`altar.bayesian.Annealer`

Module Contents

Classes

class altar.bayesian.Annealer.Annealer(name, locator, **kwds)

Bases: altar.component

A Bayesian controller that uses an annealing schedule and MCMC to approximate the posterior distribution of a model

sampler

doc = the sampler of the posterior distribution

scheduler

doc = the generator of the annealing schedule

dispatcher

doc = the event dispatcher that activates the registered handlers

archiver

doc = the archiver of simulation state

model

worker

info

warning

error

debug

firewall

initialize(self, application): Initialize me and my parts given an {application} context

posterior(self, model): Sample the posterior distribution

deduceAnnealingMethod(self, job): Instantiate an annealing method compatible the user choices

sequential(self): Instantiate the plain sequential annealing method

cuda(self): Instantiate a CUDA aware annealing method

threaded(self, threads, worker): Instantiate the multi-threaded annealing method

mpi(self, worker): Instantiate the MPI aware annealing method