Best practices to go from 1000s of netcdf files to analyses on a HPC cluster?

Hi @selipot. You’ve already got fab advice here and hopefully seen how this is a challenge that faces much of the community. Encase it helps my colleague @kaedonkers has written a post about our recent attempt to do the same on a medium-sized dataset. To the cloud and back again,
Building a Zarr from bespoke data formats using Xarray and Iris
.

1 Like