carolyner/imabc: Incremental Mixture Approximate Bayesian Computation (IMABC)

Provides functionality to perform a likelihood-free method for estimating the parameters of complex models that results in a simulated sample from the posterior distribution of model parameters given targets. The method begins with a accept/reject approximate bayes computation (ABC) step applied to a sample of points from the prior distribution of model parameters. Accepted points result in model predictions that are within the initially specified tolerance intervals around the target points. The sample is iteratively updated by drawing additional points from a mixture of multivariate normal distributions, accepting points within tolerance intervals. As the algorithm proceeds, the acceptance intervals are narrowed. The algorithm returns a set of points and sampling weights that account for the adaptive sampling scheme. For more details see Rutter, Ozik, DeYoreo, and Collier (2018) <arXiv:1804.02090>.

install.packages("remotes")
remotes::install_github("carolyner/imabc")