simulate_data: Simulate social interaction datasets

Description

This function generates datasets where individual phenotypes are influenced by both direct and indirect (social) effects, under a specified sampling design.

Usage

simulate_data(
  ind = 200,
  partners = 4,
  repeats = 1,
  iterations = 100,
  B_0 = 0,
  psi = NULL,
  Valpha,
  Vepsilon = NULL,
  Vpsi = 0,
  Vx = 1,
  Ve = 0.6,
  Vxe = 0,
  r_alpha_epsilon = 0,
  r_alpha_psi = 0,
  r_epsilon_psi = 0,
  r_alpha_x = 0,
  r_psi_x = 0,
  r_epsilon_x = 0,
  fix_total_var = TRUE
)

Value

A list with:

data: list of datasets
params: named list of effect sizes
design: sample design (n_ind, partners, repeats, iterations)

Arguments

ind: Number of individuals.
partners: Partners per individual.
repeats: Repeats per unique dyad.
iterations: Number of datasets to simulate.
B_0: Population intercept.
psi: Population-level responsiveness (social slope).
Valpha: Direct effect (focal variance).
Vepsilon: Indirect effect (partner variance).
Vpsi: Social responsiveness (among individual variance in slopes).
Vx: Partner trait variance.
Ve: Residual variance.
Vxe: Measurement error/within-individual variation in partner trait.
r_alpha_epsilon: Corr(alpha, epsilon).
r_alpha_psi: Corr(alpha, psi).
r_epsilon_psi: Corr(epsilon, psi).
r_alpha_x: Corr(alpha, x).
r_psi_x: Corr(psi, x).
r_epsilon_x: Corr(epsilon, x).
fix_total_var: Logical; if TRUE (default), residual variance is adjusted so total phenotypic variance is approx. 1.

Examples

Run this code

sim <- simulate_data(ind =50, partners = 4, iterations = 5,
                     B_0 = 1, Valpha=0.2, Vepsilon = 0.1)

Run the code above in your browser using DataLab