spark
Runs a timer task to periodically clean up metadata (e.
A class for tracking the statistics of a set of numbers (count, mean and variance) in a numerically robust way.
This is a custom implementation of scala.