simu.change.regression {changepoints} R Documentation

## Simulate a sparse regression model with change points in coefficients.

### Description

Simulate a sparse regression model with change points in coefficients under temporal dependence.

### Usage

simu.change.regression(
d0,
cpt_true,
p,
n,
sigma,
kappa,
cov_type = "I",
mod_X = "IID",
mod_e = "IID"
)


### Arguments

 d0 A numeric scalar stands for the number of nonzero coefficients. cpt_true An integer vector contains true change points (sorted in strictly increasing order). p An integer scalar stands for the dimensionality. n An integer scalar stands for the sample size. sigma A numeric scalar stands for error standard deviation. kappa A numeric scalar stands for the minimum jump size of coefficient vector in l_2 norm. cov_type A character string stands for the type of covariance matrix of covariates. 'I': Identity; 'T': Toeplitz; 'E': Equal-correlation. mod_X A character string stands for the time series model followed by the covariates. 'IID': IID multivariate Gaussian; 'AR': Multivariate AR1 with rho = 0.3; Multivariate MA1 theta = 0.3. mod_e A character string stands for the time series model followed by the errors 'IID': IID univariate Gaussian; 'AR': Univariate AR1 with rho = 0.3; Univariate MA1 theta = 0.3.

### Value

A list with the following structure:

 cpt_true A vector of true changepoints (sorted in strictly increasing order). X An n-by-p design matrix. y An n-dim vector of response variable. betafullmat A p-by-n matrix of coefficients.

### Author(s)

Daren Wang, Zifeng Zhao & Haotian Xu

### References

Rinaldo, Wang, Wen, Willett and Yu (2020) <arxiv:2010.10410>; Xu, Wang, Zhao and Yu (2022) <arXiv:2207.12453>.

### Examples

d0 = 10
p = 30
n = 100
cpt_true = c(10, 30, 40, 70, 90)
data = simu.change.regression(d0, cpt_true, p, n, sigma = 1, kappa = 9)


[Package changepoints version 1.1.0 Index]