simmr: A package for fitting Stable Isotope Mixing Models in R
We introduce an R package for fitting Stable Isotope Mixing Models (SIMMs) via both Markov chain Monte Carlo and Variational Bayes. The package is mainly used for estimating dietary contributions from food sources taken via measurements of stable isotope ratios from animals. It can also be used to estimate proportional contributions of a mixture from known sources, for example apportionment of river sediment, amongst many other use cases. The package contains a simple structure which allows non-expert users to interface with the package, with most of the computational complexity hidden behind the main fitting functions. In this paper we detail the background to these functions and provide case studies on how the package should be used. Further examples are available in the online package vignettes.
READ FULL TEXT