The rEDM software package written by Hao Ye (Ye et al. in prep) is a toolbox of commonly used methods in empirical dynamics. It is based on laboratory research code developed in the early 90’s and contains methods to identify and quantify nonlinear predictability (Sugihara & May 1990, Sugihara 1994) as well as multivariate extensions thereof (Dixon et al. 1999, Ye and Sugihara 2016). In addition, rEDM implements convergent cross mapping as a test for causality (Sugihara et al. 2012), and a method for measuring changing interactions among causal variables (Deyle et al. 2016).
Access the software package here at github.