Evaluating vector autoregressive (VAR) models
An important step when building a VAR model is to understand the model in terms of the interactions between the different endogenous variables. The statsmodels VAR implementation provides key plots to help you analyze the complex dynamic relationship between these endogenous variables (multiple time series).
In this recipe, you will continue where you left off from the previous recipe, Forecasting multivariate time series data using VAR, and explore different diagnostic plots, such as the Residual Autocorrelation Function (ACF), Impulse Response Function (IRF), and Forecast Error Variance Decomposition (FEVD).
How to do it...
The following steps continue from the previous recipe. If you have not performed these steps, you can run the code from the accompanied Jupyter Notebook to follow along.
You will focus on diagnosing the VAR model that we created using the available methods:
- The
resultsobject is of theVARResultsWrapper...
