Hi everyone. I would like to know if someone can direct me to a information on how to use nixtla’s model architecture in relative time problems. Basically I have a set of geographically different series, but each in respect to a device recording that starts given a certain ecological cue. So I have the event, the lag before, and the lag after (this one that goes up to a point of “convergence” to another cue. I already imagine adapting the current structure I see in the documentation without much problems, but I wonder if someone may point any issues that I should be aware once dealing with this type of modelling.