Create API for reusing subset indexing on other datasets

[mpiannucci]

This is basically ripping out the core of the existing subset methods and exposing the ability to only get the indexing logic or routine so it can be cached or reused by a user at a later date.