ghc-flat-mcmc

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 qualified Data.Vector.Unboxed as U
(unsafeIndex) > > rosenbrock :: Particle -> Double > rosenbrock xs = negate (5

(x1 - x0 ^ 2) ^ 2 + 0.05 * (1 - x0) ^ 2) where > x0 = U.unsafeIndex xs 0 > x1

= U.unsafeIndex xs 1 > > origin :: Ensemble > origin = ensemble [ > particle
[negate 1.0, negate 1.0] > , particle [negate 1.0, 1.0] > , particle [1.0,
negate 1.0] > , particle [1.0, 1.0] > ] > > main :: IO () > main =
withSystemRandom . asGenIO $ mcmc 12500 origin rosenbrock.

          

---

License: MIT
URL: https://hackage.haskell.org/package/flat-mcmc