likelihood.model R package

The R packge likelihood.model provides an API for specifying likelihood models for statistical inference.

The basic likelihood model is a concept that, in order for your object to satisfy, must implement a number of functions (generic methods). The package provides two different implementations of the concept:

  1. likelihood_contr_model is a flexible framework for specifying likelihood models based on the idea of independent likelihood contributions for different types of observations, e.g., right-censored versus exact observations, or other kinds observations. This model is designed to accomodate more specialized likelihood models, such as series systems with latent commponents which includes ambiguous data about the components, such as masked failure causes.

  2. likelihood_name_model provides a convenient wrapper for distribution functions that follow the naming and argument conventions in the R ecosystem, e.g., if we have some distribution norm (normal), then it has dnorm, pnorm, rnorm, and qnorm, respectively for the density function, probability function, sampler, and quantile function for the normal distribution. They also have standard paramenter arguments, like pnorm has a lower.tail Boolean parameter that computes either the CDF if TRUE and otherwise the survival function. Note that this model may be used to provide contributions to likelihood_contr_model.

The package is designed to be used with the algebraic.mle package, which provides a framework for performing maximum likelihood estimation (MLE).

comments powered by Disqus

Related