func (ra *RmaxFSSSAgent) Forget() {
ra.mdpo = discrete.NewMDPOracle(ra.rmdp, ra.lastState)
ra.s = fsss.New()
ra.s.Cfg = ra.Cfg.FS3
ra.s.NumActions = ra.rmdp.NumActions()
ra.s.Gamma = ra.task.DiscountFactor
if ra.s.Gamma == 1 {
ra.s.Gamma = 0.9
}
ra.s.Vmin = ra.task.Reward.Min / (1 - ra.s.Gamma)
ra.s.Vmax = 5
ra.stepsWithPlanner = 0
}
func New(cfg Config, mdp discrete.MDP) (this *Agent) {
this = new(Agent)
this.cfg = cfg
this.mdp = mdp
this.mdpo = discrete.NewMDPOracle(this.mdp, 0)
this.s = fsss.New()
this.s.Cfg = this.cfg.FS3
this.s.NumActions = this.mdp.NumActions()
this.s.Gamma = mdp.GetGamma()
this.s.Vmin = this.mdp.GetTask().Reward.Min / (1 - this.s.Gamma)
this.s.Vmax = this.mdp.GetTask().Reward.Max / (1 - this.s.Gamma)
this.stepsWithPlanner = 0
return
}
func (this *BFS3Agent) ResetPlanner() {
this.fs3 = fsss.New()
this.fs3.Cfg = this.Cfg.FS3
this.fs3.NumActions = uint64(this.task.Act.Ints.Count())
if this.Cfg.CustomGammaV {
this.fs3.Gamma = this.Cfg.Gamma
this.fs3.Vmin = this.Cfg.Vmin
this.fs3.Vmax = this.Cfg.Vmax
} else {
this.fs3.Gamma = this.task.DiscountFactor
if this.fs3.Gamma == 1 {
this.fs3.Gamma = 0.95
}
this.fs3.Vmin = this.task.Reward.Min / (1 - this.fs3.Gamma)
this.fs3.Vmax = this.task.Reward.Max / (1 - this.fs3.Gamma)
}
}