Autoregressive Distributive Lag (ADL) model

I want to implement an autoregressive distributive lag model and from what I can tell at the moment, an ADL cannot be implemented with brms at the moment. If this is the case, would the best way to achieve one be to use the cor_ar function and modify the code from the make_stancode function in the same way for the AR(p) model?

Is there currently any prospects that @paul.buerkner might be so gracious to add the ADL model to the package? Perhaps including it in the cor_ar function? If it already is my apologies, but a note on the cor_ar function would be brilliant to direct people.

… and by ADL(p,q) model I mean a model which assumes that a time series, Y(t) , can be represented by a linear function of p of its lagged values and q lags of another time series, X(t) .

  • Operating System: MacOS Catalina 10.15
  • brms Version: 2.10.0

Feel free to open an issue on github proposing such a model, but not sure when I find the time to look at this in more detail.

1 Like

Sorry to revive this, but I just wanted to ask you @mansillo if you managed to find a nice way to do this in brms? I have opened an issue here Ability to use stats::filter or other distributed lag models in brms formulas · Issue #1127 · paul-buerkner/brms · GitHub

I have tried to do it but couldn’t make it work so I just wanted to make sure I am not missing something obvious.

Also interested in having this implemented in brms.