To whom It May Concern,
I am working on a dimensional reduction problem using Smoothly Clipped
Asolute Deviation (SCAD) Penalty according to "Variable Selection via
Nonconcave Penalized Likelihood and its Oracle Properties, J.Fan and R. Li,
JASA, Dec 2001". I found an R package named *"ncvreg*" which is
capable
to perform this variable selection procedure. I noctice, in
function*ncvreg(x, y),
* an intercept is automatically added to the model fitting step. I wonder,
is there anyway to remove the intercept, something similar to
*"lars()"*where you can specify the option and let
*"intercept=FALSE".* Or, does intercept has to be included for SCAD
penalized regression?
Could someone give a clue? Thanks,
-Jiayi
--
Jiayi Hou
Ph.D Candidate
Department of Biostatistics
School of Medicine
Virginia Commonwealth University
Tel:(804)-828-2879(office)
(804)-274-8757(cell)
[[alternative HTML version deleted]]
