Function _bandit_bonus

uncommon_route/router/selector.py:712–732

(
    *,
    enabled: bool,
    bandit_config: BanditConfig,
    candidate_cost: float,
    cheapest_cost: float,
    reliability: float,
    samples: int,
    bucket_pulls: int,
)

def _bandit_bonus(
    *,
    enabled: bool,
    bandit_config: BanditConfig,
    candidate_cost: float,
    cheapest_cost: float,
    reliability: float,
    samples: int,
    bucket_pulls: int,
) -> float:
    if not enabled:
        return 0.0
    if cheapest_cost > 0 and candidate_cost > (cheapest_cost * bandit_config.max_cost_ratio):
        return 0.0
    if samples >= bandit_config.min_samples_for_guardrail and reliability < bandit_config.min_reliability:
        return 0.0
    if samples < bandit_config.warmup_pulls:
        return bandit_config.exploration_weight
    return bandit_config.exploration_weight * math.sqrt(
        math.log(max(2, bucket_pulls + 1)) / (samples + 1),
    )


# ---------------------------------------------------------------------------