Copyright	(c) George Ungureanu 2020
License	BSD-style (see the file LICENSE)
Maintainer	ugeorge@kth.se
Stability	experimental
Portability	portable
Safe Haskell	Safe
Language	Haskell2010

ForSyDe.Atom.Prob.Normal

Contents

Recipes
Layer

Description

This module defines recipes for normally distributed random variables, i.e. Gaussian distribution. Wraps utilities imported from Data.Random.Normal.

Synopsis

normal :: (Random a, Floating a) => a -> a -> Dist a
normal' :: (Random a, Floating a) => Dist a
module ForSyDe.Atom.Prob

Recipes

normal Source #

Arguments

:: (Random a, Floating a)
=> a	standard deviation
-> a	mean
-> Dist a

Recipe for normally distributed variables.

>>> gen <- getStdGen
>>> let x1 = normal 2 (0.2) :: Dist Float
>>> let x2 = normal 1 (0.5) :: Dist Float
>>> let x3 = normal 2 (0.6) :: Dist Float
>>> let hx = map (histogram 0 3 0.2 . samplesn gen 10000) [x1,x2,x3]
>>> dumpDat $ prepareL defaultCfg hx
Dumped hist1, hist2, hist3 in ./fig
["./fig/hist1.dat","./fig/hist2.dat","./fig/hist3.dat"]

normal' :: (Random a, Floating a) => Dist a Source #

Recipe for normally distributed variables using the default parameters, i.e. mean = 0 and standard deviation = 1.

>>> gen <- getStdGen
>>> let x = normal' :: Dist Float
>>> let hx = histogram (-1) 1 0.2 $ samplesn gen 10000 x
>>> dumpDat $ prepare defaultCfg hx
Dumped hist1 in ./fig
["./fig/hist1.dat"]

Layer

The ForSyDe.Atom.Prob module is re-exported for convenience, so it does not need to be imported explicitly.

module ForSyDe.Atom.Prob