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
combine_dataset function in hysplit.py has a couple bugs.
It does not properly combine datasets in which the time coordinates are not the same. In the hysplit DataArrays the time coordinate are datetime64 objects. The xarray align method does not seem to be able to align these properly. The fix is to change the time coordinate to integers before aligning and then convert back to datetime objects after the datasets are merged.
It also is not properly combining datasets when both the source and ensemble dimensions are greater than 1. Most use cases have focused on when we are looking at a source term ensemble with one meteorological dataset (source dimension > 1 and ens dimension=1) or a meteorological ensemble (source dimension=1 and ensemble dimension > 1). However it should also be able to produce a combined dataset when both these dimensions > 1.
The text was updated successfully, but these errors were encountered:
combine_dataset function in hysplit.py has a couple bugs.
It does not properly combine datasets in which the time coordinates are not the same. In the hysplit DataArrays the time coordinate are datetime64 objects. The xarray align method does not seem to be able to align these properly. The fix is to change the time coordinate to integers before aligning and then convert back to datetime objects after the datasets are merged.
It also is not properly combining datasets when both the source and ensemble dimensions are greater than 1. Most use cases have focused on when we are looking at a source term ensemble with one meteorological dataset (source dimension > 1 and ens dimension=1) or a meteorological ensemble (source dimension=1 and ensemble dimension > 1). However it should also be able to produce a combined dataset when both these dimensions > 1.
The text was updated successfully, but these errors were encountered: