You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am wondering why convert_trajectory_to_xarray outputs a xarray with dimensions [prediction_timedelta, time, ...]. This seems odd to me as we can concatenate along the time dimension. The timedelta between two steps can be computed via time[i + k] - time[i], so I do not see a reason to have this distinction between prediction_timedelta and time. This is also in contrast to the data stored for training (original ERA5 as well as predictions from deterministic models).