ECLasso-package: ECLasso: Solving Equality-Constrained Generalized Linear...

In umich-biostatistics/genepADMM: Solving Equality-Constrained Generalized Linear Models or Cox Regression with Lasso Penalty via Alternating Direction Method of Multipliers (ADMM)

Description

An inplementation of the Alternating Direction Method of Multipliers in R. Given data from survival (Cox) certain GLMs (Gaussian, poisson, binomial), it can estimate parameters or parameter pairs for each covariate. It includes options on whether to include an intercept effect, to solve exactly or use an inexact update in each iteration (which greatly speeds up computation time), or solve over a user-defined lambda sequence. At each iteration, it outputs the parameter estimates and the lambda at that iteration, as well as the solution path upon completion.

Details

The only function you will actually run from ECLasso is ECLasso.fit. Please refer to the ECLasso.fit documentation for more details about function options and a short example.

Author(s)

• Lam Tran <[e-mail]>

• Hui Jiang <[e-mail]>

References

• ECLasso: Fitting equality-constrained,L1-penalized models with ADMM to find gene pairs

• https://github.com/umich-biostatistics/genepADMM

umich-biostatistics/genepADMM documentation built on Jan. 29, 2020, 9:12 p.m.