The netcdf reading in database.py is very slow, at least with the version that I have (latest version available on fedora 38).
netCDF4.__version__: '1.6.3'
I will try to use xarray in the Netcdf._load() method to speed up the process. There is also a heavy loop over all lead times and ensemble members which can be replaced by stacking dimensions.
The netcdf reading in
database.pyis very slow, at least with the version that I have (latest version available on fedora 38).netCDF4.__version__: '1.6.3'I will try to use xarray in the
Netcdf._load()method to speed up the process. There is also a heavy loop over all lead times and ensemble members which can be replaced by stacking dimensions.