com.eignex.kumulant/bandit/univariate/GammaScalePosterior

GammaScalePosterior

@Serializable

@SerialName(value = "GammaScalePosterior")

data class GammaScalePosterior(val fixedShape: Double) : Posterior<WeightedMeanResult> (source)

Gamma posterior over the scale of a Gamma likelihood with fixed shape - the shape is a posterior parameter rather than something we infer from data. Not an object because of that parameter.

Constructors

GammaScalePosterior

constructor(fixedShape: Double)(source)

Properties

fixedShape

val fixedShape: Double(source)

Known shape parameter of the Gamma likelihood.

Functions

sample

open override fun sample(snapshot: WeightedMeanResult, rng: Random): Double(source)

Draw a single Monte Carlo sample of the arm's reward from its posterior.