geeVerse (version 0.3.1)

generate_data: Generate Data for Simulation

Description

This function generates simulated data including the predictor matrix `X` and the response vector `y`, based on the specified parameters. The function allows for the simulation of data under different settings of correlation, distribution, and the number of observations and subjects.

Usage

generate_data(
  nsub,
  nobs,
  p,
  beta0,
  rho,
  corstr = "AR1",
  dis = "normal",
  ka = 0,
  SNPs = NULL
)

Value

A list containing two elements: `X`, the matrix of predictors, and `y`, the response vector.

Arguments

nsub

Integer, the number of subjects.

nobs

Integer or numeric vector, the number of observations per subject.

p

Integer, the number of predictors.

beta0

Numeric vector, initial coefficients for the first few predictors.

rho

Numeric, the correlation coefficient used in generating correlated errors.

corstr

Character, specifies the correlation of correlation structure for the covariance matrix. Options are "cs" or "exchangeable" for compound symmetry, "AR1" for autoregressive, and any other input will result in an identity matrix.

dis

Character, the distribution of errors ("normal" or "t").

ka

1 for heterogeneous errors and 0 for homogeneous errors.

SNPs

User can provide simulated or real SNPs for genetic data simulation.

Examples

set.seed(123)
sim_data <- generate_data(
  nsub = 50, nobs = rep(5, 50), p = 10,
  beta0 = c(rep(1, 5), rep(0, 5)), rho = 0.3
)