)
else:
return (
- factor * pnl * (1 + lambda1 * duration_fraction)
- - 2 * lambda2 * duration_fraction
+ factor * pnl * (1.0 + lambda1 * duration_fraction)
+ - 2.0 * lambda2 * duration_fraction
- drawdown_penalty
)
"tick" not in _history_df.columns
or "tick" not in _trade_history_df.columns
):
- logger.warning("'tick' column is missing from history or trade history")
+ logger.warning("'tick' column is missing from history or trade_history")
return DataFrame()
_rollout_history = merge(
"exploration_rate": float(self.model.exploration_rate),
}
)
- if "QRDQN" in self.model.__class__.__name__:
- hparam_dict.update({"n_quantiles": int(self.model.n_quantiles)})
+ if "QRDQN" in self.model.__class__.__name__:
+ hparam_dict.update({"n_quantiles": int(self.model.n_quantiles)})
metric_dict = {
"info/total_reward": 0.0,
"info/total_profit": 1.0,
side: str,
order: Literal["entry", "exit"],
rate: float,
- min_natr_ratio_percent: float = 0.0099,
+ min_natr_ratio_percent: float = 0.0095,
max_natr_ratio_percent: float = 0.095,
lookback_period: int = 1,
decay_ratio: float = 0.5,