genDat: Function to generate data

In ShojiTaniguchi/timsync: Mixed model for time series with synchronization

Description

genDat function is the simulation function of the package timsync. genDat generates any variable for the time series data with synchrony.

Usage

genDat(x, beta, n_time, n_grp, z = NULL, OmegaMat = NULL, sig = 1, rho, omega)

Arguments

x	Independent variable. Vector or matrix.
beta	Coefficients of x.
n_time	Length of the time series
n_grp	Number of groups.
z	Random variables. Vector or matrix.
OmegaMat	List of Variance-Covariance Matrix (Omega matrix) for the random effects.
sig	Vector of the standard deviation of the random variables or error term. The last element should be that of the error term.
rho	Parameter of serial correlation of the error term.
omega	Parameter of synchronization of the error term.

Examples

## Generate independent variable.
y0 <- genDat(x = 0, beta = 0, n_time = 20, n_grp = 3, rho = 0.3, omega = 0.8)
grp <- makeGrp(n_time = 20, n_grp = 3)
dat <- data.frame(y0 = y0$objVar, time = rep(1:20, 3), grp = grp)
ggplot(dat, aes(x = time, y = y0, color = grp)) + geom_line()

## Generate independent variable with random group means.
Omega <- makeOmega(n_time = 20, n_grp = 3, rho = 1, omega = 0)
y0 <- genDat(x = 0, beta = 0, n_time = 20, n_grp = 3,
 z = 1, OmegaMat = list(Omega), sig = c(1, 1),
 rho = 0.3, omega = 0.8)
grp <- makeGrp(n_time = 20, n_grp = 3)
dat <- data.frame(y0 = y0$objVar, time = rep(1:20, 3), grp = grp)
ggplot(dat, aes(x = time, y = y0, color = grp)) + geom_line()

## Generate dependent variable with random effects.
x1 <- rnorm(20 * 3, 0, 1)
Omega_0 <- makeOmega(n_time = 20, n_grp = 3, rho = 0.5, omega = 0.5)
Omega_1 <- makeOmega(n_time = 20, n_grp = 3, rho = 0.5, omega = 0.5)
Z <- cbind(rep(1, 20 * 3), x1)
y0 <- genDat(x = x1, beta = 5, n_time = 20, n_grp = 3,
 z = Z, OmegaMat = list(Omega_0, Omega_1), sig = c(1, 1, 1),
 rho = 0.3, omega = 0.8)

ShojiTaniguchi/timsync documentation built on Oct. 10, 2020, 3:33 p.m.