We are working with the development team of NIMBLE, a system for building and sharing analysis methods for statistical models, especially for hierarchical models and computationally-intensive methods. NIMBLE is built in R but compiles your models and algorithms using C++ for speed.
We are working on using NIMBLE to build common hierarchical models used in ecology, mainly occupancy models. We are building different flavors of occupancy models and looking at efficiency gains when using NIMBLE vs. JAGS. We are also exploring alternative methods to data augmentation used to estimate unseen species in BUGS/JAGS. Finally, we are developing methods for model assessment and selection.