djgroen
commented
1 year ago After discussions with Irene, we have decided to descope this task, due to concerns for emerging bias in the model.
Instead, we will run ensemble simulations to account for uncertainty, and make our forecasts reproducible.
Add a flag that changes the random choice to simply picking the highest weight.