Skip to contents

Implements a Newton-Raphson step to update the regression parameter beta using the partial derivatives of the log-likelihood. Internally calls helper functions for computing the score and information matrix components (e.g. xi1G.f).

Usage

NR2(Hf, beta, y, delta, gamma, index, Z, eps = 1e-04, maxiter = 20)

Arguments

Hf

A list of functions (H, H1, H2) defining the baseline hazard (or similar).

beta

Numeric vector of regression coefficients (initial/current guess).

y

Numeric vector of baseline-related parameters (often updated by profile1).

delta, gamma

Numeric 0/1 vectors indicating left and interval censoring status.

index

A 2-column integer matrix mapping each subject's interval to indices in y.

Z

A matrix or data frame of covariates.

eps

Convergence tolerance for updating beta.

maxiter

Maximum number of Newton-Raphson iterations.

Value

Updated beta as a numeric vector.