vthorrf/birm: Fitting Bayesian Item Response Models

Functions for fitting Bayesian versions of Item Response Models. It was based on the edstan package, but instead of using Stan for Bayesian calculations, it uses LaplacesDemon. Stan uses the NUTS MCMC algorithm for fitting the data. In our package, the user can choose from the five estimation methods implemented on LaplacesDemon (Variational Bayes; Laplace Approximation; HARM MCMC; PMC; and Iterative Quadrature), as well as from four methods of MAP estimation: Simulated Annealing; Genetic Algorithm; Steepest Gradient Descent; or Steepest ADAM.

install.packages("remotes")
remotes::install_github("vthorrf/birm")