Readme file
SERIES
B
 Statistical
Methodology
Smoothing parameter selection for a class of semiparametric linear models, by P. T. Reiss and R. T. Ogden, pages 505–523
The R code provided can be used to implement the functional principal component regression method described in the paper. Five functions are included:
get.data Reads in data in a suitable format
fpcr Main function for functional principal component regression
make.basis Creates a cubic B-spline basis
make.halfpen Creates a matrix used to define the roughness penalty
penmod Internal function called by fpcr
Details and some examples are given in the comments in the code.
Related work and (possibly) future updates to the R code can be found at
http://works.bepress.com/phil_reiss/
Philip T. Reiss
Department of Child and Adolescent Psychiatry
New York University
16th Floor
215 Lexington Avenue
New York
NY 10016
USA
E-mail: phil.reiss@nyumc.org
|
