PgaMsgl is an R package on Proximal Gradient Algorithm for Multi-variate Sparse Group Lasso. It is mainly implemented with Rcpp and RcppEigen.
Download the whole package and Build within R.
OR
Use install from the devtools package and run
install("directory of PgaMsgl", dependencies = TRUE)
OR
Use command install.packages if the dependicies have been installed on your devices.
install.packages("/path_to_PgaMsgl/PgaMsgl_version.tar.gz", repos=NULL, type="sourse", INSTALL_opts=c('--no-lock'))
Later on it will be released on CRAN.
The package includes two demo datasets for testing usage. One is with relative low dimension, and another is with relative high dimension.
For example, to run the low dimension dataset:
data(lowD)
system.time(lowD_result <- PgaMsgl(lowD$X, lowD$Y, lowD$B0, model="L121", lowD$Gm, lowD$mi, lowD$mg, lowD$mc))
Yiming Qin - Initial work - TriangularCell
This project is licensed under the GPL-3.0 License - see the LICENSE.md file for details
- Yaohua Hu
- Jing Qin
- Xinlin Hu
- Sifan Liu
Hu, Yaohua, et al. "Group sparse optimization via lp, q regularization." J Mach Learn Res 18 (2017): 1-52.
Beck, Amir, and Marc Teboulle. "A fast iterative shrinkage-thresholding algorithm for linear inverse problems." SIAM journal on imaging sciences 2.1 (2009): 183-202.