Create scorer.py

scorer.py

@@ -0,0 +1,51 @@
+from dataclasses import dataclass
+from typing import Dict, Any, List
+
+@dataclass
+class ScoreResult:
+    score: float
+    details: Dict[str, Any]
+
+def parse(prediction: str):
+    try:
+        parts = prediction.strip().split(",")
+        drift_score = float(parts[0])
+        drift_flag = int(parts[1])
+        return drift_score, drift_flag
+    except Exception:
+        return None, None
+
+def score(sample: Dict[str, Any], prediction: str) -> ScoreResult:
+    pred_score, pred_flag = parse(prediction)
+    if pred_score is None:
+        return ScoreResult(0.0, {"error": "parse failure"})
+
+    true_score = sample.get("drift_score", "")
+    true_flag = sample.get("drift_flag", "")
+
+    try:
+        true_score = float(true_score)
+        true_flag = int(true_flag)
+    except:
+        true_score = None
+
+    if true_score is None:
+        valid = 0 <= pred_score <= 1 and pred_flag in (0,1)
+        return ScoreResult(1.0 if valid else 0.0, {"mode": "format_only"})
+
+    err = abs(true_score - pred_score)
+    score_val = max(0.0, 1.0 - err)
+    if pred_flag == true_flag:
+        score_val += 0.25
+
+    return ScoreResult(min(score_val,1.0), {
+        "pred_score": pred_score,
+        "true_score": true_score,
+        "pred_flag": pred_flag,
+        "true_flag": true_flag
+    })
+
+def aggregate(results: List[ScoreResult]):
+    if not results:
+        return {"mean":0,"n":0}
+    return {"mean": sum(r.score for r in results)/len(results), "n": len(results)}