William

I see that `skein.Service()` allows for a `nodes` parameter, but `skein.Master()` does not. Is there any way to limit which nodes the ApplicationMaster runs on?

Add better partition string escaping

3

Based on the https://github.com/nils-braun/dask-sql/issues/179 this PR adds string escaping to non-integer partition values. Partitions like: ``` forecasted_business_day=2021-05-27 ``` will now be wrapped like ``` forecasted_business_day="2021-05-27" ``` when executing the SQL

random vs. zero initialization of weights

4

Perhaps I'm mistaken, but it seems you initialize (the DBN) weights to all zeros before training, rather than doing a random weight initialization. Is there a reason for this? It...

YarnCluster.shutdown() Won't Work on EMR, results in `concurrent.futures._base.CancelledError`

1

As I was working on #135 I've encountered an issue where the YARN cluster fails to shutdown when asked. Code to reproduce: ``` from dask_yarn import YarnCluster cluster = YarnCluster("python:///usr/bin/python3")...

Dask Scheduler host/port Not Written to Skein Key-Value Storage When YARN Application Restarts

5

We are running a Dask cluster using Dask-Yarn on AWS EMR. I am trying to make it more robust so we're running the Scheduler in local mode. To make the...