Control.Monad.Bayes.Population

Reexports

Definitions

record ListT : (Type -> Type) -> Type -> Type

  List transformer

Totality: total
Visibility: public export
Constructor: 

MkListT : m (List a) -> ListT m a

Projection: 

.runListT : ListT m a -> m (List a)

Hints:

Applicative m => Applicative (ListT m)

Functor m => Functor (ListT m)

Monad m => Monad (ListT m)

MonadCond m => MonadCond (ListT m)

MonadInfer m => MonadInfer (ListT m)

MonadSample m => MonadSample (ListT m)

MonadTrans ListT

.runListT : ListT m a -> m (List a)

Visibility: public export

runListT : ListT m a -> m (List a)

Visibility: public export

record Population : (Type -> Type) -> Type -> Type

  A collection of weighted samples, or particles.

Totality: total
Visibility: public export
Constructor: 

MkPopulation : Weighted (ListT m) a -> Population m a

Projection: 

.runPopulation' : Population m a -> Weighted (ListT m) a

Hints:

Monad m => Applicative (Population m)

Functor m => Functor (Population m)

Monad m => Monad (Population m)

Monad m => MonadCond (Population m)

MonadSample m => MonadInfer (Population m)

MonadSample m => MonadSample (Population m)

MonadTrans Population

.runPopulation' : Population m a -> Weighted (ListT m) a

Visibility: public export

runPopulation' : Population m a -> Weighted (ListT m) a

Visibility: public export

runPopulation : Population m a -> m (List (Log Double, a))

  Explicit representation of the weighted sample with weights in the log domain.

Visibility: export

explicitPopulation : Functor m => Population m a -> m (List (Double, a))

  Explicit representation of the weighted sample.

Visibility: export

fromWeightedList : Monad m => m (List (Log Double, a)) -> Population m a

  Initialize 'Population' with a concrete weighted sample.

Visibility: export

hoist : Monad m2 => (m1 x -> m2 x) -> Population m1 a -> Population m2 a

  Applies a transformation to the inner monad.

Visibility: export

spawn : Monad m => Nat -> Population m ()

  Increase the sample size by a given factor.
  The weights are adjusted such that their sum is preserved. It is therefore
  safe to use 'spawn' in arbitrary places in the program without introducing bias.

Visibility: export

resampleGeneric : MonadSample m => (Vect k Double -> m (List (Fin k))) -> Population m a -> Population m a

Visibility: export

systematic : Double -> Vect n Double -> List (Fin n)

  Systematic sampler.

Visibility: export

resampleSystematic : MonadSample m => Population m a -> Population m a

  Resample the population using the underlying monad and a systematic resampling scheme.
  The total weight is preserved.

Visibility: export

stratified : MonadSample m => Vect n Double -> m (List (Fin n))

  The conditional variance of stratified sampling is always smaller than that of multinomial
  sampling and it is also unbiased - see [Comparison of Resampling Schemes for Particle
  Filtering](https://arxiv.org/abs/cs/0507025).

Visibility: export

resampleStratified : MonadSample m => Population m a -> Population m a

  Resample the population using the underlying monad and a stratified resampling scheme.
  The total weight is preserved.

Visibility: export

multinomial : MonadSample m => Vect n Double -> m (List (Fin n))

  Multinomial sampler.  Sample from \(0, \ldots, n - 1\) \(n\)
  times drawn at random according to the weights where \(n\) is the
  length of vector of weights.

Visibility: export

resampleMultinomial : MonadSample m => Population m a -> Population m a

  Resample the population using the underlying monad and a multinomial resampling scheme.
  The total weight is preserved.

Visibility: export

extractEvidence : Monad m => Population m a -> Population (Weighted m) a

  Separate the sum of weights into the 'Weighted' transformer.
  Weights are normalized after this operation.

Visibility: export

pushEvidence : MonadCond m => Population m a -> Population m a

  Push the evidence estimator as a score to the transformed monad.
  Weights are normalized after this operation.

Visibility: export

proper : MonadSample m => Population m a -> Weighted m a

  A properly weighted single sample, that is one picked at random according
  to the weights, with the sum of all weights.

Visibility: export

evidence : Monad m => Population m a -> m (Log Double)

  Model evidence estimator, also known as pseudo-marginal likelihood.

Visibility: export

collapse : MonadInfer m => Population m a -> m a

  Picks one point from the population and uses model evidence as a 'score' in the transformed monad.
  This way a single sample can be selected from a population without introducing bias.

Visibility: export