Newton-DR Estimation for Survival Data

The `Newton_DR` function is an iterative algorithm to estimate the parameters in a transformation model for survival data.

Newton_DR(
  phi.t,
  phi.v,
  weights = 1,
  max.iter = 100,
  tol = 1e-05,
  initial = rep(0, ncol(phi.t)),
  lambda0 = NULL
)

Arguments

phi.t

A matrix with labeled data, where rows are observations and columns are covariates.

phi.v

A matrix with unlabeled data, where rows are observations and columns are covariates.

weights

An optional vector of weights for each observation. Default is 1 for equal weights.

max.iter

Maximum number of iterations for the Newton-Raphson algorithm. Default is 100.

tol

Tolerance for convergence of the Newton-Raphson algorithm. Default is 1e-5.

initial

An optional vector of initial values for the algorithm. Default is a vector of zeros with length equal to the number of columns in phi.t.

lambda0

An optional initial value for the regularization parameter. If NULL, lambda0 is set to log(ncol(phi.t))/n.t^1.5. Default is NULL.

Value

A vector of parameter estimates for the transformation model.