Averaging over molecules of the same type

[skfegan]

We would like the entropy values to be averaged over molecules of the same type.

The difficulties are (1) defining "same type" and (2) automating the finding of the types.

(1) "same type" molecules should be of the same chemistry - for example it is obvious that water and benzene are two different types. But the environment could also play a role, for example water in the bulk could be a different type to water next to a protein. There is the question of how detailed you consider the environment (i.e. are all waters next to proteins considered the same, or are there different types for waters next to different amino acid residues)

(2) Classifying and searching for molecules of the same type can be computationally expensive (neighbour lists, etc) and clearly identifying the different environments is easier in chemical intuition than code.