GNU R depth-based classification and calculation of data depth

 Contains procedures for depth-based supervised learning, which are
 entirely non-parametric, in particular the DDalpha-procedure (Lange,
 Mosler and Mozharovskyi, 2014). The training data sample is transformed
 by a statistical depth function to a compact low-dimensional space,
 where the final classification is done. It also offers an extension to
 functional data and routines for calculating certain notions of
 statistical depth functions. 50 multivariate and 5 functional
 classification problems are included.