I run a Markov switching model and now I want to run Log-Likelihood ratio test to test the non-switching hypothesis (one state model) against a two-state model. How could I run this test in eviews 8. Markov chain is stationary with constant transition probabilities.