Hi All,

I was replicating the examples on https://github.com/antoinecarme/pyaf/blob/master/notebooks_sandbox/croston.ipynb

But is looks like to get a forecast, we need the q (the “demand”) and a (the “inter-arrival time”) of the day-of-interest.

df2['forecast'] = coeff * df2['q_est'] / df2['a_est']

Does this mean if I want to do a 7-day-out forecast, I need to forecast the q and a first (using arima/ets/bsts...) and then use this formula to calculate the forecast value?

All the best,

Kathy