make_ard: Generate simulated ARD

In networkscaleup: Network Scale-Up Models for Aggregated Relational Data

Generate simulated ARD

Description

Generate simulated ARD

Usage

make_ard(
  n_i = 500,
  n_k = 20,
  N = 1e+06,
  p = 0,
  p_global_nonzero = 0,
  p_local_nonzero = 0,
  group_corr = FALSE,
  degree_corr = FALSE,
  family = c("poisson", "nbinomial"),
  omega_range = c(1, 5),
  alpha_mean = 5,
  alpha_sd = 0.15,
  eta = 3,
  seed = NULL
)

Arguments

n_i	number of respondents (rows)
n_k	number of groups (columns)
N	total population size
p	number of collected covariates
p_global_nonzero	number of non-zero global covariates
p_local_nonzero	number of non-zero local covariates
group_corr	group correlation
degree_corr	degree correlation
family	sampling distribution
omega_range	minimum and maximum omega for negative binomial overdispersion
alpha_mean	mean of alphas
alpha_sd	variance of alphas
eta	correlation hyperparameter for LKJ prior
seed	random seed

Value

simulated ARD along with all true parameters

Examples

make_ard(N = 10000, family = "poisson")

networkscaleup documentation built on Dec. 11, 2025, 1:07 a.m.