Create scorer.py

scorer.py

@@ -0,0 +1,61 @@
+from dataclasses import dataclass
+from typing import Dict, Any, List
+import re
+
+REQ = [
+    "regime_coherence_score",
+    "dominant_driver_label",
+    "cross_asset_alignment_index",
+    "dispersion_vs_concentration_index",
+    "regime_stability_band",
+]
+
+LABELS = [
+    "macro_rates",
+    "risk_on_liquidity",
+    "growth_beta",
+    "concentration_equity",
+    "shock_repricing",
+    "uncertain_transition",
+]
+
+BANDS = ["stable", "moderate", "unstable", "fragile"]
+
+@dataclass
+class ScoreResult:
+    score: float
+    details: Dict[str, Any]
+
+def _has_float(p: str) -> bool:
+    return bool(re.search(r"\b0\.\d+\b", p)) or "1.0" in p
+
+def _has_driver(p: str) -> bool:
+    return any(l in p for l in LABELS)
+
+def _has_band(p: str) -> bool:
+    return any(b in p for b in BANDS)
+
+def score(sample: Dict[str, Any], prediction: str) -> ScoreResult:
+    p = (prediction or "").lower()
+    words_ok = len(p.split()) <= 900
+
+    hits = sum(1 for k in REQ if k.lower() in p)
+
+    has_numbers = _has_float(p)
+    has_driver = _has_driver(p)
+    has_band = _has_band(p)
+
+    raw = (
+        0.20 * int(words_ok) +
+        0.50 * (hits / len(REQ)) +
+        0.10 * int(has_numbers) +
+        0.10 * int(has_driver) +
+        0.10 * int(has_band)
+    )
+
+    return ScoreResult(score=min(1.0, raw), details={"id": sample.get("id"), "hits": hits})
+
+def aggregate(results: List[ScoreResult]) -> Dict[str, Any]:
+    if not results:
+        return {"mean": 0.0, "n": 0}
+    return {"mean": sum(r.score for r in results)/len(results), "n": len(results)}