grouplasso2pop_logreg_slower: Minimize the objective function of the 2-population group...

In gregorkb/semipadd2pop: Combine two data sets to fit semiparametric additive models

Minimize the objective function of the 2-population group lasso problem with a binary response

Description

Minimize the objective function of the 2-population group lasso problem with a binary response

Usage

grouplasso2pop_logreg_slower(
  rY1,
  rX1,
  groups1,
  rY2,
  rX2,
  groups2,
  rho1,
  rho2,
  lambda,
  eta,
  w1,
  w2,
  w,
  rAA1,
  rAA2,
  rCom,
  tol,
  maxiter,
  beta1_init = as.numeric(c()),
  beta2_init = as.numeric(c())
)

Arguments

groups1	a vector of integers indicating to which group each covariate in data set 1 belongs
groups2	a vector of integers indicating to which group each covariate in data set 1 belongs
rho1	weight placed on the first data set
rho2	weight placed on the second data set
lambda	the level of sparsity penalization
eta	the level of penalization towards model similarity
w1	group-specific weights for different penalization across groups in data set 1
w2	group-specific weights for different penalization across groups in data set 2
w	group-specific weights for different penalization toward similarity for different groups
tol	a convergence criterion
beta1_init	optional starting value for beta1
beta2_init	optional starting value for beta2
Y1	the binary response vector of data set 1
X1	matrix containing the design matrices for data set 1
Y2	the binary response vector of data set 2
X2	matrix containing the design matrices for data set 2
AA1	a list of the matrices A2j
eigen1	a list of eigen info on groups from data set 1
eigen2	a list of eigen info on groups from data set 2
Com	the indices of the covariate groups which are common in the two datasets
max.iter	the maximum allowed number of iterations
return_obj	a logical indicating whether the value of the objection function should be recorded after every step of the algorithm

Value

Returns the minimizers of the 2-population group lasso objective function for the two data sets.

Examples

data <- get_grouplasso2pop_data(n1 = 400,n2 = 600, response = "binary")

grouplasso2pop_logreg.out <- grouplasso2pop_logreg(rY1 = data$Y1,
                                                   rX1 = data$X1,
                                                   groups1 = data$groups1,
                                                   rY2 = data$Y2,
                                                   rX2 = data$X2,
                                                   groups2 = data$groups2,
                                                   rho1 = 2,
                                                   rho2 = 1,
                                                   lambda = 1,
                                                   eta = 1,
                                                   w1 = data$w1,
                                                   w2 = data$w2,
                                                   w = data$w,
                                                   rAA1 = data$AA1,
                                                   rAA2 = data$AA2,
                                                   rCom = data$Com,
                                                   tol = 1e-4,
                                                   maxiter = 500)

gregorkb/semipadd2pop documentation built on Oct. 2, 2022, 1:37 p.m.