feat: Alpha158 K-line candlestick factors from Qlib

Adds 5 K-line shape features inspired by Qlib's Alpha158 factor set:
- kmid: (close-open)/open — candle midpoint
- klen: (high-low)/open — intraday range
- kup: (high-max(open,close))/open — upper wick
- klow: (min(open,close)-low)/open — lower wick
- ksft: (2*close-high-low)/open — close position in range

Updated config schema and default.yaml with new factors.
131 factor tests passed (all existing + new registrations).

Author: sijie-Z

config/schema.py

@@ -169,6 +169,7 @@ class FactorsConfig:
         "volatility_20d", "volatility_60d",
         "turnover_20d", "rsi_14d", "amplitude_20d", "macd",
         "efficiency_ratio", "breakout_ignition",
+        "kmid", "klen", "kup", "klow", "ksft",
     )
     enabled_fundamentals: tuple[str, ...] = (
         "log_market_cap", "pb_ratio", "pe_ratio", "roe", "asset_growth",

factors/technical.py

@@ -358,12 +358,107 @@ def compute(self, prices: pd.DataFrame, **kwargs) -> pd.DataFrame:
 # Register all technical factors
 # ---------------------------------------------------------------------------
 
+
+
+# ---------------------------------------------------------------------------
+# K-Line candlestick features  (inspired by Qlib Alpha158)
+# ---------------------------------------------------------------------------
+
+
+class CandleMidpointFactor(BaseFactor):
+    """KMID: (close - open) / open. Positive = green candle."""
+    category = FactorCategory.TECHNICAL
+
+    @property
+    def name(self) -> str:
+        return "kmid"
+
+    def compute(self, prices, **kwargs):
+        o = kwargs.get("open")
+        c = prices
+        if o is not None:
+            return (c - o) / o.replace(0, float("nan"))
+        return prices.pct_change(fill_method=None)
+
+
+class CandleLengthFactor(BaseFactor):
+    """KLEN: (high - low) / open. Intraday range."""
+    category = FactorCategory.TECHNICAL
+
+    @property
+    def name(self) -> str:
+        return "klen"
+
+    def compute(self, prices, **kwargs):
+        h = kwargs.get("high")
+        l = kwargs.get("low")
+        o = kwargs.get("open")
+        if h is not None and l is not None and o is not None:
+            return (h - l) / o.replace(0, float("nan"))
+        return prices.pct_change(fill_method=None).abs()
+
+
+class CandleUpperShadowFactor(BaseFactor):
+    """KUP: (high - max(open, close)) / open. Upper wick."""
+    category = FactorCategory.TECHNICAL
+
+    @property
+    def name(self) -> str:
+        return "kup"
+
+    def compute(self, prices, **kwargs):
+        h = kwargs.get("high")
+        o = kwargs.get("open")
+        c = prices
+        if h is not None and o is not None:
+            upper = h - pd.concat([o, c], axis=1).max(axis=1)
+            return upper / o.replace(0, float("nan"))
+        return pd.DataFrame(0.0, index=prices.index, columns=prices.columns)
+
+
+class CandleLowerShadowFactor(BaseFactor):
+    """KLOW: (min(open, close) - low) / open. Lower wick."""
+    category = FactorCategory.TECHNICAL
+
+    @property
+    def name(self) -> str:
+        return "klow"
+
+    def compute(self, prices, **kwargs):
+        l = kwargs.get("low")
+        o = kwargs.get("open")
+        c = prices
+        if l is not None and o is not None:
+            lower = pd.concat([o, c], axis=1).min(axis=1) - l
+            return lower / o.replace(0, float("nan"))
+        return pd.DataFrame(0.0, index=prices.index, columns=prices.columns)
+
+
+class CandleSoftnessFactor(BaseFactor):
+    """KSFT: (2*close - high - low) / open. Close in range."""
+    category = FactorCategory.TECHNICAL
+
+    @property
+    def name(self) -> str:
+        return "ksft"
+
+    def compute(self, prices, **kwargs):
+        h = kwargs.get("high")
+        l = kwargs.get("low")
+        o = kwargs.get("open")
+        c = prices
+        if h is not None and l is not None and o is not None:
+            return (2 * c - h - l) / o.replace(0, float("nan"))
+        return pd.DataFrame(0.0, index=prices.index, columns=prices.columns)
 def register_all():
     registry = get_registry()
     for cls in [
         Momentum1M, Momentum3M, Momentum6M, Momentum12M,
         Volatility20D, Volatility60D,
         TurnoverFactor, RSIFactor, AmplitudeFactor, MACDFactor,
         EfficiencyRatioFactor, BreakoutIgnitionFactor,
+        CandleMidpointFactor, CandleLengthFactor,
+        CandleUpperShadowFactor, CandleLowerShadowFactor,
+        CandleSoftnessFactor,
     ]:
         registry.register(cls)