/// <summary>
/// 获取当前actor下的action和reward
/// </summary>
/// <param name="step"></param>
/// <param name="state"></param>
/// <returns></returns>
public (double[] action, double q) EpsilonGreedy(int step, double[] state)
{
int totalEpochs = Convert.ToInt32(_epoches * 0.9);
var epsion = EpsilonCalcute(step, eps_total: totalEpochs);
if (NP.Random() < epsion)
{
return(_env.RandomAction(), 0);
}
else
{
var(action, q) = ChooseAction(state);
return(action, q);
}
}