Forecasting out-of-sample discrete_markov_chain.ipynb

Thank you very much, you have been really kind. Thanks for the work you programmers do, the pymc library is something great. :heart::tada:
One last question, are there any examples of ARX/VARX models in pymc by any chance?
Speaking of statespace I am not very clear about the difference with in Markov switch type models. That is, instead of having two states the models with statespace have the parameters moving every t. Have I said something vaguely correct?