| Type: | Package |
| Title: | Robust Selection Algorithm |
| Version: | 0.1.0 |
| Date: | 2021-05-23 |
| Description: | An implementation of algorithms for estimation of the graphical lasso regularization parameter described in Pedro Cisneros-Velarde, Alexander Petersen and Sang-Yun Oh (2020) http://proceedings.mlr.press/v108/cisneros20a.html. |
| BugReports: | https://github.com/dddlab/robust-selection/issues |
| License: | GPL-2 |
| Encoding: | UTF-8 |
| Imports: | glasso, Rcpp |
| Suggests: | knitr, rmarkdown |
| VignetteBuilder: | knitr |
| RoxygenNote: | 7.1.1 |
| LinkingTo: | Rcpp, RcppEigen |
| NeedsCompilation: | yes |
| Packaged: | 2021-05-24 16:11:48 UTC; chautran |
| Author: | Chau Tran [aut, cre], Sang-Yun Oh [aut], Pedro Cisneros-Velarde [aut], Alexander Petersen [aut] |
| Maintainer: | Chau Tran <chautran@ucsb.edu> |
| Repository: | CRAN |
| Date/Publication: | 2021-05-25 12:00:02 UTC |
Robust Selection
Description
Robust Selection algorithm for estimation of the regularization parameter for Graphical Lasso
Usage
robsel(x, alpha = 0.9, B = 200)
Arguments
x |
A |
alpha |
Prespecified confidence level. Default 0.9 |
B |
Number of bootstrap sample. Default 200 |
Value
lambda Estimation of the regularization parameter for Graphical Lasso. A vector of lambda will be return if more than 1 value of alpha is provided.
References
P Cisneros-Velarde, A Petersen and S-Y Oh (2020). Distributionally Robust Formulation and Model Selection for the Graphical Lasso. Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics.
See Also
robsel.glasso for using Graphical Lasso algorithm with estimate regularization parameter lambda from Robust Selection.
Examples
set.seed(17)
library(robsel)
x <-matrix(rnorm(50*20),ncol=20)
#Compute estimation of lambda at confidence level alpha
lambda <- robsel(x = x, alpha = 0.9, B = 200)
Fit Graphical Lasso with RobSel
Description
Fit Graphical Lasso with estimate regularization parameter from Robust Selection
Usage
robsel.glasso(x, alpha = 0.9, B = 200, ...)
Arguments
x |
A |
alpha |
Prespecified confidence level. Default 0.9 |
B |
Number of bootstrap sample. Default 200 |
... |
Optional arguments passed on to glasso. |
Value
A list with components:
alpha |
A list of prespecified confidence level |
lambda |
A list of estimate regularization parameter for Graphical Lasso |
Omega |
A list of estimated inverse covariance matrix |
Sigma |
A list of estimated covariance matrix |
Note
Each item in each component corresponds to a prespecified level alpha.
References
P Cisneros-Velarde, A Petersen and S-Y Oh (2020). Distributionally Robust Formulation and Model Selection for the Graphical Lasso. Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics.
Friedman, Jerome, Trevor Hastie, and Robert Tibshirani. 'Sparse inverse covariance estimation with the graphical lasso.' Biostatistics 9.3 (2008): 432-441.
Meinshausen, Nicolai and Buhlmann, Peter. 2006. 'High-Dimensional Graphs and Variable Selection with the Lasso.' The Annals of Statistics. JSTOR: 1436-1462.
Witten, Daniela M, Friedman, Jerome H, and Simon, Noah. 2011. 'New Insights and Faster computations for the Graphical Lasso.' Journal of Computation and Graphical Statistics. Taylor and Francis: 892-900.
See Also
robsel for Robust Selection algorithm, glasso for Graphical Lasso algorithm.
Examples
set.seed(17)
library(robsel)
x <-matrix(rnorm(50*20),ncol=20)
#Use Graphical Lasso with estimate regularization parameter lambda from RobSel
fit <- robsel.glasso(x = x, alpha = 0.9, B = 200)