Group Splittings Based on Multiplicity

Each splitting should only depend on the multiplicity and flavor of the jets on each side of the splitting, not on the subsequent branching pattern.

So splittings can be grouped based on the flavor and multiplicity of the two childer

eg:

1b0j/1b0j: bb 1b1j/1b0j: (bj)b 1b2j/0b1j ((bj)j)j : ((jj)b)j 1b2j/1b0j ((jj)b)b: ((bj)j)b (1,2)/(1,1) ((bj)j)(bj) (0,3)/(1,0) ((jj)j)b

This is a good way to go ! Implemented it in the Declustering algorithm