dglars: Differential Geometric Least Angle Regression

Differential geometric least angle regression method for fitting sparse generalized linear models. In this version of the package, the user can fit models specifying Gaussian, Poisson, Binomial, Gamma and Inverse Gaussian family. Furthermore, several link functions can be used to model the relationship between the conditional expected value of the response variable and the linear predictor. The solution curve can be computed using an efficient predictor-corrector or a cyclic coordinate descent algorithm, as described in the paper linked to via the URL below.

Package details

AuthorLuigi Augugliaro [aut, cre], Angelo Mineo [aut], Ernst Wit [aut], Hassan Pazira [aut], Michael Wichura [ctb, cph], John Burkardt [ctb, cph]
MaintainerLuigi Augugliaro <luigi.augugliaro@unipa.it>
LicenseGPL (>= 2)
URL http://www.jstatsoft.org/v59/i08/.
Package repositoryView on CRAN
Installation Install the latest version of this package by entering the following in R:

Try the dglars package in your browser

Any scripts or data that you put into this service are public.

dglars documentation built on Feb. 26, 2020, 5:14 p.m.