README.md

In poworoznek/infinitefactor: Bayesian Infinite Factor Models

infinitefactor

Bayesian infinite factor modelling

This package for R was developed for modular construction of Bayesian factor samplers using a variety of data models and priors. Bayesian factor models are key tools for performing reproducible and robust dimension reduction and modeling complex relationships among sets of regressors. Accounting for structure with factor models allows both efficiencies in estimation, and inference on that structure. Popular shrinkage priors (MGSP, Dirichlet-Laplace, others) have been formulated for the factor loadings matrix that do not require prespecified covariate grouping or order-dependence. This package aims to implement those priors for use with multiple data models in an easy, straightforward, and fast environment for application, as well as encourage further model and prior development through use of modular and exchangeable parameter updates. Install the most recent version from this GitHub repository using:

devtools::install_github('poworoznek/infinitefactor')

or in release version from CRAN using:

install.packages('infinitefactor')

Augmentation

The augment sampler input argument can be used to modify sampler behavior for non-standard applications. This argument takes an R expression to be evaluated every sampler iteration after the standard parameter updates and before samples are stored. A straightforward application of this argument is to change a sampling parameter from a single value to a random variable. This can be done to place further hyperpriors on shrinkage parameters, or to perform data augmentation. Refer to the sample source code for parameter names.

Missingness

A common form of data augmentation is imputation due to missingness. In the included factor models the posterior predictive distributions of the factorized data matrix X and the interaction response y have a simple form. Here we provide example code to sample missing entries of X under a Missing at Random (MAR) assumption.

# for data matrix "data" with NA missing values

completeX = function(X, Xmiss, lambda, eta, ps){
  noise = t(replicate(nrow(X), rnorm(ncol(X), 0, sqrt(1/ps))))
  X[is.na(Xmiss)] = (tcrossprod(eta, lambda) +noise)[is.na(Xmiss)]
  return(X)}

missing = expression({X = completeX(X, data, lambda, eta, ps)})

X = data
X[is.na(X)] = rnorm(sum(is.na(X)))

sample = linearDL(X, 10000, 5000, augment = missing)

Acknowledgements

This package is based upon work supported by the National Institute for Environmental Health Sciences under grant 1R01ES028804-01. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Institute for Environmental Health Sciences.

poworoznek/infinitefactor documentation built on April 7, 2020, 1:10 p.m.