ml_drought
ml_drought copied to clipboard
init commit forecast engineer ...
Create an engineer for specifically dealing with forecast data:
This takes the forecast variables for the target timestep and the target variables for all previous timesteps (not the target timestep).
This produces a dataset:
t-3 t-2 t-1 t=0 | y (t=0)
--------------------------------
X Y X Y X Y fcast | Y
| Y