0

I am looking to implement HyperLogLog algorithm to count distinct users for different audience segments (or filters). I user Cassandra + Spark batch. Wondering if Cassandra provides any support for HyperLogLog type.

I could not find any plugin or anything related except http://vilkeliskis.com/blog/2013/12/28/hacking_cassandra.html which is a great attempt but unfinished.

Thanks for any possible hints!

Sammy
  • 151
  • 2
  • 6

1 Answers1

1

Spark 2.0 has a native implementation of HyperLogLog++.

Omer van Kloeten
  • 11,800
  • 9
  • 42
  • 53