Mcmc.Algorithm.MHG

Creation date: Tue May 5 20:11:30 2020.

The Metropolis-Hastings-Green (MHG) algorithm.

For example, see Geyer, C. J., Introduction to Markov chain Monte Carlo, In Handbook of Markov Chain Monte Carlo (pp. 45) (2011). CRC press.

Documentation

newtype MHG a

The MHG algorithm.

Constructors

MHG

Fields

• fromMHG :: Chain a

Instances

ToJSON a => Algorithm (MHG a)

Defined in Mcmc.Algorithm.MHG

Methods

aName :: MHG a -> String

aIteration :: MHG a -> Int

aIsInValidState :: MHG a -> Bool

aIterate :: IterationMode -> ParallelizationMode -> MHG a -> IO (MHG a)

aAutoTune :: Int -> MHG a -> IO (MHG a)

aResetAcceptance :: MHG a -> MHG a

aSummarizeCycle :: IterationMode -> MHG a -> ByteString

aOpenMonitors :: AnalysisName -> ExecutionMode -> MHG a -> IO (MHG a)

aExecuteMonitors :: Verbosity -> UTCTime -> Int -> MHG a -> IO (Maybe ByteString)

aStdMonitorHeader :: MHG a -> ByteString

aCloseMonitors :: MHG a -> IO (MHG a)

aSave :: AnalysisName -> MHG a -> IO ()

mhg :: Settings -> PriorFunction a -> LikelihoodFunction a -> Cycle a -> Monitor a -> InitialState a -> GenIO -> IO (MHG a)

Initialize an MHG algorithm.

NOTE: Computation in the IO Monad is necessary because the trace is mutable.

mhgSave :: ToJSON a => AnalysisName -> MHG a -> IO ()

Save an MHG algorithm.

mhgLoad :: FromJSON a => PriorFunction a -> LikelihoodFunction a -> Cycle a -> Monitor a -> AnalysisName -> IO (MHG a)

Load an MHG algorithm.

See mcmcContinue.

mhgLoadUnsafe :: FromJSON a => PriorFunction a -> LikelihoodFunction a -> Cycle a -> Monitor a -> AnalysisName -> IO (MHG a)

See mhgLoad but do not perform sanity checks.

Useful when restarting a run with changed prior function, likelihood function or proposals. Use with care!

type MHGRatio = Log Double

MHG ratios are stored in log domain.

mhgAccept :: MHGRatio -> GenIO -> IO Bool

Accept or reject a proposal with given MHG ratio?