Skip to contents

Simulate subject-level parameters with a known Model 2a covariance

Source: R/sim_params_2a.R
sim_params_2a.Rd

Draws n subjects' log-scale kinetic parameters for two biomarkers from a Model 2a multivariate normal: full within-biomarker blocks Sigma_G, Sigma_A plus a diagonal cross-biomarker block C = diag(c_vec) (assembled by build_sigma_2a()). This gives ground-truth correlated parameters for validating run_mod_2a().

The parameter order matches model.jags: 1 = log(y0), 2 = log(y1 - y0), 3 = log(t1), 4 = log(alpha), 5 = log(shape - 1). Uses a base-R Cholesky draw (no extra dependency).

Usage

sim_params_2a(n, mu_g, mu_a, sigma_g, sigma_a, c_vec, seed = NULL)

Arguments

n

integer number of subjects.

mu_g, mu_a

Length-P log-scale mean vectors for biomarker 1 / 2.

sigma_g, sigma_a

P x P within-biomarker covariances.

c_vec

Length-P same-parameter cross-biomarker covariances.

seed

Optional RNG seed.

Value

A list with:

  • log_par: an n x (2P) matrix of draws, columns ordered G1..GP, A1..AP;

  • sigma: the 2P x 2P true covariance;

  • rho: the length-P true cross-biomarker correlations.