Hi,
Is it possible to run a Mahout Hadoop job directly from a column family instead of CFS? For example, I have my item,user,preferences data stored in column in a keyspace and would like to read those as the input vs having to write those out to CFS first.
Thanks,
Thomas
