nll: Compute the negative log-likelihood

In dkaschek/dMod: Dynamic Modeling and Parameter Estimation in ODE Models

Compute the negative log-likelihood

Description

Gaussian Log-likelihood. Supports NONMEM-like BLOQ handling methods M1, M3 and M4 and estimation of error models

Usage

nll(
  nout,
  pars,
  deriv,
  opt.BLOQ = "M3",
  opt.hessian = c(ALOQ_part1 = TRUE, ALOQ_part2 = TRUE, ALOQ_part3 = TRUE, BLOQ_part1 =
    TRUE, BLOQ_part2 = TRUE, BLOQ_part3 = TRUE, PD = TRUE)
)

Arguments

nout	data.frame (result of res) or object of class res.
pars	Example named vector of outer parameters to construct the objlist
deriv	TRUE or FALSE
opt.BLOQ	see normIndiv
opt.hessian	see normIndiv

Value

list with entries value (numeric, the weighted residual sum of squares), gradient (numeric, gradient) and hessian (matrix of type numeric).

dkaschek/dMod documentation built on July 27, 2023, 11:45 p.m.