Distributions

All of these distributions are found in com.stripe.rainier.core.

In all cases where distributions are parameterized by a Real, numeric values will also work.

Continuous

These distributions all extend Distribution[Double].

They all implement:

  • latent: Real
  • latentVec(k: Int): Vec[Real]
  • generator: Generator[Double]

Normal

  • Normal.standard
  • Normal(location: Real, scale: Real)

Cauchy

  • Cauchy.standard
  • Cauchy(location: Real, scale: Real)

Laplace

  • Laplace.standard
  • Laplace(location: Real, scale: Real)

Gamma

  • Gamma.standard(shape: Real)
  • Gamma(shape: Real, scale: Real)
  • Gamma.meanAndScale(mean: Real, scale: Real)

Exponential

  • Exponential.standard
  • Exponential(rate: Real)

Beta

  • Beta(a: Real, b: Real)
  • Beta.meanAndPrecision(mean: Real, precision: Real)
  • Beta.meanAndVariance(mean: Real, variance: Real)

LogNormal

  • LogNormal(location: Real, scale: Real)

Uniform

  • Uniform.standard
  • Uniform(from: Real, to: Real)

Mixture

  • Mixture(components: Map[Continuous, Real])

Discrete

These distributions all extend Distribution[Long].

They all implement:

  • generator: Generator[Long]
  • zeroInflated(psi: Real): Distribution[Long]

Bernoulli

  • Bernouilli(p: Real)

Geometric

  • Geometric(p: Real)

NegativeBinomial

  • NegativeBinomial(p: Real, n: Real)

Poisson

  • Poisson(lambda: Real)

Binomial

  • Binomial(p: Real, k: Real)

BetaBinomial

  • BetaBinomial(a: Real, b: Real, k: Real)
  • BetaBinomial.meanAndPrecision(mean: Real, precision: Real, k: Real)

DiscreteMixture

  • DiscreteMixture(components: Map[Discrete, Real])

Multinomial

extends Distribution[T,Long]

  • Multinomial[T](pmf: Map[T, Real], k: Real)
← ModulesModel and Trace →