R-Forge Project: RobASt - Robust Asymptotic Statistics - News

RobASt release 1.1

ruckdeschel@users.r-forge.r-project.org (Peter Ruckdeschel) — Mon, 20 Aug 2018 07:26:01 GMT

Updates for the packages of the RobASt family are now avaialable on CRAN in
version >= 1.1.0

Most importantly, we have (finally) released on CRAN a (long announced) new
package

"RobExtremes"

in the RobASt family of packages.

+ It provides (speeded up) optimally-robust estimators [MBRE, OMSE, RMXE]
for Generalized Extreme Value [GEV] distributions, Generalized Pareto
distributions [GPD], Pareto distributions,
+ As other examples of L2 differentiable Scale-shape families, it also
provides these (speeded up) estimators for Weibull and Gamma
distributions.
+ It has robust (high-breakdown) starting estimators for
- GPD (PickandsEstimator, medkMAD, medSn, medQn)
- GEV (PickandsEstimator)
- Pareto (Cramér-von-Mises-Minimum-Distance-Estimator)
- Weibull (the quantile based estimator of Boudt/Caliskan/Croux)
+ For all these families, of course, MLEs and Minimum-Distance-Estimators
are also available through package distrMod
+ We bridge to the diagnostics provided by package ismev, i.e. our
return objects can be plugged into the diagnostics of this package
+ We have the usual diagnostic plots from package RobAStBase, i.e.
- Outylingness plots
- IC plots
- Information plots
- compareIC plots
- Cniperpoint plots (from ROptEst)
but also (adopted from package distrMod)
- qqplots (with confidence bands)
- returnlevel plots
+ As a starting point you may look at the included script
"RobFitsAtRealData.R" in the scripts folder of the package,
accessible by
file.path(system.file(package="RobExtremes"),
"scripts/RobFitsAtRealData.R")

This is joint work with Nataliya Horbenko (whose PhD thesis went into this
package to a large extent), nataliya.horbenko@gmail.de, with contributions
by Dasha Pupashenko, Misha Pupashenko, Gerald Kroisandt, Eugen Massini,
Sascha Desmettre and Bernhard Spangl in the framework of project
"Robust Risk Estimation" (2011-2016) funded by Volkswagen foundation
(and gratefully ackknowledged). Thanks also goes to the maintainers of CRAN,
in particully to Uwe Ligges who greatly helped us with finding an appropriate
way to store the database of interpolating functions which allow the speed up
-- this is now package RobAStRDA on CRAN.

References
N. Horbenko, P. Ruckdeschel, T. Bae (2011): Robust Estimation of Operational
Risk. Journal of Operational Risk 6(2), 3-30.
Ruckdeschel, P. and Horbenko, N. (2011): Optimally-Robust Estimators in
Generalized Pareto Models. Statistics. 47(4), 762–791.
Ruckdeschel, P. and Horbenko, N. (2012): Yet another breakdown point notion:
EFSBP –illustrated at scale-shape models. Metrika, 75(8), 1025–1047.

=================================================================================
In the other packages of the RobASt family of pkgs, the most important changes are:

As in distr 2.7, wherever possible we now use q.l internally instead of q to
provide functionality in IRKernel

RobAStBase:
- we enhanced our diagnostic plots:
+ all diagnostics (including qqplot and returnlevelplot) have adopted the same
argument naming (and selection paradigm)
the suffix is .lbs instead of .lbl,
the attributes of shown points have ending .pts
the observations are classed into three groups:
- the labelled observations selected through which.lbs and which.Order
- the shown non labelled observations (which are not in the previous set)
selected by which.nonlbs
- the non-shown observations (the remaining ones not contained in the former 2 grps)
-> point attributes may either refer to prior selection or to post-selection in
which case we have .npts variants
+ wherever possible arguments are vectorized to allow point - individual attributes
+ plot methods now return an S3 object of class \code{c("plotInfo","DiagnInfo")},
i.e., a list containing the information needed to produce the respective plot,
which at a later stage could be used by different graphic engines (like, e.g.
\code{ggplot}) to produce the plot in a different framework.
+ new methods for returnlevelplot for RobModel, InfRobModel, kStepEstimate (as qqplot)
ROptEst:
+ new wrapper functions RMXEstimator, OBREstimator, MBREstimator, OMSEstimator
+ several tweaks to speed up things:
- optIC gains argument withMakeIC
- roptest gains argument withMakeIC
- getStartIC-methods gain argument withMakeIC
- getRiskIC and getBiasIC gain argument withCheck
RobAStRDA:
+ the Lagrange multiplier interpolaters allowing for speed up in our opt-robust
estimators have been re-built as the current .rda file was corrupted

For details please see the NEWS files in the packages, available as
NEWS("<pkgname>").

Best regards from the main developpers & maintainers,
Peter Ruckdeschel (peter.ruckdeschel@uni-oldenburg.de) &
Matthias Kohl (matthias.kohl@stamats.de)

Version 0.8 of RobASt-family of packages on CRAN soon

ruckdeschel@users.r-forge.r-project.org (Peter Ruckdeschel) — Thu, 20 Jan 2011 09:53:50 GMT

New versions 0.8 of our RobASt-family of packages are now available on CRAN.
[we have just uploaded them to CRAN]

Most importantly, we have included:

+ a quasi-MC trick by Nataliya Horbenko to better produce
random variables under complicated not necessarily
monotone transformations

+ enhanced functions
infoPlot, (plots relative information used for coordinates
of a parameter estimator)
ddPlot, (distance-distance plot)
cniperPointPlot, (cniper concept for seemingly harmless
contamination behavior)
qqplot (now gets outlier corrected versions)

+ new risks: asAnscombe, asL1, asL4
for asymptotic L1 L4 risk, and optimal bias robust estimator,
to given efficiency loss in ideal model

+ new helper methods makeIC
to apply to functions or list of functions
for easily producing (suboptimal) ICs

+ new function getReq for two ICs IC1 and IC2
to compute a radius interval where IC1 is better
than IC2 acc. to G-Risk

+ new function getMaxIneff() to compute,
for any IC of class 'IC', the maximal inefficiency
for radius r varying in [0,Inf)

+ as well as several bug fixes

For more details see the corresponding NEWS files
(e.g. news(package = "RobAStBase")
or using function NEWS from package startupmsg
i.e. NEWS("RobAStBase")).

