Consider refactoring calculation of total VHF from partial data

[rmatsum836]

Per a conversation with @lisankim0321, the dictionary used to compute the total VHF from the partial data is a little clunky. Currently the key of the dictionary is a tuple of MDTraj.Atom objects, which was decided upon as these objects carry both atom name and element name information.

Perhaps it may be easier to have the user pass in two separate dictionaries. The first with atom name as key and g_r_t data as value. The second with atom name as key and element name as value