Q: Multicore support?

[dahnj]

Q: Multicore support?

[DahnJ]

Hi @diehl in general the answer is no, H3-Pandas only adds h3-related functionality to Pandas Dataframes and as such does not suport concurrency.

However, you can use Dask and Dask-Geopandas to achieve this fairly easily. See the section "Dask-GeoPandas example" in this example.

[DahnJ]

Sounds good!

By the way, with Dask distributed, you don't need to embed the import into your functions: since you have a stable set of workers, you can simply import h3pandas once on them after startup.

Sample code

from dask.distributed import Client

client = Client()

def import_h3pandas():
    import h3pandas

client.run(import_h3pandas)

[diehl]

Thanks @DahnJ - That is exactly the route I'm heading down now. Going to use Coiled to parallelize the computation in the cloud.

[diehl]

I appreciate that pointer @DahnJ!