mighty-metropolis

The classic Metropolis algorithm.
git clone git://git.jtobin.io/mighty-metropolis.git
Log | Files | Refs | README | LICENSE

README.md (1271B)

      1 # mighty-metropolis
      2 
      3 [![Build Status](https://secure.travis-ci.org/jtobin/mighty-metropolis.png)](http://travis-ci.org/jtobin/mighty-metropolis)
      4 [![Hackage Version](https://img.shields.io/hackage/v/mighty-metropolis.svg)](http://hackage.haskell.org/package/mighty-metropolis)
      5 [![MIT License](https://img.shields.io/badge/license-MIT-blue.svg)](https://github.com/jtobin/mighty-metropolis/blob/master/LICENSE)
      6 
      7 The classic Metropolis algorithm.  Wander around parameter space according to a
      8 simple spherical Gaussian distribution.
      9 
     10 Exports a `mcmc` function that prints a trace to stdout, a `chain` function for
     11 collecting results in memory, and a `metropolis` transition operator that can
     12 be used more generally.
     13 
     14 See the *test* directory for example usage.
     15 
     16     import Numeric.MCMC.Metropolis
     17 
     18     rosenbrock :: [Double] -> Double
     19     rosenbrock [x0, x1] = negate (5  *(x1 - x0 ^ 2) ^ 2 + 0.05 * (1 - x0) ^ 2)
     20 
     21     main :: IO ()
     22     main = withSystemRandom . asGenIO $ mcmc 10000 1 [0, 0] rosenbrock
     23 
     24 ![trace](https://dl.dropboxusercontent.com/spa/u0s6617yxinm2ca/osecfv_w.png)
     25 
     26 *mighty-metropolis* is a member of the [declarative][decl] suite of libraries,
     27 containing a bunch of MCMC algorithms that play nicely together.
     28 
     29 [decl]: https://github.com/jtobin/declarative