The repository includes the development version of R package rmx
Radius-minimax (rmx) estimators are a special class of optimally-robust statistical procedures (Kohl (2005), https://epub.uni-bayreuth.de/839/2/DissMKohl.pdf; Rieder et al. (2008), doi:10.1007/s10260-007-0047-7). It is a class of asymptotically linear estimators where infinitesimal (shrinking) neighborhoods around parametric models are assumed (Rieder (1994), ISBN:978-1-4684-0624-5; Ruckdeschel (2006), doi:10.1007/s00184-005-0020-0; Kohl et al. (2010), doi:10.1007/s10260-010-0133-0; Kohl (2012), doi:10.1080/02331888.2010.540668). In addition, various diagnostic plots and finite-sample corrections are included.
The rmx estimators are also implemented in the RobASt-family of packages (i.e., packages RobAStBase, ROptEst, RobLox, RobExtremes, ROptRegTS, RobRex). However, package rmx provides a much simpler interface and lower computation times at the cost of being less general and flexible.
## Development version from GitHub
# install.packages("remotes")
remotes::install_github("stamats/rmx")
library(rmx)