SAS Procedures for Variable Selection

Our software resources are permanently archived at https://scholarsphere.psu.edu/collections/v41687m23q. Our software is no longer supported by the Help Desk. For the latest software and research updates, visit Runze Li’ s website on variable screening and variable selection.

 

Select a Version

1. Are you are using a 32-bit or 64-bit machine?

2. Are you using 32-bit or 64-bit SAS?

  • In SAS, go to HELP > ABOUT SAS 9.
  • Under Software Information “64” = 64-bit SAS.
  • “64” not displayed = 32-bit SAS.

Overview

Variable selection is a classic and very important problem in applied statistics. Methodology Center researcher Dr. Runze Li and his collaborators have been studying penalized likelihood approaches to automatic variable selection for predictive modeling when the number of potential predictors is very large. This involves adjusting the likelihood function using a penalty that rewards parsimony. One such penalty function, which has some favorable theoretical properties, is the smoothly clipped absolute deviation (SCAD) penalty (see Fan & Li, 2001).

The Methodology Center released two SAS procedures, PROC SCADLS, which does linear regression using penalized least squares with the SCAD penalty and SCADGLIM which implements the SCAD variable selection method in the context of penalized maximum-likelihood generalized linear regression models. This approach automatically selects and estimates a model at the same time.

Reference

Fan, J., & Li, R. (2001). Variable selection via nonconcave penalized likelihood and its oracle properties. Journal of the American Statistical Association, 96(456), 1348-1360.


Recommended Citations

PROC SCADLS & PROC SCADGLIM (Version 1.2.1) [Software]. (2014). University Park: The Methodology Center, Penn State. Retrieved from methodology.psu.edu

Dziak, J. J., Lemmon, D. R., Li, R., & Huang, L. (2014). PROC SCADLS users’ guide (Version 1.2). University Park: The Methodology Center, Penn State. Retrieved from methodology.psu.edu

Dziak, J. J., Lemmon, D. R., Li, R., & Huang, L. (2014). PROC SCADGLIM users’ guide (Version 1.2). University Park: The Methodology Center, Penn State. Retrieved from methodology.psu.edu