功能细节优化

backend/limit_analysis.py

@@ -1,10 +1,4 @@
-"""涨跌停分析 — 连板股追踪、炸板率统计、敢死队排行。
-
-功能：
-1. 连板股追踪器
-2. 炸板率统计
-3. 涨停敢死队排行
-"""
+"""涨跌停分析 — 连板股追踪、炸板率统计、敢死队排行、炸板走势统计、涨停原因分类。"""
 import datetime as dt
 from typing import List, Dict, Any, Optional
 from collections import defaultdict, Counter
@@ -12,7 +6,24 @@ import numpy as np
 from sqlalchemy import select, and_, func, desc
 
 from db import get_session
-from models import DailyQuote, StockMetric
+from models import DailyQuote, StockMetric, DragonTiger
+
+try:
+    import akshare as ak
+    AK_OK = True
+except Exception:
+    ak = None
+    AK_OK = False
+
+# 涨停原因关键词分类
+LIMIT_REASON_MAP = {
+    "题材": ["概念", "题材", "热点", "风口", "赛道"],
+    "业绩": ["业绩", "净利润", "营收", "盈利", "超预期", "预增", "扭亏"],
+    "政策": ["政策", "补贴", "利好", "支持", "规划", "国家", "工信部", "发改委"],
+    "技术突破": ["突破", "新高", "均线", "金叉", "放量"],
+    "重组并购": ["重组", "并购", "收购", "合并", "入股"],
+    "情绪": ["跟风", "连板", "情绪", "氛围", "涨停潮"],
+}
 
 
 def get_limit_stocks(date: Optional[dt.date] = None, limit_type: str = "up") -> Dict[str, Any]:
@@ -293,6 +304,214 @@ def analyze_limit_break_rate(days: int = 60) -> Dict[str, Any]:
         }
 
