Fix bugs related to handling of transformations without Coulomb changes and (certain) plots of free energy versus time

[davidlmobley]

This should resolve issues #64 and issue #70 by:

• Fixing an issue with the number of time points in cumulative free energy plots being incorrectly computed (causing an IndexError) as in index out of bounds error with some values of bForwrev #64
• Fixing an issue where the maximum lambda index for the vdW free energy component was incorrectly assigned when no Coulomb transformation was present (making the data in Lambda index out of bounds when accumulating free energies #70 at least process without an IndexError).

The larger issue raised by the data in #70 - a transformation which changes masses, van der Waals, and bonded interactions but not Coulomb interactions - will still need to be dealt with and is mentioned in #65 . Specifically, the code needs to be generalized to handle the case that other free energy components are changing. This will be dealt with in a separate PR.

@Matteoa and @driesvr - could you check that this resolves your issues?

[davidlmobley]

@gduarter - can you check that this resolves your issue?

[davidlmobley]

@gduarter told me this is resolved, and the authors of #64 and #70 also responded favorably, so I'm merging.