ParallelDataCollection.get_parallel_datacollections

coinmetrics._data_collection.ParallelDataCollection.get_parallel_datacollections()

Materialize the cartesian product of every parallelized URL parameter (and, optionally, every time / height increment) into one DataCollection per combination.

For example, calling:

client.get_asset_metrics(assets="btc,eth,algo", ...) \
    .parallel() \
    .get_parallel_datacollections()

returns three data collections split by asset. Adding a second parallelization axis:

client.get_asset_metrics(
    assets="btc,eth,algo",
    metrics="volume_reported_spot_usd_1d,volume_trusted_spot_usd_1d",
).parallel(parallelize_on=["metrics", "assets"]) \
 .get_parallel_datacollections()

returns six combinations. Layering a time_increment on top:

client.get_asset_metrics(
    assets="btc,eth,algo",
    metrics="volume_reported_spot_usd_1d,volume_trusted_spot_usd_1d",
    start_time="2023-01-01",
    end_time="2023-02-01",
).parallel(
    parallelize_on=["metrics", "assets"],
    time_increment=timedelta(weeks=2),
).get_parallel_datacollections()

returns twelve, because each metric / asset pair is also split into two-week windows.

• Returns: One DataCollection per combination of parallelized parameters and time / height increment.

Returns:

• List[DataCollection]