Safe Haskell	Safe-Inferred
Language	Haskell2010

Control.Monad.Bayes.Sampler.Lazy

Description

This is a port of the implementation of LazyPPL: https://lazyppl.bitbucket.io/

Synopsis

data Tree = Tree {
- currentUniform :: Double
- lazyUniforms :: Trees
}
data Trees = Trees {
- headTree :: Tree
- tailTrees :: Trees
}
newtype Sampler a = Sampler {
- runSampler :: Tree -> a
}
splitTree :: Tree -> (Tree, Tree)
randomTree :: RandomGen g => g -> Tree
randomTrees :: RandomGen g => g -> Trees
sampler :: Sampler a -> IO a
independent :: Monad m => m a -> m [a]
weightedsamples :: Weighted Sampler a -> IO [(a, Log Double)]

Documentation

data Tree Source #

A Tree is a lazy, infinitely wide and infinitely deep tree, labelled by Doubles | Our source of randomness will be a Tree, populated by uniform [0,1] choices for each label. | Often people just use a list or stream instead of a tree. | But a tree allows us to be lazy about how far we are going all the time.

Constructors

Tree
Fields currentUniform :: Double lazyUniforms :: Trees

data Trees Source #

An infinite stream of Trees.

Constructors

Trees
Fields headTree :: Tree tailTrees :: Trees

newtype Sampler a Source #

A probability distribution over a is | a function 'Tree -> a' | The idea is that it uses up bits of the tree as it runs

Constructors

Sampler
Fields runSampler :: Tree -> a

Instances

Instances details

Applicative Sampler Source #
Instance details Defined in Control.Monad.Bayes.Sampler.Lazy Methods pure :: a -> Sampler a # (<>) :: Sampler (a -> b) -> Sampler a -> Sampler b # liftA2 :: (a -> b -> c) -> Sampler a -> Sampler b -> Sampler c # (>) :: Sampler a -> Sampler b -> Sampler b # (<*) :: Sampler a -> Sampler b -> Sampler a #
Functor Sampler Source #
Instance details Defined in Control.Monad.Bayes.Sampler.Lazy Methods fmap :: (a -> b) -> Sampler a -> Sampler b # (<$) :: a -> Sampler b -> Sampler a #
Monad Sampler Source #	probabilities for a monad. \| Sequencing is done by splitting the tree \| and using different bits for different computations.
Instance details Defined in Control.Monad.Bayes.Sampler.Lazy Methods (>>=) :: Sampler a -> (a -> Sampler b) -> Sampler b # (>>) :: Sampler a -> Sampler b -> Sampler b # return :: a -> Sampler a #
MonadDistribution Sampler Source #
Instance details Defined in Control.Monad.Bayes.Sampler.Lazy Methods random :: Sampler Double Source # uniform :: Double -> Double -> Sampler Double Source # normal :: Double -> Double -> Sampler Double Source # gamma :: Double -> Double -> Sampler Double Source # beta :: Double -> Double -> Sampler Double Source # bernoulli :: Double -> Sampler Bool Source # categorical :: Vector v Double => v Double -> Sampler Int Source # logCategorical :: (Vector v (Log Double), Vector v Double) => v (Log Double) -> Sampler Int Source # uniformD :: [a] -> Sampler a Source # geometric :: Double -> Sampler Int Source # poisson :: Double -> Sampler Int Source # dirichlet :: Vector v Double => v Double -> Sampler (v Double) Source #

splitTree :: Tree -> (Tree, Tree) Source #

Two key things to do with trees: | Split tree splits a tree in two (bijectively) | Get the label at the head of the tree and discard the rest

randomTree :: RandomGen g => g -> Tree Source #

Preliminaries for the simulation methods. Generate a tree with uniform random labels. This uses split to split a random seed

randomTrees :: RandomGen g => g -> Trees Source #

sampler :: Sampler a -> IO a Source #

independent :: Monad m => m a -> m [a] Source #

weightedsamples :: Weighted Sampler a -> IO [(a, Log Double)] Source #

weightedsamples runs a probability measure and gets out a stream of (result,weight) pairs