09-30-2015 10:53 AM
I want to know what is the right way to create a feature list for each element of input stream. I have more than 100 featues that I want to create for each input data that we are getting from the Kafka spark stream. Since the features are a lot I want to run these features in parallel. Becasue I cannot call RDD inside another RDD. Should I write my own multithreaded application to process features.