scorer.py

from dataclasses import dataclass
from typing import Dict, Any, List
import re

REQ = ["niche_definition","safe_dose_band","relapse_risk_index","paradox_zone_flags","systemic_resilience_gain","subgroup_sensitivity"]
BANDS = ["avoid","narrow","moderate","broad"]

@dataclass
class ScoreResult:
    score: float
    details: Dict[str, Any]

def _f(p: str, key: str):
    m = re.search(rf"{key}\s*[:=]\s*(0\.\d+|1\.0)\b", p)
    return float(m.group(1)) if m else None

def score(sample: Dict[str, Any], prediction: str) -> ScoreResult:
    p = (prediction or "").lower()
    words_ok = len(p.split()) <= 950
    hits = sum(1 for k in REQ if k in p)

    rel = _f(p, "relapse_risk_index")
    gain = _f(p, "systemic_resilience_gain")
    num_ok = int(rel is not None and 0 <= rel <= 1 and gain is not None and 0 <= gain <= 1)

    band_ok = int("safe_dose_band" in p and any(b in p for b in BANDS))
    niche_ok = int("niche_definition" in p)
    paradox_ok = int("paradox_zone_flags" in p)
    sens_ok = int("subgroup_sensitivity" in p)

    raw = (
        0.18 * int(words_ok) +
        0.44 * (hits / len(REQ)) +
        0.22 * num_ok +
        0.06 * band_ok +
        0.04 * niche_ok +
        0.03 * paradox_ok +
        0.03 * sens_ok
    )
    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)}