timixCovRho: Mixed model for the synchronized time sereis by optimization,...

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

Description

timixCovRho function is a mixed model function to solve synchronized time sereis when omega = 0. The effects of serial correlation is treated as the random effects. The variance-covariance matrix of random effects are prescribed by the parameter rho. timixCovRho function estimate the parameter through oprimization.

Usage

timixCovRho(dat, y, f, r, grp, grid = T, trace = F)

Arguments

dat	data for the analysis
y	column number of the dependent variable
f	column number of the fixed effects
r	column number of the random effects
grp	column number of the group

Examples

n_time <- 50
n_grp <- 5
y <- simuOneCol(n_time, n_grp, rho = 0.5, omega = 0.5, mean = 0)
x0 <- rep(1, n_time * n_grp) # Intercept
x1 <- simuOneCol(n_time, n_grp, rho = 0.5, omega = 1, mean = 0)
x2 <- simuOneCol(n_time, n_grp, rho = 0.5, omega = 0, mean = 0)
area <- makeGrp(n_time, n_grp)
dat <- data.frame(y, x0, x1, x2, area)
res <- timixCovRho(dat, y = 1, f = 2:4, r = 2:4, grp = 5)

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