使用 r 模拟多变量时间序列数据

Question

I want to simulate a multivariate time series data which fits to the year given in the table. How can I do that?

I use the following code

Z <- rnorm(100, mean = 0, sd = 1.5)

# process simulation
X <- c()
for (i in 1:length(Z)) {

# my table
my_table <- 1982:2008 

Answer 1

You will need the covariance or correlation of series. If there is not correlation it would be easy.

# No correlation: 
my_table <- 1982:2008
num.series <- 4

sdata=matrix(rnorm(length(my_table)*num.series,0,sd=1.5),ncol=num.series)
sdf=data.frame(my_table,sdata)
summary(sdf)

# with covariance
library(MASS)
Sigma <- matrix(c(10,3,3,3,3,5,3,3,3,3,8,3,3,3,3,15)/10,ncol=4)
Sigma
smdata=mvrnorm(n = length(my_table), rep(0, num.series), Sigma)
scdf=data.frame(my_table,smdata)
summary(scdf)