Update documentation for ParticleSet.from_particlefile

[claudiofgcardoso]

Hello all,

The addition of zarr as an output format (#1165) is a great step forward, as it significantly improves the storage efficiency of my large particle files! It is also important to note that the initialisation of a ParticleSet from a .zarr particle file is also possible. So I suggest to update the parcels documentation for ParticleSet.from_particlefile() so that other users know that this functionality is also possible with zarr.

Cláudio

[erikvansebille]

Hi @claudiofgcardoso, thanks for the comment. Yes, it would be a good addition to also support zarr in ParticleSet.from_particlefile(). But I'll do that as part of the larger refactoring project this summer, where I'm implementing dumping directly to zarr during a pset.execute() (instead of dumping to temporary npy files).

There are at least three advantages of using zarr directly:

1. It doesn't require a conversion step at the end of the run anymore (and thus fixes issues like Adding particles to pset in MPI run #1193 (comment)).
2. It will be much easier to inspect data when the execute is still running.
3. It will hopefully also reduce file size of the output

See the branch at https://github.com/OceanParcels/parcels/tree/dump_to_zarr to check progress. It's still a bit too early for a Pull Request, but hopefully this is a complete feature by the end of August!

[claudiofgcardoso]

Hi @erikvansebille, glad to read that Parcels is migrating to zarr and I'm looking forward for this merge! There's no need to keep this issue open so I'll close it :)

[erikvansebille]

Thanks @claudiofgcardoso! FYI, the branch is now also a PR, see #1199