mmlasso-package: Robust and Sparse Estimators for Linear Regression Models

In esmucler/mmlasso: Robust and Sparse Estimators for Linear Regression Models

Description

Functions to calculate the MM-Lasso and adaptive MM-Lasso estimators proposed in Smucler and Yohai (2015). The S-Ridge estimator of Maronna (2011) is used as the initial estimator.

Details

Package:	mmlasso
Type:	Package
Version:	1.3.4
Date:	2016-2-26
License:	GPL (>= 2)
Imports:	Rcpp, robustHD, robustbase, parallel, doParallel, foreach, MASS
LinkingTo:	Rcpp, RcppArmadillo
NeedsCompilation:	yes

Index:

mmlasso        Function to calculate the adaptive MM-Lasso
               and the MM-Lasso

sridge         Function to calculate the S-Ridge

Author(s)

Ezequiel Smucler <[e-mail]>

Maintainer: Ezequiel Smucler <[e-mail]>

References

Ezequiel Smucler and Victor J. Yohai. Robust and sparse estimators for linear regression models (2015). Available at http://arxiv.org/abs/1508.01967.

Maronna, R.A. (2011). Robust Ridge Regression for High-Dimensional Data. Technometrics 53 44-53.

Examples

require(MASS)
p <- 8
n <- 60
rho <- 0.5
desv <- 1
beta.true <- c(rep(0,p+1))
beta.true[2] <- 3
beta.true[3] <- 1.5
beta.true[7] <- 2
mu <- rep(0,p)
sigma <- rho^t(sapply(1:p, function(i, j) abs(i-j), 1:p))
set.seed(1234)
x <- mvrnorm(n,mu,sigma)
u <- rnorm(n)*desv
y <- x%*%beta.true[2:(p+1)]+beta.true[1]+u

###Calculate estimators

set.seed(1234)
RobSparse <- mmlasso(x,y)

esmucler/mmlasso documentation built on May 16, 2019, 8:52 a.m.