Robust (or "resistant") methods for statistics modelling have been
available in S from the start, in R in package
mean(*, trim = . ),
(the statistic behind
package graphics) or
lowess(), which had been complemented
Much further important functionality has been made available in package
(by Bill Venables and Brian
Modern Applied Statistics with S
for robust regression and
robust multivariate scatter and covariance.
This task view is about R add-on packages providing newer or faster,
more efficient algorithms and notably for (robustification of) new models.
Please send suggestions for additions and extensions to the
task view maintainer
An international group of scientists working in the field of robust
statistics has made efforts (since October 2005) to coordinate several of
the scattered developments and make the important ones available
through a set of R packages complementing each other.
These should build on a basic package with "Essentials",
with (potentially many) other packages
building on top and extending the essential functionality to particular
models and or applications.
Further, there is the quite comprehensive package
robust, a version of the robust library of S-PLUS,
as an R package now GPLicensed thanks to Insightful and Kjell Konis.
Whereas there is currently quite a bit of overlap (between 'robustbase'
and 'robust'), the intent (as of Nov.2007) is that
robustbase, the former
providing convenient routines for the casual user where the latter
will contain the underlying functionality, and provide the more
advanced statistician with a considerable choice of methodology.
We structure the packages roughly into the following topics, and
typically will first mention functionality in packages
Regression (Linear, Generalized Linear, Nonlinear Models)
(robust) where the former uses the latest of the
fast-S algorithms and heteroscedasticity and autocorrelation corrected
(HAC) standard errors, the latter makes use of the M-S algorithm of
Maronna and Yohai (2000), automatically when there are factors
among the predictors (where S-estimators (and hence MM-estimators)
based on resampling typically badly fail).
are available in
robustbase, but rather for comparison
Note that Koenker's quantile regression package
contains L1 (aka LAD, least absolute deviations)-regression as a
special case, doing so also for nonparametric regression via
median-based (Theil-Sen or Siegel's repeated) simple linear models.
Generalized linear models (GLMs) are provided both via
Robust Nonlinear model fitting is available through
fits overdispersed multinomial regression
models for count data.
fits robust GAMs, i.e., robust Generalized Additive
package which builds ("
provides nice S4 class based methods,
that are planned to be moved into
robustbase, and additionally
robust PCA methodology.
contains a slightly more flexible
fastmcd(), and similarly for
has automatically chosen
for large dimensionality p.
sign methods for outlier identification in high dimensions.
performs robust inference based on
ootstrap on robust estimators, including
multivarate regression, PCA and Hotelling tests.
nearest neighbor variance estimation (NNVE) method of Wang and
Note that robust PCA can be performed by using standard
X <- stackloss; pc.rob <- princomp(X, covmat= MASS::cov.rob(X))
See also the CRAN task views
Large Data Sets
(See also the CRAN task view
Descriptive Statistics / Exploratory Data Analysis
Note however that these (last two items) are not yet available from CRAN.
contains robust regression and
filtering methods for univariate time series, typically based on
repeated (weighted) median regressions.
Peter Ruckdeschel has started to lead an effort for a robust
time-series package, see
"Routines for Robust Kalman
Filtering --- the ACM- and rLS-filter"
, is being developed, see
Econometricians tend to like HAC (heteroscedasticity and
autocorrelation corrected) standard errors. For a broad class of
models, these are provided by package
also uses a version of HAC
standard errors for its robustly estimated linear models.
See also the CRAN task view
Robust Methods for Bioinformatics
There are several packages in the
providing specialized robust methods.
Other approaches to robust and resistant methodology
and its several child packages
also allow to explore robust estimation concepts, see e.g.,
Notably, based on these,
aims for the implementation of R
packages for the computation of optimally robust estimators and
tests as well as the necessary infrastructure (mainly S4 classes
and methods) and diagnostics; cf. M. Kohl (2005).
It includes the R packages
computes Robust Accelerated Failure
Time Regression for Gaussian and logWeibull errors.