The flat-mcmc package

[Tags:library, mit, test]

flat-mcmc is a Haskell library for painless, efficient, general-purpose sampling from continuous distributions.

flat-mcmc uses an ensemble sampler that is invariant to affine transformations of space. It wanders a target probability distribution's parameter space as if it had been flattened or unstretched in some sense, allowing many particles to explore it locally and in parallel.

In general this sampler is useful when you want decent performance without dealing with any tuning parameters or local proposal distributions.

flat-mcmc exports an mcmc function that prints a trace to stdout, as well as a flat transition operator that can be used more generally.

 import Numeric.MCMC.Flat
 import Data.Vector (Vector, toList, fromList)

 rosenbrock :: Vector Double -> Double
 rosenbrock xs = negate (5  *(x1 - x0 ^ 2) ^ 2 + 0.05 * (1 - x0) ^ 2) where
   [x0, x1] = toList xs

 ensemble :: Ensemble
 ensemble = fromList [
     fromList [negate 1.0, negate 1.0]
   , fromList [negate 1.0, 1.0]
   , fromList [1.0, negate 1.0]
   , fromList [1.0, 1.0]
   ]

 main :: IO ()
 main = withSystemRandom . asGenIO $ mcmc 12500 ensemble rosenbrock

Properties

Versions 0.1.0.0, 1.0.0, 1.0.1
Dependencies base (<5), mcmc-types (>=1.0.1 && <2), monad-par, monad-par-extras, mwc-probability (>=1.0.1 && <2), pipes (==4.*), primitive, transformers, vector [details]
License MIT
Author Jared Tobin
Maintainer jared@jtobin.ca
Stability Unknown
Category Math
Home page http://jtobin.github.com/flat-mcmc
Source repository head: git clone http://github.com/jtobin/flat-mcmc.git
Uploaded Wed Apr 6 13:42:23 UTC 2016 by JaredTobin
Distributions LTSHaskell:1.0.1, NixOS:1.0.1, Stackage:1.0.1, Tumbleweed:1.0.1
Downloads 369 total (8 in the last 30 days)
Votes
0 []
Status Docs pending
Build status unknown [no reports yet]

Modules

  • Numeric
    • MCMC
      • Numeric.MCMC.Flat

Downloads

Maintainer's Corner

For package maintainers and hackage trustees