where
DataArray.
persist
Trigger computation in constituent dask arrays
This keeps them as dask arrays but encourages them to keep data in memory. This is particularly useful when on a distributed machine. When on a single machine consider using .compute() instead.
.compute()