Sicotte, Hugues, Ph.D.
2014-May-01 12:58 UTC
[R] Tobit model with L1 and L2 penalties (elastic net)
Does anyone know of package that implements a tobit model with L1 and L2 (lasso and ridge) penalties aka elastic net. I am interested in an elastic net implementation of tobit because my variables are SNPs genotypes/dosage in Linkage disequilibrium and elastic net type models to well for that. The closest I found was The R Brq package implement the Bayesian Tobit quantile regression model with adaptive lasso penalty and will limit itself to an L1 penalty. p.s. I cannot do this by combining the penalized package and survreg as the penalized package does not yet allow for survreg models (as per Terry Therneau). Hugues Sicotte, Ph.D. Assistant Professor in Medical Informatics Bioinformatics Core Division of Biomedical Statistics and Informatics Phone: 507-538-3866 Fax: 507-284-0360 Email: sicotte.hugues@mayo.edu<mailto:sicotte.hugues@mayo.edu> ____________________________________________ Mayo Clinic 200 First Street SW Rochester, MN 55905 www.mayoclinic.org [[alternative HTML version deleted]]