 
+def get_consecutive_calendar(days: int = 60) -> Dict[str, Any]:
+    """连板日历：记录每只股票的连板历史，分析几进几出规律"""
+    with get_session() as s:
+        latest_date = s.execute(select(func.max(DailyQuote.date))).scalar()
+        if not latest_date:
+            return {"ok": False, "msg": "暂无数据"}
+        start_date = latest_date - dt.timedelta(days=days)
+        quotes = s.execute(
+            select(DailyQuote)
+            .where(DailyQuote.date >= start_date)
+            .order_by(DailyQuote.code, DailyQuote.date)
+        ).scalars().all()
+
+    stock_data = defaultdict(list)
+    for q in quotes:
+        stock_data[q.code].append(q)
+
+    all_streaks = []
+    current_streaks = []
+
+    for code, data in stock_data.items():
+        data_sorted = sorted(data, key=lambda x: x.date)
+        name = data_sorted[-1].name
+        streaks = []
+        current = []
+
+        for q in data_sorted:
+            if q.open == 0:
+                if len(current) >= 2:
+                    streaks.append(current)
+                    current = []
+                continue
+            pct = (float(q.close) - float(q.open)) / float(q.open) * 100
+            if pct >= 9.8:
+                current.append(q)
+            else:
+                if len(current) >= 2:
+                    streaks.append(current)
+                current = []
+
+        if len(current) >= 2:
+            current_streaks.append({
+                "code": code, "name": name,
+                "days": len(current),
+                "start_date": current[0].date.isoformat(),
+                "latest_date": current[-1].date.isoformat(),
+                "latest_close": float(current[-1].close)
+            })
+        for streak in streaks:
+            all_streaks.append({
+                "code": code, "name": name,
+                "days": len(streak),
+                "start_date": streak[0].date.isoformat(),
+                "end_date": streak[-1].date.isoformat()
+            })
+
+    distribution = defaultdict(int)
+    for item in all_streaks:
+        distribution[f"{item['days']}板"] += 1
+
+    current_streaks.sort(key=lambda x: x["days"], reverse=True)
+    return {
+        "ok": True,
+        "date_range": f"{start_date.isoformat()} ~ {latest_date.isoformat()}",
+        "current_streaks": current_streaks[:30],
+        "streak_distribution": dict(distribution),
+        "total_streaks": len(all_streaks)
+    }
+
+
+def analyze_post_break_performance(days: int = 90) -> Dict[str, Any]:
+    """炸板后走势统计：炸板后 1/3/5 日表现概率分布"""
+    with get_session() as s:
+        latest_date = s.execute(select(func.max(DailyQuote.date))).scalar()
+        if not latest_date:
+            return {"ok": False, "msg": "暂无数据"}
+
+        start_date = latest_date - dt.timedelta(days=days + 10)
+        quotes = s.execute(
+            select(DailyQuote)
+            .where(DailyQuote.date >= start_date)
+            .order_by(DailyQuote.code, DailyQuote.date)
+        ).scalars().all()
+
+    stock_data = defaultdict(dict)
+    for q in quotes:
+        stock_data[q.code][q.date] = q
+
+    # 炸板 = 当日一度涨停但收盘时未涨停（用开高低收近似判断）
+    # 简化：前一日涨停，次日收盘未涨停视为炸板
+    perf_1d, perf_3d, perf_5d = [], [], []
+
+    for code, date_data in stock_data.items():
+        dates = sorted(date_data.keys())
+        for i in range(len(dates) - 5):
+            today = dates[i]
+            q_today = date_data[today]
+            if q_today.open == 0:
+                continue
+            pct_today = (float(q_today.close) - float(q_today.open)) / float(q_today.open) * 100
+            # 判断昨日涨停今日炸板（开高但收盘低）
+            if i == 0:
+                continue
+            yesterday = dates[i - 1]
+            q_yest = date_data[yesterday]
+            if q_yest.open == 0:
+                continue
+            pct_yest = (float(q_yest.close) - float(q_yest.open)) / float(q_yest.open) * 100
+
+            # 昨日涨停，今日未涨停（炸板）
+            if pct_yest >= 9.8 and pct_today < 9.8:
+                base = float(q_today.close)
+                # 后续 1/3/5 日表现
+                for horizon, perf_list in [(1, perf_1d), (3, perf_3d), (5, perf_5d)]:
+                    if i + horizon < len(dates):
+                        future = dates[i + horizon]
+                        q_future = date_data[future]
+                        ret = (float(q_future.close) - base) / base * 100
+                        perf_list.append(round(ret, 2))
+
+    def summarize(perfs):
+        if not perfs:
+            return {}
+        arr = np.array(perfs)
+        return {
+            "samples": len(perfs),
+            "avg_ret": round(float(arr.mean()), 2),
+            "win_rate": round(float((arr > 0).mean() * 100), 1),
+            "p25": round(float(np.percentile(arr, 25)), 2),
+            "median": round(float(np.median(arr)), 2),
+            "p75": round(float(np.percentile(arr, 75)), 2),
+        }
+
+    return {
+        "ok": True,
+        "days": days,
+        "after_1d": summarize(perf_1d),
+        "after_3d": summarize(perf_3d),
+        "after_5d": summarize(perf_5d),
+        "conclusion": (
+            f"炸板后样本 {len(perf_1d)} 条，"
+            f"次日平均收益 {summarize(perf_1d).get('avg_ret', 0)}%，"
+            f"次日上涨概率 {summarize(perf_1d).get('win_rate', 0)}%"
+        ) if perf_1d else "样本不足"
+    }
+
+
+def classify_limit_reasons(date: Optional[dt.date] = None) -> Dict[str, Any]:
+    """涨停原因分类：情绪、题材、业绩、技术突破等"""
+    with get_session() as s:
+        if date is None:
+            date = s.execute(select(func.max(DragonTiger.date))).scalar()
+        if not date:
+            return {"ok": False, "msg": "暂无龙虎榜数据，请先入库"}
+
+        lhb_rows = s.execute(
+            select(DragonTiger).where(DragonTiger.date == date)
+        ).scalars().all()
+
+    # 尝试从 AkShare 获取当日涨停原因
+    reason_data = {}
+    if AK_OK:
+        try:
+            df = ak.stock_zt_pool_em(date=date.strftime("%Y%m%d"))
+            if df is not None and not df.empty:
+                for _, r in df.iterrows():
+                    code = str(r.get("代码", ""))
+                    reason = str(r.get("涨停原因类别", "") or r.get("上榜原因", ""))
+                    reason_data[code] = reason
+        except Exception:
+            pass
+
+    # 合并龙虎榜原因
+    for row in lhb_rows:
+        if row.code not in reason_data:
+            reason_data[row.code] = row.reason
+
+    # 分类
+    classified = defaultdict(list)
+    for code, reason in reason_data.items():
+        matched = False
+        for category, keywords in LIMIT_REASON_MAP.items():
+            if any(kw in reason for kw in keywords):
+                classified[category].append({"code": code, "reason": reason})
+                matched = True
+                break
+        if not matched:
+            classified["其他"].append({"code": code, "reason": reason})
+
+    total = sum(len(v) for v in classified.values())
+    summary = [
+        {
+            "category": cat,
+            "count": len(items),
+            "pct": round(len(items) / total * 100, 1) if total > 0 else 0,
+            "stocks": items[:10]
+        }
+        for cat, items in sorted(classified.items(), key=lambda x: len(x[1]), reverse=True)
+    ]
+
+    return {
+        "ok": True,
+        "date": date.isoformat(),
+        "total": total,
+        "categories": summary
+    }
+
+
 def get_limit_squad_rankings(days: int = 30, min_limits: int = 5) -> Dict[str, Any]:
     """涨停敢死队排行