04-04-Forecast-VAR.Rmd

## VAR Modeling {-}

```{r a1, comment=NA, warning=FALSE, message=FALSE, fig.width=9}

library(vars)

# Generated Time Series
x1  = 10 + arima.sim(n = 100,model = list(order = c(1, 1, 1), ar = .1, ma = .2))
x2  = 10 + arima.sim(n = 100,model = list(order = c(1, 1, 1), ar = .1, ma = .2))
x3  = 10 + arima.sim(n = 100,model = list(order = c(1, 1, 1), ar = .1, ma = .2))
x4  = 10 + arima.sim(n = 100,model = list(order = c(1, 1, 1), ar = .1, ma = .2))

# Combine into a multivariate time series
x = data.frame(x1, x2, x3, x4)

# Create a variable auto-regression modle
mdl = VAR(x)

# Model summary
summary(mdl)

# Plot model
plot(mdl)

# Plot forecast
plot(predict(mdl, h = 12))

```