mighty-metropolis

The Metropolis algorithm.

http://github.com/jtobin/mighty-metropolis

LTS Haskell 22.37:	2.0.0
Stackage Nightly 2024-10-11:	2.0.0
Latest on Hackage:	2.0.0

See all snapshots mighty-metropolis appears in

MIT licensed by Jared Tobin

Maintained by [email protected]

This version can be pinned in stack with:mighty-metropolis-2.0.0@sha256:0d97141389077346326a684d929f31c31e6306c5f48e6b9197676366bd08049e,2526

Module documentation for 2.0.0

Numeric
- Numeric.MCMC
  - Numeric.MCMC.Metropolis

Depends on 7 packages(full list with versions):

base, kan-extensions, mcmc-types, mwc-probability, pipes, primitive, transformers

Used by 1 package in lts-21.21(full list with versions):

declarative

The classic Metropolis algorithm.

Wander around parameter space according to a simple spherical Gaussian distribution.

Exports a mcmc function that prints a trace to stdout, a chain function for collecting results in-memory, and a metropolis transition operator that can be used more generally.

import Numeric.MCMC.Metropolis

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

main :: IO ()
main = withSystemRandom . asGenIO $ mcmc 10000 1 [0, 0] rosenbrock