Closed DEHewitt closed 3 years ago
Currently using default setting of dask to control the chunksizes which may not be the most efficient (for time or memory). This can be controlled via the argument field_chunksize when the field is defined in FieldSet.from_netcdf(). See https://nbviewer.jupyter.org/github/OceanParcels/parcels/blob/master/parcels/examples/documentation_MPI.ipynb for a full explanation.
dask
field_chunksize
FieldSet.from_netcdf()
Currently using default setting of
daskto control the chunksizes which may not be the most efficient (for time or memory). This can be controlled via the argumentfield_chunksizewhen the field is defined inFieldSet.from_netcdf(). See https://nbviewer.jupyter.org/github/OceanParcels/parcels/blob/master/parcels/examples/documentation_MPI.ipynb for a full explanation.