Bayesian Aoristic Analyses

Provides an alternative approach to aoristic analyses for archaeological datasets by fitting Bayesian parametric growth models and non-parametric random-walk Intrinsic Conditional Autoregressive (ICAR) models on time frequency data (Crema (2024)). It handles event typo-chronology based timespans defined by start/end date as well as more complex user-provided vector of probabilities.

baorista

baorista is an R package that provides a Bayesian inferential tool for analysing time-frequencies of archaeological events associated with time spans typically obtained from relative chronological sequences (e.g. periods and phases) and often analyses using aoristic sums. At its core baorista is a frontend for fitting Bayesian models via the NIMBLE probabilistic programming language.

Please note that baorista is based on Nimble, which requires a working C++ compiler. For more information please read the dedicated section of the nimble manual.

For a quick introduction to baorista check the package vignette as well as the associated paper and its github repo

Funding

The development of this package was funded by a Philip Leverhulme Prize (PLP-2019-304) awarded to E.Crema.