R package that finds efficient designs for choice experiments with mixtures. It uses a coordinate-exchange algorithm and allows the computation of locally D-optimal, Bayesian D-optimal, locally I-optimal, and Bayesian I-optimal designs for first-order, second-order, and special-cubic Scheffé models.
Work in progress. Further instructions and details will be added to this README in the near future.
The article Bayesian I-optimal designs for choice experiments with mixtures by Mario Becerra and Peter Goos uses this package. The paper can be found on Science Direct (with paywall) or Arxiv (open access).
To reproduce the results in the paper, go to this repository.