Penalized Generalized Estimating Equations for Bivariate Mixed Outcomes

Perform simultaneous estimation and variable selection for correlated bivariate mixed outcomes (one continuous outcome and one binary outcome per cluster) using penalized generalized estimating equations. In addition, clustered Gaussian and binary outcomes can also be modeled. The SCAD, MCP, and LASSO penalties are supported. Cross-validation can be performed to find the optimal regularization parameter(s).

Installation from GitHub:

if (!require("devtools"))

Installation note for OS X users:

This package uses Rcpp and RcppArmadillo. Mac OS X users may see the following error in their console while trying to install the package:

ld: warning: directory not found for option '-L/usr/local/lib/gcc/x86_64-apple-darwin13.0.0/4.8.2'  
ld: library not found for -lgfortran

The solution is documented here. In a nutshell, type the following in a terminal:

curl -O  
sudo tar fvxz gfortran-4.8.2-darwin13.tar.bz2 -C /

To do list

If there is sufficient interest in this package, the following features could be added:

Try the pgee.mixed package in your browser

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

pgee.mixed documentation built on May 2, 2019, 8:35 a.m.