Copyright	(c) Adam Scibior 2015-2020
License	MIT
Maintainer	leonhard.markert@tweag.io
Stability	experimental
Portability	GHC
Safe Haskell	Safe-Inferred
Language	Haskell2010

Control.Monad.Bayes.Weighted

Description

Weighted is an instance of MonadFactor. Apply a MonadDistribution transformer to obtain a MonadMeasure that can execute probabilistic models.

Synopsis

data Weighted m a
withWeight :: Monad m => m (a, Log Double) -> Weighted m a
weighted :: Weighted m a -> m (a, Log Double)
extractWeight :: Functor m => Weighted m a -> m (Log Double)
unweighted :: Functor m => Weighted m a -> m a
applyWeight :: MonadFactor m => Weighted m a -> m a
hoist :: (forall x. m x -> n x) -> Weighted m a -> Weighted n a
runWeighted :: Weighted m a -> m (a, Log Double)

Documentation

data Weighted m a Source #

Execute the program using the prior distribution, while accumulating likelihood.

Instances

Instances details

MonadTrans Weighted Source #
Instance details Defined in Control.Monad.Bayes.Weighted Methods lift :: Monad m => m a -> Weighted m a #
MonadIO m => MonadIO (Weighted m) Source #
Instance details Defined in Control.Monad.Bayes.Weighted Methods liftIO :: IO a -> Weighted m a #
Monad m => Applicative (Weighted m) Source #
Instance details Defined in Control.Monad.Bayes.Weighted Methods pure :: a -> Weighted m a # (<>) :: Weighted m (a -> b) -> Weighted m a -> Weighted m b # liftA2 :: (a -> b -> c) -> Weighted m a -> Weighted m b -> Weighted m c # (>) :: Weighted m a -> Weighted m b -> Weighted m b # (<*) :: Weighted m a -> Weighted m b -> Weighted m a #
Functor m => Functor (Weighted m) Source #
Instance details Defined in Control.Monad.Bayes.Weighted Methods fmap :: (a -> b) -> Weighted m a -> Weighted m b # (<$) :: a -> Weighted m b -> Weighted m a #
Monad m => Monad (Weighted m) Source #
Instance details Defined in Control.Monad.Bayes.Weighted Methods (>>=) :: Weighted m a -> (a -> Weighted m b) -> Weighted m b # (>>) :: Weighted m a -> Weighted m b -> Weighted m b # return :: a -> Weighted m a #
MonadDistribution m => MonadDistribution (Weighted m) Source #
Instance details Defined in Control.Monad.Bayes.Weighted Methods random :: Weighted m Double Source # uniform :: Double -> Double -> Weighted m Double Source # normal :: Double -> Double -> Weighted m Double Source # gamma :: Double -> Double -> Weighted m Double Source # beta :: Double -> Double -> Weighted m Double Source # bernoulli :: Double -> Weighted m Bool Source # categorical :: Vector v Double => v Double -> Weighted m Int Source # logCategorical :: (Vector v (Log Double), Vector v Double) => v (Log Double) -> Weighted m Int Source # uniformD :: [a] -> Weighted m a Source # geometric :: Double -> Weighted m Int Source # poisson :: Double -> Weighted m Int Source # dirichlet :: Vector v Double => v Double -> Weighted m (v Double) Source #
Monad m => MonadFactor (Weighted m) Source #
Instance details Defined in Control.Monad.Bayes.Weighted Methods score :: Log Double -> Weighted m () Source #
MonadDistribution m => MonadMeasure (Weighted m) Source #
Instance details Defined in Control.Monad.Bayes.Weighted

withWeight :: Monad m => m (a, Log Double) -> Weighted m a Source #

Embed a random variable with explicitly given likelihood.

weighted . withWeight = id

weighted :: Weighted m a -> m (a, Log Double) Source #

Obtain an explicit value of the likelihood for a given value.

extractWeight :: Functor m => Weighted m a -> m (Log Double) Source #

Compute the weight and discard the sample.

unweighted :: Functor m => Weighted m a -> m a Source #

Compute the sample and discard the weight.

This operation introduces bias.

applyWeight :: MonadFactor m => Weighted m a -> m a Source #

Use the weight as a factor in the transformed monad.

hoist :: (forall x. m x -> n x) -> Weighted m a -> Weighted n a Source #

Apply a transformation to the transformed monad.

runWeighted :: Weighted m a -> m (a, Log Double) Source #

Obtain an explicit value of the likelihood for a given value.