Skip to contents

Bagging for KL-Integrated Cox Elastic-Net Models

coxkl_enet_bagging.Rd

Performs bootstrap aggregation (bagging) for the KL-integrated Cox elastic-net model by repeatedly applying cv.coxkl_enet on bootstrap resamples of the data. The procedure aggregates fitted coefficient vectors across replicates to produce a more stable estimate that is less sensitive to sampling variation or a single data split.

Usage

coxkl_enet_bagging(
  z,
  delta,
  time,
  stratum = NULL,
  RS = NULL,
  beta = NULL,
  etas,
  alpha = 1,
  B = 100,
  lambda = NULL,
  nlambda = 100,
  lambda.min.ratio = ifelse(nrow(z) < ncol(z), 0.05, 0.001),
  nfolds = 5,
  cv.criteria = c("V&VH", "LinPred", "CIndex_pooled", "CIndex_foldaverage"),
  c_index_stratum = NULL,
  message = FALSE,
  seed = NULL,
  ncores = 1,
  ...
)

Arguments

z

Matrix of predictors of dimension n x p.

delta

Event indicator vector.

time

Survival time vector.

stratum

Optional stratum indicator vector for stratified Cox models.

RS

Optional matrix or vector of external risk scores. If provided, it is resampled within each bootstrap replicate.

beta

Optional vector of external coefficients. If provided, it is treated as fixed and not resampled.

etas

Vector of eta values for transfer-learning shrinkage.

alpha

Elastic-net mixing parameter (between 0 and 1).

B

Number of bootstrap replicates. Default is 100.

lambda

Optional user-specified lambda sequence for the underlying elastic-net fit.

nlambda

Number of lambda values to generate if lambda is not supplied.

lambda.min.ratio

Ratio of smallest to largest lambda value when generating a lambda sequence.

nfolds

Number of folds for cross-validation in cv.coxkl_enet.

cv.criteria

Cross-validation criterion used for selecting eta–lambda pairs.

c_index_stratum

Optional stratum assignment for stratified C-index evaluation.

message

Logical indicating whether to print progress.

seed

Optional seed for reproducibility.

ncores

Integer. Number of parallel cores. Default 1 (sequential execution).

...

Additional arguments passed to cv.coxkl_enet.

Value

An object of class "bagging", which is a list containing:

  • best_beta — aggregated coefficient estimate obtained via averaging across valid replicates.

  • all_betas — matrix of dimension p x B_valid containing coefficient vectors from each successful bootstrap fit.

  • B — total number of bootstrap replicates.

  • seed — seed used (if any).

  • valid_replicates — number of successful (non-error) bootstrap fits used in aggregation.

Details

External information may be supplied either as a fixed coefficient vector (beta) or as pre-computed external risk scores (RS). When RS is provided, it is resampled along with the bootstrap replicates; when beta is provided, it is treated as fixed across replicates and not resampled.

Examples

if (FALSE) { # \dontrun{
data(ExampleData_highdim)
train_dat_highdim     <- ExampleData_highdim$train
beta_external_highdim <- ExampleData_highdim$beta_external
etas <- generate_eta(method = "exponential", n = 10, max_eta = 100)

bag.out <- coxkl_enet_bagging(
  z     = train_dat_highdim$z,
  delta = train_dat_highdim$status,
  time  = train_dat_highdim$time,
  stratum = train_dat_highdim$stratum,
  beta    = beta_external_highdim,
  etas    = etas,
  B       = 5,
  seed    = 1
)
} # }