Learning from Feedback and Iteration Questions

from typing import List, Dict

def compute_rice(items: List[Dict]) -> List[Dict]:
    """
    Compute RICE score for each feedback item and return items sorted by descending score.
    Handles missing keys with defaults and zero/negative effort using small epsilon.
    """
    EPS = 1e-6  # small value to avoid division by zero; treats near-zero effort as very high priority
    result = []
    for it in items:
        reach = int(it.get('reach', 0)) if it.get('reach') is not None else 0
        impact = float(it.get('impact', 0.0)) if it.get('impact') is not None else 0.0
        confidence = float(it.get('confidence', 0.0)) if it.get('confidence') is not None else 0.0
        effort = it.get('effort')
        # If effort is missing, None, or <=0, use EPS
        try:
            effort = float(effort)
            if effort <= 0:
                effort = EPS
        except (TypeError, ValueError):
            effort = EPS
        rice = (reach * impact * confidence) / effort
        copy = dict(it)  # avoid mutating original
        copy['rice'] = rice
        result.append(copy)
    # sort descending by rice, tie-break by id for deterministic order
    return sorted(result, key=lambda x: (-x['rice'], x.get('id')))

# Short test
if __name__ == "__main__":
    items = [
        {'id': 'A', 'reach': 1000, 'impact': 3.0, 'confidence': 0.8, 'effort': 10},
        {'id': 'B', 'reach': 200,  'impact': 5.0, 'confidence': 0.9, 'effort': 0},   # zero effort
        {'id': 'C', 'reach': 500,  'impact': 2.0, 'confidence': 0.5, 'effort': 5},
        {'id': 'D', 'reach': None, 'impact': 4.0, 'confidence': 0.7},               # missing effort & reach
    ]
    for item in compute_rice(items):
        print(item)