Hi, I have provided a PR around 2 months back to improve the performance of decision tree by allowing flexible user provided storage class for intermediate data. I have posted few questions about handling backward compatibility but there is no answers from long.
Can anybody help me to move this forward? The below is the link to PR https://github.com/apache/spark/pull/17972 -- Regards, Madhukara Phatak http://datamantra.io/