markkoob

I would like to build a high level graph which performs several RandomizedSearchCVs across different estimator types in parallel, possibly with different data sets, using a distributed system. What I...

I'm not certain I've tried fitting with dask.DataFrame types on the distributed client with a delayed fit, but I'm pretty sure I've tried most other combinations. I'll give that one...

Okay I've been able to reproduce a couple of things with the LocalCluster. Hopefully that means I'm just doing something wrong! ``` import numpy as np import pandas as pd...