Another feature request: Rather than just calling

StatsForecast().forecast(duration)

, it would be handy to be able to pass in the specific values of

ds

that I want to predict for. for instance, I might have trained a weekly forecaster, but want a prediction mid-week.

Maybe I'm getting this wrong, but wouldn't be possible to just set duration for a period in the future and then filter the values ds values you need ?

not if you have trained a model on one time scale and want a prediction for a timescale that is shorter than that. ie you have monthly data and want a weekly forecast

Thanks! Now I get it.