level": "medium",
+        "risk_emoji": "🟡",
+        "airport_name": "成都双流国际机场",
+        "distance_km": 16.0,
+        "warning": "盆地地形，夜间降温较慢，多云雾影响日照升温。",
+    },
+    "chongqing": {
+        "name": "Chongqing",
+        "lat": 29.7196,
+        "lon": 106.6416,
+        "icao": "ZUCK",
+        "tz_offset": 28800,
+        "use_fahrenheit": False,
+        "is_major": True,
+        "risk_level": "high",
+        "risk_emoji": "🔴",
+        "airport_name": "重庆江北国际机场",
+        "distance_km": 19.0,
+        "warning": "四大火炉之一，夏季持续高温，夜间温度偏高，湿度大。",
+    },
+    "shenzhen": {
+        "name": "Shenzhen",
+        "lat": 22.6393,
+        "lon": 113.8107,
+        "icao": "ZGSZ",
+        "tz_offset": 28800,
+        "use_fahrenheit": False,
+        "is_major": True,
+        "risk_level": "medium",
+        "risk_emoji": "🟡",
+        "airport_name": "深圳宝安国际机场",
+        "distance_km": 32.0,
+        "warning": "近海受海洋调节，热岛效应强，日夜温差较小。",
+    },
+    "beijing": {
+        "name": "Beijing",
+        "lat": 40.0801,
+        "lon": 116.5846,
+        "icao": "ZBAA",
+        "tz_offset": 28800,
+        "use_fahrenheit": False,
+        "is_major": True,
+        "risk_level": "medium",
+        "risk_emoji": "🟡",
+        "airport_name": "北京首都国际机场",
+        "distance_km": 25.0,
+        "warning": "北方内陆，夏季干热，冬季寒冷，春秋易受风沙影响。",
+    },
+    "wuhan": {
+        "name": "Wuhan",
+        "lat": 30.7838,
+        "lon": 114.2081,
+        "icao": "ZHHH",
+        "tz_offset": 28800,
+        "use_fahrenheit": False,
+        "is_major": True,
+        "risk_level": "high",
+        "risk_emoji": "🔴",
+        "airport_name": "武汉天河国际机场",
+        "distance_km": 26.0,
+        "warning": "江汉平原，水汽充足，夏季常现长时闷热高温。",
+    },
 }
 
 ALIASES = {
@@ -406,4 +477,14 @@ ALIASES = {
     "華沙": "warsaw",
     "马德里": "madrid",
     "馬德里": "madrid",
+    "ctu": "chengdu", "cd": "chengdu",
+    "ckg": "chongqing", "cq": "chongqing",
+    "szx": "shenzhen", "sz": "shenzhen",
+    "pek": "beijing", "bjs": "beijing", "bj": "beijing",
+    "wuh": "wuhan", "wh": "wuhan",
+    "成都": "chengdu",
+    "重庆": "chongqing",
+    "深圳": "shenzhen",
+    "北京": "beijing",
+    "武汉": "wuhan",
 }

src/data_collection/weather_sources.py

@@ -34,6 +34,11 @@ class WeatherDataCollector:
         "seoul": ["RKSI", "RKSS"],
         "hong kong": ["VHHH", "VMMC", "ZGSZ"],
         "taipei": ["RCSS", "RCTP"],
+        "chengdu": ["ZUUU", "ZUTF"],
+        "chongqing": ["ZUCK", "ZUPS"],
+        "shenzhen": ["ZGSZ", "ZGGG"],
+        "beijing": ["ZBAA", "ZBAD"],
+        "wuhan": ["ZHHH", "ZHES"],
         "shanghai": ["ZSPD", "ZSSS", "ZSNB", "ZSHC"],
         "singapore": ["WSSS", "WSAP", "WMKK"],
         "tokyo": ["RJTT", "RJAA", "RJAH", "RJTJ"],
@@ -1960,6 +1965,16 @@ class WeatherDataCollector:
             "taipei": "Taipei",
             "台北": "Taipei",
             "臺北": "Taipei",
+            "chengdu": "Chengdu",
+            "成都": "Chengdu",
+            "chongqing": "Chongqing",
+            "重庆": "Chongqing",
+            "shenzhen": "Shenzhen",
+            "深圳": "Shenzhen",
+            "beijing": "Beijing",
+            "北京": "Beijing",
+            "wuhan": "Wuhan",
+            "武汉": "Wuhan",
             "shanghai": "Shanghai",
             "上海": "Shanghai",
             "singapore": "Singapore",