lsgl.algorithm.config: Create a new algorithm configuration

In nielsrhansen/lsgl: Multivariate Linear Sparse Group Lasso

Create a new algorithm configuration

Description

With the exception of verbose it is not recommended to change any of the default values.

Usage

lsgl.algorithm.config(tolerance_penalized_main_equation_loop = 1e-10,
  tolerance_penalized_inner_loop_alpha = 1e-04,
  tolerance_penalized_inner_loop_beta = 1,
  tolerance_penalized_middel_loop_alpha = 0.01,
  tolerance_penalized_outer_loop_alpha = 0.01,
  tolerance_penalized_outer_loop_beta = 0,
  tolerance_penalized_outer_loop_gamma = 1e-05,
  use_bound_optimization = FALSE,
  use_stepsize_optimization_in_penalizeed_loop = TRUE,
  stepsize_opt_penalized_initial_t = 1, stepsize_opt_penalized_a = 0.1,
  stepsize_opt_penalized_b = 0.1, max_iterations_outer = 10000,
  inner_loop_convergence_limit = 1e+05, verbose = TRUE)

Arguments

tolerance_penalized_main_equation_loop	tolerance threshold.
tolerance_penalized_inner_loop_alpha	tolerance threshold.
tolerance_penalized_inner_loop_beta	tolerance threshold.
tolerance_penalized_middel_loop_alpha	tolerance threshold.
tolerance_penalized_outer_loop_alpha	tolerance threshold.
tolerance_penalized_outer_loop_beta	tolerance threshold.
tolerance_penalized_outer_loop_gamma	tolerance threshold.
use_bound_optimization	if TRUE hessian bound check will be used.
use_stepsize_optimization_in_penalizeed_loop	if TRUE step-size optimization will be used.
stepsize_opt_penalized_initial_t	initial step-size.
stepsize_opt_penalized_a	step-size optimization parameter.
stepsize_opt_penalized_b	step-size optimization parameter.
max_iterations_outer	max iteration of outer loop
inner_loop_convergence_limit	inner loop convergence limit.
verbose	If TRUE some information, regarding the status of the algorithm, will be printed in the R terminal.

Value

A configuration.

Author(s)

Martin Vincent

Examples

set.seed(100) # This may be removed, it ensures consistency of the daily tests

## Simulate from Y=XB+E, the dimension of Y is N x K, X is N x p, B is p x K

N <- 50 #number of samples
p <- 50 #number of features
K <- 25  #number of groups

B<-matrix(sample(c(rep(1,p*K*0.1),rep(0, p*K-as.integer(p*K*0.1)))),nrow=p,ncol=K)

X<-matrix(rnorm(N*p,1,1),nrow=N,ncol=p)
Y<-X%*%B+matrix(rnorm(N*K,0,1),N,K)

# Create configuration
config <- lsgl.algorithm.config(verbose = FALSE)

lambda<-lsgl::lambda(X,Y, alpha=1, lambda.min=.5, intercept=FALSE, algorithm.config = config)

fit <-lsgl::fit(X,Y, alpha=1, lambda = lambda, intercept=FALSE, algorithm.config = config)

nielsrhansen/lsgl documentation built on Feb. 11, 2024, 8:07 a.m.