fix: report direction probability as signal confidence

README.md

@@ -21,7 +21,7 @@ pinned: false
 
 - 请求/响应主字段由 `ts_code` 统一为 `symbol`
 - 数据源切换回 Tushare Pro（`pro_bar`，前复权 `qfq`）
-- 方向概率 `direction.probability` 定义为“预测区间内看涨概率（0–1）”
+- 方向概率 `direction.probability` 定义为“当前 `direction.signal` 的概率（0–1）”
 - 推理路径改为分批采样（可通过 `MC_BATCH_SIZE` 调整批大小）
 - 新增阶段耗时日志（`fetch/calendar/infer/build/cache/total`）
 
@@ -122,7 +122,7 @@ print(r["result"])
   "cache_expires_at": "2026-03-16 15:00:00 UTC+08:00",
   "direction": {
     "signal": "bearish",
-    "probability": 0.4333
+    "probability": 0.5667
   },
   "summary": {
     "mean_close": 36.35,
@@ -281,8 +281,8 @@ curl "https://yingfeng64-kronos-api.hf.space/api/v1/cache"
 | 字段 | 含义 |
 |---|---|
 | `base_date` | 预测所基于的最后一个历史 K 线日期 |
-| `direction.signal` | `"bullish"` / `"bearish"`，由 `direction.probability >= 0.5` 决定 |
-| `direction.probability` | 预测区间内看涨概率（0–1） |
+| `direction.signal` | `"bullish"` / `"bearish"`，由内部看涨概率是否 >= 0.5 决定 |
+| `direction.probability` | 当前 `direction.signal` 的概率（0–1） |
 | `trading_low` | 该日预测最低价的 q2.5 分位数（95% 交易区间下沿） |
 | `trading_high` | 该日预测最高价的 q97.5 分位数（95% 交易区间上沿） |
 | `uncertainty` | `(trading_high − trading_low) / last_close`，无量纲不确定性 |

app.py

@@ -162,6 +162,10 @@ class PredictRequest(BaseModel):
 def _build_response(req: PredictRequest, base_date: str, pred_mean, ci,
                     trading_low, trading_high, direction_prob, last_close,
                     y_timestamp) -> dict:
+    bullish_prob = float(direction_prob)
+    direction_signal = "bullish" if bullish_prob >= 0.5 else "bearish"
+    signal_prob = bullish_prob if direction_signal == "bullish" else (1 - bullish_prob)
+
     bands = []
     for i in range(req.pred_len):
         band: dict = {
@@ -191,8 +195,8 @@ def _build_response(req: PredictRequest, base_date: str, pred_mean, ci,
         "confidence":          95,
         "confidence_warning":  req.pred_len > 30,
         "direction": {
-            "signal":      "bullish" if direction_prob >= 0.5 else "bearish",
-            "probability": round(direction_prob, 4),
+            "signal":      direction_signal,
+            "probability": round(signal_prob, 4),
         },
         "summary": {
             "mean_close":  round(float(pred_mean["close"].iloc[-1]), 4),