feat: split portfolio and opportunity decision models

src/coinhunter/services/signal_service.py

@@ -0,0 +1,78 @@
+"""Shared market signal scoring."""
+
+from __future__ import annotations
+
+from statistics import mean
+from typing import Any
+
+
+def _safe_pct(new: float, old: float) -> float:
+    if old == 0:
+        return 0.0
+    return (new - old) / old
+
+
+def get_signal_weights(config: dict[str, Any]) -> dict[str, float]:
+    signal_config = config.get("signal", {})
+    return {
+        "trend": float(signal_config.get("trend", 1.0)),
+        "momentum": float(signal_config.get("momentum", 1.0)),
+        "breakout": float(signal_config.get("breakout", 0.8)),
+        "volume": float(signal_config.get("volume", 0.7)),
+        "volatility_penalty": float(signal_config.get("volatility_penalty", 0.5)),
+    }
+
+
+def get_signal_interval(config: dict[str, Any]) -> str:
+    signal_config = config.get("signal", {})
+    if signal_config.get("lookback_interval"):
+        return str(signal_config["lookback_interval"])
+    return "1h"
+
+
+def score_market_signal(
+    closes: list[float],
+    volumes: list[float],
+    ticker: dict[str, Any],
+    weights: dict[str, float],
+) -> tuple[float, dict[str, float]]:
+    if len(closes) < 2 or not volumes:
+        return 0.0, {
+            "trend": 0.0,
+            "momentum": 0.0,
+            "breakout": 0.0,
+            "volume_confirmation": 1.0,
+            "volatility": 0.0,
+        }
+
+    current = closes[-1]
+    sma_short = mean(closes[-5:]) if len(closes) >= 5 else current
+    sma_long = mean(closes[-20:]) if len(closes) >= 20 else mean(closes)
+    trend = 1.0 if current >= sma_short >= sma_long else -1.0 if current < sma_short < sma_long else 0.0
+    momentum = (
+        _safe_pct(closes[-1], closes[-2]) * 0.5
+        + (_safe_pct(closes[-1], closes[-5]) * 0.3 if len(closes) >= 5 else 0.0)
+        + float(ticker.get("price_change_pct", 0.0)) / 100.0 * 0.2
+    )
+    recent_high = max(closes[-20:]) if len(closes) >= 20 else max(closes)
+    breakout = 1.0 - max((recent_high - current) / recent_high, 0.0)
+    avg_volume = mean(volumes[:-1]) if len(volumes) > 1 else volumes[-1]
+    volume_confirmation = volumes[-1] / avg_volume if avg_volume else 1.0
+    volume_score = min(max(volume_confirmation - 1.0, -1.0), 2.0)
+    volatility = (max(closes[-10:]) - min(closes[-10:])) / current if len(closes) >= 10 and current else 0.0
+
+    score = (
+        weights.get("trend", 1.0) * trend
+        + weights.get("momentum", 1.0) * momentum
+        + weights.get("breakout", 0.8) * breakout
+        + weights.get("volume", 0.7) * volume_score
+        - weights.get("volatility_penalty", 0.5) * volatility
+    )
+    metrics = {
+        "trend": round(trend, 4),
+        "momentum": round(momentum, 4),
+        "breakout": round(breakout, 4),
+        "volume_confirmation": round(volume_confirmation, 4),
+        "volatility": round(volatility, 4),
+    }
+    return score, metrics