`altar.cuda.bayesian`

Submodules

Package Contents

Classes

Functions

class altar.cuda.bayesian.controller

Bases: altar.protocol

The protocol that all AlTar controllers must implement

dispatcher

doc = the event dispatcher that activates the registered handlers

archiver

doc = the archiver of simulation state

posterior(self, model): Sample the posterior distribution of the given {model}

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

classmethod pyre_default(cls, **kwds): Supply a default implementation

class altar.cuda.bayesian.sampler

Bases: altar.protocol

The protocol that all AlTar samplers must implement

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

samplePosterior(self, controller, step): Sample the posterior distribution

resample(self, controller, statistics): Update my statistics based on the results of walking my Markov chains

classmethod pyre_default(cls, **kwds): Supply a default implementation

class altar.cuda.bayesian.scheduler

Bases: altar.protocol

The protocol that all AlTar schedulers must implement

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

update(self, step): Push {step} forward along the annealing schedule

updateTemperature(self, step): Generate the next temperature increment

computeCovariance(self, step): Compute the parameter covariance of the sample in the {step}

rank(self, step): Rebuild the sample and its statistics sorted by the likelihood of the parameter values

classmethod pyre_default(cls, **kwds): Supply a default implementation

altar.cuda.bayesian.metropolis()

altar.cuda.bayesian.metropolisvaryingsteps()

altar.cuda.bayesian.adaptivemetropolis()