AEnetCC.aft: Adaptive elastic net for AFT models with censoring...

In AdapEnetClass: A Class of Adaptive Elastic Net Methods for Censored Data

Description

This function is used to fit AFT models by using adaptive elastic net.

Usage

AEnetCC.aft(X, Y, delta, weight, C = 1, s = 1, lambda2)

Arguments

X	covariate matrix under study, particularly for AFT modelling. The order of matrix covariate is typically n by p.
Y	typically the logarithmic of the survival time under AFT models. Otherwise survival time.
delta	status. it includes value 1 for uncensored and value 0 for censored subject.
weight	vector of observation weights.
C	this is a positive value that accounts for the penalties of violations of constraints. C is typically allowed to take values in a grid such as (0, 0.5, 1, 1.5, ..., 10).
s	this is the optimal equivalent specification for lasso in terms of fraction of the L1 norm. This is obtained from the AEnet.aft function.
lambda2	regularization parameter for the L2 norm of the coefficients. This is typically assumed to take values in a relatively small grid, say (0, 0.5, 1.0, 1.5, 2.0, ..., 5).

Details

This function is used to fit AFT models using adaptive elastic net with censorig constraints (Khan and Shaw, 2015). This is an extension of the adaptive elastic approach that allows the censoring constraints to be implemented into the optimization framework. The method works for both cases of datasets high-dimensional where the number of variables (p) is greater than the number of subjects (n) and low-dimensional.

This method use the same optimal pair of (lambda1, lambda2) as found in AEnet.aft. Then C is typically allowed to take values in a grid such as (0, 0.5, 1.0, 1.5, 2.0,...,5), and the optimal value for C obtained by 5-fold cross-validation. Here C typically depends upon how stringently one wants the model to satisfy the censoring constraints compared to how good is the prediction for uncensored data.

Value

An object of type numeric is returned that provides the coefficient estimates.

Author(s)

Hasinur Rahaman Khan and Ewart Shaw

References

Ghosh, S. (2007). Adaptive Elastic Net: An Improvement of Elastic Net to achieve Oracle Properties. Technical Reports, Indiana University- Purdue University, Indianapolis, USA. PR no. 07-01.

Khan and Shaw (2015) imputeYn: Imputing the last largest censored observation/observations under weighted least squares. R package version 1.3, https://cran.r-project.org/package=imputeYn.

Khan and Shaw (2015). Variable Selection for Survival Data with a Class of Adaptive Elastic Net Techniques. Statistics and Computing (published online; DOI: 10.1007/s11222-015-9555-8). Also available in http://arxiv.org/abs/1312.2079.

See Also

cv.AWEnetCC

Examples

#For full data typically used for AFT models (using imputeYn (2015) package).
dat<-data(n=100, p=10, r=0, b1=c(rep(5,5), rep(0, 5)), sig=1, Cper=0)

#This needs to run for generating weights of the observations
l<-mrbj(cbind(dat$y, dat$delta) ~ dat$x, mcsize=100, trace=FALSE, gehanonly=FALSE)

#AEnetCC.aft: adaptive elastic net with censoring constraints
wt<-round(l$enet,3)
ft<-AEnetCC.aft(dat$x, dat$y, dat$delta, weight=wt, C=1, s=959596, lambda2=0.5)
ft

Example output

Loading required package: imputeYn
Loading required package: quadprog
Loading required package: emplik
Loading required package: mvtnorm
Loading required package: survival
Loading required package: boot

Attaching package: 'boot'

The following object is masked from 'package:survival':

    aml


Attaching package: 'imputeYn'

The following object is masked from 'package:utils':

    data

Loading required package: glmnet
Loading required package: Matrix
Loading required package: foreach
Loaded glmnet 2.0-16

Loading required package: lars
Loaded lars 1.2

 [1]  1.086276591  0.594557074  0.088600164  0.887376727  0.676101501
 [6]  0.132631656  0.007688453 -0.585497890  0.227647941 -0.306973181

AdapEnetClass documentation built on May 2, 2019, 7:55 a.m.