func (this *CRPReward) Next(s discrete.State, a discrete.Action) (r float64) {
index := s.Hashcode() + this.NumStates*a.Hashcode()
if this.Known[index] {
r = this.R[index]
return
}
if this.chooser == nil {
if len(this.Counts) == 0 {
this.chooser = func() int64 { return 0 }
} else {
normalizer := 1.0 / (float64(this.Total) + this.Alpha)
weights := make([]float64, len(this.Counts))
for i := range weights {
weights[i] = float64(this.Counts[i]) * normalizer
}
this.chooser = stat.Choice(weights)
}
}
which := int(this.chooser())
if which == len(this.SeenRewards) {
r = this.BaseSampler()
} else {
r = this.SeenRewards[which]
}
return
}
func (this Histogram) GetChoice() (chooser func() uint64) {
var sum float64
for _, v := range this {
sum += float64(v)
}
norm := 1 / sum
weights := make([]float64, len(this))
for i, v := range this {
weights[i] = float64(v) * norm
}
chooser = func() uint64 { chooser := stat.Choice(weights); return uint64(chooser()) }
return
}