Best
Peter
Matthias
Nataliya

version 0.7 of robast-family on CRAN

stamats@users.r-forge.r-project.org (Matthias Kohl) — Sat, 07 Nov 2009 13:23:37 GMT

We recently submitted version 0.7 of our robast-family of packages to CRAN where the new package RobLoxBioC, which provides robust methods for preprocessing omics-data, was added.
Peter
Matthias

New package RobLoxBioC in project RobASt

stamats@users.r-forge.r-project.org (Matthias Kohl) — Sun, 26 Apr 2009 16:18:57 GMT

there is a new package called RobLoxBioC in project RobASt which can be used for preprocessing Affymetrix and Illumina gene expression data.

New versions of RobASt-packages on CRAN

stamats@users.r-forge.r-project.org (Matthias Kohl) — Tue, 16 Sep 2008 19:04:30 GMT

-----------------------------------------------------------------------------------------
Packages for the computation of optimally robust estimators
-----------------------------------------------------------------------------------------

We would like to announce the availability on CRAN (with possibly a
minor delay until on every mirror) of new versions of our packages for
the computation of optimally robust estimators; i.e., "RandVar",
"ROptEst", "RobLox" as well as a new package "RobAStBase" (not yet:
ROptRegTS and RobRex).

-----------------------------------------------------------------------------------------
Devel versions on R-forge
-----------------------------------------------------------------------------------------
The development of these packages is under r-forge project RobASt
(Robust Asymptotic Statistics):

http://r-forge.r-project.org/projects/robast/
http://robast.r-forge.r-project.org/

If you find this project interesting and would like to collaborate, you
are warmly welcome.

We look forward to receiving questions, comments and suggestions.

Matthias Kohl
Peter Ruckdeschel

-----------------------------------------------------------------------------------------
RandVar - Implementation of random variables (version 0.6.3)
-----------------------------------------------------------------------------------------
The package RandVar which includes an S4 implementation of random
variables together with the packages distr, distrEx and distrMod form
the basis of our packages on robust statistics.

-----------------------------------------------------------------------------------------
RobAStBase - Robust Asymptotic Statistics (version 0.1.0)
-----------------------------------------------------------------------------------------
This is a new package including some necessary S4 class infrastructure
like neighborhoods, influence curves and robust models.

-----------------------------------------------------------------------------------------
ROptEst - Optimally robust estimation (version 0.6.0)
-----------------------------------------------------------------------------------------
This is the main package for the optimally robust estimation in smoothly
(L2-differentiable) parametric models [optimal in the sense of the
shrinking neighborhood setup]. By using S4 classes and methods
the implementation so far covers the optimally robust estimation for
all(!) smoothly (L2-differentiable/differentiable in quadratic mean)
parametric models which are based on a univariate distribution. Many
well-known parametric (in particular, exponential) families (Binomial,
Poission, Normal, Gamma, Gumbel, ...) are L2-differentiable.
We include several
+neighborhood types (convex contamination, total variation)
+risks (MSE, Hampel, overshoot/undershoot),
+bias-types (symmetric, one-sided, asymmetric)
+norms (unstandardized, self-standardized, information-standardized)
for all these models.
After installation you find a folder "scripts" in the package directory
which includes many example scripts.
As the computation of optimally robust estimators involves several
steps, we -- in this new version -- added an interface function
"roptest" which can be used to perform all steps via one function.

-----------------------------------------------------------------------------------------
RobLox - Optimally robust influence curves for location and scale
(version 0.6.0)
-----------------------------------------------------------------------------------------
This package includes functions for the computation of many well known
influence curves (e.g., Huber-, Hampel-, Tukey-, Andrews-type) for
normal location and scale in the framework of our asymptotic setup.
Moreover, (and for us, more importantly) it includes the functions
"roblox", "rowRoblox" and "colRoblox" which can be used to compute
optimally robust estimators in case of normal location and scale. These
functions are optimized for speed and can be applied to large scale
problems like for instance gene expression data. Using rowRobLox the
computation for a 50000 x 20 matrix takes about 2 sec. on a Centrino Duo
with 1.66 GHz. As a comparison (all on the same system): using apply and
huberM (robustbase), resp. huber (MASS) takes about 168 sec. resp 197
sec., using apply and roblox takes about 16 minutes and using apply and
roptest (ROptEst) takes about 1 month.

-----------------------------------------------------------------------------------------
ROptRegTS - Optimally robust estimation for regression-type models
RobRex - Optimally robust influence curves for regression and scale
-----------------------------------------------------------------------------------------
These two packages which provide S4 classes and methods for the
computation of optimally robust estimators in regression-type models are
not yet adapted to the new implementation. If you are interested in
working with these packages you have to use the old versions of the
above packages which we are pleased to provide on request (the sources
can also be found in the CRAN archives). But, of course, we will try to
update these packages as soon as possible.

Start of Project RobASt

stamats@users.r-forge.r-project.org (Matthias Kohl) — Sat, 02 Feb 2008 14:30:49 GMT

The project RobASt 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). So far, it includes the R packages
RandVar, ROptEst, RobLox, ROptRegTS, RobRex.