Hi folks. Just a bit confused with what's required.

The sample submission outline in the "Evaluation" section outlines a bunch of columns (1 per beam, it seems), and the "Data" tab mentions that we should be forecasting DLThpVol. However, the sample submission file has only 2 columns (not many), and there are placeholders for forecasts for the other time series (DLThpTime, DLPRB, and MR_number).

Should we follow the sample submission's format and forecast for all the time series, or the format mentioned in the Evaluation's "Data and Code Submission Requirements" section and just forecast DLThpVol?

Then if we forecast all the time series, is the MAE averaged evenly across the different time series?