golden_section: Golden section grid search on a lmmen penalty

Description Usage Arguments Value Examples

Description

Solve for local minimum with one dimensional golden section one of the regularization parameters of the lmmen penalty.

Usage

1
2
3
golden_section(dat, init.beta, pen.effect = "FE.L1", opt.lb = 0,
  opt.ub = 1, opt.maxiter = 100, opt.tol = 0.1, opt.tau = (sqrt(5) -
  1)/2)

Arguments

dat

matrix, matrix that includes y (response),X (population covariates), Z (random effects covariates (not incl random intercept))

init.beta

numeric, initial fixed effects estimates

pen.effect

character,which penalty to search on c('FE.L1','RE.L1',FE.L2”,'RE.L2'), Default: 'FE.L1'

opt.lb

numeric, start of search interval, Default: 0

opt.ub

numeric, end of search interval, Default: 1

opt.maxiter

numeric, maximum iterations to search, Default: 100

opt.tol

numeric, accuracy value, Default: 0.1

opt.tau

numeric, golden proportion coefficient (~0.618) Default: (sqrt(5) - 1)/2

Value

lmmen list object inluding lmmen fit object of min BIC solution and summary statistics from the grid searc

Examples

1
2
3
4
5
6
## Not run:  
 dat <- initialize_example(n.i = 5,n = 30,q=4,seed=1)
 init <- init.beta(dat,method='glmnet')
 golden_section(dat,init,pen.effect = 'FE.L1')
 
## End(Not run)

lmmen documentation built on May 1, 2019, 8:53 p.m.