search for: undifferenc

...e bottom of this note. The steps include (1) an initial linear regression model, (2) building an ARIMA model, (3) creating 12 months of simulated 'real' data - for various values of projected changes in slope from the linear regression - from the ARIMA model using arima.sim with subsequent undifferencing and appending to historical data, (3) one-step-ahead forecasting for 12 months using the 'fixed' arima model and simulated data, (4) taking the residuals from the forecasting (simulated minus forecast for each of the 12 months) and applying QC charting, and (5) counting the number (if an...

Hi, All, This is my program ts1.sim <- arima.sim(list(order = c(1,1,0), ar = c(0.7)), n = 200) ts2.sim <- arima.sim(list(order = c(1,1,0), ar = c(0.5)), n = 200) tdata<-ts(c(ts1.sim[-1],ts2.sim[-1])) tre<-c(rep(0,200),rep(1,200)) gender<-rbinom(400,1,.5) x<-matrix(0,2,400) x[1,]<-tre x[2,]<-gender fit <- arima(tdata, c(1, 1, 0), method = "CSS",xreg=t(x))

...es the standard recipe (as far as I know) is to take first differences of the individual series and apply the VAR program to the new bivariate difference series. The VAR program seems to derive a reasonable set of coefficients. My Question: How do I simulate future paths for the original pair of (undifferenced) series? Thanks for your time and help, KW ________________________________ The rest of this note is just how I have done this up until now... The method I have been using is to take a simulated path of the differences (note that in my case there are 2 simulated paths) say > x.d [1] 9 9...