core

Type Members

final case class Bernoulli(p: Real) extends Discrete with Product with Serializable

Bernoulli distribution with expectation p
Bernoulli distribution with expectation p
p
The probability of success
final case class Beta(a: Real, b: Real) extends StandardContinuous with Product with Serializable

A Beta distribution with expectation a/(a + b) and variance ab/((a + b)^2 (1 + a + b)).
final case class BetaBinomial(a: Real, b: Real, k: Real) extends Discrete with Product with Serializable

A Beta-binomial distribution with expectation a/(a + b) * k
final case class Binomial(p: Real, k: Real) extends Discrete with Product with Serializable

A Binomial distribution with expectation k*p
A Binomial distribution with expectation k*p
p
The probability of success
k
The number of trials
case class BoundedAboveSupport(max: Real = Real.zero) extends Support with Product with Serializable

A support representing an open-below {r < k} interval.
A support representing an open-below {r < k} interval.
max
The upper bound of the distribution
case class BoundedBelowSupport(min: Real = Real.zero) extends Support with Product with Serializable

A support representing an open-above {r > k} interval.
A support representing an open-above {r > k} interval.
min
The lower bound of the distribution
case class BoundedSupport(min: Real, max: Real) extends Support with Product with Serializable

A support representing a bounded (min, max) interval.
final case class Categorical[T](pmf: Map[T, Real]) extends Distribution[T] with Product with Serializable

A finite discrete distribution
A finite discrete distribution
pmf
A map with keys corresponding to the possible outcomes and values corresponding to the probabilities of those outcomes
trait Continuous extends Distribution[Double]

A Continuous Distribution, with method param allowing conversion to a RandomVariable.
trait Discrete extends Distribution[Int]
final case class DiscreteConstant(constant: Real) extends Discrete with Product with Serializable

Discrete Constant (point mass) with expecation constant
Discrete Constant (point mass) with expecation constant
constant
The integer value of the point mass
final case class DiscreteMixture(components: Map[Discrete, Real]) extends Discrete with Product with Serializable

Discrete Mixture Distribution
Discrete Mixture Distribution
components
Map of Discrete distribution and probabilities
trait Distribution[T] extends AnyRef
trait Generator[T] extends AnyRef

Generator trait, for posterior predictive distributions to be forwards sampled during sampling
final case class Geometric(p: Real) extends Discrete with Product with Serializable

Geometric distribution with expectation 1/p
Geometric distribution with expectation 1/p
p
The probability of success
trait Likelihood[T] extends AnyRef
trait LocationScaleFamily extends AnyRef

Location-scale family distribution
case class Mixture(components: Map[Continuous, Real]) extends Continuous with Product with Serializable
final case class Multinomial[T](pmf: Map[T, Real], k: Real) extends Distribution[Map[T, Int]] with Product with Serializable

A Multinomial distribution
A Multinomial distribution
pmf
A map with keys corresponding to the possible outcomes of a single multinomial trial and values corresponding to the probabilities of those outcomes
k
The number of multinomial trials
final case class NegativeBinomial(p: Real, n: Real) extends Discrete with Product with Serializable

Negative Binomial distribution with expectation n*p/(1-p)
Negative Binomial distribution with expectation n*p/(1-p)
p
Probability of success
n
Total number of failures
final case class Poisson(lambda: Real) extends Discrete with Product with Serializable

Poisson distribution with expectation lambda
Poisson distribution with expectation lambda
lambda
The mean of the Poisson distribution
sealed trait Predictor[X, L] extends AnyRef

Predictor class, for fitting data with covariates
class RandomVariable[+T] extends AnyRef

The main probability monad used in Rainier for constructing probabilistic programs which can be sampled
final case class SBC[T, L](priors: Seq[Continuous], fn: (Seq[Real]) ⇒ (L, Real))(implicit lh: ToLikelihood[L, T], gen: ToGenerator[L, T]) extends Product with Serializable
final case class Scale(a: Real) extends Injection with Product with Serializable

Class to scale a distribution under multiplication by a positive scale factor.
Class to scale a distribution under multiplication by a positive scale factor. We assume that (a > 0).
trait ToGenerator[-T, U] extends AnyRef
trait ToLikelihood[L, T] extends AnyRef
final case class Translate(b: Real) extends Injection with Product with Serializable

Class to translate a distribution by adding a constant.

