grouplasso2pop_logreg_cv_fixedgrid: Choose tuning parameters by crossvalidation for...

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

Choose tuning parameters by crossvalidation for grouplasso2pop logreg when given a fixed grid of lambda and eta values

Description

Choose tuning parameters by crossvalidation for grouplasso2pop logreg when given a fixed grid of lambda and eta values

Usage

grouplasso2pop_logreg_cv_fixedgrid(
  Y1,
  X1,
  groups1,
  Y2,
  X2,
  groups2,
  rho1,
  rho2,
  lambda.seq,
  eta.seq,
  n.folds,
  b1.init.arr,
  b2.init.arr,
  w1,
  w2,
  w,
  AA1,
  AA2,
  Com,
  tol = 0.001,
  maxiter = 500,
  report.prog = FALSE
)

Arguments

Y1	the binary response vector of data set 1
X1	matrix containing the design matrices for data set 1
groups1	a vector indicating to which group each covariate of data set 1 belongs
Y2	the binary response vector of data set 2
X2	matrix containing the design matrices for data set 2
groups2	a vector indicating to which group each covariate of data set 2 belongs
rho1	weight placed on the first data set
rho2	weight placed on the second data set
lambda.seq	sequence of lambda values
eta.seq	sequence of eta values
n.folds	the number of crossvalidation folds
b1.init.arr	array of initial values for beta1
b2.init.arr	array of initial values for beta2
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
AA1	a list of the matrices A2j
Com	the indices of the covariate groups which are common in the two datasets
tol	the convergence tolerance
maxiter	the maximum number of iterations allowed for each fit
report.prog	a logical indicating whether the progress of the algorithm should be printed to the console

Value

a list containing the fits over a grid of lambda and eta values as well as the vector of lambda values and the vector of eta values

Examples

grouplasso2pop_logreg_data <- get_grouplasso2pop_data(n1 = 1000, n2 = 800, response = "binary")

grouplasso2pop_logreg_grid.out <- grouplasso2pop_logreg_grid(Y1 = grouplasso2pop_logreg_data$Y1,
                                                             X1 = grouplasso2pop_logreg_data$X1,
                                                             groups1 = grouplasso2pop_logreg_data$groups1,
                                                             Y2 = grouplasso2pop_logreg_data$Y2,
                                                             X2 = grouplasso2pop_logreg_data$X2,
                                                             groups2 = grouplasso2pop_logreg_data$groups2,
                                                             rho1 = 1,
                                                             rho2 = 1,
                                                             n.lambda = 5,
                                                             n.eta = 3,
                                                             lambda.min.ratio = 0.001,
                                                             lambda.max.ratio = 0.50,
                                                             w1 = grouplasso2pop_logreg_data$w1,
                                                             w2 = grouplasso2pop_logreg_data$w2,
                                                             w = grouplasso2pop_logreg_data$w,
                                                             AA1 = grouplasso2pop_logreg_data$AA1,
                                                             AA2 = grouplasso2pop_logreg_data$AA2,
                                                             Com = grouplasso2pop_logreg_data$Com,
                                                             tol = 1e-3,
                                                             maxiter = 500,
                                                             report.prog = TRUE)

grouplasso2pop_logreg_cv_fixedgrid.out <- grouplasso2pop_logreg_cv_fixedgrid(Y1 = grouplasso2pop_logreg_data$Y1,
                                                                             X1 = grouplasso2pop_logreg_data$X1,
                                                                             groups1 = grouplasso2pop_logreg_data$groups1,
                                                                             Y2 = grouplasso2pop_logreg_data$Y2,
                                                                             X2 = grouplasso2pop_logreg_data$X2,
                                                                             groups2 = grouplasso2pop_logreg_data$groups2,
                                                                             rho1 = 1,
                                                                             rho2 = 1,
                                                                             lambda.seq = grouplasso2pop_logreg_grid.out$lambda.seq,
                                                                             eta.seq = grouplasso2pop_logreg_grid.out$eta.seq,
                                                                             n.folds = 5,
                                                                             b1.init.arr = grouplasso2pop_logreg_grid.out$b1.arr,
                                                                             b2.init.arr = grouplasso2pop_logreg_grid.out$b2.arr,
                                                                             w1 = grouplasso2pop_logreg_data$w1,
                                                                             w2 = grouplasso2pop_logreg_data$w2,
                                                                             w = grouplasso2pop_logreg_data$w,
                                                                             AA1 = grouplasso2pop_logreg_data$AA1,
                                                                             AA2 = grouplasso2pop_logreg_data$AA2,
                                                                             Com = grouplasso2pop_logreg_data$Com,
                                                                             tol = 1e-3,
                                                                             maxiter = 500,
                                                                             report.prog = TRUE)

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