MixModelBayes-class: Class '"MixModelBayes"'

Description Objects from the Class Slots Extends Methods Note Author(s) See Also Examples

Description

This class stores a Bayesian mixture model fitted by MCMC methods.

Objects from the Class

Objects can be created by calls of the form new("MixModelBayes", ...).

Slots

chains:

Object of class "list" storing the course of the Markov chains for each parameter.

mmData:

Object of class "numeric" storing the data.

configuration:

Object of class "list" storing configuration. See notes for details.

results:

Object of class "list" storing results. See notes for details.

Extends

Class "MixModel", directly.

Methods

chains

signature(object = "MixModelBayes"): Gives access to the chains slot of the object.

acceptanceRate

signature(object = "MixModelBayes"): Gives the acceptance rate for the parameter of the Dirichlet distribution. Acceptance rates between 0.3 and 0.7 are usually desired. Values not smaller than 0.1 (not larger than 0.9) might still be acceptable. The acceptance rate is only meaningful if the option weightsPrior was set to the Finite-dimensional Dirichlet prior.

Note

In addition to the content described in MixModel, the following elements are present: Slot configuration:

  1. initsAs in MixModel.

  2. priorsA list specifying the prior distributions for the parameters of the components and the parameter of the Dirichlet process.

  3. chainA list with the technical specifications for the Markov Chains.

Slot results is exactly like in MixModel. Slot chains:

  1. componentsA list giving the values for the parameters of the components in each iteration after burn-in and application of thinning.

  2. piA matrix giving the values for the weights pi of the components in each iteration after burn-in and application of thinning.

  3. dirichletParameterA vector giving the values for dirichlet Parameter in each iteration after burn-in and application of thinning.

  4. classificationA matrix giving the number of genes classified to each components in each iteration after burn-in and application of thinning.

Author(s)

Hans-Ulrich Klein ([email protected])

See Also

bayesMixModel MixModel

Examples

1
showClass("MixModelBayes")

epigenomix documentation built on May 6, 2019, 2:17 a.m.