Value Members

object Beta extends Serializable
object Categorical extends Serializable
object Cauchy extends LocationScaleFamily

A Cauchy distribution with mode location and scaling relative to standard Cauchy of scale
object Combinatorics

Combinatoric functions useful in log density calculations.
Combinatoric functions useful in log density calculations. Note that they all return the log of the function described.
object Continuous
object Discrete
object Distribution
object Exp extends Injection

Object to exponentiate a distribution.
object Exponential

An Exponential distribution with expectation 1/rate
object Gamma

A Gamma distribution with expectation shape*scale and variance shape*scale*scale.
A Gamma distribution with expectation shape*scale and variance shape*scale*scale. N.B. It is parameterised with *scale* rather than *rate*, as is more typical in statistics texts.
object Generator

Generator object, for posterior predictive distributions to be forwards sampled during sampling
object Laplace extends LocationScaleFamily

A Laplace distribution with expectation location and variance 2*scale*scale
object Likelihood
object Log extends Logger
object LogNormal

A LogNormal distribution representing the exponential of a Gaussian random variable with expectation location and standard deviation scale.
A LogNormal distribution representing the exponential of a Gaussian random variable with expectation location and standard deviation scale. It therefore has expectation exp(location + scale*scale/2).
object Multinomial extends Serializable
object Normal extends LocationScaleFamily

A Gaussian distribution with expectation location and standard deviation scale
object Predictor
object RandomVariable

The main probability monad used in Rainier for constructing probabilistic programs which can be sampled
object SBC extends Serializable
object Support
object ToGenerator
object UnboundedSupport extends Support

A support representing the whole real line.
object Uniform

A Uniform distribution over [from,to] with expectation (to-from)/2.

package core

Type Members

final case class Bernoulli(p: Real) extends Discrete with Product with Serializable

final case class Beta(a: Real, b: Real) extends StandardContinuous with Product with Serializable

final case class BetaBinomial(a: Real, b: Real, k: Real) extends Discrete with Product with Serializable

final case class Binomial(p: Real, k: Real) extends Discrete with Product with Serializable

case class BoundedAboveSupport(max: Real = Real.zero) extends Support with Product with Serializable

case class BoundedBelowSupport(min: Real = Real.zero) extends Support with Product with Serializable

case class BoundedSupport(min: Real, max: Real) extends Support with Product with Serializable

final case class Categorical[T](pmf: Map[T, Real]) extends Distribution[T] with Product with Serializable

trait Continuous extends Distribution[Double]

trait Discrete extends Distribution[Int]

final case class DiscreteConstant(constant: Real) extends Discrete with Product with Serializable

final case class DiscreteMixture(components: Map[Discrete, Real]) extends Discrete with Product with Serializable

trait Distribution[T] extends AnyRef

trait Generator[T] extends AnyRef

final case class Geometric(p: Real) extends Discrete with Product with Serializable

trait Likelihood[T] extends AnyRef

trait LocationScaleFamily extends AnyRef

case class Mixture(components: Map[Continuous, Real]) extends Continuous with Product with Serializable

final case class Multinomial[T](pmf: Map[T, Real], k: Real) extends Distribution[Map[T, Int]] with Product with Serializable

final case class NegativeBinomial(p: Real, n: Real) extends Discrete with Product with Serializable

final case class Poisson(lambda: Real) extends Discrete with Product with Serializable

sealed trait Predictor[X, L] extends AnyRef

class RandomVariable[+T] extends AnyRef

final case class SBC[T, L](priors: Seq[Continuous], fn: (Seq[Real]) ⇒ (L, Real))(implicit lh: ToLikelihood[L, T], gen: ToGenerator[L, T]) extends Product with Serializable

final case class Scale(a: Real) extends Injection with Product with Serializable

trait ToGenerator[-T, U] extends AnyRef

trait ToLikelihood[L, T] extends AnyRef

final case class Translate(b: Real) extends Injection with Product with Serializable

Value Members

object Beta extends Serializable

object Categorical extends Serializable

object Cauchy extends LocationScaleFamily

object Combinatorics

object Continuous

object Discrete

object Distribution

object Exp extends Injection

object Exponential

object Gamma

object Generator

object Laplace extends LocationScaleFamily

object Likelihood

object Log extends Logger

object LogNormal

object Multinomial extends Serializable

object Normal extends LocationScaleFamily

object Predictor

object RandomVariable

object SBC extends Serializable

object Support

object ToGenerator

object UnboundedSupport extends Support

object Uniform

Ungrouped