VectorAutoregressions.jl
Estimating Vector Autoregressions (VARs) using Julia.
VectorAutoregressions.jl is an in-development package for estimating VARs using Julia. The initial
goal is to provide functionality comparable to the vars
package in R. Longer-term goals include adding Bayesian estimation.
Currently implemented:
- Calculate lag selection criteria.
- Estimate an unrestricted VAR using OLS.
- Check the stability of an estimated VAR.
- Portmanteau test for autocorrelation in the residuals.
- LM (Breusch-Godfrey) test for autocorrelation in the residuals.
To do:
- Test for normally-distributed residuals.
- Calculate simple and orthogonalized impulse responses.
- Add support for linear restrictions.
- Add support for structural VARs.