Need to do an analysis on internal SparkML methods to decide if it's ok to call `LocalTransformer.transform` from multiple threads
Need to do an analysis on internal SparkML methods to decide if it's ok to call
LocalTransformer.transformfrom multiple threads