Compute Score Components for Interval-Censored Data

Computes partial derivatives or score components for interval-censored data, given current estimates of baseline functions and a regression parameter vector beta. This function is typically part of an iterative algorithm for semiparametric estimation.

Usage

CSD.f2(Hf, beta, y, index, delta, gamma, Z)

Arguments

Hf

A list containing baseline functions or related objects (e.g., H, H1, H2).

beta

A numeric vector of regression coefficients for the covariates in Z.

y

A numeric vector (often the current estimate of the cumulative baseline hazard or a similar function) of length p.

index

An n x 2 matrix of indices referencing positions in y for each subject.

delta

Numeric (0/1) event indicator for the left endpoint U.

gamma

Numeric (0/1) event indicator for the right endpoint V.

Z

A matrix or data frame of covariates, with n rows (one per subject).

Value

A list with the following components:

G

A numeric vector (length p) of cumulative increments related to gamma.

Q

A numeric vector (length p) of combined increments (e.g., dW + y[i] * dG).

Details

This function uses auxiliary routines (e.g., FdG.f(), eta1G.f(), eta2G.f(), zeta1.f(), etc.) to compute partial derivatives of the log-likelihood or estimating equations with respect to y.

Examples

# (Not run) Example with placeholder data:
# Hf <- list(H = NULL, H1 = NULL, H2 = NULL)
# beta <- c(0.1, -0.2)
# y <- rep(0.05, 5)
# index <- matrix(c(1, 2, 2, 3), nrow = 2, byrow = TRUE)
# delta <- c(1, 0)
# gamma <- c(0, 1)
# Z <- matrix(c(1, 0, 0, 1), nrow = 2)
# CSD.f2(Hf, beta, y, index, delta, gamma, Z)