More efficient set cardinality measurement

[cespare]

Can use hyperloglog like statsite.

Other links:

http://static.googleusercontent.com/external_content/untrusted_dlcp/research.google.com/en/us/pubs/archive/40671.pdf http://metamarkets.com/2012/fast-cheap-and-98-right-cardinality-estimation-for-big-data/ http://blog.aggregateknowledge.com/2012/10/25/sketch-of-the-day-hyperloglog-cornerstone-of-a-big-data-infrastructure/