Initial dask implementation

[kreczko]

Adding the ability to execute via Dask, will require modifications to the original workflow:

• add ability to configure Dask from the config file
• add extra stage that submits the workflow to dask (this can be done by injecting such a stage into the YAML before it is parsed)
• add a different executor, parsing the configuration (e.g. n_workers, threads_per_worker)

This example code demonstrates the necessary wrappers to make Dask work with prefect

from prefect import flow, task
from prefect_dask import DaskTaskRunner

....

def dask_flow():
    prefect_future = stage.submit()
    return prefect_future.result()

client = Client()
    wrapper = flow(
        dask_flow,
        task_runner=DaskTaskRunner(
            cluster_kwargs={"n_workers": 2, "threads_per_worker": 1},
        )
    )
    wrapper()

[kreczko]

Fixed in #34