Taken from Stephanie Wang's gist: https://gist.github.com/stephanie-wang/7400fd3b50bef0c8ce2d5b29f58c2530#file-pyspark_on_ray-py

dask_on_ray.py

import ray
from ray.util.dask import ray_dask_get

import dask
import dask.dataframe as dd 
import pandas as pd
import numpy as np

dask.config.set(scheduler=ray_dask_get) # Sets Ray as the default backend.

ray.init()
df = pd.DataFrame(np.random.randint(0, 100, size=(2**10, 2**8)))
df = dd.from_pandas(df, npartitions=10)
print(df.head(10))